zotero/scrapers.sql

5614 lines
175 KiB
MySQL
Raw Normal View History

2006-08-19 23:15:38 +00:00
-- 53
-- Set the following timestamp to the most recent scraper update date
REPLACE INTO "version" VALUES ('repository', STRFTIME('%s', '2006-08-15 15:42:00'));
REPLACE INTO "translators" VALUES ('96b9f483-c44d-5784-cdad-ce21b984fe01', '2006-08-11 11:18:00', 4, 'Amazon.com', 'Simon Kornblith', '^http://www\.amazon\.com/(?:gp/(?:product|search)/|exec/obidos/search-handle-url/|s/)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var searchRe = new RegExp(''^http://www\.amazon\.com/(gp/search/|exec/obidos/search-handle-url/|s/)'');
if(searchRe.test(doc.location.href)) {
return "multiple";
} else {
return "book";
}
}
',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'function scrape(doc) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item("book");
// Retrieve authors
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
try {
var xpath = ''/html/body/table/tbody/tr/td[2]/form/div[@class="buying"]/a/text()[1]'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
var elmt;
while(elmt = elmts.iterateNext()) {
newItem.creators.push(Scholar.Utilities.cleanAuthor(elmt.nodeValue, "author"));
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
} catch(ex) {Scholar.Utilities.debug(ex);}
// Retrieve data from "Product Details" box
var xpath = ''/html/body/table/tbody/tr/td[2]/table/tbody/tr/td[@class="bucket"]/div[@class="content"]/ul/li'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
var elmt;
newItem.extra = "";
while(elmt = elmts.iterateNext()) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
try {
var attribute = Scholar.Utilities.cleanString(doc.evaluate(''./B[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue);
var value = Scholar.Utilities.getNodeString(doc, elmt, ''./descendant-or-self::*[name() != "B"]/text()'', nsResolver);
if(value) {
value = Scholar.Utilities.cleanString(value);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(attribute == "Publisher:") {
if(value.lastIndexOf("(") != -1) {
var date = value.substring(value.lastIndexOf("(")+1, value.length-1);
jsDate = new Date(date);
if(!isNaN(jsDate.valueOf())) {
date = Scholar.Utilities.dateToSQL(jsDate);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
newItem.date = date;
value = value.substring(0, value.lastIndexOf("(")-1);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(value.lastIndexOf(";") != -1) {
newItem.edition = value.substring(value.lastIndexOf(";")+2, value.length);
value = value.substring(0, value.lastIndexOf(";"));
}
newItem.publisher = value;
} else if(attribute == "ISBN:") {
newItem.ISBN = value;
} else if(value.substring(value.indexOf(" ")+1, value.length) == "pages") {
newItem.pages = value.substring(0, value.indexOf(" "));
} else if(attribute != "Average Customer Review:") {
if(attribute == "In-Print Editions:") {
value = value.replace(" | All Editions", "");
} else {
value = value.replace(/\([^)]*\)/g, "");
}
newItem.extra += attribute+" "+value+"\n";
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} catch(ex) {}
}
if(newItem.extra) {
newItem.extra = newItem.extra.substr(0, newItem.extra.length-1);
}
newItem.attachments.push({title:"Amazon.com Product Page", document:doc});
var xpath = ''/html/body/table/tbody/tr/td[2]/form/div[@class="buying"]/b[@class="sans"]/text()[1]'';
var title = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue;
title = Scholar.Utilities.cleanString(title);
if(title.lastIndexOf("(") != -1 && title.lastIndexOf(")") == title.length-1) {
title = title.substring(0, title.lastIndexOf("(")-1);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.title = title;
newItem.complete();
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doWeb(doc, url) {
var searchRe = new RegExp(''^http://www\.amazon\.com/(gp/search/|exec/obidos/search-handle-url/|s/)'');
var m = searchRe.exec(doc.location.href)
if(m) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
// Why can''t amazon use the same stylesheets
var xpath;
if(m == "exec/obidos/search-handle-url/") {
xpath = ''//table[@cellpadding="3"]'';
} else {
xpath = ''//table[@class="searchresults"]'';
}
var searchresults = Scholar.Utilities.gatherElementsOnXPath(doc, doc, xpath, nsResolver);
var items = Scholar.Utilities.getItemArray(doc, searchresults, ''^http://www\.amazon\.com/(gp/product/|exec/obidos/tg/detail/)'', ''^(Buy new|Hardcover|Paperback|Digital)$'');
items = Scholar.selectItems(items);
if(!items) {
return true;
}
var uris = new Array();
for(var i in items) {
uris.push(i);
}
Scholar.Utilities.processDocuments(uris, function(doc) { scrape(doc) },
function() { Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
} else {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
scrape(doc);
}
}');
REPLACE INTO "translators" VALUES ('838d8849-4ffb-9f44-3d0d-aa8a0a079afe', '2006-06-26 16:01:00', 4, 'WorldCat', 'Simon Kornblith', '^http://(?:new)?firstsearch\.oclc\.org/WebZ/',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.title == ''FirstSearch: WorldCat Detailed Record'') {
return "book";
} else if(doc.title == ''FirstSearch: WorldCat List of Records'') {
return "multiple";
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}',
'function processURLs(urls) {
if(!urls.length) { // last url
Scholar.done();
return;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
var newUrl = urls.shift();
Scholar.Utilities.HTTP.doPost(newUrl,
''exportselect=record&exporttype=plaintext'', function(text) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var lineRegexp = new RegExp();
lineRegexp.compile("^([\\w() ]+): *(.*)$");
var newItem = new Scholar.Item("book");
newItem.extra = "";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var lines = text.split(''\n'');
for(var i=0;i<lines.length;i++) {
var testMatch = lineRegexp.exec(lines[i]);
if(testMatch) {
var match = newMatch;
var newMatch = testMatch
} else {
var match = false;
}
if(match) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// is a useful match
if(match[1] == ''Title'') {
var title = match[2];
if(!lineRegexp.test(lines[i+1])) {
i++;
title += '' ''+lines[i];
}
if(title.substring(title.length-2) == " /") {
title = title.substring(0, title.length-2);
}
newItem.title = title;
} else if(match[1] == ''Author(s)'') {
var yearRegexp = /[0-9]{4}-([0-9]{4})?/;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var authors = match[2].split('';'');
if(authors) {
newItem.creators.push(Scholar.Utilities.cleanAuthor(authors[0], "author", true));
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var j=1; j<authors.length; j+=2) {
if(authors[j-1].substring(0, 1) != ''('' && !yearRegexp.test(authors[j])) {
// ignore places where there are parentheses
newItem.creators.push(Scholar.Utilities.cleanAuthor(authors[j], "author", true));
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
2006-06-07 17:44:55 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else {
newItem.creators.push(Scholar.Utilities.cleanString(match[2]));
2006-06-07 17:44:55 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(match[1] == ''Publication'') {
// Don''t even try to deal with this. The WorldCat metadata is of poor enough quality that this isn''t worth it.
match[2] = Scholar.Utilities.cleanString(match[2]);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(match[2].substring(match[2].length-1) == '','') {
match[2] = match[2].substring(0, match[2].length-1);
}
newItem.publisher = match[2];
} else if(match[1] == ''Standard No'') {
var identifiers = match[2].split(/ +/);
var j=0;
while(j<(identifiers.length-1)) {
var type = identifiers[j].substring(0, identifiers[j].length-1);
var lastChar;
var value;
j++;
while(j<identifiers.length && (lastChar = identifiers[j].substring(identifiers[j].length-1)) != '':'') {
if(identifiers[j].substring(0, 1) != ''('') {
if(lastChar == '';'') {
value = identifiers[j].substring(0, identifiers[j].length-1);
} else {
value = identifiers[j];
}
if(type == "ISBN" || type == "ISSN") {
newItem[type] = value;
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
j++;
2006-06-07 17:44:55 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
} else if(match[1] == ''Year'') {
newItem.date = match[2];
} else if(match[1] == "Descriptor") {
if(match[2][match[2].length-1] == ".") {
match[2] = match[2].substr(0, match[2].length-1);
}
var tags = match[2].split("--");
for(var j in tags) {
newItem.tags.push(Scholar.Utilities.cleanString(tags[j]));
}
} else if(match[1] == "Accession No") {
newItem.accessionNumber = Scholar.Utilities.superCleanString(match[2]);
} else if(match[1] != "Database") {
newItem.extra += match[1]+": "+match[2]+"\n";
}
} else {
if(lines[i] != "" && lines[i] != "SUBJECT(S)") {
newMatch[2] += " "+lines[i];
}
2006-06-07 17:44:55 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(newItem.extra) {
newItem.extra = newItem.extra.substr(0, newItem.extra.length-1);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
processURLs(urls);
});
}
function doWeb(doc, url) {
var sessionRegexp = /(?:\?|\:)sessionid=([^?:]+)(?:\?|\:|$)/;
var numberRegexp = /(?:\?|\:)recno=([^?:]+)(?:\?|\:|$)/;
var resultsetRegexp = /(?:\?|\:)resultset=([^?:]+)(?:\?|\:|$)/;
var hostRegexp = new RegExp("http://([^/]+)/");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var sMatch = sessionRegexp.exec(url);
var sessionid = sMatch[1];
var hMatch = hostRegexp.exec(url);
var host = hMatch[1];
var newUri, exportselect;
if(doc.title == ''FirstSearch: WorldCat Detailed Record'') {
var publisherRegexp = /^(.*), (.*?),?$/;
var nMatch = numberRegexp.exec(url);
if(nMatch) {
var number = nMatch[1];
} else {
number = 1;
}
var rMatch = resultsetRegexp.exec(url);
if(rMatch) {
var resultset = rMatch[1];
} else {
// It''s in an XPCNativeWrapper, so we have to do this black magic
resultset = doc.forms.namedItem(''main'').elements.namedItem(''resultset'').value;
}
urls = [''http://''+host+''/WebZ/DirectExport?numrecs=10:smartpage=directexport:entityexportnumrecs=10:entityexportresultset='' + resultset + '':entityexportrecno='' + number + '':sessionid='' + sessionid + '':entitypagenum=35:0''];
} else {
var items = Scholar.Utilities.getItemArray(doc, doc, ''/WebZ/FSFETCH\\?fetchtype=fullrecord'', ''^(See more details for locating this item|Detailed Record)$'');
items = Scholar.selectItems(items);
if(!items) {
return true;
}
var urls = new Array();
for(var i in items) {
var nMatch = numberRegexp.exec(i);
var rMatch = resultsetRegexp.exec(i);
if(rMatch && nMatch) {
var number = nMatch[1];
var resultset = rMatch[1];
urls.push(''http://''+host+''/WebZ/DirectExport?numrecs=10:smartpage=directexport:entityexportnumrecs=10:entityexportresultset='' + resultset + '':entityexportrecno='' + number + '':sessionid='' + sessionid + '':entitypagenum=35:0'');
}
}
}
processURLs(urls);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('88915634-1af6-c134-0171-56fd198235ed', '2006-06-26 21:40:00', 4, 'LOC/Voyager WebVoyage', 'Simon Kornblith', 'Pwebrecon\.cgi',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var export_options = doc.forms.namedItem(''frm'').elements.namedItem(''RD'').options;
for(var i in export_options) {
if(export_options[i].text == ''Latin1 MARC''
|| export_options[i].text == ''Raw MARC''
|| export_options[i].text == ''UTF-8''
|| export_options[i].text == ''MARC (Unicode/UTF-8)''
|| export_options[i].text == ''MARC (non-Unicode/MARC-8)'') {
// We have an exportable single record
if(doc.forms.namedItem(''frm'').elements.namedItem(''RC'')) {
return "multiple";
} else {
return "book";
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}',
'function doWeb(doc, url) {
var postString = '''';
var form = doc.forms.namedItem(''frm'');
var newUri = form.action;
var multiple = false;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.forms.namedItem(''frm'').elements.namedItem(''RC'')) {
multiple = true;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var availableItems = new Object(); // Technically, associative arrays are objects
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
// Require link to match this
var tagRegexp = new RegExp();
tagRegexp.compile(''Pwebrecon\\.cgi\\?.*v1=[0-9]+\\&.*ti='');
// Do not allow text to match this
var rejectRegexp = new RegExp();
rejectRegexp.compile(''\[ [0-9]+ \]'');
var checkboxes = new Array();
var urls = new Array();
var tableRows = doc.evaluate(''/html/body/form/table/tbody/tr[td/input[@type="checkbox"]]'', doc, nsResolver, XPathResult.ANY_TYPE, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through table rows
var tableRow;
var i = 0;
while(tableRow = tableRows.iterateNext()) {
i++;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// CHK is what we need to get it all as one file
var input = doc.evaluate(''./td/input[@name="CHK"]'', tableRow, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
checkboxes[i] = input.value;
var links = tableRow.getElementsByTagName("a");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through links
for(var j=0; j<links.length; j++) {
if(tagRegexp.test(links[j].href)) {
var text = Scholar.Utilities.getNodeString(doc, links[j], ".//text()", null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(text) {
text = Scholar.Utilities.cleanString(text);
if(!rejectRegexp.test(text)) {
if(availableItems[i]) {
availableItems[i] += " "+text;
} else {
availableItems[i] = text;
}
}
}
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var items = Scholar.selectItems(availableItems);
if(!items) {
return true;
}
// add arguments for items we need to grab
for(var i in items) {
postString += "CHK="+checkboxes[i]+"&";
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var raw, unicode, latin1;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i=0; i<form.elements.length; i++) {
if(form.elements[i].type && form.elements[i].type.toLowerCase() == ''hidden'') {
postString += escape(form.elements[i].name)+''=''+escape(form.elements[i].value)+''&'';
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var export_options = form.elements.namedItem(''RD'').options;
for(var i=0; i<export_options.length; i++) {
if(export_options[i].text == ''Raw MARC''
|| export_options[i].text == ''MARC (non-Unicode/MARC-8)'') {
raw = i;
} if(export_options[i].text == ''Latin1 MARC'') {
latin1 = i;
} else if(export_options[i].text == ''UTF-8''
|| export_options[i].text == ''MARC (Unicode/UTF-8)'') {
unicode = i;
2006-06-25 18:17:00 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(unicode) {
var rd = unicode;
} else if(latin1) {
var rd = latin1;
} else if(raw) {
var rd = raw;
} else {
return false;
2006-06-25 18:17:00 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
postString += ''RD=''+rd+''&MAILADDY=&SAVE=Press+to+SAVE+or+PRINT'';
// No idea why this doesn''t work as post
Scholar.Utilities.HTTP.doGet(newUri+''?''+postString, function(text) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// load translator for MARC
var marc = Scholar.loadTranslator("import");
marc.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
marc.setString(text);
marc.translate();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.done();
})
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('d921155f-0186-1684-615c-ca57682ced9b', '2006-06-26 16:01:00', 4, 'JSTOR', 'Simon Kornblith', '^http://www\.jstor\.org/(?:view|browse|search/)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
// See if this is a seach results page
if(doc.title == "JSTOR: Search Results") {
return "multiple";
2006-06-25 18:17:00 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
2006-06-25 18:17:00 +00:00
// If this is a view page, find the link to the citation
var xpath = ''/html/body/div[@class="indent"]/center//a[@class="nav"]'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
if(elmts.iterateNext()) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return "journalArticle";
}
}',
'function getList(urls, each, done) {
2006-06-25 18:17:00 +00:00
var url = urls.shift();
Scholar.Utilities.HTTP.doGet(url, function(text) {
2006-06-25 18:17:00 +00:00
if(each) {
each(text);
}
if(urls.length) {
getList(urls, each, done);
2006-06-25 18:17:00 +00:00
} else if(done) {
done(text);
}
});
}
function getJSTORAttachment(viewURL) {
var viewRe = new RegExp("(^http://[^/]+/)view([^?]+)");
var m = viewRe.exec(viewURL);
if(m) {
return {url:m[1]+"cgi-bin/jstor/printpage"+m[2]+".pdf?dowhat=Acrobat",
mimeType:"application/pdf", title:"JSTOR Full Text PDF",
downloadable:true};
} else {
return false;
}
}
2006-06-20 17:06:41 +00:00
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function itemComplete(newItem, url) {
if(newItem.url) {
newItem.attachments.push({url:newItem.url, mimeType:"text/html",
title:"JSTOR Web-Readable Version"});
} else {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(newItem.ISSN) {
newItem.url = "http://www.jstor.org/browse/"+newItem.ISSN;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else {
newItem.url = url;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
2006-06-25 18:17:00 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var saveCitations = new Array();
var viewPages = new Array();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.title == "JSTOR: Search Results") {
var availableItems = new Object();
// Require link to match this
var tagRegexp = new RegExp();
tagRegexp.compile(''citationAction='');
var tableRows = doc.evaluate(''/html/body/div[@class="indent"]/table/tbody/tr[td/span[@class="printDownloadSaveLinks"]]'', doc, nsResolver, XPathResult.ANY_TYPE, null);
var tableRow;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through table rows
var tableView = new Array();
var tableSave = new Array();
var i = 0;
while(tableRow = tableRows.iterateNext()) {
i++;
var links = tableRow.getElementsByTagName("a");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through links
for(var j=0; j<links.length; j++) {
if(links[j].href.indexOf("citationAction=") != -1) {
tableSave[i] = links[j].href;
var link = doc.evaluate(''.//a[strong]'', tableRow, null, XPathResult.ANY_TYPE, null).iterateNext();
if(link) {
tableView[i] = link.href;
}
var text = doc.evaluate(''.//strong/text()'', tableRow, null, XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(text && text.nodeValue) {
text = Scholar.Utilities.cleanString(text.nodeValue);
if(availableItems[i]) {
availableItems[i] += " "+text;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else {
availableItems[i] = text;
2006-06-25 18:17:00 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
}
var items = Scholar.selectItems(availableItems);
if(!items) {
return true;
}
for(var i in items) {
viewPages.push(tableView[i]);
saveCitations.push(tableSave[i].replace(''citationAction=remove'', ''citationAction=save''));
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
} else {
// If this is a view page, find the link to the citation
var xpath = ''/html/body/div[@class="indent"]/center//a[@class="nav"]'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
var saveCitation = elmts.iterateNext();
var viewSavedCitations = elmts.iterateNext();
if(saveCitation && viewSavedCitations) {
viewPages.push(url);
saveCitations.push(saveCitation.href.replace(''citationAction=remove'', ''citationAction=save''));
} else {
throw("Could not find citation save links");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
Scholar.Utilities.HTTP.doGet(''http://www.jstor.org/browse?citationAction=removeAll&confirmRemAll=on&viewCitations=1'', function() { // clear marked
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Mark all our citations
getList(saveCitations, null, function() { // mark this
Scholar.Utilities.HTTP.doGet(''http://www.jstor.org/browse/citations.txt?exportAction=Save+as+Text+File&exportFormat=cm&viewCitations=1'', function(text) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// get marked
var k = 0;
var lines = text.split("\n");
var haveStarted = false;
var newItemRe = /^<[0-9]+>/;
var newItem = new Scholar.Item("journalArticle");
newItem.attachments.push(getJSTORAttachment(viewPages[k]));
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i in lines) {
if(lines[i].substring(0,3) == "<1>") {
haveStarted = true;
} else if(newItemRe.test(lines[i])) {
itemComplete(newItem, url);
k++;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem = new Scholar.Item("journalArticle");
newItem.attachments.push(getJSTORAttachment(viewPages[k]));
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(lines[i].substring(2, 5) == " : " && haveStarted) {
var fieldCode = lines[i].substring(0, 2);
var fieldContent = Scholar.Utilities.cleanString(lines[i].substring(5))
if(fieldCode == "TI") {
if(fieldContent) {
newItem.title = fieldContent;
} else {
newItem.title = "[untitled]";
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(fieldCode == "AU") {
var authors = fieldContent.split(";");
for(j in authors) {
if(authors[j]) {
newItem.creators.push(Scholar.Utilities.cleanAuthor(authors[j], "author", true));
2006-06-07 16:48:03 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(fieldCode == "SO") {
newItem.publicationTitle = fieldContent;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(fieldCode == "VO") {
newItem.volume = fieldContent;
} else if(fieldCode == "NO") {
newItem.issue = fieldContent;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(fieldCode == "SE") {
newItem.seriesTitle = fieldContent;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(fieldCode == "DA") {
var date = new Date(fieldContent.replace(".", ""));
if(isNaN(date.valueOf())) {
newItem.date = fieldContent;
} else {
newItem.date = Scholar.Utilities.dateToSQL(date);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
} else if(fieldCode == "PP") {
newItem.pages = fieldContent;
} else if(fieldCode == "EI") {
newItem.url = fieldContent;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(fieldCode == "IN") {
newItem.ISSN = fieldContent;
} else if(fieldCode == "PB") {
newItem.publisher = fieldContent;
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// last item is complete
if(haveStarted) {
itemComplete(newItem, url);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.done();
});
});
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
});
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('e85a3134-8c1a-8644-6926-584c8565f23e', '2006-06-26 16:01:00', 4, 'History Cooperative', 'Simon Kornblith', '^http://www\.historycooperative\.org/(?:journals/.+/.+/.+\.s?html$|cgi-bin/search.cgi)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.title == "History Cooperative: Search Results") {
return "multiple";
} else {
return "journalArticle";
}
}',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'function associateMeta(newItem, metaTags, field, scholarField) {
var field = metaTags.namedItem(field);
if(field) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem[scholarField] = field.getAttribute("content");
}
}
function scrape(doc) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item("journalArticle");
newItem.url = doc.location.href;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var month, year;
var metaTags = doc.getElementsByTagName("meta");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
associateMeta(newItem, metaTags, "Title", "title");
associateMeta(newItem, metaTags, "Journal", "publicationTitle");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
associateMeta(newItem, metaTags, "Volume", "volume");
associateMeta(newItem, metaTags, "Issue", "issue");
var author = metaTags.namedItem("Author");
if(author) {
var authors = author.getAttribute("content").split(" and ");
for(j in authors) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.creators.push(Scholar.Utilities.cleanAuthor(authors[j], "author"));
}
}
newItem.attachments.push({document:doc, title:"History Cooperative Full Text",
downloadable:true});
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
// don''t actually need date info for a journal article
var month = metaTags.namedItem("PublicationMonth");
var year = metaTags.namedItem("PublicationYear");
if(month && year) {
newItem.date = month.getAttribute("content")+" "+year.getAttribute("content");
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doWeb(doc, url) {
if(doc.title == "History Cooperative: Search Results") {
var items = Scholar.Utilities.getItemArray(doc, doc, ''^http://[^/]+/journals/.+/.+/.+\.html$'');
items = Scholar.selectItems(items);
if(!items) {
return true;
}
var uris = new Array();
for(var i in items) {
uris.push(i);
}
Scholar.Utilities.processDocuments(uris, function(doc) { scrape(doc) },
function() { Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
} else {
scrape(doc);
}
2006-06-25 19:32:49 +00:00
}');
REPLACE INTO "translators" VALUES ('4fd6b89b-2316-2dc4-fd87-61a97dd941e8', '2006-08-06 21:45:00', 4, 'InnoPAC', 'Simon Kornblith', '^http://[^/]+/(?:search/|record=)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// First, check to see if the URL alone reveals InnoPAC, since some sites don''t reveal the MARC button
var matchRegexp = new RegExp(''^(http://[^/]+/search/[^/]+/[^/]+/1\%2C[^/]+/)frameset(.+)$'');
if(matchRegexp.test(doc.location.href)) {
return "book";
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Next, look for the MARC button
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var xpath = ''//a[img[@src="/screens/marcdisp.gif" or @alt="MARC Display" or @src="/screens/regdisp.gif" or @alt="REGULAR RECORD DISPLAY"]]'';
var elmt = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(elmt) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return "book";
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Also, check for links to an item display page
var tags = doc.getElementsByTagName("a");
for(var i=0; i<tags.length; i++) {
if(matchRegexp.test(tags[i].href)) {
return "multiple";
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return false;
}',
'function scrape(marc, newDoc) {
var namespace = newDoc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var xpath = ''//pre/text()[1]'';
var text = newDoc.evaluate(xpath, newDoc, nsResolver,
XPathResult.ANY_TYPE, null).iterateNext().nodeValue;
var newItem = new Scholar.Item();
var record = new marc.record();
var linee = text.split("\n");
for (var i=0; i<linee.length; i++) {
if(!linee[i]) {
continue;
}
linee[i] = linee[i].replace(/[\xA0_\t]/g, " ");
var value = linee[i].substr(7);
if(linee[i].substr(0, 6) == " ") {
// add this onto previous value
tagValue += value;
} else {
if(linee[i].substr(0, 6) == "LEADER") {
// trap leader
record.leader = value;
} else {
if(tagValue) { // finish last tag
tagValue = tagValue.replace(/\|(.)/g, marc.subfieldDelimiter+"$1");
if(tagValue[0] != marc.subfieldDelimiter) {
tagValue = marc.subfieldDelimiter+"a"+tagValue;
}
// add previous tag
record.addField(tag, ind, tagValue);
}
var tag = linee[i].substr(0, 3);
var ind = linee[i].substr(4, 2);
var tagValue = value;
}
}
}
if(tagValue) {
tagValue = tagValue.replace(/\|(.)/g, marc.subfieldDelimiter+"$1");
if(tagValue[0] != marc.subfieldDelimiter) {
tagValue = marc.subfieldDelimiter+"a"+tagValue;
}
// add previous tag
record.addField(tag, ind, tagValue);
}
record.translate(newItem);
newItem.complete();
}
function pageByPage(marc, urls) {
Scholar.Utilities.processDocuments(urls, function(newDoc) {
scrape(marc.getTranslatorObject(), newDoc);
}, function() { Scholar.done() });
}
function doWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = doc.location.href;
var newUri;
// load translator for MARC
var marc = Scholar.loadTranslator("import");
marc.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var matchRegexp = new RegExp(''^(http://[^/]+/search/[^/]+/[^/]+/1\%2C[^/]+/)frameset(.+)$'');
var m = matchRegexp.exec(uri);
if(m) {
newUri = m[1]+''marc''+m[2];
} else {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var xpath = ''//a[img[@src="/screens/marcdisp.gif" or @alt="MARC Display"]]'';
var aTag = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(aTag) {
newUri = aTag.href;
} else {
var xpath = ''//a[img[@src="/screens/regdisp.gif" or @alt="REGULAR RECORD DISPLAY"]]'';
var aTag = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(aTag) {
scrape(marc.getTranslatorObject(), doc);
return;
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(newUri) { // single page
pageByPage(marc, [newUri]);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else { // Search results page
// Require link to match this
var tagRegexp = new RegExp();
tagRegexp.compile(''^http://[^/]+/search/[^/]+/[^/]+/1\%2C[^/]+/frameset'');
var checkboxes = new Array();
var urls = new Array();
var availableItems = new Array();
var tableRows = doc.evaluate(''//table[@class="browseScreen"]//tr[@class="browseEntry" or @class="briefCitRow" or td/input[@type="checkbox"]]'',
doc, nsResolver, XPathResult.ANY_TYPE, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through table rows
var i = 0;
while(tableRow = tableRows.iterateNext()) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// CHK is what we need to get it all as one file
var input = doc.evaluate(''./td/input[@type="checkbox"]'', tableRow,
nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(input) {
checkboxes[i] = input.name+"="+escape(input.value);
}
// get link
var links = doc.evaluate(''.//span[@class="briefcitTitle"]/a'', tableRow,
nsResolver, XPathResult.ANY_TYPE, null);
var link = links.iterateNext();
if(!link) {
var links = doc.evaluate(".//a", tableRow, nsResolver,
XPathResult.ANY_TYPE, null);
link = links.iterateNext();
}
if(link) {
urls[i] = link.href;
// Go through links
while(link) {
if(tagRegexp.test(link.href)) {
var text = Scholar.Utilities.getNodeString(doc, link,
".//text()", null);
if(text) {
text = Scholar.Utilities.cleanString(text);
if(availableItems[i]) {
availableItems[i] += " "+text;
} else {
availableItems[i] = text;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
link = links.iterateNext();
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
i++;
};
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var items = Scholar.selectItems(availableItems);
if(!items) {
return true;
}
var urlRe = new RegExp("^(https?://[^/]+(/search/[^/]+(?:/|$)))");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var m = urlRe.exec(urls[0]);
if(!m) {
throw("urlRe choked on "+urls[0]);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var clearUrl = m[0]+"?clear_saves=1";
var postUrl = m[0];
var exportUrl = m[1]+"++export/1,-1,-1,B/export";
var newUrls = new Array();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var postString = "";
var number = 0;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i in items) {
if(checkboxes[i]) {
postString += checkboxes[i]+"&";
number++;
}
var m = matchRegexp.exec(urls[i]);
if(!m) {
throw("matchRegexp choked on "+urls[i]);
}
newUrls.push(m[1]+"marc"+m[2]);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
if(postString && number > 1) {
postString += "save_func=save_marked";
Scholar.Utilities.HTTP.doGet(clearUrl, function() {
Scholar.Utilities.HTTP.doPost(postUrl, postString, function() {
Scholar.Utilities.HTTP.doPost(exportUrl, "ex_format=50&ex_device=45&SUBMIT=Submit", function(text) {
var notSpace = /[^\s]/
if(notSpace.test(text)) {
marc.setString(text);
marc.translate();
Scholar.done();
} else {
pageByPage(marc, newUrls);
}
});
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
});
});
} else {
pageByPage(marc, newUrls);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('add7c71c-21f3-ee14-d188-caf9da12728b', '2006-06-26 16:01:00', 4, 'SIRSI 2003+', 'Simon Kornblith', '/uhtbin/cgisirsi',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var xpath = ''//tr[th[@class="viewmarctags"]][td[@class="viewmarctags"]]'';
if(doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext()) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return "book";
}
var xpath = ''//td[@class="searchsum"]/table'';
if(doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext()) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return "multiple";
}
}',
'function scrape(doc) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var xpath = ''//tr[th[@class="viewmarctags"]][td[@class="viewmarctags"]]'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
var elmt = elmts.iterateNext();
if(!elmt) {
return false;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item("book");
newItem.extra = "";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
while(elmt) {
try {
var node = doc.evaluate(''./TD[1]/A[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(!node) {
var node = doc.evaluate(''./TD[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
}
if(node) {
var casedField = Scholar.Utilities.superCleanString(doc.evaluate(''./TH[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue);
field = casedField.toLowerCase();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var value = Scholar.Utilities.superCleanString(node.nodeValue);
if(field == "publisher") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.publisher = value;
} else if(field == "pub date") {
var re = /[0-9]+/;
var m = re.exec(value);
newItem.date = m[0];
} else if(field == "isbn") {
var re = /^[0-9](?:[0-9X]+)/;
var m = re.exec(value);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.ISBN = m[0];
} else if(field == "title") {
var titleParts = value.split(" / ");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.title = titleParts[0];
} else if(field == "publication info") {
var pubParts = value.split(" : ");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.place = pubParts[0];
} else if(field == "personal author") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.creators.push(Scholar.Utilities.cleanAuthor(value, "author", true));
} else if(field == "added author") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.creators.push(Scholar.Utilities.cleanAuthor(value, "contributor", true));
} else if(field == "corporate author") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.creators.push({lastName:author});
} else if(field == "subject term" || field == "corporate subject" || field == "geographic term") {
var subjects = value.split("--");
newItem.tags = newItem.tags.concat(subjects);
} else if(field == "personal subject") {
var subjects = value.split(", ");
newItem.tags = newItem.tags.push(value[0]+", "+value[1]);
} else if(value && field != "http") {
newItem.extra += casedField+": "+value+"\n";
}
}
} catch (e) {}
elmt = elmts.iterateNext();
}
if(newItem.extra) {
newItem.extra = newItem.extra.substr(0, newItem.extra.length-1);
}
var callNumber = doc.evaluate(''//tr/td[1][@class="holdingslist"]/text()'', doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
2006-06-26 01:08:59 +00:00
if(callNumber && callNumber.nodeValue) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.callNumber = callNumber.nodeValue;
2006-06-26 01:08:59 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
return true;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
if(!scrape(doc)) {
var checkboxes = new Array();
var urls = new Array();
var availableItems = new Array();
var tableRows = doc.evaluate(''//td[@class="searchsum"]/table[//input[@value="Details"]]'', doc, nsResolver, XPathResult.ANY_TYPE, null);
var tableRow = tableRows.iterateNext(); // skip first row
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through table rows
while(tableRow = tableRows.iterateNext()) {
var input = doc.evaluate(''.//input[@value="Details"]'', tableRow, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
var text = Scholar.Utilities.getNodeString(doc, tableRow, ''.//label/strong//text()'', nsResolver);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(text) {
availableItems[input.name] = text;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var items = Scholar.selectItems(availableItems);
if(!items) {
return true;
}
var hostRe = new RegExp("^http://[^/]+");
var m = hostRe.exec(doc.location.href);
var hitlist = doc.forms.namedItem("hitlist");
var baseUrl = m[0]+hitlist.getAttribute("action")+"?first_hit="+hitlist.elements.namedItem("first_hit").value+"&last_hit="+hitlist.elements.namedItem("last_hit").value;
var uris = new Array();
for(var i in items) {
uris.push(baseUrl+"&"+i+"=Details");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
Scholar.Utilities.processDocuments(uris, function(doc) { scrape(doc) },
function() { Scholar.done() }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}
}');
REPLACE INTO "translators" VALUES ('a77690cf-c5d1-8fc4-110f-d1fc765dcf88', '2006-06-26 16:01:00', 4, 'ProQuest', 'Simon Kornblith', '^http://proquest\.umi\.com/pqdweb\?((?:.*\&)?did=.*&Fmt=[0-9]|(?:.*\&)Fmt=[0-9].*&did=|(?:.*\&)searchInterface=)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.title == "Results") {
return "multiple";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else {
return "magazineArticle";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'function scrape(doc) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item();
var elmt;
2006-06-25 19:32:49 +00:00
// Title
var xpath = ''/html/body/span[@class="textMedium"]/table/tbody/tr/td[@class="headerBlack"]/strong//text()'';
newItem.title = Scholar.Utilities.getNodeString(doc, doc, xpath, nsResolver);
2006-06-25 19:32:49 +00:00
// Authors
var xpath = ''/html/body/span[@class="textMedium"]/table/tbody/tr/td[@class="textMedium"]/a/em'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
while(elmt = elmts.iterateNext()) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// there are sometimes additional tags representing higlighting
var author = Scholar.Utilities.getNodeString(doc, elmt, ''.//text()'', nsResolver);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(author) {
newItem.creators.push(Scholar.Utilities.cleanAuthor(author, "author", true));
}
2006-06-25 19:32:49 +00:00
}
// Other info
var xpath = ''/html/body/span[@class="textMedium"]/font/table/tbody/tr'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
while(elmt = elmts.iterateNext()) {
var field = Scholar.Utilities.superCleanString(doc.evaluate(''./TD[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue).toLowerCase();
2006-06-25 19:32:49 +00:00
if(field == "publication title") {
var publication = doc.evaluate(''./TD[2]/A[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
2006-06-25 19:32:49 +00:00
if(publication.nodeValue) {
newItem.publicationTitle = Scholar.Utilities.superCleanString(publication.nodeValue);
2006-06-25 19:32:49 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var place = doc.evaluate(''./TD[2]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
2006-06-25 19:32:49 +00:00
if(place.nodeValue) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.place = Scholar.Utilities.superCleanString(place.nodeValue);
2006-06-25 19:32:49 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var date = doc.evaluate(''./TD[2]/A[2]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
2006-06-25 19:32:49 +00:00
if(date.nodeValue) {
date = date.nodeValue;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var jsDate = new Date(Scholar.Utilities.superCleanString(date));
2006-06-25 19:32:49 +00:00
if(!isNaN(jsDate.valueOf())) {
date = Scholar.Utilities.dateToSQL(jsDate);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.date = date;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var moreInfo = doc.evaluate(''./TD[2]/text()[2]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
2006-06-25 19:32:49 +00:00
if(moreInfo.nodeValue) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
moreInfo = Scholar.Utilities.superCleanString(moreInfo.nodeValue);
2006-06-25 19:32:49 +00:00
var parts = moreInfo.split(";\xA0");
2006-06-25 19:32:49 +00:00
var issueRegexp = /^(\w+)\.(?: |\xA0)?(.+)$/
var issueInfo = parts[0].split(",\xA0");
for(j in issueInfo) {
var m = issueRegexp.exec(issueInfo[j]);
if(m) {
var info = m[1].toLowerCase();
if(info == "vol") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.volume = Scholar.Utilities.superCleanString(m[2]);
2006-06-25 19:32:49 +00:00
} else if(info == "iss" || info == "no") {
newItem.issue = Scholar.Utilities.superCleanString(m[2]);
2006-06-25 19:32:49 +00:00
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(parts[1] && Scholar.Utilities.superCleanString(parts[1]).substring(0, 3).toLowerCase() == "pg.") {
2006-06-25 19:32:49 +00:00
var re = /[0-9\-]+/;
var m = re.exec(parts[1]);
if(m) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.pages = m[0];
2006-06-25 19:32:49 +00:00
}
}
}
2006-06-25 19:32:49 +00:00
} else if(field == "source type") {
var value = doc.evaluate(''./TD[2]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
2006-06-25 19:32:49 +00:00
if(value.nodeValue) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
value = Scholar.Utilities.superCleanString(value.nodeValue).toLowerCase();
2006-06-25 19:32:49 +00:00
if(value.indexOf("periodical") >= 0) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.itemType = "magazineArticle";
2006-06-25 19:32:49 +00:00
} else if(value.indexOf("newspaper") >= 0) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.itemType = "newspaperArticle";
} else { // TODO: support thesis
newItem.itemType = "book";
2006-06-25 19:32:49 +00:00
}
}
2006-06-25 19:32:49 +00:00
} else if(field == "isbn" || field == "issn" || field == "issn/isbn") {
var value = doc.evaluate(''./TD[2]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
2006-06-25 19:32:49 +00:00
if(value) {
var type;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
value = Scholar.Utilities.superCleanString(value.nodeValue);
2006-06-25 19:32:49 +00:00
if(value.length == 10 || value.length == 13) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.ISBN = value;
2006-06-25 19:32:49 +00:00
} else if(value.length == 8) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.ISSN = value;
2006-06-25 19:32:49 +00:00
}
}
} else if(field == "document url") {
var value = doc.evaluate(''./TD[2]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(value) {
newItem.url = Scholar.Utilities.cleanString(value.nodeValue);
}
} else if(field == "proquest document id") {
var value = doc.evaluate(''./TD[2]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(value) {
newItem.accessionNumber = Scholar.Utilities.cleanString(value.nodeValue);
}
} else if(field == "subjects" || field == "people" || field == "locations") {
var subjects = doc.evaluate(".//a", elmt, nsResolver, XPathResult.ANY_TYPE, null);
var currentSubject;
while(currentSubject = subjects.iterateNext()) {
var subjectValue = Scholar.Utilities.getNodeString(doc, currentSubject, ".//text()", nsResolver);
subjectValue = Scholar.Utilities.superCleanString(subjectValue);
if(subjectValue) {
newItem.tags.push(subjectValue);
}
}
}
}
// magazineArticle -> journalArticle if issue and volume exist
if(newItem.itemType == "magazineArticle" && (newItem.issue || newItem.volume)) {
newItem.itemType = "journalArticle";
}
// figure out what we can attach
var attachArray = {
''//td[@class="textSmall"]//img[@alt="Full Text - PDF"]'':"ProQuest Full Text (PDF)",
''//td[@class="textSmall"]//img[@alt="Text+Graphics"]'':"ProQuest Full Text (HTML with Graphics)",
''//td[@class="textSmall"]//img[@alt="Full Text"]'':"ProQuest Full Text (HTML)",
''//td[@class="textSmall"]//img[@alt="Abstract"]'':"ProQuest Abstract"
}
for(var xpath in attachArray) {
var item = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(item) {
var title = attachArray[xpath];
if(item.parentNode.tagName.toLowerCase() == "a") {
// item is not this page
newItem.attachments.push({url:item.parentNode.href,
title:title, mimeType:(title == "ProQuest Full Text (PDF)" ? "application/pdf" : "text/html"),
downloadable:true});
} else {
// item is this page
newItem.attachments.push({document:doc, title:title, downloadable:true});
}
2006-06-25 19:32:49 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
2006-06-25 19:32:49 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
if(doc.title == "Results") {
var items = new Object();
// Require link to match this
var tagRegexp = new RegExp();
tagRegexp.compile(''^http://[^/]+/pqdweb\\?((?:.*&)?did=.*&Fmt=[12]|(?:.*&)Fmt=[12].*&did=)'');
var tableRows = doc.evaluate(''//tr[@class="rowUnMarked"]'',
doc, nsResolver, XPathResult.ANY_TYPE, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through table rows
var tableRow;
while(tableRow = tableRows.iterateNext()) {
var links = tableRow.getElementsByTagName("a");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through links
for(var j=0; j<links.length; j++) {
if(tagRegexp.test(links[j].href)) {
var text = doc.evaluate(''.//a[@class="bold"]/text()'', tableRow, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(text && text.nodeValue) {
text = Scholar.Utilities.cleanString(text.nodeValue);
items[links[j].href] = text;
}
break;
2006-06-25 19:32:49 +00:00
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
items = Scholar.selectItems(items);
if(!items) {
return true;
}
var uris = new Array();
for(var i in items) {
uris.push(i);
}
Scholar.Utilities.processDocuments(uris, function(doc) { scrape(doc) },
function() { Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
} else {
var fmtCheck = /(?:\&|\?)Fmt=([0-9]+)/i
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var m = fmtCheck.exec(doc.location.href);
if(m && (m[1] == "1" || m[1] == "2" || m[1] == "3")) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
scrape(doc);
} else if(m) {
Scholar.Utilities.loadDocument(doc.location.href.replace("Fmt="+m[1], "Fmt=1"), function(doc) { scrape(doc); Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}
2006-06-25 19:32:49 +00:00
}
}');
REPLACE INTO "translators" VALUES ('6773a9af-5375-3224-d148-d32793884dec', '2006-06-26 16:01:00', 4, 'InfoTrac College Edition', 'Simon Kornblith', '^http://infotrac-college\.thomsonlearning\.com/itw/infomark/',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.title.substring(0, 8) == "Article ") {
return "magazineArticle";
} else if(doc.title.substring(0, 10) == "Citations ") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return "multiple";
}
}',
'function extractCitation(url, elmts, title, doc) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item();
newItem.url = url;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(title) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.title = Scholar.Utilities.superCleanString(title);
}
while(elmt = elmts.iterateNext()) {
var colon = elmt.nodeValue.indexOf(":");
var field = elmt.nodeValue.substring(1, colon).toLowerCase();
var value = elmt.nodeValue.substring(colon+1, elmt.nodeValue.length-1);
if(field == "title") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.title = Scholar.Utilities.superCleanString(value);
} else if(field == "journal") {
newItem.publicationTitle = value;
} else if(field == "pi") {
parts = value.split(" ");
var date = "";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var field = null;
for(j in parts) {
firstChar = parts[j].substring(0, 1);
if(firstChar == "v") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.itemType = "journalArticle";
field = "volume";
} else if(firstChar == "i") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
field = "issue";
} else if(firstChar == "p") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
field = "pages";
var pagesRegexp = /p(\w+)\((\w+)\)/; // weird looking page range
var match = pagesRegexp.exec(parts[j]);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(match) { // yup, it''s weird
var finalPage = parseInt(match[1])+parseInt(match[2])
parts[j] = "p"+match[1]+"-"+finalPage.toString();
} else if(!newItem.itemType) { // no, it''s normal
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// check to see if it''s numeric, bc newspaper pages aren''t
var justPageNumber = parts[j].substr(1);
if(parseInt(justPageNumber).toString() != justPageNumber) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.itemType = "newspaperArticle";
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(!field) { // date parts at the beginning, before
// anything else
date += " "+parts[j];
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(field) {
isDate = false;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(parts[j] != "pNA") { // make sure it''s not an invalid
// page number
// chop of letter
newItem[field] = parts[j].substring(1);
} else if(!newItem.itemType) { // only newspapers are missing
// page numbers on infotrac
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.itemType = "newspaperArticle";
}
}
}
// Set type
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(!newItem.itemType) {
newItem.itemType = "magazineArticle";
}
if(date != "") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.date = date.substring(1);
}
} else if(field == "author") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.creators.push(Scholar.Utilities.cleanAuthor(value, "author", true));
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc) {
newItem.attachments.push({document:doc, title:"InfoTrac Full Text",
downloadable:true});
} else {
newItem.attachments.push({url:url, title:"InfoTrac Full Text",
mimeType:"text/html", downloadable:true});
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = doc.location.href;
if(doc.title.substring(0, 8) == "Article ") { // article
var xpath = ''/html/body//comment()'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
extractCitation(uri, elmts);
} else { // search results
var items = new Array();
var uris = new Array();
var elmts = new Array();
var tableRows = doc.evaluate(''/html/body//table/tbody/tr/td[a/b]'', doc, nsResolver,
XPathResult.ANY_TYPE, null);
var tableRow;
var javaScriptRe = /''([^'']*)'' *, *''([^'']*)''/
var i = 0;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through table rows
while(tableRow = tableRows.iterateNext()) {
var link = doc.evaluate(''./a'', tableRow, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
var m = javaScriptRe.exec(link.href);
if(m) {
uris[i] = "http://infotrac-college.thomsonlearning.com/itw/infomark/192/215/90714844w6"+m[1]+"?sw_aep=olr_wad"+m[2];
}
var article = doc.evaluate(''./b/text()'', link, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
items[i] = article.nodeValue;
// Chop off final period
if(items[i].substr(items[i].length-1) == ".") {
items[i] = items[i].substr(0, items[i].length-1);
}
elmts[i] = doc.evaluate(".//comment()", tableRow, nsResolver, XPathResult.ANY_TYPE, null);
i++;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
items = Scholar.selectItems(items);
if(!items) {
return true;
}
for(var i in items) {
extractCitation(uris[i], elmts[i], items[i]);
}
}
}');
REPLACE INTO "translators" VALUES ('b047a13c-fe5c-6604-c997-bef15e502b09', '2006-06-26 16:01:00', 4, 'LexisNexis', 'Simon Kornblith', '^http://web\.lexis-?nexis\.com/universe/(?:document|doclist)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var detailRe = new RegExp("^http://[^/]+/universe/document");
if(detailRe.test(doc.location.href)) {
return "newspaperArticle";
} else {
return "multiple";
}
}',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'function scrape(doc) {
var newItem = new Scholar.Item();
newItem.attachments.push({document:doc, title:"LexisNexis Full Text",
downloadable:true});
2006-06-25 20:09:27 +00:00
var citationDataDiv;
var divs = doc.getElementsByTagName("div");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i=0; i<divs.length; i++) {
2006-06-25 20:09:27 +00:00
if(divs[i].className == "bodytext") {
citationDataDiv = divs[i];
break;
}
}
centerElements = citationDataDiv.getElementsByTagName("center");
var elementParts = centerElements[0].innerHTML.split(/<br[^>]*>/gi);
newItem.publicationTitle = elementParts[elementParts.length-1];
2006-06-25 20:09:27 +00:00
var dateRegexp = /<br[^>]*>(?:<b>)?([A-Z][a-z]+)(?:<\/b>)? ([0-9]+, [0-9]{4})/;
var m = dateRegexp.exec(centerElements[centerElements.length-1].innerHTML);
if(m) {
var jsDate = new Date(m[1]+" "+m[2]);
newItem.date = Scholar.Utilities.dateToSQL(jsDate);
2006-06-25 20:09:27 +00:00
} else {
var elementParts = centerElements[centerElements.length-1].innerHTML.split(/<br[^>]*>/gi);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.date = elementParts[1];
2006-06-25 20:09:27 +00:00
}
var cutIndex = citationDataDiv.innerHTML.indexOf("<b>BODY:</b>");
if(cutIndex < 0) {
cutIndex = citationDataDiv.innerHTML.indexOf("<b>TEXT:</b>");
}
if(cutIndex > 0) {
citationData = citationDataDiv.innerHTML.substring(0, cutIndex);
} else {
citationData = citationDataDiv.innerHTML;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
citationData = Scholar.Utilities.cleanTags(citationData);
2006-06-25 20:09:27 +00:00
var headlineRegexp = /\n(?:HEADLINE|TITLE|ARTICLE): ([^\n]+)\n/;
var m = headlineRegexp.exec(citationData);
if(m) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.title = Scholar.Utilities.cleanTags(m[1]);
2006-06-25 20:09:27 +00:00
}
var bylineRegexp = /\nBYLINE: *(\w[\w\- ]+)/;
var m = bylineRegexp.exec(citationData);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(m) { // there is a byline; use it as an author
2006-06-25 20:09:27 +00:00
if(m[1].substring(0, 3).toLowerCase() == "by ") {
m[1] = m[1].substring(3);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.creators.push(Scholar.Utilities.cleanAuthor(m[1], "author"));
newItem.itemType = "newspaperArticle";
} else { // no byline; must be a journal
newItem.itemType = "journalArticle";
2006-06-25 20:09:27 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// other ways authors could be encoded
var authorRegexp = /\n(?:AUTHOR|NAME): ([^\n]+)\n/;
2006-06-25 20:09:27 +00:00
var m = authorRegexp.exec(citationData);
if(m) {
var authors = m[1].split(/, (?:and )?/);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i in authors) {
newItem.creators.push(Scholar.Utilities.cleanAuthor(authors[i].replace(" *", ""), "author"));
2006-06-25 20:09:27 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doWeb(doc, url) {
var detailRe = new RegExp("^http://[^/]+/universe/document");
if(detailRe.test(doc.location.href)) {
scrape(doc);
} else {
var items = Scholar.Utilities.getItemArray(doc, doc, "^http://[^/]+/universe/document");
items = Scholar.selectItems(items);
if(!items) {
return true;
}
var uris = new Array();
for(var i in items) {
uris.push(i);
}
Scholar.Utilities.processDocuments(uris, function(doc) { scrape(doc) },
function() { Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}
}');
REPLACE INTO "translators" VALUES ('cf87eca8-041d-b954-795a-2d86348999d5', '2006-06-26 16:01:00', 4, 'Aleph', 'Simon Kornblith', '^http://[^/]+/F(?:/[A-Z0-9\-]+(?:\?.*)?$|\?func=find)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var singleRe = new RegExp("^http://[^/]+/F/[A-Z0-9\-]+\?.*func=full-set-set.*\&format=[0-9]{3}");
if(singleRe.test(doc.location.href)) {
return "book";
} else {
var tags = doc.getElementsByTagName("a");
for(var i=0; i<tags.length; i++) {
if(singleRe.test(tags[i].href)) {
return "multiple";
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}',
'function doWeb(doc, url) {
var detailRe = new RegExp("^http://[^/]+/F/[A-Z0-9\-]+\?.*func=full-set-set.*\&format=[0-9]{3}");
var uri = doc.location.href;
var newUris = new Array();
if(detailRe.test(uri)) {
newUris.push(uri.replace(/\&format=[0-9]{3}/, "&format=001"))
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else {
var items = Scholar.Utilities.getItemArray(doc, doc, ''^http://[^/]+/F/[A-Z0-9\-]+\?.*func=full-set-set.*\&format=999'', ''^[0-9]+$'');
// ugly hack to see if we have any items
var haveItems = false;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i in items) {
haveItems = true;
break;
}
// If we don''t have any items otherwise, let us use the numbers
if(!haveItems) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var items = Scholar.Utilities.getItemArray(doc, doc, ''^http://[^/]+/F/[A-Z0-9\-]+\?.*func=full-set-set.*\&format=999'');
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
items = Scholar.selectItems(items);
if(!items) {
return true;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i in items) {
newUris.push(i.replace("&format=999", "&format=001"));
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
var translator = Scholar.loadTranslator("import");
translator.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
var marc = translator.getTranslatorObject();
Scholar.Utilities.processDocuments(newUris, function(newDoc) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = newDoc.location.href;
var namespace = newDoc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var xpath = ''/html/body/table/tbody/tr[td[1][@id="bold"]][td[2]]'';
var elmts = newDoc.evaluate(xpath, newDoc, nsResolver, XPathResult.ANY_TYPE, null);
var elmt;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var record = new marc.record();
while(elmt = elmts.iterateNext()) {
var field = Scholar.Utilities.superCleanString(doc.evaluate(''./TD[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var value = Scholar.Utilities.getNodeString(doc, elmt, ''./TD[2]//text()'', nsResolver);
if(field == "LDR") {
record.leader = value;
} else if(field != "FMT") {
value = value.replace(/\|([a-z]) /g, marc.subfieldDelimiter+"$1");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var code = field.substring(0, 3);
var ind = "";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(field.length > 3) {
ind = field[3];
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(field.length > 4) {
ind += field[4];
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
record.addField(code, ind, value);
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item();
record.translate(newItem);
newItem.complete();
}, function() { Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('774d7dc2-3474-2684-392c-f787789ec63d', '2006-06-26 16:01:00', 4, 'Dynix', 'Simon Kornblith', 'ipac\.jsp\?.*(?:uri=full=[0-9]|menu=search)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var detailsRe = new RegExp(''ipac\.jsp\?.*uri=full=[0-9]'');
if(detailsRe.test(doc.location.href)) {
return "book";
} else {
return "multiple";
}
}',
'function doWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = doc.location.href;
var detailsRe = new RegExp(''ipac\.jsp\?.*uri=full=[0-9]'');
2006-06-23 20:53:29 +00:00
var uris = new Array();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(detailsRe.test(uri)) {
uris.push(uri+''&fullmarc=true'');
} else {
var items = Scholar.Utilities.getItemArray(doc, doc, "ipac\.jsp\?.*uri=full=[0-9]|^javascript:buildNewList\\(''.*uri%3Dfull%3D[0-9]");
items = Scholar.selectItems(items);
if(!items) {
return true;
2006-06-23 20:53:29 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var buildNewList = new RegExp("^javascript:buildNewList\\(''([^'']+)");
var uris = new Array();
for(var i in items) {
var m = buildNewList.exec(i);
if(m) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
uris.push(unescape(m[1]+''&fullmarc=true''));
} else {
uris.push(i+''&fullmarc=true'');
}
}
}
var translator = Scholar.loadTranslator("import");
translator.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
var marc = translator.getTranslatorObject();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.Utilities.processDocuments(uris, function(newDoc) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = newDoc.location.href;
var namespace = newDoc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var xpath = ''//form/table[@class="tableBackground"]/tbody/tr/td/table[@class="tableBackground"]/tbody/tr[td[1]/a[@class="normalBlackFont1"]]'';
var elmts = newDoc.evaluate(xpath, newDoc, nsResolver, XPathResult.ANY_TYPE, null);
var elmt;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var record = new marc.record();
while(elmt = elmts.iterateNext()) {
var field = Scholar.Utilities.superCleanString(newDoc.evaluate(''./TD[1]/A[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var value = Scholar.Utilities.getNodeString(newDoc, elmt, ''./TD[2]/TABLE[1]/TBODY[1]/TR[1]/TD[1]/A[1]//text()'', nsResolver);
if(field == "LDR") {
record.leader = value;
} else if(field != "FMT") {
value = value.replace(/\$([a-z]) /g, marc.subfieldDelimiter+"$1");
var code = field.substring(0, 3);
var ind = "";
if(field.length > 3) {
ind = field[3];
if(field.length > 4) {
ind += field[4];
2006-06-23 19:22:24 +00:00
}
}
record.addField(code, ind, value);
2006-06-23 19:22:24 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item();
record.translate(newItem);
newItem.complete();
}, function() { Scholar.done() }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}');
2006-06-07 16:48:03 +00:00
REPLACE INTO "translators" VALUES ('63a0a351-3131-18f4-21aa-f46b9ac51d87', '2006-06-26 16:01:00', 4, 'VTLS', 'Simon Kornblith', '/chameleon(?:\?|$)',
'function detectWeb(doc, url) {
var node = doc.evaluate(''//tr[@class="intrRow"]/td/table/tbody/tr[th]'', doc, null, XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(node) {
return "multiple";
}
var node = doc.evaluate(''//a[text()="marc"]'', doc, null, XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(node) {
return "book";
}
}',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'function doWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = doc.location.href;
var newUris = new Array();
var marcs = doc.evaluate(''//a[text()="marc"]'', doc, nsResolver,
XPathResult.ANY_TYPE, null);
var record = marcs.iterateNext();
if(record && !marcs.iterateNext()) {
newUris.push(record.href);
} else {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Require link to match this
var tagRegexp = new RegExp();
tagRegexp.compile("/chameleon\?.*function=CARDSCR");
var items = new Array();
var tableRows = doc.evaluate(''//tr[@class="intrRow"]'', doc, nsResolver,
XPathResult.ANY_TYPE, null);
var tableRow
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through table rows
while(tableRow = tableRows.iterateNext()) {
var links = tableRow.getElementsByTagName("a");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through links
var url;
for(var j=0; j<links.length; j++) {
if(tagRegexp.test(links[j].href)) {
url = links[j].href;
break;
}
}
if(url) {
// Collect title information
var fields = doc.evaluate(''./td/table/tbody/tr[th]'', tableRow,
nsResolver, XPathResult.ANY_TYPE, null);
var field;
while(field = fields.iterateNext()) {
var header = doc.evaluate(''./th/text()'', fields[j], nsResolver,
XPathResult.ANY_TYPE, null).iterateNext();
if(header.nodeValue == "Title") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var value = Scholar.Utilities.getNodeString(doc, fields[j], ''./td//text()'', nsResolver);
if(value) {
items[url] = Scholar.Utilities.cleanString(value);
}
}
}
}
}
items = Scholar.selectItems(items);
if(!items) {
return true;
}
for(var i in items) {
Scholar.Utilities.debug(i.replace(/function=[A-Z]{7}/, "function=MARCSCR"));
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newUris.push(i.replace(/function=[A-Z]{7}/, "function=MARCSCR"));
}
}
var translator = Scholar.loadTranslator("import");
translator.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
var marc = translator.getTranslatorObject();
Scholar.Utilities.processDocuments(newUris, function(newDoc) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = newDoc.location.href
var namespace = newDoc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var record = new marc.record();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var xpath = ''//table[@class="outertable"]/tbody/tr[td[4]]'';
var elmts = newDoc.evaluate(xpath, newDoc, nsResolver,
XPathResult.ANY_TYPE, null);
while(elmt = elmts.iterateNext()) {
var field = doc.evaluate(''./TD[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue;
var ind1 = doc.evaluate(''./TD[2]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue;
var ind2 = doc.evaluate(''./TD[3]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue;
var value = doc.evaluate(''./TD[4]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue;
value = value.replace(/\\([a-z]) /g, marc.subfieldDelimiter+"$1");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
record.addField(field, ind1+ind2, value);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item();
record.translate(newItem);
newItem.complete();
}, function(){ Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}');
2006-06-07 16:48:03 +00:00
REPLACE INTO "translators" VALUES ('fb12ae9e-f473-cab4-0546-27ab88c64101', '2006-06-26 16:01:00', 4, 'DRA', 'Simon Kornblith', '/web2/tramp2\.exe/(?:see\_record/|authority\_hits/|goto/.*\?.*screen=Record\.html)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.location.href.indexOf("/authority_hits") > 0) {
return "multiple";
} else {
return "book";
}
}',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'function doWeb(doc, url) {
var checkItems = false;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.location.href.indexOf("/authority_hits") > 0) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
checkItems = Scholar.Utilities.gatherElementsOnXPath(doc, doc, "/html/body//ol/li", nsResolver);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(checkItems && checkItems.length) {
var items = Scholar.Utilities.getItemArray(doc, checkItems, ''https?://.*/web2/tramp2\.exe/see_record'');
items = Scholar.selectItems(items);
if(!items) {
return true;
}
var uris = new Array();
for(var i in items) {
uris.push(i);
}
} else {
var uris = new Array(doc.location.href);
}
2006-06-07 16:48:03 +00:00
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i in uris) {
var uri = uris[i];
var uriRegexp = /^(https?:\/\/.*\/web2\/tramp2\.exe\/)(?:goto|see\_record|authority\_hits)(\/.*)\?(?:screen=Record\.html\&)?(.*)$/i;
var m = uriRegexp.exec(uri);
if(uri.indexOf("/authority_hits") < 0) {
var newUri = m[1]+"download_record"+m[2]+"/RECORD.MRC?format=marc&"+m[3];
2006-06-07 16:48:03 +00:00
} else {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newUri = m[1]+"download_record"+m[2]+"/RECORD.MRC?format=marc";
2006-06-07 16:48:03 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Keep track of how many requests have been completed
var j = 0;
2006-06-07 16:48:03 +00:00
var translator = Scholar.loadTranslator("import");
translator.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
2006-06-07 16:48:03 +00:00
Scholar.Utilities.HTTP.doGet(newUri, function(text) {
translator.setString(text);
translator.translate();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
j++;
if(j == uris.length) {
Scholar.done();
}
});
2006-06-07 16:48:03 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}');
2006-06-07 17:44:55 +00:00
REPLACE INTO "translators" VALUES ('c0e6fda6-0ecd-e4f4-39ca-37a4de436e15', '2006-06-26 16:01:00', 4, 'GEAC', 'Simon Kornblith', '/(?:GeacQUERY|GeacFETCH[\:\?].*[&:]next=html/(?:record\.html|geacnffull\.html))',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.location.href.indexOf("/GeacQUERY") > 0) {
return "multiple";
} else {
return "book";
2006-06-07 17:44:55 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}',
'function doWeb(doc, url) {
var uri = doc.location.href;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uris = new Array();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(uri.indexOf("/GeacQUERY") > 0) {
var items = Scholar.Utilities.getItemArray(doc, doc, "(?:Geac)?FETCH[\:\?].*[&:]next=html/(?:record\.html|geacnffull\.html)");
items = Scholar.selectItems(items);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(!items) {
return true;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uris = new Array();
for(var i in items) {
var newUri = i.replace(/([:&])next=html\/geacnffull.html/, "$1next=html/marc.html");
newUri = newUri.replace(/([:&])next=html\/record.html/, "$1next=html/marc.html");
uris.push(newUri);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else {
var newUri = uri.replace(/([:&])next=html\/geacnffull.html/, "$1next=html/marc.html");
newUri = newUri.replace(/([:&])next=html\/record.html/, "$1next=html/marc.html");
uris.push(newUri);
2006-06-07 17:44:55 +00:00
}
var translator = Scholar.loadTranslator("import");
translator.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
var marc = translator.getTranslatorObject();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.Utilities.processDocuments(uris, function(newDoc) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = newDoc.location.href;
var namespace = newDoc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var record = new marc.record();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var elmts = newDoc.evaluate(''//pre/text()'', newDoc, nsResolver,
XPathResult.ANY_TYPE, null);
var elmt, tag, content;
var ind = "";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
while(elmt = elmts.iterateNext()) {
var line = elmt.nodeValue;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(line.substring(0, 6) == " ") {
content += " "+line.substring(6);
continue;
} else {
if(tag) {
record.addField(tag, ind, content);
}
}
line = line.replace(/[_\t\xA0]/g," "); // nbsp
tag = line.substr(0, 3);
if(tag[0] != "0" || tag[1] != "0") {
ind = line.substr(4, 2);
content = line.substr(7).replace(/\$([a-z])(?: |$)/g, marc.subfieldDelimiter+"$1");
} else {
if(tag == "000") {
tag = undefined;
record.leader = "00000"+line.substr(4);
} else {
content = line.substr(4);
}
2006-06-07 17:44:55 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
2006-06-07 17:44:55 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item();
record.translate(newItem);
newItem.complete();
}, function() { Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}');
2006-06-07 18:44:27 +00:00
REPLACE INTO "translators" VALUES ('5287d20c-8a13-6004-4dcb-5bb2b66a9cc9', '2006-06-26 16:01:00', 4, 'SIRSI -2003', 'Simon Kornblith', '/uhtbin/cgisirsi',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var elmts = doc.evaluate(''/html/body/form/p/text()[1]'', doc, nsResolver,
XPathResult.ANY_TYPE, null);
var elmt;
while(elmt = elmts.iterateNext()) {
if(Scholar.Utilities.superCleanString(elmt.nodeValue) == "Viewing record") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return "book";
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var xpath = ''//form[@name="hitlist"]/table/tbody/tr'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
if(elmts.iterateNext()) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return "multiple";
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}',
'function doWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
2006-06-07 18:44:27 +00:00
var nsResolver = namespace ? function(prefix) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if (prefix == ''x'') return namespace; else return null;
2006-06-07 18:44:27 +00:00
} : null;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = doc.location.href;
var recNumbers = new Array();
2006-06-07 18:44:27 +00:00
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var xpath = ''//form[@name="hitlist"]/table/tbody/tr'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
var elmt = elmts.iterateNext();
if(elmt) { // Search results page
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uriRegexp = /^http:\/\/[^\/]+/;
var m = uriRegexp.exec(uri);
var postAction = doc.forms.namedItem("hitlist").getAttribute("action");
var newUri = m[0]+postAction.substr(0, postAction.length-1)+"40"
2006-06-07 18:44:27 +00:00
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var titleRe = /<br>\s*(.*[^\s])\s*<br>/i;
2006-06-07 18:44:27 +00:00
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var items = new Array();
do {
var checkbox = doc.evaluate(''.//input[@type="checkbox"]'', elmt, nsResolver,
XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Collect title
var title = Scholar.Utilities.getNodeString(doc, elmt, "./td[2]/text()", nsResolver);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(checkbox && title) {
items[checkbox.name] = Scholar.Utilities.cleanString(title);
}
} while(elmt = elmts.iterateNext());
2006-06-07 18:44:27 +00:00
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
items = Scholar.selectItems(items);
if(!items) {
return true;
2006-06-07 18:44:27 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i in items) {
recNumbers.push(i);
2006-06-07 18:44:27 +00:00
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else { // Normal page
var uriRegexp = /^(.*)(\/[0-9]+)$/;
var m = uriRegexp.exec(uri);
var newUri = m[1]+"/40"
2006-06-07 18:44:27 +00:00
var elmts = doc.evaluate(''/html/body/form/p'', doc, nsResolver,
XPathResult.ANY_TYPE, null);
while(elmt = elmts.iterateNext()) {
var initialText = doc.evaluate(''./text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(initialText && initialText.nodeValue && Scholar.Utilities.superCleanString(initialText.nodeValue) == "Viewing record") {
recNumbers.push(doc.evaluate(''./b[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
break;
}
}
2006-06-07 21:26:55 +00:00
}
var translator = Scholar.loadTranslator("import");
translator.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
var marc = translator.getTranslatorObject();
Scholar.Utilities.loadDocument(newUri+''?marks=''+recNumbers.join(",")+''&shadow=NO&format=FLAT+ASCII&sort=TITLE&vopt_elst=ALL&library=ALL&display_rule=ASCENDING&duedate_code=l&holdcount_code=t&DOWNLOAD_x=22&DOWNLOAD_y=12&address=&form_type='', function(doc) {
var pre = doc.getElementsByTagName("pre");
var text = pre[0].textContent;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var documents = text.split("*** DOCUMENT BOUNDARY ***");
for(var j=1; j<documents.length; j++) {
var uri = newUri+"?marks="+recNumbers[j]+"&shadow=NO&format=FLAT+ASCII&sort=TITLE&vopt_elst=ALL&library=ALL&display_rule=ASCENDING&duedate_code=l&holdcount_code=t&DOWNLOAD_x=22&DOWNLOAD_y=12&address=&form_type=";
var lines = documents[j].split("\n");
var record = new marc.record();
var tag, content;
var ind = "";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i=0; i<lines.length; i++) {
var line = lines[i];
if(line[0] == "." && line.substr(4,2) == ". ") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(tag) {
content = content.replace(/\|([a-z])/g, marc.subfieldDelimiter+"$1");
record.addField(tag, ind, content);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else {
content += " "+line.substr(6);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
continue;
}
tag = line.substr(1, 3);
if(tag[0] != "0" || tag[1] != "0") {
ind = line.substr(6, 2);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
content = line.substr(8);
} else {
content = line.substr(7);
if(tag == "000") {
tag = undefined;
record.leader = "00000"+content;
Scholar.Utilities.debug("the leader is: "+record.leader);
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item();
record.translate(newItem);
newItem.complete();
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.done();
});
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('0f9fc2fc-306e-5204-1117-25bca009dffc', '2006-06-26 16:01:00', 4, 'TLC/YouSeeMore', 'Simon Kornblith', 'TLCScripts/interpac\.dll\?(?:.*LabelDisplay.*RecordNumber=[0-9]|Search|ItemTitles)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var detailRe = new RegExp("TLCScripts/interpac\.dll\?.*LabelDisplay.*RecordNumber=[0-9]");
if(detailRe.test(doc.location.href)) {
return "book";
} else {
return "multiple";
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}',
'function doWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var detailRe = new RegExp("TLCScripts/interpac\.dll\?.*LabelDisplay.*RecordNumber=[0-9]");
var uri = doc.location.href;
var newUris = new Array();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(detailRe.test(uri)) {
newUris.push(uri.replace("LabelDisplay", "MARCDisplay"));
} else {
var items = Scholar.Utilities.getItemArray(doc, doc, ''TLCScripts/interpac\.dll\?.*LabelDisplay.*RecordNumber=[0-9]'');
items = Scholar.selectItems(items);
if(!items) {
return true;
}
for(var i in items) {
newUris.push(i.replace("LabelDisplay", "MARCDisplay"));
}
}
var translator = Scholar.loadTranslator("import");
translator.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
var marc = translator.getTranslatorObject();
Scholar.Utilities.processDocuments(newUris, function(newDoc) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var uri = newDoc.location.href;
var namespace = newDoc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var record = new marc.record();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var elmts = newDoc.evaluate(''/html/body/table/tbody/tr[td[4]]'', newDoc, nsResolver,
XPathResult.ANY_TYPE, null);
var tag, ind, content, elmt;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
while(elmt = elmts.iterateNext()) {
tag = newDoc.evaluate(''./td[2]/tt[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue;
var inds = newDoc.evaluate(''./td[3]/tt[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
tag = tag.replace(/[\r\n]/g, "");
inds = inds.replace(/[\r\n\xA0]/g, "");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var children = newDoc.evaluate(''./td[4]/tt[1]//text()'', elmt, nsResolver,
XPathResult.ANY_TYPE, null);
var subfield = children.iterateNext();
var fieldContent = children.iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(tag == "LDR") {
record.leader = "00000"+subfield.nodeValue;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else {
content = "";
if(!fieldContent) {
content = subfield.nodeValue;
} else {
while(subfield && fieldContent) {
content += marc.subfieldDelimiter+subfield.nodeValue.substr(1, 1)+fieldContent.nodeValue;
var subfield = children.iterateNext();
var fieldContent = children.iterateNext();
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
record.addField(tag, inds, content);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item();
record.translate(newItem);
newItem.complete();
}, function() {Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('c54d1932-73ce-dfd4-a943-109380e06574', '2006-06-26 16:01:00', 4, 'Project MUSE', 'Simon Kornblith', '^http://muse\.jhu\.edu/(?:journals/[^/]+/[^/]+/[^/]+\.html|search/pia.cgi)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var searchRe = new RegExp("^http://[^/]+/search/pia\.cgi");
if(searchRe.test(url)) {
return "multiple";
} else {
return "journalArticle";
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}',
'function doWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var searchRe = new RegExp("^http://[^/]+/search/pia\.cgi");
if(searchRe.test(doc.location.href)) {
var items = new Array();
var attachments = new Array();
var pdfRe = /\.pdf$/i;
var htmlRe = /\.html$/i;
var tableRows = doc.evaluate(''/html/body/table[@class="navbar"]/tbody/tr/td/form/table'',
doc, nsResolver, XPathResult.ANY_TYPE, null);
var tableRow;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Go through table rows
while(tableRow = tableRows.iterateNext()) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// article_id is what we need to get it all as one file
var input = doc.evaluate(''./tbody/tr/td/input[@name="article_id"]'', tableRow, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
var link = doc.evaluate(''.//b/i/a/text()'', tableRow, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(input && input.value && link && link.nodeValue) {
items[input.value] = link.nodeValue;
var aTags = tableRow.getElementsByTagName("a");
// get attachments
attachments[input.value] = new Array();
for(var i=0; i<aTags.length; i++) {
if(pdfRe.test(aTags[i].href)) {
attachments[input.value].push({url:aTags[i].href,
title:"Project MUSE Full Text (PDF)",
mimeType:"application/pdf",
downloadable:true});
} else if(htmlRe.test(aTags[i].href)) {
attachments[input.value].push({url:aTags[i].href,
title:"Project MUSE Full Text (HTML)",
mimeType:"text/html",
downloadable:true});
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
items = Scholar.selectItems(items);
if(!items) {
return true;
}
try {
var search_id = doc.forms.namedItem("results").elements.namedItem("search_id").value;
} catch(e) {
var search_id = "";
}
var articleString = "";
var newAttachments = new Array();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i in items) {
articleString += "&article_id="+i;
newAttachments.push(attachments[i]);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
var savePostString = "actiontype=save&search_id="+search_id+articleString;
Scholar.Utilities.HTTP.doGet("http://muse.jhu.edu/search/save.cgi?"+savePostString, function() {
Scholar.Utilities.HTTP.doGet("http://muse.jhu.edu/search/export.cgi?exporttype=endnote"+articleString, function(text) {
Scholar.Utilities.debug(text);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// load translator for RIS
var translator = Scholar.loadTranslator("import");
translator.setTranslator("32d59d2d-b65a-4da4-b0a3-bdd3cfb979e7");
translator.setString(text);
translator.setHandler("itemDone", function(obj, item) {
if(item.notes && item.notes[0]) {
Scholar.Utilities.debug(item.notes);
item.extra = item.notes[0].note;
delete item.notes;
item.notes = undefined;
}
item.attachments = newAttachments.shift();
Scholar.Utilities.debug(item.attachments);
item.complete();
});
translator.translate();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.done();
}, function() {});
}, function() {});
Scholar.wait();
} else {
var newItem = new Scholar.Item("journalArticle");
newItem.url = url;
newItem.attachments.push({title:"Project MUSE Full Text (HTML)", mimeType:"text/html",
url:url, downloadable:true});
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var getPDF = doc.evaluate(''//a[text() = "[Access article in PDF]"]'', doc,
nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(getPDF) {
newItem.attachments.push({title:"Project MUSE Full Text (PDF)", mimeType:"application/pdf",
url:getPDF.href, downloadable:true});
}
var elmts = doc.evaluate(''//comment()'', doc, nsResolver,
XPathResult.ANY_TYPE, null);
var headerRegexp = /HeaderData((?:.|\n)*)\#\#EndHeaders/i
while(elmt = elmts.iterateNext()) {
if(elmt.nodeValue.substr(0, 10) == "HeaderData") {
var m = headerRegexp.exec(elmt.nodeValue);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var headerData = m[1];
}
}
// Use E4X rather than DOM/XPath, because the Mozilla gods have decided not to
// expose DOM/XPath to sandboxed scripts
var newDOM = new XML(headerData);
newItem.publicationTitle = newDOM.journal.text();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.volume = newDOM.volume.text();
newItem.issue = newDOM.issue.text();
newItem.date = newDOM.pubdate.text().toString();
if(!newItem.date) {
newItem.date = newDOM.year.text();
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.title = newDOM.doctitle.text();
newItem.ISSN = newDOM.issn.text();
// Do pages
var fpage = newDOM.fpage.text();
var lpage = newDOM.lpage.text();
if(fpage != "") {
newItem.pages = fpage;
if(lpage) {
newItem.pages += "-"+lpage;
}
}
// Do authors
var elmts = newDOM.docauthor;
for(var i in elmts) {
var fname = elmts[i].fname.text();
var surname = elmts[i].surname.text();
newItem.creators.push({firstName:fname, lastName:surname, creatorType:"author"});
}
newItem.complete();
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}');
REPLACE INTO "translators" VALUES ('fcf41bed-0cbc-3704-85c7-8062a0068a7a', '2006-08-07 21:55:00', 12, 'PubMed', 'Simon Kornblith', '^http://www\.ncbi\.nlm\.nih\.gov/entrez/query\.fcgi\?(?:.*db=PubMed.*list_uids=[0-9]|.*list_uids=[0-9].*db=PubMed|.*db=PubMed.*CMD=search|.*CMD=search.*db=PubMed)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(doc.location.href.indexOf("list_uids=") >= 0) {
return "journalArticle";
} else {
return "multiple";
}
}
function getPMID(co) {
var coParts = co.split("&");
for each(part in coParts) {
if(part.substr(0, 7) == "rft_id=") {
var value = unescape(part.substr(7));
if(value.substr(0, 10) == "info:pmid/") {
return value.substr(10);
}
}
}
}
function detectSearch(item) {
if(item.contextObject) {
if(getPMID(item.contextObject)) {
return "journalArticle";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
return false;
}',
'function lookupPMIDs(ids, doc) {
Scholar.wait();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newUri = "http://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=PubMed&retmode=xml&rettype=citation&id="+ids.join(",");
Scholar.Utilities.HTTP.doGet(newUri, function(text) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Remove xml parse instruction and doctype
text = text.replace(/<!DOCTYPE[^>]*>/, "").replace(/<\?xml[^>]*\?>/, "");
var xml = new XML(text);
for(var i=0; i<xml.PubmedArticle.length(); i++) {
var newItem = new Scholar.Item("journalArticle");
var citation = xml.PubmedArticle[i].MedlineCitation;
var PMID = citation.PMID.text().toString();
newItem.accessionNumber = "PMID "+PMID;
// add attachments
if(doc) {
newItem.attachments.push({document:doc, title:"PubMed Abstract",
downloadable:true});
} else {
var url = "http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Retrieve&dopt=AbstractPlus&list_uids="+PMID;
newItem.attachments.push({url:url, title:"PubMed Abstract (HTML)",
mimeType:"text/html", downloadable:true});
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var article = citation.Article;
if(article.ArticleTitle.length()) {
var title = article.ArticleTitle.text().toString();
if(title.substr(-1) == ".") {
title = title.substring(0, title.length-1);
}
newItem.title = title;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(article.Journal.length()) {
var issn = article.Journal.ISSN.text();
if(issn) {
newItem.ISSN = issn.replace(/[^0-9]/g, "");
}
if(article.Journal.Title.length()) {
newItem.publicationTitle = Scholar.Utilities.superCleanString(article.Journal.Title.text().toString());
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(citation.MedlineJournalInfo.MedlineTA.length()) {
newItem.publicationTitle = Scholar.Utilities.superCleanString(citation.MedlineJournalInfo.MedlineTA.text().toString());
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
if(article.Journal.JournalIssue.length()) {
newItem.volume = article.Journal.JournalIssue.Volume.text();
newItem.issue = article.Journal.JournalIssue.Issue.text();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(article.Journal.JournalIssue.PubDate.length()) { // try to get the date
if(article.Journal.JournalIssue.PubDate.Day.text().toString() != "") {
var date = article.Journal.JournalIssue.PubDate.Month.text()+" "+article.Journal.JournalIssue.PubDate.Day.text()+", "+article.Journal.JournalIssue.PubDate.Year.text();
var jsDate = new Date(date);
if(!isNaN(jsDate.valueOf())) {
date = Scholar.Utilities.dateToSQL(jsDate);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
} else if(article.Journal.JournalIssue.PubDate.Month.text().toString() != "") {
var date = article.Journal.JournalIssue.PubDate.Month.text()+" "+article.Journal.JournalIssue.PubDate.Year.text();
} else if(article.Journal.JournalIssue.PubDate.Year.text().toString() != "") {
var date = article.Journal.JournalIssue.PubDate.Year.text();
}
if(date) {
newItem.date = date;
}
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(article.AuthorList.length() && article.AuthorList.Author.length()) {
var authors = article.AuthorList.Author;
for(var j=0; j<authors.length(); j++) {
var lastName = authors[j].LastName.text().toString();
var firstName = authors[j].FirstName.text().toString();
if(firstName == "") {
var firstName = authors[j].ForeName.text().toString();
}
if(firstName || lastName) {
newItem.creators.push({lastName:lastName, firstName:firstName});
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.done();
});
}
function doWeb(doc, url) {
var uri = doc.location.href;
var ids = new Array();
var idRegexp = /[\?\&]list_uids=([0-9\,]+)/;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var m = idRegexp.exec(uri);
if(m) {
ids.push(m[1]);
lookupPMIDs(ids, doc);
} else {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var items = new Array();
var tableRows = doc.evaluate(''//div[@class="ResultSet"]/table/tbody'', doc,
nsResolver, XPathResult.ANY_TYPE, null);
var tableRow;
// Go through table rows
while(tableRow = tableRows.iterateNext()) {
var link = doc.evaluate(''.//a'', tableRow, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
var article = doc.evaluate(''./tr[2]/td[2]/text()[1]'', tableRow, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
items[link.href] = article.nodeValue;
}
items = Scholar.selectItems(items);
if(!items) {
return true;
}
for(var i in items) {
var m = idRegexp.exec(i);
ids.push(m[1]);
}
lookupPMIDs(ids);
}
}
function doSearch(item) {
// pmid was defined earlier in detectSearch
lookupPMIDs([getPMID(item.contextObject)]);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}');
REPLACE INTO "translators" VALUES ('951c027d-74ac-47d4-a107-9c3069ab7b48', '2006-06-26 16:41:00', 4, 'Embedded RDF', 'Simon Kornblith', NULL,
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var metaTags = doc.getElementsByTagName("meta");
for(var i=0; i<metaTags.length; i++) {
var tag = metaTags[i].getAttribute("name");
if(tag && tag.substr(0, 3).toLowerCase() == "dc.") {
return "website";
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return false;
}',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'function doWeb(doc, url) {
var dc = "http://purl.org/dc/elements/1.1/";
2006-06-21 14:28:51 +00:00
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// load RDF translator
var translator = Scholar.loadTranslator("import", "5e3ad958-ac79-463d-812b-a86a9235c28f");
var metaTags = doc.getElementsByTagName("meta");
var foundTitle = false; // We can use the page title if necessary
for(var i=0; i<metaTags.length; i++) {
var tag = metaTags[i].getAttribute("name");
var value = metaTags[i].getAttribute("content");
if(tag && value && tag.substr(0, 3).toLowerCase() == "dc.") {
if(tag == "dc.title") {
foundTitle = true;
}
translator.Scholar.RDF.addStatement(url, dc + tag.substr(3), value, true);
Scholar.Utilities.debug(tag.substr(3) + " = " + value);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(tag && value && (tag == "author" || tag == "author-personal")) {
translator.Scholar.RDF.addStatement(url, dc + "creator", value, true);
} else if(tag && value && tag == "author-corporate") {
translator.Scholar.RDF.addStatement(url, dc + "creator", value, true);
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(!foundTitle) {
translator.Scholar.RDF.addStatement(url, dc + "title", doc.title, true);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
translator.doImport();
}');
REPLACE INTO "translators" VALUES ('05d07af9-105a-4572-99f6-a8e231c0daef', '2006-08-07 01:09:00', 4, 'COinS', 'Simon Kornblith', NULL,
'function detectWeb(doc, url) {
var spanTags = doc.getElementsByTagName("span");
var encounteredType = false;
for(var i=0; i<spanTags.length; i++) {
var spanClass = spanTags[i].getAttribute("class");
if(spanClass) {
var spanClasses = spanClass.split(" ");
if(Scholar.Utilities.inArray("Z3988", spanClasses)) {
var spanTitle = spanTags[i].getAttribute("title");
// determine if it''s a valid type
var coParts = spanTitle.split("&");
var type = null
for(var j in coParts) {
if(coParts[j].substr(0, 12) == "rft_val_fmt=") {
var format = unescape(coParts[j].substr(12));
if(format == "info:ofi/fmt:kev:mtx:journal") {
var type = "journalArticle";
} else if(format == "info:ofi/fmt:kev:mtx:book") {
if(Scholar.Utilities.inArray("rft.genre=bookitem", coParts)) {
var type = "bookSection";
} else {
var type = "book";
}
break;
}
}
}
if(type) {
if(encounteredType) {
return "multiple";
} else {
encounteredType = type;
}
}
}
}
}
return encounteredType;
}',
'// used to retrieve next COinS object when asynchronously parsing COinS objects
// on a page
function retrieveNextCOinS(needFullItems, newItems, doc) {
if(needFullItems.length) {
var item = needFullItems.shift();
Scholar.Utilities.debug("looking up contextObject");
var search = Scholar.loadTranslator("search");
search.setHandler("itemDone", function(obj, item) {
newItems.push(item);
});
search.setHandler("done", function() {
retrieveNextCOinS(needFullItems, newItems, doc);
});
search.setSearch(item);
// look for translators
var translators = search.getTranslators();
if(translators) {
search.setTranslator(translators);
search.translate();
} else {
retrieveNextCOinS(needFullItems, newItems, doc);
}
} else {
completeCOinS(newItems, doc);
Scholar.done(true);
}
}
// saves all COinS objects
function completeCOinS(newItems, doc) {
if(newItems.length > 1) {
var selectArray = new Array();
for(var i in newItems) {
selectArray[i] = newItems[i].title;
}
selectArray = Scholar.selectItems(selectArray);
for(var i in selectArray) {
// add doc as attachment
newItems[i].attachments.push({document:doc});
newItems[i].complete();
}
} else if(newItems.length) {
newItems[0].attachments.push({document:doc});
newItems[0].complete();
}
}
function doWeb(doc, url) {
var newItems = new Array();
var needFullItems = new Array();
var spanTags = doc.getElementsByTagName("span");
for(var i=0; i<spanTags.length; i++) {
var spanClass = spanTags[i].getAttribute("class");
if(spanClass) {
var spanClasses = spanClass.split(" ");
if(Scholar.Utilities.inArray("Z3988", spanClasses)) {
var spanTitle = spanTags[i].getAttribute("title");
var newItem = new Scholar.Item();
if(Scholar.Utilities.parseContextObject(spanTitle, newItem)) {
if(newItem.title && newItem.creators.length) {
// title and creators are minimum data to avoid looking up
newItems.push(newItem);
} else {
// retrieve full item
newItem.contextObject = spanTitle;
needFullItems.push(newItem);
}
}
}
}
}
if(needFullItems.length) {
// retrieve full items asynchronously
Scholar.wait();
retrieveNextCOinS(needFullItems, newItems, doc);
} else {
completeCOinS(newItems, doc);
}
}');
REPLACE INTO "translators" VALUES ('3e684d82-73a3-9a34-095f-19b112d88bbf', '2006-06-26 16:01:00', 4, 'Google Books', 'Simon Kornblith', '^http://books\.google\.com/books\?(.*vid=.*\&id=.*|.*q=.*)',
'function detectWeb(doc, url) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var re = new RegExp(''^http://books\\.google\\.com/books\\?vid=([^&]+).*\\&id=([^&]+)'', ''i'');
if(re.test(doc.location.href)) {
return "book";
} else {
return "multiple";
}
}',
'function doWeb(doc, url) {
var uri = doc.location.href;
var newUris = new Array();
var re = new RegExp(''^http://books\\.google\\.com/books\\?vid=([^&]+).*\\&id=([^&]+)'', ''i'');
var m = re.exec(uri);
if(m) {
newUris.push(''http://books.google.com/books?vid=''+m[1]+''&id=''+m[2]);
} else {
var items = Scholar.Utilities.getItemArray(doc, doc, ''http://books\\.google\\.com/books\\?vid=([^&]+).*\\&id=([^&]+)'', ''^(?:All matching pages|About this Book|Table of Contents|Index)'');
// Drop " - Page" thing
for(var i in items) {
items[i] = items[i].replace(/- Page [0-9]+\s*$/, "");
}
items = Scholar.selectItems(items);
if(!items) {
return true;
}
for(var i in items) {
var m = re.exec(i);
newUris.push(''http://books.google.com/books?vid=''+m[1]+''&id=''+m[2]);
}
}
2006-06-21 14:28:51 +00:00
Scholar.Utilities.processDocuments(newUris, function(newDoc) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item("book");
newItem.extra = "";
newItem.attachments.push({title:"Google Books Information Page", document:newDoc});
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var namespace = newDoc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var xpath = ''//table[@id="bib"]/tbody/tr'';
var elmts = newDoc.evaluate(xpath, newDoc, nsResolver,
XPathResult.ANY_TYPE, null);
var elmt;
while(elmt = elmts.iterateNext()) {
var field = newDoc.evaluate(''./td[1]//text()'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
var value = newDoc.evaluate(''./td[2]//text()'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(field && value) {
field = Scholar.Utilities.superCleanString(field.nodeValue);
value = Scholar.Utilities.cleanString(value.nodeValue);
if(field == "Title") {
newItem.title = value;
} else if(field == "Author(s)") {
var authors = value.split(", ");
for(j in authors) {
newItem.creators.push(Scholar.Utilities.cleanAuthor(authors[j], "author"));
}
} else if(field == "Editor(s)") {
var authors = value.split(", ");
for(j in authors) {
newItem.creators.push(Scholar.Utilities.cleanAuthor(authors[j], "editor"));
}
} else if(field == "Publisher") {
newItem.publisher = value;
} else if(field == "Publication Date") {
var date = value;
jsDate = new Date(value);
if(!isNaN(jsDate.valueOf())) {
date = Scholar.Utilities.dateToSQL(jsDate);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
newItem.date = date;
/*} else if(field == "Format") {
.addStatement(uri, prefixDC + ''medium'', value);*/
} else if(field == "ISBN") {
newItem.ISBN = value;
} else if(field == "Pages") {
newItem.pages = value;
} else {
newItem.extra += field+": "+value+"\n";
}
2006-06-21 14:28:51 +00:00
}
}
if(newItem.extra) {
newItem.extra = newItem.extra.substr(newItem.extra, newItem.extra.length-1);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
}, function() { Scholar.done(); }, null);
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('9c335444-a562-4f88-b291-607e8f46a9bb', '2006-08-15 15:42:00', 4, 'Berkeley Library', 'Simon Kornblith', '^http://[^/]*berkeley.edu[^/]*/WebZ/(?:html/results.html|FETCH)\?.*sessionid=',
'function detectWeb(doc, url) {
var resultsRegexp = /\/WebZ\/html\/results.html/i
if(resultsRegexp.test(url)) {
return "multiple";
} else {
return "book";
}
}',
'function reformURL(url) {
return url.replace(/fmtclass=[^&]*/, "")+":fmtclass=marc";
}
function doWeb(doc, url) {
var resultsRegexp = /\/WebZ\/html\/results.html/i
if(resultsRegexp.test(url)) {
var items = Scholar.Utilities.getItemArray(doc, doc, "/WebZ/FETCH", "^[0-9]*$");
items = Scholar.selectItems(items);
if(!items) {
return true;
}
var urls = new Array();
for(var i in items) {
urls.push(reformURL(i));
}
} else {
var urls = [reformURL(url)];
}
var translator = Scholar.loadTranslator("import");
translator.setTranslator("a6ee60df-1ddc-4aae-bb25-45e0537be973");
var marc = translator.getTranslatorObject();
Scholar.Utilities.processDocuments(urls, function(newDoc) {
var uri = newDoc.location.href;
var namespace = newDoc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var elmts = newDoc.evaluate(''//table/tbody/tr[@valign="top"]'',
newDoc, nsResolver, XPathResult.ANY_TYPE, null);
var record = new marc.record();
while(elmt = elmts.iterateNext()) {
var field = Scholar.Utilities.superCleanString(doc.evaluate(''./TD[1]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue);
var value = doc.evaluate(''./TD[2]/text()[1]'', elmt, nsResolver, XPathResult.ANY_TYPE, null).iterateNext().nodeValue;
// remove spacing
value = value.replace(/^\s+/, "");
value = value.replace(/\s+$/, "");
if(field == 0) {
record.leader = "00000"+value;
} else {
var ind = value[3]+value[5];
value = Scholar.Utilities.cleanString(value.substr(5)).
replace(/\$([a-z0-9]) /g, marc.subfieldDelimiter+"$1");
if(value[0] != marc.subfieldDelimiter) {
value = marc.subfieldDelimiter+"a"+value;
}
record.addField(field, ind, value);
}
}
var newItem = new Scholar.Item();
record.translate(newItem);
newItem.complete();
}, function() { Scholar.done(); }, null);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('d0b1914a-11f1-4dd7-8557-b32fe8a3dd47', '2006-08-18 18:03:00', 4, 'EBSCOhost', 'Simon Kornblith', '^http://web\.ebscohost\.com/ehost/(?:results|detail)',
'function detectWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var searchRe = new RegExp("^http://web\\.ebscohost\\.com/ehost/results", "i");
// See if this is a seach results page
if(searchRe.test(url)) {
return "multiple";
} else {
var persistentLink = doc.evaluate(''//tr[td[@class="left-content-ft"]/text() = "Persistent link to this record:"]/td[@class="right-content-ft"]'',
doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext();
if(persistentLink) {
return "journalArticle";
}
}
}',
'function fullEscape(text) {
return escape(text).replace(/\//g, "%2F").replace(/\+/g, "%2B");
}
function doWeb(doc, url) {
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
var queryRe = /\?(.*)$/;
var m = queryRe.exec(url);
var queryString = m[1];
var eventValidation = doc.evaluate(''//input[@name="__EVENTVALIDATION"]'', doc, nsResolver,
XPathResult.ANY_TYPE, null).iterateNext();
eventValidation = fullEscape(eventValidation.value);
var viewState = doc.evaluate(''//input[@name="__VIEWSTATE"]'', doc, nsResolver,
XPathResult.ANY_TYPE, null).iterateNext();
viewState = fullEscape(viewState.value);
var searchRe = new RegExp("^http://web\\.ebscohost\\.com/ehost/results", "i");
if(searchRe.test(url)) {
var items = new Object();
var tableRows = doc.evaluate(''//table[@class="cluster-result-record-table"]/tbody/tr'',
doc, nsResolver, XPathResult.ANY_TYPE, null);
var tableRow;
// Go through table rows
while(tableRow = tableRows.iterateNext()) {
var title = doc.evaluate(''.//a[@class="title-link"]'', tableRow, nsResolver,
XPathResult.ANY_TYPE, null).iterateNext();
var addLink = doc.evaluate(''.//a[substring(@id, 1, 11)="addToFolder"]'', tableRow, nsResolver,
XPathResult.ANY_TYPE, null).iterateNext();
if(title && addLink) {
items[addLink.href] = title.textContent;
}
}
var items = Scholar.selectItems(items);
if(!items) {
return true;
}
var citations = new Array();
var argRe = /''([^'']+)''/;
for(var i in items) {
var m = argRe.exec(i);
citations.push(m[1]);
}
var saveString = "__EVENTTARGET=FolderItem:AddItem&IsCallBack=true&SearchTerm1=test&listDatabaseGroupings=pdh&SortOptionDropDown=date&__EVENTVALIDATION="+eventValidation+"&__EVENTARGUMENT="+citations.join(",")+"&";
} else {
// If this is a view page, find the link to the citation
var xpath = ''/html/body/div[@class="indent"]/center//a[@class="nav"]'';
var elmts = doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null);
var saveCitation = elmts.iterateNext();
var viewSavedCitations = elmts.iterateNext();
var saveString = "__EVENTTARGET=ctl00%24ctl00%24MainContentArea%24MainContentArea%24topAddToFolderControl%24lnkAddToFolder&__EVENTARGUMENT=&__VIEWSTATE="+viewState+"&__EVENTVALIDATION="+eventValidation;
}
var folderString = "__EVENTTARGET=ctl00%24ctl00%24ToolbarArea%24toolbar%24folderControl%24lnkFolder&__EVENTARGUMENT=&__VIEWSTATE="+viewState+"&__EVENTVALIDATION="+eventValidation;
var getString = "__EVENTTARGET=Tabs&IsCallBack=true&chkRemoveFromFolder=true&chkIncludeHTMLFT=true&chkIncludeHTMLLinks=true&CitationFormat=standard&lstFormatStandard=1&lstFormatIndustry=4&cfCommonAb=false&cfCommonAu=true&cfCommonTypDoc=true&cfCommonID=true&cfCommonISSN=true&cfCommonNote=false&cfCommonRevInfo=false&cfCommonSrc=true&cfCommonTi=true&__EVENTARGUMENT=1&"
var viewStateMatch = /<input type="hidden" name="__VIEWSTATE" id="__VIEWSTATE" value="([^"]+)" \/>/
var eventValidationMatch = /<input type="hidden" name="__EVENTVALIDATION" id="__EVENTVALIDATION" value="([^"]+)" \/>/
Scholar.Utilities.HTTP.doPost(url, saveString, function() { // mark records
Scholar.Utilities.HTTP.doPost(url, folderString, function(text) {
var postLocation = /<form name="aspnetForm" method="post" action="([^"]+)"/
var m = postLocation.exec(text);
var folderURL = m[1].replace(/&amp;/g, "&");
m = viewStateMatch.exec(text);
var folderViewState = m[1];
var folderBase = "__EVENTARGUMENT=&__VIEWSTATE="+fullEscape(folderViewState);
m = eventValidationMatch.exec(text);
var folderEventValidation = m[1];
folderBase += "&__EVENTVALIDATION="+fullEscape(folderEventValidation);
var deliverString = "__EVENTTARGET=ctl00%24ctl00%24MainContentArea%24MainContentArea%24btnDelivery%24lnkSave&"+folderBase
Scholar.Utilities.HTTP.doPost("http://web.ebscohost.com/ehost/"+folderURL,
deliverString, function(text) {
var postLocation = /<form name="aspnetForm" method="post" action="([^"]+)"/
var m = postLocation.exec(text);
var deliveryURL = m[1].replace(/&amp;/g, "&");
var m = viewStateMatch.exec(text);
var downloadString = "__EVENTTARGET=&__EVENTARGUMENT=&__VIEWSTATE="+fullEscape(m[1])+"&ctl00%24ctl00%24MainContentArea%24MainContentArea%24ctl01%24chkRemoveFromFolder=on&ctl00%24ctl00%24MainContentArea%24MainContentArea%24ctl01%24btnSubmit=Save&ctl00%24ctl00%24MainContentArea%24MainContentArea%24ctl01%24BibFormat=1";
Scholar.Utilities.HTTP.doPost("http://web.ebscohost.com/ehost/"+deliveryURL,
getString, function(text) {
Scholar.Utilities.HTTP.doPost("http://web.ebscohost.com/ehost/"+deliveryURL,
downloadString, function(text) { // get marked
var form = doc.createElement("form");
form.setAttribute("method", "post");
form.setAttribute("action", "http://web.ebscohost.com/ehost/"+folderURL);
var args = [
["__EVENTARGUMENT", ""],
["__VIEWSTATE", folderViewState],
["__EVENTVALIDATION", folderEventValidation],
["__EVENTTARGET", "ctl00$ctl00$MainContentArea$MainContentArea$btnBack$lnkBack"]
];
for(var i in args) {
var input = doc.createElement("input");
input.setAttribute("type", "hidden");
input.setAttribute("name", args[i][0]);
input.setAttribute("value", args[i][1]);
form.appendChild(input);
}
var body = doc.getElementsByTagName("body");
body[0].appendChild(form);
form.submit();
// load translator for RIS
var translator = Scholar.loadTranslator("import");
translator.setTranslator("32d59d2d-b65a-4da4-b0a3-bdd3cfb979e7");
translator.setString(text);
translator.setHandler("itemDone", function(obj, item) {
if(item.notes && item.notes[0]) {
item.extra = item.notes[0].note;
delete item.notes;
item.notes = undefined;
}
item.complete();
});
translator.translate();
Scholar.done();
});
});
});
});
});
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('e07e9b8c-0e98-4915-bb5a-32a08cb2f365', '2006-08-07 11:36:00', 8, 'Open WorldCat', 'Simon Kornblith', 'http://partneraccess.oclc.org/',
'function detectSearch(item) {
if(item.itemType == "book" || item.itemType == "bookSection") {
return true;
}
return false;
}',
'// creates an item from an Open WorldCat document
function processOWC(doc) {
var spanTags = doc.getElementsByTagName("span");
for(var i=0; i<spanTags.length; i++) {
var spanClass = spanTags[i].getAttribute("class");
if(spanClass) {
var spanClasses = spanClass.split(" ");
if(Scholar.Utilities.inArray("Z3988", spanClasses)) {
var spanTitle = spanTags[i].getAttribute("title");
var item = new Scholar.Item();
if(Scholar.Utilities.parseContextObject(spanTitle, item)) {
item.complete();
return true;
} else {
return false;
}
}
}
}
return false;
}
function doSearch(item) {
if(item.contextObject) {
var co = item.contextObject;
} else {
var co = Scholar.Utilities.createContextObject(item);
}
Scholar.Utilities.loadDocument("http://partneraccess.oclc.org/wcpa/servlet/OpenUrl?"+co, function(doc) {
// find new COinS in the Open WorldCat page
if(processOWC(doc)) { // we got a single item page
Scholar.done();
} else { // assume we have a search results page
var items = new Array();
var namespace = doc.documentElement.namespaceURI;
var nsResolver = namespace ? function(prefix) {
if (prefix == ''x'') return namespace; else return null;
} : null;
// first try to get only books
var elmts = doc.evaluate(''//table[@class="tableLayout"]/tbody/tr/td[@class="content"]/table[@class="tableResults"]/tbody/tr[td/img[@alt="Book"]]/td/div[@class="title"]/a'', doc, nsResolver, Components.interfaces.nsIDOMXPathResult.ANY_TYPE,null);
var elmt = elmts.iterateNext();
if(!elmt) { // if that fails, look for other options
var elmts = doc.evaluate(''//table[@class="tableLayout"]/tbody/tr/td[@class="content"]/table[@class="tableResults"]/tbody/tr[td/img[@alt="Book"]]/td/div[@class="title"]/a'', doc, nsResolver, Components.interfaces.nsIDOMXPathResult.ANY_TYPE,null);
elmt = elmts.iterateNext()
}
var urlsToProcess = new Array();
do {
urlsToProcess.push(elmt.href);
} while(elmt = elmts.iterateNext());
Scholar.Utilities.processDocuments(urlsToProcess, function(doc) {
// per URL
processOWC(doc);
}, function() { // done
Scholar.done();
}, function() { // error
Scholar.done(false);
});
}
}, null);
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('11645bd1-0420-45c1-badb-53fb41eeb753', '2006-08-07 18:17:00', 8, 'CrossRef', 'Simon Kornblith', 'http://partneraccess.oclc.org/',
'function detectSearch(item) {
if(item.itemType == "journal") {
return true;
}
return false;
}',
'function processCrossRef(xmlOutput) {
xmlOutput = xmlOutput.replace(/<\?xml[^>]*\?>/, "");
// parse XML with E4X
var qr = new Namespace("http://www.crossref.org/qrschema/2.0");
try {
var xml = new XML(xmlOutput);
} catch(e) {
return false;
}
// ensure status is valid
var status = xml.qr::query_result.qr::body.qr::query.@status.toString();
if(status != "resolved" && status != "multiresolved") {
return false;
}
var query = xml.qr::query_result.qr::body.qr::query;
var item = new Scholar.Item("journalArticle");
// try to get a DOI
item.DOI = query.qr::doi.(@type=="journal_article").text().toString();
if(!item.DOI) {
item.DOI = query.qr::doi.(@type=="book_title").text().toString();
}
if(!item.DOI) {
item.DOI = query.qr::doi.(@type=="book_content").text().toString();
}
// try to get an ISSN (no print/electronic preferences)
item.ISSN = query.qr::issn[0].text().toString();
// get title
item.title = query.qr::article_title.text().toString();
// get publicationTitle
item.publicationTitle = query.qr::journal_title.text().toString();
// get author
item.creators.push(Scholar.Utilities.cleanAuthor(query.qr::author.text().toString(), "author", true));
// get volume
item.volume = query.qr::volume.text().toString();
// get issue
item.issue = query.qr::issue.text().toString();
// get year
item.date = query.qr::year.text().toString();
// get edition
item.edition = query.qr::edition_number.text().toString();
// get first page
item.pages = query.qr::first_page.text().toString();
item.complete();
return true;
}
function doSearch(item) {
if(item.contextObject) {
var co = item.contextObject;
if(co.indexOf("url_ver=") == -1) {
co = "url_ver=Z39.88-2004"+co;
}
} else {
var co = Scholar.Utilities.createContextObject(item);
}
Scholar.Utilities.HTTP.doGet("http://www.crossref.org/openurl/?"+co+"&noredirect=true", function(responseText) {
processCrossRef(responseText);
Scholar.done();
});
Scholar.wait();
}');
REPLACE INTO "translators" VALUES ('0e2235e7-babf-413c-9acf-f27cce5f059c', '2006-07-05 23:40:00', 3, 'MODS (XML)', 'Simon Kornblith', 'xml',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'Scholar.addOption("exportNotes", true);
function detectImport() {
var read = Scholar.read(512);
var modsTagRegexp = /<mods[^>]+>/
if(modsTagRegexp.test(read)) {
return true;
}
}',
'var partialItemTypes = ["bookSection", "journalArticle", "magazineArticle", "newspaperArticle"];
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doExport() {
var modsCollection = <modsCollection xmlns="http://www.loc.gov/mods/v3" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-2.xsd" />;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var item;
while(item = Scholar.nextItem()) {
var isPartialItem = Scholar.Utilities.inArray(item.itemType, partialItemTypes);
var mods = <mods />;
/** CORE FIELDS **/
// XML tag titleInfo; object field title
if(item.title) {
mods.titleInfo.title = item.title;
}
// XML tag typeOfResource/genre; object field type
var modsType, marcGenre;
if(item.itemType == "book" || item.itemType == "bookSection") {
modsType = "text";
marcGenre = "book";
} else if(item.itemType == "journalArticle" || item.itemType == "magazineArticle") {
modsType = "text";
marcGenre = "periodical";
} else if(item.itemType == "newspaperArticle") {
modsType = "text";
marcGenre = "newspaper";
} else if(item.itemType == "thesis") {
modsType = "text";
marcGenre = "theses";
} else if(item.itemType == "letter") {
modsType = "text";
marcGenre = "letter";
} else if(item.itemType == "manuscript") {
modsType = "text";
modsType.@manuscript = "yes";
} else if(item.itemType == "interview") {
modsType = "text";
marcGenre = "interview";
} else if(item.itemType == "film") {
modsType = "moving image";
marcGenre = "motion picture";
} else if(item.itemType == "artwork") {
modsType = "still image";
marcGenre = "art original";
} else if(item.itemType == "website") {
modsType = "multimedia";
marcGenre = "web site";
} else if(item.itemType == "note") {
continue;
}
mods.typeOfResource = modsType;
mods.genre += <genre authority="local">{item.itemType}</genre>;
if(marcGenre) {
mods.genre += <genre authority="marcgt">{marcGenre}</genre>;
}
// XML tag genre; object field thesisType, type
if(item.thesisType) {
mods.genre += <genre>{item.thesisType}</genre>;
}
if(item.type) {
mods.genre += <genre>{item.type}</genre>;
}
// XML tag name; object field creators
for(var j in item.creators) {
var roleTerm = "";
if(item.creators[j].creatorType == "author") {
roleTerm = "aut";
} else if(item.creators[j].creatorType == "editor") {
roleTerm = "edt";
} else if(item.creators[j].creatorType == "creator") {
roleTerm = "ctb";
}
// FIXME - currently all names are personal
mods.name += <name type="personal">
<namePart type="family">{item.creators[j].lastName}</namePart>
<namePart type="given">{item.creators[j].firstName}</namePart>
<role><roleTerm type="code" authority="marcrelator">{roleTerm}</roleTerm></role>
</name>;
}
// XML tag recordInfo.recordOrigin; used to store our generator note
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
//mods.recordInfo.recordOrigin = "Scholar for Firefox "+Scholar.Utilities.getVersion();
/** FIELDS ON NEARLY EVERYTHING BUT NOT A PART OF THE CORE **/
// XML tag recordInfo.recordContentSource; object field source
if(item.source) {
mods.recordInfo.recordContentSource = item.source;
}
// XML tag recordInfo.recordIdentifier; object field accessionNumber
if(item.accessionNumber) {
mods.recordInfo.recordIdentifier = item.accessionNumber;
}
// XML tag accessCondition; object field rights
if(item.rights) {
mods.accessCondition = item.rights;
}
/** SUPPLEMENTAL FIELDS **/
// XML tag relatedItem.titleInfo; object field series
if(item.seriesTitle) {
var series = <relatedItem type="series">
<titleInfo><title>{item.seriesTitle}</title></titleInfo>
</relatedItem>;
if(item.itemType == "bookSection") {
// For a book section, series info must go inside host tag
mods.relatedItem.relatedItem = series;
} else {
mods.relatedItem += series;
}
}
// Make part its own tag so we can figure out where it goes later
var part = new XML();
// XML tag detail; object field volume
if(item.volume) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(Scholar.Utilities.isInt(item.volume)) {
part += <detail type="volume"><number>{item.volume}</number></detail>;
} else {
part += <detail type="volume"><text>{item.volume}</text></detail>;
}
}
// XML tag detail; object field number
if(item.issue) {
if(Scholar.Utilities.isInt(item.issue)) {
part += <detail type="issue"><number>{item.issue}</number></detail>;
} else {
part += <detail type="issue"><text>{item.issue}</text></detail>;
}
}
// XML tag detail; object field section
if(item.section) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(Scholar.Utilities.isInt(item.section)) {
part += <detail type="section"><number>{item.section}</number></detail>;
} else {
part += <detail type="section"><text>{item.section}</text></detail>;
}
}
// XML tag detail; object field pages
if(item.pages) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var range = Scholar.Utilities.getPageRange(item.pages);
part += <extent unit="pages"><start>{range[0]}</start><end>{range[1]}</end></extent>;
}
// Assign part if something was assigned
if(part.length() != 1) {
if(isPartialItem) {
// For a journal article, bookSection, etc., the part is the host
mods.relatedItem.part += <part>{part}</part>;
} else {
mods.part += <part>{part}</part>;
}
}
// XML tag originInfo; object fields edition, place, publisher, year, date
var originInfo = new XML();
if(item.edition) {
originInfo += <edition>{item.edition}</edition>;
}
if(item.place) {
originInfo += <place><placeTerm type="text">{item.place}</placeTerm></place>;
}
if(item.publisher) {
originInfo += <publisher>{item.publisher}</publisher>;
} else if(item.distributor) {
originInfo += <publisher>{item.distributor}</publisher>;
}
if(item.date) {
if(Scholar.Utilities.inArray(item.itemType, ["book", "bookSection"])) {
// Assume year is copyright date
var dateType = "copyrightDate";
} else if(Scholar.Utilities.inArray(item.itemType, ["journalArticle", "magazineArticle", "newspaperArticle"])) {
// Assume date is date issued
var dateType = "dateIssued";
} else {
// Assume date is date created
var dateType = "dateCreated";
}
var tag = <{dateType}>{item.date}</{dateType}>;
tag.@encoding = "iso8601";
originInfo += tag;
}
if(item.accessDate) {
originInfo += <dateCaptured encoding="iso8601">{item.accessDate}</dateCaptured>;
}
if(originInfo.length() != 1) {
if(isPartialItem) {
// For a journal article, bookSection, etc., this goes under the host
mods.relatedItem.originInfo += <originInfo>{originInfo}</originInfo>;
} else {
mods.originInfo += <originInfo>{originInfo}</originInfo>;
}
}
// XML tag identifier; object fields ISBN, ISSN
if(isPartialItem) {
var identifier = mods.relatedItem;
} else {
var identifier = mods;
}
if(item.ISBN) {
identifier.identifier += <identifier type="isbn">{item.ISBN}</identifier>;
}
if(item.ISSN) {
identifier.identifier += <identifier type="issn">{item.ISSN}</identifier>;
}
if(item.DOI) {
identifier.identifier += <identifier type="doi">{item.DOI}</identifier>;
}
// XML tag relatedItem.titleInfo; object field publication
if(item.publicationTitle) {
mods.relatedItem.titleInfo += <titleInfo><title>{item.publicationTitle}</title></titleInfo>;
}
// XML tag classification; object field callNumber
if(item.callNumber) {
mods.classification = item.callNumber;
}
// XML tag location.physicalLocation; object field archiveLocation
if(item.archiveLocation) {
mods.location.physicalLocation = item.archiveLocation;
}
// XML tag location.url; object field archiveLocation
if(item.url) {
mods.location.url = item.url;
}
// XML tag title.titleInfo; object field journalAbbreviation
if(item.journalAbbreviation) {
mods.relatedItem.titleInfo += <titleInfo type="abbreviated"><title>{item.journalAbbreviation}</title></titleInfo>;
}
if(mods.relatedItem.length() == 1 && isPartialItem) {
mods.relatedItem.@type = "host";
}
/** NOTES **/
2006-08-08 23:00:33 +00:00
if(Scholar.getOption("exportNotes")) {
for(var j in item.notes) {
// Add note tag
var note = <note type="content">{item.notes[j].note}</note>;
mods.note += note;
}
}
/** TAGS **/
for(var j in item.tags) {
mods.subject += <subject>{item.tags[j]}</subject>;
}
modsCollection.mods += mods;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.write(''<?xml version="1.0"?>''+"\n");
Scholar.write(modsCollection.toXMLString());
}
function doImport() {
var text = "";
var read;
// read in 16384 byte increments
while(read = Scholar.read(16384)) {
text += read;
}
Scholar.Utilities.debug("read in");
// eliminate <?xml ?> heading so we can parse as XML
text = text.replace(/<\?xml[^?]+\?>/, "");
// parse with E4X
var m = new Namespace("http://www.loc.gov/mods/v3");
// why does this default namespace declaration not work!?
default xml namespace = m;
var xml = new XML(text);
for each(var mods in xml.m::mods) {
Scholar.Utilities.debug("item is: ");
for(var i in mods) {
Scholar.Utilities.debug(i+" = "+mods[i].toString());
}
var newItem = new Scholar.Item();
// title
newItem.title = mods.m::titleInfo.(m::title.@type!="abbreviated").m::title;
// try to get genre from local genre
var localGenre = mods.m::genre.(@authority=="local").text().toString();
if(localGenre && Scholar.Utilities.itemTypeExists(localGenre)) {
newItem.itemType = localGenre;
} else {
// otherwise, look at the marc genre
var marcGenre = mods.m::genre.(@authority=="marcgt").text().toString();
if(marcGenre) {
if(marcGenre == "book") {
newItem.itemType = "book";
} else if(marcGenre == "periodical") {
newItem.itemType = "magazineArticle";
} else if(marcGenre == "newspaper") {
newItem.itemType = "newspaperArticle";
} else if(marcGenre == "theses") {
newItem.itemType = "thesis";
} else if(marcGenre == "letter") {
newItem.itemType = "letter";
} else if(marcGenre == "interview") {
newItem.itemType = "interview";
} else if(marcGenre == "motion picture") {
newItem.itemType = "film";
} else if(marcGenre == "art original") {
newItem.itemType = "artwork";
} else if(marcGenre == "web site") {
newItem.itemType = "website";
}
}
if(!newItem.itemType) {
newItem.itemType = "book";
}
}
var isPartialItem = Scholar.Utilities.inArray(newItem.itemType, partialItemTypes);
// TODO: thesisType, type
for each(var name in mods.m::name) {
// TODO: institutional authors
var creator = new Array();
creator.firstName = name.m::namePart.(@type=="given").text().toString();
creator.lastName = name.m::namePart.(@type=="family").text().toString();
// look for roles
var role = name.m::role.m::roleTerm.(@type=="code").(@authority=="marcrelator").text().toString();
if(role == "edt") {
creator.creatorType = "editor";
} else if(role == "ctb") {
creator.creatorType = "contributor";
} else {
creator.creatorType = "author";
}
newItem.creators.push(creator);
}
// source
newItem.source = mods.m::recordInfo.m::recordContentSource.text().toString();
// accessionNumber
newItem.accessionNumber = mods.m::recordInfo.m::recordIdentifier.text().toString();
// rights
newItem.rights = mods.m::accessCondition.text().toString();
/** SUPPLEMENTAL FIELDS **/
// series
if(newItem.itemType == "bookSection") {
newItem.seriesTitle = mods.m::relatedItem.(@type=="host").m::relatedItem.(@type=="series").m::titleInfo.m::title.text().toString();
} else {
newItem.seriesTitle = mods.m::relatedItem.(@type=="series").m::titleInfo.m::title.text().toString();
}
// get part
if(isPartialItem) {
var part = mods.m::relatedItem.m::part;
var originInfo = mods.m::relatedItem.m::originInfo;
var identifier = mods.m::relatedItem.m::identifier;
} else {
var part = mods.m::part;
var originInfo = mods.m::originInfo;
var identifier = mods.m::identifier;
}
// volume
newItem.volume = part.m::detail.(@type=="volume").m::number.text().toString();
if(!newItem.volume) {
newItem.volume = part.m::detail.(@type=="volume").m::text.text().toString();
}
// number
newItem.issue = part.m::detail.(@type=="issue").m::number.text().toString();
if(!newItem.issue) {
newItem.issue = part.m::detail.(@type=="issue").m::text.text().toString();
}
// section
newItem.section = part.m::detail.(@type=="section").m::number.text().toString();
if(!newItem.section) {
newItem.section = part.m::detail.(@type=="section").m::text.text().toString();
}
// pages
var pagesStart = part.m::extent.(@unit=="pages").m::start.text().toString();
var pagesEnd = part.m::extent.(@unit=="pages").m::end.text().toString();
if(pagesStart || pagesEnd) {
if(pagesStart && pagesEnd && pagesStart != pagesEnd) {
newItem.pages = pagesStart+"-"+pagesEnd;
} else {
newItem.pages = pagesStart+pagesEnd;
}
}
// edition
newItem.edition = originInfo.m::edition.text().toString();
// place
newItem.place = originInfo.m::place.m::placeTerm.text().toString();
// publisher/distributor
newItem.publisher = newItem.distributor = originInfo.m::publisher.text().toString();
// date
newItem.date = originInfo.m::copyrightDate.text().toString();
if(!newItem.date) {
newItem.date = originInfo.m::dateIssued.text().toString();
if(!newItem.date) {
newItem.date = originInfo.dateCreated.text().toString();
}
}
// lastModified
newItem.lastModified = originInfo.m::dateModified.text().toString();
// accessDate
newItem.accessDate = originInfo.m::dateCaptured.text().toString();
// ISBN
newItem.ISBN = identifier.(@type=="isbn").text().toString()
// ISSN
newItem.ISSN = identifier.(@type=="issn").text().toString()
// DOI
newItem.DOI = identifier.(@type=="doi").text().toString()
// publication
newItem.publicationTitle = mods.m::relatedItem.m::publication.text().toString();
// call number
newItem.callNumber = mods.m::classification.text().toString();
// archiveLocation
newItem.archiveLocation = mods.m::location.m::physicalLocation.text().toString();
// url
newItem.url = mods.m::location.m::url.text().toString();
// journalAbbreviation
newItem.journalAbbreviation = mods.m::relatedItem.(m::titleInfo.@type=="abbreviated").m::titleInfo.m::title.text().toString();
/** NOTES **/
for each(var note in mods.m::note) {
newItem.notes.push({note:note.text().toString()});
}
/** TAGS **/
for each(var subject in mods.m::subject) {
newItem.tags.push(subject.text().toString());
}
newItem.complete();
}
}');
REPLACE INTO "translators" VALUES ('14763d24-8ba0-45df-8f52-b8d1108e7ac9', '2006-07-07 12:44:00', 2, 'Biblio/DC/FOAF/PRISM/VCard (RDF/XML)', 'Simon Kornblith', 'rdf',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'Scholar.configure("getCollections", true);
Scholar.configure("dataMode", "rdf");
Scholar.addOption("exportNotes", true);
Scholar.addOption("exportFileData", true);',
'function generateSeeAlso(resource, seeAlso) {
for(var i in seeAlso) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"relation", itemResources[seeAlso[i]], false);
}
}
function generateCollection(collection) {
var collectionResource = "#collection:"+collection.id;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(collectionResource, rdf+"type", n.bib+"Collection", false);
Scholar.RDF.addStatement(collectionResource, n.dc+"title", collection.name, true);
for each(var child in collection.children) {
// add child list items
if(child.type == "collection") {
Scholar.RDF.addStatement(collectionResource, n.dcterms+"hasPart", "#collection:"+child.id, false);
// do recursive processing of collections
generateCollection(child);
} else if(itemResources[child.id]) {
Scholar.RDF.addStatement(collectionResource, n.dcterms+"hasPart", itemResources[child.id], false);
}
}
}
function handleAttachment(attachmentResource, attachment) {
Scholar.RDF.addStatement(attachmentResource, rdf+"type", n.fs+"File", false);
if(attachment.url) {
// add url as identifier
var term = Scholar.RDF.newResource();
// set term type
Scholar.RDF.addStatement(term, rdf+"type", n.dcterms+"URI", false);
// set url value
Scholar.RDF.addStatement(term, rdf+"value", attachment.url, true);
// add relationship to resource
Scholar.RDF.addStatement(attachmentResource, n.dc+"identifier", term, false);
}
// add mime type
var term = Scholar.RDF.newResource();
// set term type
Scholar.RDF.addStatement(term, rdf+"type", n.dcterms+"IMT", false);
// set mime type value
Scholar.RDF.addStatement(term, rdf+"value", attachment.mimeType, true);
// add relationship to resource
Scholar.RDF.addStatement(attachmentResource, n.dc+"format", term, false);
// add title
Scholar.RDF.addStatement(attachmentResource, n.dc+"title", attachment.title, true);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doExport() {
rdf = "http://www.w3.org/1999/02/22-rdf-syntax-ns#";
n = {
bib:"http://purl.org/net/biblio#",
dc:"http://purl.org/dc/elements/1.1/",
dcterms:"http://purl.org/dc/terms/",
prism:"http://prismstandard.org/namespaces/1.2/basic/",
foaf:"http://xmlns.com/foaf/0.1/",
vcard:"http://nwalsh.com/rdf/vCard#",
fs:"http://chnm.gmu.edu/firefoxscholar/rdf#"
};
// add namespaces
for(var i in n) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addNamespace(i, n[i]);
}
// leave as global
itemResources = new Array();
// keep track of resources already assigned (in case two book items have the
// same ISBN, or something like that)
var usedResources = new Array();
var items = new Array();
// first, map each ID to a resource
while(item = Scholar.nextItem()) {
items.push(item);
if(item.itemType == "attachment" && item.path) {
// file is stored locally (paths are always unique)
itemResources[item.itemID] = item.path+item.filename;
} else if(item.ISBN && !usedResources["urn:isbn:"+item.ISBN]) {
itemResources[item.itemID] = "urn:isbn:"+item.ISBN;
usedResources[itemResources[item.itemID]] = true;
} else if(item.url && !usedResources[item.url]) {
itemResources[item.itemID] = item.url;
usedResources[itemResources[item.itemID]] = true;
} else {
// just specify a node ID
itemResources[item.itemID] = "#item:"+item.itemID;
}
for(var j in item.notes) {
itemResources[item.notes[j].itemID] = "#item:"+item.notes[j].itemID;
}
for each(var attachment in item.attachments) {
if(attachment.path) {
// file is stored locally (paths are always unique)
itemResources[attachment.itemID] = attachment.path+attachment.filename;
} else if(!usedResources[attachment.url]) {
// file is referenced via url, and no other item has this url
itemResources[attachment.itemID] = attachment.url;
usedResources[attachment.url] = true;
} else {
// just specify a node ID
itemResources[attachment.itemID] = "#item:"+attachment.itemID;
}
}
}
for each(item in items) {
// these items are global
resource = itemResources[item.itemID];
container = null;
containerElement = null;
section = null;
/** CORE FIELDS **/
// title
if(item.title) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"title", item.title, true);
}
// type
var type = null;
if(item.itemType == "book") {
type = "Book";
} else if (item.itemType == "bookSection") {
type = "BookSection";
container = "Book";
} else if(item.itemType == "journalArticle") {
type = "Article";
container = "Journal";
} else if(item.itemType == "magazineArticle") {
type = "Article";
container = "Periodical";
} else if(item.itemType == "newspaperArticle") {
type = "Article";
container = "Newspaper";
} else if(item.itemType == "thesis") {
type = "Thesis";
} else if(item.itemType == "letter") {
type = "Letter";
} else if(item.itemType == "manuscript") {
type = "Manuscript";
} else if(item.itemType == "interview") {
type = "Interview";
} else if(item.itemType == "film") {
type = "MotionPicture";
} else if(item.itemType == "artwork") {
type = "Illustration";
} else if(item.itemType == "website") {
type = "Document";
} else if(item.itemType == "note") {
type = "Memo";
2006-08-08 23:00:33 +00:00
if(!Scholar.getOption("exportNotes")) {
continue;
}
} else if(item.itemType == "attachment") {
handleAttachment(resource, item);
continue;
}
if(type) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, rdf+"type", n.bib+type, false);
}
// authors/editors/contributors
var creatorContainers = new Object();
for(var j in item.creators) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var creator = Scholar.RDF.newResource();
Scholar.RDF.addStatement(creator, rdf+"type", n.foaf+"Person", false);
// gee. an entire vocabulary for describing people, and these aren''t even
// standardized in it. oh well. using them anyway.
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(creator, n.foaf+"surname", item.creators[j].lastName, true);
Scholar.RDF.addStatement(creator, n.foaf+"givenname", item.creators[j].firstName, true);
// in addition, these tags are not yet in Biblio, but Bruce D''Arcus
// says they will be.
if(item.creators[j].creatorType == "author") {
var cTag = "authors";
} else if(item.creators[j].creatorType == "editor") {
var cTag = "editors";
} else {
var cTag = "contributors";
}
if(!creatorContainers[cTag]) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var creatorResource = Scholar.RDF.newResource();
// create new seq for author type
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
creatorContainers[cTag] = Scholar.RDF.newContainer("seq", creatorResource);
// attach container to resource
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.bib+cTag, creatorResource, false);
}
Scholar.RDF.addContainerElement(creatorContainers[cTag], creator, false);
}
/** FIELDS ON NEARLY EVERYTHING BUT NOT A PART OF THE CORE **/
// source
if(item.source) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"source", item.source, true);
}
// url
if(item.url) {
// add url as identifier
var term = Scholar.RDF.newResource();
// set term type
Scholar.RDF.addStatement(term, rdf+"type", n.dcterms+"URI", false);
// set url value
Scholar.RDF.addStatement(term, rdf+"value", attachment.url, true);
// add relationship to resource
Scholar.RDF.addStatement(resource, n.dc+"identifier", term, false);
}
// accessionNumber as generic ID
if(item.accessionNumber) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"identifier", item.accessionNumber, true);
}
// rights
if(item.rights) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"rights", item.rights, true);
}
/** SUPPLEMENTAL FIELDS **/
// use section to set up another container element
if(item.section) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
section = Scholar.RDF.newResource(); // leave as global
// set section type
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(section, rdf+"type", n.bib+"Part", false);
// set section title
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(section, n.dc+"title", item.section, true);
// add relationship to resource
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"isPartOf", section, false);
}
// generate container
if(container) {
if(item.ISSN && !Scholar.RDF.getArcsIn("urn:issn:"+item.ISSN)) {
// use ISSN as container URI if no other item is
containerElement = "urn:issn:"+item.ISSN
} else {
containerElement = Scholar.RDF.newResource();
}
// attach container to section (if exists) or resource
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement((section ? section : resource), n.dcterms+"isPartOf", containerElement, false);
// add container type
Scholar.RDF.addStatement(containerElement, rdf+"type", n.bib+container, false);
}
// ISSN
if(item.ISSN) {
Scholar.RDF.addStatement((containerElement ? containerElement : resource), n.dc+"identifier", "ISSN "+item.ISSN, true);
}
// ISBN
if(item.ISBN) {
Scholar.RDF.addStatement((containerElement ? containerElement : resource), n.dc+"identifier", "ISBN "+item.ISBN, true);
}
// DOI
if(item.DOI) {
Scholar.RDF.addStatement((containerElement ? containerElement : resource), n.dc+"identifier", "DOI "+item.DOI, true);
}
// publication gets linked to container via isPartOf
if(item.publicationTitle) {
Scholar.RDF.addStatement((containerElement ? containerElement : resource), n.dc+"title", item.publicationTitle, true);
}
// series also linked in
if(item.seriesTitle) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var series = Scholar.RDF.newResource();
// set series type
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(series, rdf+"type", n.bib+"Series", false);
// set series title
Scholar.RDF.addStatement(series, n.dc+"title", item.seriesTitle, true);
// add relationship to resource
Scholar.RDF.addStatement((containerElement ? containerElement : resource), n.dcterms+"isPartOf", series, false);
}
// volume
if(item.volume) {
Scholar.RDF.addStatement((containerElement ? containerElement : resource), n.prism+"volume", item.volume, true);
}
// number
if(item.issue) {
Scholar.RDF.addStatement((containerElement ? containerElement : resource), n.prism+"number", item.issue, true);
}
// edition
if(item.edition) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.prism+"edition", item.edition, true);
}
// publisher/distributor and place
if(item.publisher || item.distributor || item.place) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var organization = Scholar.RDF.newResource();
// set organization type
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(organization, rdf+"type", n.foaf+"Organization", false);
// add relationship to resource
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"publisher", organization, false);
// add publisher/distributor
if(item.publisher) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(organization, n.foaf+"name", item.publisher, true);
} else if(item.distributor) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(organization, n.foaf+"name", item.distributor, true);
}
// add place
if(item.place) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var address = Scholar.RDF.newResource();
// set address type
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(address, rdf+"type", n.vcard+"Address", false);
// set address locality
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(address, n.vcard+"locality", item.place, true);
// add relationship to organization
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(organization, n.vcard+"adr", address, false);
}
}
// date/year
if(item.date) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"date", item.date, true);
}
if(item.accessDate) { // use date submitted for access date?
Scholar.RDF.addStatement(resource, n.dcterms+"dateSubmitted", item.accessDate, true);
}
// callNumber
if(item.callNumber) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var term = Scholar.RDF.newResource();
// set term type
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(term, rdf+"type", n.dcterms+"LCC", false);
// set callNumber value
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(term, rdf+"value", item.callNumber, true);
// add relationship to resource
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"subject", term, false);
}
// archiveLocation
if(item.archiveLocation) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"coverage", item.archiveLocation, true);
}
// type (not itemType)
if(item.type) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"type", item.type, true);
} else if(item.thesisType) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"type", item.thesisType, true);
}
// THIS IS NOT YET IN THE BIBLIO NAMESPACE, BUT BRUCE D''ARCUS HAS SAID
// IT WILL BE SOON
if(item.pages) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.bib+"pages", item.pages, true);
}
// journalAbbreviation
if(item.journalAbbreviation) {
Scholar.RDF.addStatement((containerElement ? containerElement : resource), n.dcterms+"alternative", item.journalAbbreviation, true);
}
/** NOTES **/
2006-08-08 23:00:33 +00:00
if(Scholar.getOption("exportNotes")) {
for(var j in item.notes) {
var noteResource = itemResources[item.notes[j].itemID];
// add note tag
Scholar.RDF.addStatement(noteResource, rdf+"type", n.bib+"Memo", false);
// add note value
Scholar.RDF.addStatement(noteResource, rdf+"value", item.notes[j].note, true);
// add relationship between resource and note
Scholar.RDF.addStatement(resource, n.dcterms+"isReferencedBy", noteResource, false);
// Add see also info to RDF
generateSeeAlso(resource, item.notes[j].seeAlso);
}
2006-08-08 23:00:33 +00:00
if(item.note) {
Scholar.RDF.addStatement(resource, rdf+"value", item.note, true);
}
}
/** FILES **/
for each(var attachment in item.attachments) {
var attachmentResource = itemResources[attachment.itemID];
Scholar.RDF.addStatement(resource, n.dc+"relation", attachmentResource, false);
handleAttachment(attachmentResource, attachment);
}
/** TAGS **/
for(var j in item.tags) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, n.dc+"subject", item.tags[j], true);
}
// Add see also info to RDF
generateSeeAlso(resource, item.seeAlso);
}
/** RDF COLLECTION STRUCTURE **/
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var collection;
while(collection = Scholar.nextCollection()) {
generateCollection(collection);
}
}');
REPLACE INTO "translators" VALUES ('6e372642-ed9d-4934-b5d1-c11ac758ebb7', '2006-07-05 23:40:00', 2, 'Unqualified Dublin Core (RDF/XML)', 'Simon Kornblith', 'rdf',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'Scholar.configure("dataMode", "rdf");',
'function doExport() {
var dc = "http://purl.org/dc/elements/1.1/";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addNamespace("dc", dc);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var item;
while(item = Scholar.nextItem()) {
if(item.itemType == "note") {
continue;
}
var resource;
if(item.ISBN) {
resource = "urn:isbn:"+item.ISBN;
} else if(item.url) {
resource = item.url;
} else {
// just specify a node ID
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
resource = Scholar.RDF.newResource();
}
/** CORE FIELDS **/
// title
if(item.title) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"title", item.title, true);
}
// type
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"type", item.itemType, true);
// creators
for(var j in item.creators) {
// put creators in lastName, firstName format (although DC doesn''t specify)
var creator = item.creators[j].lastName;
if(item.creators[j].firstName) {
creator += ", "+item.creators[j].firstName;
}
if(item.creators[j].creatorType == "author") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"creator", creator, true);
} else {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"contributor", creator, true);
}
}
/** FIELDS ON NEARLY EVERYTHING BUT NOT A PART OF THE CORE **/
// source
if(item.source) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"source", item.source, true);
}
// accessionNumber as generic ID
if(item.accessionNumber) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"identifier", item.accessionNumber, true);
}
// rights
if(item.rights) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"rights", item.rights, true);
}
/** SUPPLEMENTAL FIELDS **/
// TODO - create text citation and OpenURL citation to handle volume, number, pages, issue, place
// publisher/distributor
if(item.publisher) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"publisher", item.publisher, true);
} else if(item.distributor) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"publisher", item.distributor, true);
}
// date/year
if(item.date) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"date", item.date, true);
}
// ISBN/ISSN/DOI
if(item.ISBN) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"identifier", "ISBN "+item.ISBN, true);
}
if(item.ISSN) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"identifier", "ISSN "+item.ISSN, true);
}
if(item.DOI) {
Scholar.RDF.addStatement(resource, dc+"identifier", "DOI "+item.DOI, true);
}
// callNumber
if(item.callNumber) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"identifier", item.callNumber, true);
}
// archiveLocation
if(item.archiveLocation) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.RDF.addStatement(resource, dc+"coverage", item.archiveLocation, true);
}
}
}');
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
REPLACE INTO "translators" VALUES ('5e3ad958-ac79-463d-812b-a86a9235c28f', '2006-07-15 17:09:00', 1, 'RDF', 'Simon Kornblith', 'rdf',
'Scholar.configure("dataMode", "rdf");
function detectImport() {
// unfortunately, Mozilla will let you create a data source from any type
// of XML, so we need to make sure there are actually nodes
var nodes = Scholar.RDF.getAllResources();
if(nodes) {
return true;
}
}',
'// gets the first result set for a property that can be encoded in multiple
// ontologies
function getFirstResults(node, properties, onlyOneString) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i=0; i<properties.length; i++) {
var result = Scholar.RDF.getTargets(node, properties[i]);
if(result) {
if(onlyOneString) {
// onlyOneString means we won''t return nsIRDFResources, only
// actual literals
if(typeof(result[0]) != "object") {
return result[0];
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else {
return result;
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
return; // return undefined on failure
}
// adds creators to an item given a list of creator nodes
function handleCreators(newItem, creators, creatorType) {
if(!creators) {
return;
}
if(typeof(creators[0]) != "string") { // see if creators are in a container
try {
var creators = Scholar.RDF.getContainerElements(creators[0]);
} catch(e) {}
}
if(typeof(creators[0]) == "string") { // support creators encoded as strings
for(var i in creators) {
if(typeof(creators[i]) != "object") {
newItem.creators.push(Scholar.Utilities.cleanAuthor(creators[i], creatorType, true));
}
}
} else { // also support foaf
for(var i in creators) {
var type = Scholar.RDF.getTargets(creators[i], rdf+"type");
if(type) {
type = Scholar.RDF.getResourceURI(type[0]);
if(type == n.foaf+"Person") { // author is FOAF type person
var creator = new Array();
creator.lastName = getFirstResults(creators[i],
[n.foaf+"surname", n.foaf+"family_name"], true);
creator.firstName = getFirstResults(creators[i],
[n.foaf+"givenname", n.foaf+"firstName"], true);
creator.creatorType = creatorType;
newItem.creators.push(creator);
}
}
}
}
}
// gets attachment info
function handleAttachment(node, attachment) {
if(!attachment) {
attachment = new Array();
}
attachment.title = getFirstResults(node, [n.dc+"title"], true);
var identifiers = getFirstResults(node, [n.dc+"identifier"]);
for each(var identifier in identifiers) {
if(typeof(identifier) != "string") {
var identifierType = Scholar.RDF.getTargets(identifier, rdf+"type");
if(identifierType) {
identifierType = Scholar.RDF.getResourceURI(identifierType[0]);
if(identifierType == n.dcterms+"URI") { // uri is url
attachment.url = getFirstResults(identifier, [rdf+"value"], true);
}
}
}
}
var formats = getFirstResults(node, [n.dc+"format"]);
for each(var format in formats) {
if(typeof(format) != "string") {
var formatType = Scholar.RDF.getTargets(format, rdf+"type");
if(formatType) {
formatType = Scholar.RDF.getResourceURI(formatType[0]);
if(formatType == n.dcterms+"IMT") { // uri is url
attachment.mimeType = getFirstResults(format, [rdf+"value"], true);
}
}
}
}
var stringNode = node;
if(typeof(stringNode) != "string") {
stringNode = Scholar.RDF.getResourceURI(stringNode);
}
if(stringNode.substr(0, 8) == "file:///") {
// not a protocol specifier; we have a path name
attachment.path = stringNode;
}
return attachment;
}
// processes collections recursively
function processCollection(node, collection) {
if(!collection) {
collection = new Array();
}
collection.type = "collection";
collection.name = getFirstResults(node, [n.dc+"title"], true);
collection.children = new Array();
// check for children
var children = getFirstResults(node, [n.dcterms+"hasPart"]);
for each(var child in children) {
var type = Scholar.RDF.getTargets(child, rdf+"type");
if(type) {
type = Scholar.RDF.getResourceURI(type[0]);
}
if(type == n.bib+"Collection") {
// for collections, process recursively
collection.children.push(processCollection(child));
} else {
// all other items are added by ID
collection.children.push({id:Scholar.RDF.getResourceURI(child), type:"item"});
}
}
return collection;
}
// gets the node with a given type from an array
function getNodeByType(nodes, type) {
if(!nodes) {
return false;
}
for each(node in nodes) {
var nodeType = Scholar.RDF.getTargets(node, rdf+"type");
if(nodeType) {
nodeType = Scholar.RDF.getResourceURI(nodeType[0]);
if(nodeType == type) { // we have a node of the correct type
return node;
}
}
}
return false;
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
function doImport() {
rdf = "http://www.w3.org/1999/02/22-rdf-syntax-ns#";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
n = {
bib:"http://purl.org/net/biblio#",
dc:"http://purl.org/dc/elements/1.1/",
dcterms:"http://purl.org/dc/terms/",
prism:"http://prismstandard.org/namespaces/1.2/basic/",
foaf:"http://xmlns.com/foaf/0.1/",
vcard:"http://nwalsh.com/rdf/vCard#",
fs:"http://chnm.gmu.edu/firefoxscholar/rdf#"
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
};
callNumberTypes = [
n.dcterms+"LCC", n.dcterms+"DDC", n.dcterms+"UDC"
];
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var nodes = Scholar.RDF.getAllResources();
if(!nodes) {
return false;
}
// keep track of collections while we''re looping through
var collections = new Array();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for each(var node in nodes) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var newItem = new Scholar.Item();
newItem.itemID = Scholar.RDF.getResourceURI(node);
var container = undefined;
// type
var type = Scholar.RDF.getTargets(node, rdf+"type");
// also deal with type detection based on parts, so we can differentiate
// magazine and journal articles, and find container elements
var isPartOf = getFirstResults(node, [n.dcterms+"isPartOf"]);
if(type) {
type = Scholar.RDF.getResourceURI(type[0]);
if(type == n.bib+"Book") {
newItem.itemType = "book";
} else if(type == n.bib+"BookSection") {
newItem.itemType = "bookSection";
container = getNodeByType(isPartOf, n.bib+"Book");
} else if(type == n.bib+"Article") { // choose between journal,
// newspaper, and magazine
// articles
if(container = getNodeByType(isPartOf, n.bib+"Journal")) {
newItem.itemType = "journalArticle";
} else if(container = getNodeByType(isPartOf, n.bib+"Periodical")) {
newItem.itemType = "magazineArticle";
} else if(container = getNodeByType(isPartOf, n.bib+"Newspaper")) {
newItem.itemType = "newspaperArticle";
}
} else if(type == n.bib+"Thesis") {
newItem.itemType = "thesis";
} else if(type == n.bib+"Letter") {
newItem.itemType = "letter";
} else if(type == n.bib+"Manuscript") {
newItem.itemType = "manuscript";
} else if(type == n.bib+"Interview") {
newItem.itemType = "interview";
} else if(type == n.bib+"MotionPicture") {
newItem.itemType = "film";
} else if(type == n.bib+"Illustration") {
newItem.itemType = "illustration";
} else if(type == n.bib+"Document") {
newItem.itemType = "website";
} else if(type == n.bib+"Memo") {
// check to see if this note is independent
var arcs = Scholar.RDF.getArcsIn(node);
var skip = false;
for each(var arc in arcs) {
arc = Scholar.RDF.getResourceURI(arc);
if(arc != n.dc+"relation" && arc != n.dcterms+"hasPart") {
// related to another item by some arc besides see also
skip = true;
}
}
if(skip) {
continue;
}
newItem.itemType = "note";
} else if(type == n.bib+"Collection") {
// skip collections until all the items are done
collections.push(node);
continue;
} else if(type == n.fs+"File") {
// check to see if file is independent
var arcs = Scholar.RDF.getArcsIn(node);
if(arcs.length) {
continue;
}
// process as file
newItem.itemType = "attachment";
handleAttachment(node, newItem);
Scholar.Utilities.debug(newItem);
newItem.complete();
continue;
} else { // default to book
newItem.itemType = "book";
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// title
newItem.title = getFirstResults(node, [n.dc+"title"], true);
if(newItem.itemType != "note" && !newItem.title) { // require the title
// (if not a note)
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
continue;
}
// regular author-type creators
var creators = getFirstResults(node, [n.bib+"authors", n.dc+"creator"]);
handleCreators(newItem, creators, "author");
// editors
var creators = getFirstResults(node, [n.bib+"editors"]);
handleCreators(newItem, creators, "editor");
// contributors
var creators = getFirstResults(node, [n.bib+"contributors"]);
handleCreators(newItem, creators, "contributor");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// source
newItem.source = getFirstResults(node, [n.dc+"source"], true);
// rights
newItem.rights = getFirstResults(node, [n.dc+"rights"], true);
// section
var section = getNodeByType(isPartOf, n.bib+"Part");
if(section) {
newItem.section = getFirstResults(section, [n.dc+"title"], true);
}
// publication
if(container) {
newItem.publicationTitle = getFirstResults(container, [n.dc+"title"], true);
}
// series
var series = getNodeByType(isPartOf, n.bib+"Series");
if(series) {
newItem.seriesTitle = getFirstResults(container, [n.dc+"title"], true);
}
// volume
newItem.volume = getFirstResults((container ? container : node), [n.prism+"volume"], true);
// number
newItem.issue = getFirstResults((container ? container : node), [n.prism+"number"], true);
// edition
newItem.edition = getFirstResults(node, [n.prism+"edition"], true);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// publisher
var publisher = getFirstResults(node, [n.dc+"publisher"]);
if(publisher) {
if(typeof(publisher[0]) == "string") {
newItem.publisher = publisher[0];
} else {
var type = Scholar.RDF.getTargets(publisher[0], rdf+"type");
if(type) {
type = Scholar.RDF.getResourceURI(type[0]);
if(type == n.foaf+"Organization") { // handle foaf organizational publishers
newItem.publisher = getFirstResults(publisher[0], [n.foaf+"name"], true);
var place = getFirstResults(publisher[0], [n.vcard+"adr"]);
if(place) {
newItem.place = getFirstResults(place[0], [n.vcard+"locality"]);
}
}
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// (this will get ignored except for films, where we encode distributor as publisher)
newItem.distributor = newItem.publisher;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// date
newItem.date = getFirstResults(node, [n.dc+"date"], true);
// accessDate
newItem.accessDate = getFirstResults(node, [n.dcterms+"dateSubmitted"], true);
// lastModified
newItem.lastModified = getFirstResults(node, [n.dcterms+"modified"], true);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// identifier
var identifiers = getFirstResults(node, [n.dc+"identifier"]);
if(container) {
var containerIdentifiers = getFirstResults(container, [n.dc+"identifier"]);
// concatenate sets of identifiers
if(containerIdentifiers) {
if(identifiers) {
identifiers = identifiers.concat(containerIdentifiers);
} else {
identifiers = containerIdentifiers;
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(identifiers) {
for(var i in identifiers) {
var beforeSpace = identifiers[i].substr(0, identifiers[i].indexOf(" ")).toUpperCase();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(beforeSpace == "ISBN") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.ISBN = identifiers[i].substr(5).toUpperCase();
} else if(beforeSpace == "ISSN") {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.ISSN = identifiers[i].substr(5).toUpperCase();
} else if(beforeSpace == "DOI") {
newItem.DOI = identifiers[i].substr(4);
} else if(!newItem.accessionNumber) {
newItem.accessionNumber = identifiers[i];
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
}
// archiveLocation
newItem.archiveLocation = getFirstResults(node, [n.dc+"coverage"], true);
// type
newItem.type = newItem.thesisType = getFirstResults(node, [n.dc+"type"], true);
// journalAbbreviation
newItem.journalAbbreviation = getFirstResults((container ? container : node), [n.dcterms+"alternative"], true);
// see also
var relations;
if(relations = getFirstResults(node, [n.dc+"relation"])) {
for each(var relation in relations) {
newItem.seeAlso.push(Scholar.RDF.getResourceURI(relation));
}
}
/** NOTES **/
var referencedBy = Scholar.RDF.getTargets(node, n.dcterms+"isReferencedBy");
for each(var referentNode in referencedBy) {
var type = Scholar.RDF.getTargets(referentNode, rdf+"type");
if(type && Scholar.RDF.getResourceURI(type[0]) == n.bib+"Memo") {
// if this is a memo
var note = new Array();
note.note = getFirstResults(referentNode, [rdf+"value", n.dc+"description"], true);
if(note.note != undefined) {
// handle see also
var relations;
if(relations = getFirstResults(referentNode, [n.dc+"relation"])) {
note.seeAlso = new Array();
for each(var relation in relations) {
note.seeAlso.push(Scholar.RDF.getResourceURI(relation));
}
}
// add note
newItem.notes.push(note);
}
}
}
if(newItem.itemType == "note") {
// add note for standalone
newItem.note = getFirstResults(node, [rdf+"value", n.dc+"description"], true);
}
/** TAGS **/
var subjects = getFirstResults(node, [n.dc+"subject"]);
for each(var subject in subjects) {
if(typeof(subject) == "string") { // a regular tag
newItem.tags.push(subject);
} else { // a call number
var type = Scholar.RDF.getTargets(subject, rdf+"type");
if(type) {
type = Scholar.RDF.getResourceURI(type[0]);
if(Scholar.Utilities.inArray(type, callNumberTypes)) {
newItem.callNumber = getFirstResults(subject, [rdf+"value"], true);
}
}
}
}
/* ATTACHMENTS */
var relations = getFirstResults(node, [n.dc+"relation"]);
for each(var relation in relations) {
var type = Scholar.RDF.getTargets(relation, rdf+"type");
if(type) {
type = Scholar.RDF.getResourceURI(type[0]);
if(type == n.fs+"File") {
newItem.attachments.push(handleAttachment(relation));
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
}
/* COLLECTIONS */
for each(var collection in collections) {
if(!Scholar.RDF.getArcsIn(collection)) {
var newCollection = new Scholar.Collection();
processCollection(collection, newCollection);
newCollection.complete();
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}');
REPLACE INTO "translators" VALUES ('32d59d2d-b65a-4da4-b0a3-bdd3cfb979e7', '2006-08-08 17:12:00', 3, 'RIS', 'Simon Kornblith', 'ris',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'Scholar.configure("dataMode", "line");
Scholar.addOption("exportNotes", true);
function detectImport() {
var line;
while(line = Scholar.read()) {
if(line.replace(/\s/g, "") != "") {
if(line.substr(0, 6) == "TY - ") {
return true;
Scholar.Utilities.debug("YES!");
} else {
return false;
}
}
}
}',
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
'var itemsWithYears = ["book", "bookSection", "thesis", "film"];
var fieldMap = {
ID:"itemID",
T1:"title",
T3:"seriesTitle",
JF:"publicationTitle",
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
VL:"volume",
IS:"issue",
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
CP:"place",
PB:"publisher"
};
var inputFieldMap = {
TI:"title",
CT:"title",
JO:"publicationTitle",
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
CY:"place"
};
// TODO: figure out if these are the best types for letter, interview, website, manuscript
var typeMap = {
book:"BOOK",
bookSection:"CHAP",
journalArticle:"JOUR",
magazineArticle:"MGZN",
newspaperArticle:"NEWS",
thesis:"THES",
letter:"PCOMM",
manuscript:"UNPB",
interview:"PCOMM",
film:"MPCT",
artwork:"ART",
website:"ELEC"
};
// supplements outputTypeMap for importing
// TODO: BILL, CASE, COMP, CONF, DATA, HEAR, MUSIC, PAT, SOUND, STAT
var inputTypeMap = {
ABST:"journalArticle",
ADVS:"film",
CTLG:"magazineArticle",
GEN:"book",
INPR:"manuscript",
JFULL:"journalArticle",
MAP:"artwork",
PAMP:"book",
RPRT:"book",
SER:"book",
SLIDE:"artwork",
UNBILL:"manuscript",
VIDEO:"film"
};
function processTag(item, tag, value) {
if(fieldMap[tag]) {
item[fieldMap[tag]] = value;
} else if(inputFieldMap[tag]) {
item[inputFieldMap[tag]] = value;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} else if(tag == "TY") {
// look for type
// first check typeMap
for(var i in typeMap) {
if(value == typeMap[i]) {
item.itemType = i;
}
}
// then check inputTypeMap
if(!item.itemType) {
if(inputTypeMap[value]) {
item.itemType = inputTypeMap[value];
} else {
// default to generic from inputTypeMap
item.itemType = inputTypeMap["GEN"];
}
}
} else if(tag == "BT") {
// ignore, unless this is a book or unpublished work, as per spec
if(item.itemType == "book" || item.itemType == "manuscript") {
item.title = value;
}
} else if(tag == "A1" || tag == "AU") {
// primary author
var names = value.split(/, ?/);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
item.creators.push({lastName:names[0], firstName:names[1], creatorType:"author"});
} else if(tag == "A2" || tag == "ED") {
// contributing author
var names = value.split(/, ?/);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
item.creators.push({lastName:names[0], firstName:names[1], creatorType:"contributor"});
} else if(tag == "Y1" || tag == "PY") {
// year or date
var dateParts = value.split("/");
if(dateParts.length == 1) {
// technically, if there''s only one date part, the file isn''t valid
// RIS, but EndNote accepts this, so we have to too
item.date = value+"-00-00";
} else if(dateParts[1].length == 0 && dateParts[2].length == 0 && dateParts[3] && dateParts[3].length != 0) {
// in the case that we have a year and other data, format that way
item.date = dateParts[3]+(dateParts[0] ? " "+dateParts[0] : "");
} else {
// standard YMD data
item.date = Scholar.Utilities.lpad(dateParts[0], "0", 4)+"-"+Scholar.Utilities.lpad(dateParts[1], "0", 2)+"-"+Scholar.Utilities.lpad(dateParts[2], "0", 2);
}
} else if(tag == "N1" || tag == "AB") {
// notes
item.notes.push({note:value});
} else if(tag == "KW") {
// keywords/tags
item.tags.push(value);
} else if(tag == "SP") {
// start page
if(!item.pages) {
item.pages = value;
} else if(item.pages[0] == "-") { // already have ending page
item.pages = value + item.pages;
} else { // multiple ranges? hey, it''s a possibility
item.pages += ", "+value;
}
} else if(tag == "EP") {
// end page
if(value) {
if(!item.pages || value != item.pages) {
if(!item.pages) {
item.pages = "";
}
item.pages += "-"+value;
}
}
} else if(tag == "SN") {
// ISSN/ISBN - just add both
if(!item.ISBN) {
item.ISBN = value;
}
if(!item.ISSN) {
item.ISSN = value;
}
} else if(tag == "UR") {
// URL
item.url = value;
}
}
function doImport(attachments) {
Scholar.Utilities.debug("hello");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var line = true;
var tag = data = false;
do { // first valid line is type
Scholar.Utilities.debug("ignoring "+line);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
line = Scholar.read();
Scholar.Utilities.debug(line);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
} while(line !== false && line.substr(0, 6) != "TY - ");
var item = new Scholar.Item();
var i = 0;
if(attachments && attachments[i]) {
item.attachments = attachments[i];
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var tag = "TY";
var data = line.substr(6);
while((line = Scholar.read()) !== false) { // until EOF
if(line.substr(2, 4) == " - ") {
// if this line is a tag, take a look at the previous line to map
// its tag
if(tag) {
processTag(item, tag, data);
}
// then fetch the tag and data from this line
tag = line.substr(0,2);
data = line.substr(6);
Scholar.Utilities.debug("tag: ''"+tag+"''; data: ''"+data+"''");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(tag == "ER") { // ER signals end of reference
// unset info
tag = data = false;
// new item
item.complete();
item = new Scholar.Item();
i++;
if(attachments && attachments[i]) {
item.attachments = attachments[i];
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
} else {
// otherwise, assume this is data from the previous line continued
if(tag) {
data += line;
}
}
}
if(tag) { // save any unprocessed tags
processTag(item, tag, data);
item.complete();
}
}
function addTag(tag, value) {
if(value) {
Scholar.write(tag+" - "+value+"\r\n");
}
}
function doExport() {
var item;
while(item = Scholar.nextItem()) {
// can''t store independent notes in RIS
if(item.itemType == "note") {
continue;
}
// type
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
addTag("TY", typeMap[item.itemType]);
// use field map
for(var j in fieldMap) {
addTag(j, item[fieldMap[j]]);
}
// creators
for(var j in item.creators) {
// only two types, primary and secondary
var risTag = "A1"
if(item.creators[j].creatorType != "author") {
risTag = "A2";
}
addTag(risTag, item.creators[j].lastName+","+item.creators[j].firstName);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// date
if(item.date) {
var isoDate = /^[0-9]{4}(-[0-9]{2}-[0-9]{2})?$/;
if(isoDate.test(item.date)) { // can directly accept ISO format with minor mods
addTag("Y1", item.date.replace("-", "/")+"/");
} else { // otherwise, extract year and attach other data
var year = /^(.*?) *([0-9]{4})/;
var m = year.exec(item.date);
if(m) {
addTag("Y1", m[2]+"///"+m[1]);
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// notes
2006-08-08 23:00:33 +00:00
if(Scholar.getOption("exportNotes")) {
for(var j in item.notes) {
addTag("N1", item.notes[j].note.replace(/[\r\n]/g, " "));
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// tags
for(var j in item.tags) {
addTag("KY", item.tags[j]);
}
// pages
if(item.pages) {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var range = Scholar.Utilities.getPageRange(item.pages);
addTag("SP", range[0]);
addTag("EP", range[1]);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// ISBN/ISSN
addTag("SN", item.ISBN);
addTag("SN", item.ISSN);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// URL
if(item.url) {
addTag("UR", item.url);
} else if(item.source && item.source.substr(0, 7) == "http://") {
addTag("UR", item.source);
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
Scholar.write("ER - \r\n\r\n");
}
}');
REPLACE INTO "translators" VALUES ('a6ee60df-1ddc-4aae-bb25-45e0537be973', '2006-07-16 17:18:00', 1, 'MARC', 'Simon Kornblith', 'marc',
'function detectImport() {
var marcRecordRegexp = /^[0-9]{5}[a-z ]{3}$/
var read = Scholar.read(8);
if(marcRecordRegexp.test(read)) {
return true;
}
}',
'var fieldTerminator = "\x1E";
var recordTerminator = "\x1D";
var subfieldDelimiter = "\x1F";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
/*
* CLEANING FUNCTIONS
*/
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// general purpose cleaning
function clean(value) {
value = value.replace(/^[\s\.\,\/\:]+/, '''');
value = value.replace(/[\s\.\,\/\:]+$/, '''');
value = value.replace(/ +/g, '' '');
var char1 = value[0];
var char2 = value[value.length-1];
if((char1 == "[" && char2 == "]") || (char1 == "(" && char2 == ")")) {
// chop of extraneous characters
return value.substr(1, value.length-2);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
return value;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
// number extraction
function pullNumber(text) {
var pullRe = /[0-9]+/;
var m = pullRe.exec(text);
if(m) {
return m[0];
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
// ISBN extraction
function pullISBN(text) {
var pullRe = /[0-9X\-]+/;
var m = pullRe.exec(text);
if(m) {
return m[0];
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
// corporate author extraction
function corpAuthor(author) {
return {lastName:author};
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
// regular author extraction
function author(author, type, useComma) {
return Scholar.Utilities.cleanAuthor(author, type, useComma);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
/*
* END CLEANING FUNCTIONS
*/
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var record = function() {
this.directory = new Object();
this.leader = "";
this.content = "";
// defaults
this.indicatorLength = 2;
this.subfieldCodeLength = 2;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
// import a binary MARC record into this record
record.prototype.importBinary = function(record) {
// get directory and leader
var directory = record.substr(0, record.indexOf(fieldTerminator));
this.leader = directory.substr(0, 24);
var directory = directory.substr(24);
// get various data
this.indicatorLength = parseInt(this.leader[10], 10);
this.subfieldCodeLength = parseInt(this.leader[11], 10);
var baseAddress = parseInt(this.leader.substr(12, 5), 10);
// get record data
this.content = record.substr(baseAddress);
// read directory
for(var i=0; i<directory.length; i+=12) {
var tag = parseInt(directory.substr(i, 3), 10);
var fieldLength = parseInt(directory.substr(i+3, 4), 10);
var fieldPosition = parseInt(directory.substr(i+7, 5), 10);
if(!this.directory[tag]) {
this.directory[tag] = new Array();
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
this.directory[tag].push([fieldPosition, fieldLength]);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
// add a field to this record
record.prototype.addField = function(field, indicator, value) {
// make sure indicator is the right length
if(indicator.length > this.indicatorLength) {
indicator = indicator.substr(0, this.indicatorLength);
} else if(indicator.length != this.indicatorLength) {
indicator = Scholar.Utilities.lpad(indicator, " ", this.indicatorLength);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
// add terminator
value = indicator+value+fieldTerminator;
// add field to directory
if(!this.directory[field]) {
this.directory[field] = new Array();
}
this.directory[field].push([this.content.length, value.length]);
// add field to record
this.content += value;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
// get all fields with a certain field number
record.prototype.getField = function(field) {
var fields = new Array();
// make sure fields exist
if(!this.directory[field]) {
return fields;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
// get fields
for(var i in this.directory[field]) {
var location = this.directory[field][i];
// add to array
fields.push([this.content.substr(location[0], this.indicatorLength),
this.content.substr(location[0]+this.indicatorLength,
location[1]-this.indicatorLength-1)]);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
return fields;
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
// get subfields from a field
record.prototype.getFieldSubfields = function(tag) { // returns a two-dimensional array of values
var fields = this.getField(tag);
var returnFields = new Array();
for(var i in fields) {
returnFields[i] = new Object();
var subfields = fields[i][1].split(subfieldDelimiter);
if (subfields.length == 1) {
returnFields[i]["?"] = fields[i][1];
} else {
for(var j in subfields) {
if(subfields[j]) {
returnFields[i][subfields[j].substr(0, this.subfieldCodeLength-1)] = subfields[j].substr(this.subfieldCodeLength-1);
}
}
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
return returnFields;
}
// add field to DB
record.prototype._associateDBField = function(item, fieldNo, part, fieldName, execMe, arg1, arg2) {
var field = this.getFieldSubfields(fieldNo);
Scholar.Utilities.debug(''found ''+field.length+'' matches for ''+fieldNo+part);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(field) {
for(var i in field) {
var value = false;
for(var j=0; j<part.length; j++) {
var myPart = part[j];
if(field[i][myPart]) {
if(value) {
value += " "+field[i][myPart];
} else {
value = field[i][myPart];
}
}
}
if(value) {
value = clean(value);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(execMe) {
value = execMe(value, arg1, arg2);
}
if(fieldName == "creator") {
item.creators.push(value);
} else {
item[fieldName] = value;
}
}
}
}
}
// add field to DB as tags
record.prototype._associateTags = function(item, fieldNo, part) {
var field = this.getFieldSubfields(fieldNo);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
for(var i in field) {
for(var j=0; j<part.length; j++) {
var myPart = part[j];
if(field[i][myPart]) {
item.tags.push(clean(field[i][myPart]));
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
}
}
// this function loads a MARC record into our database
record.prototype.translate = function(item) {
// get item type
if(this.leader) {
var marcType = this.leader[6];
if(marcType == "g") {
item.itemType = "film";
} else if(marcType == "k" || marcType == "e" || marcType == "f") {
item.itemType = "artwork";
} else if(marcType == "t") {
item.itemType = "manuscript";
} else {
item.itemType = "book";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
} else {
item.itemType = "book";
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
// Extract ISBNs
this._associateDBField(item, "020", "a", "ISBN", pullISBN);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Extract ISSNs
this._associateDBField(item, "022", "a", "ISSN", pullISBN);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Extract creators
this._associateDBField(item, "100", "a", "creator", author, "author", true);
this._associateDBField(item, "110", "a", "creator", corpAuthor, "author");
this._associateDBField(item, "111", "a", "creator", corpAuthor, "author");
this._associateDBField(item, "700", "a", "creator", author, "contributor", true);
this._associateDBField(item, "710", "a", "creator", corpAuthor, "contributor");
this._associateDBField(item, "711", "a", "creator", corpAuthor, "contributor");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(!item.creators.length) {
// some LOC entries have no listed author, but have the author in the person subject field as the first entry
var field = this.getFieldSubfields("600");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
if(field[0]) {
item.creators.push(cleanAuthor(field[0]["a"], true));
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
}
// Extract tags
// personal
this._associateTags(item, "600", "aqtxyz");
// corporate
this._associateTags(item, "611", "abtxyz");
// meeting
this._associateTags(item, "630", "acetxyz");
// uniform title
this._associateTags(item, "648", "atxyz");
// chronological
this._associateTags(item, "650", "axyz");
// topical
this._associateTags(item, "651", "abcxyz");
// geographic
this._associateTags(item, "653", "axyz");
// uncontrolled
this._associateTags(item, "653", "a");
// faceted topical term (whatever that means)
this._associateTags(item, "654", "abcyz");
// genre/form
this._associateTags(item, "655", "abcxyz");
// occupation
this._associateTags(item, "656", "axyz");
// function
this._associateTags(item, "657", "axyz");
// curriculum objective
this._associateTags(item, "658", "ab");
// hierarchical geographic place name
this._associateTags(item, "662", "abcdfgh");
// Extract title
this._associateDBField(item, "245", "ab", "title");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Extract edition
this._associateDBField(item, "250", "a", "edition");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Extract place info
this._associateDBField(item, "260", "a", "place");
// Extract publisher/distributor
if(item.itemType == "film") {
this._associateDBField(item, "260", "b", "distributor");
} else {
this._associateDBField(item, "260", "b", "publisher");
}
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Extract year
this._associateDBField(item, "260", "c", "date", pullNumber);
// Extract pages
this._associateDBField(item, "300", "a", "pages", pullNumber);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Extract series
this._associateDBField(item, "440", "a", "seriesTitle");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
// Extract call number
this._associateDBField(item, "084", "ab", "callNumber");
this._associateDBField(item, "082", "a", "callNumber");
this._associateDBField(item, "080", "ab", "callNumber");
this._associateDBField(item, "070", "ab", "callNumber");
this._associateDBField(item, "060", "ab", "callNumber");
this._associateDBField(item, "050", "ab", "callNumber");
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
}
function doImport() {
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
var text;
var holdOver = ""; // part of the text held over from the last loop
while(text = Scholar.read(4096)) { // read in 4096 byte increments
var records = text.split("\x1D");
if(records.length > 1) {
records[0] = holdOver + records[0];
holdOver = records.pop(); // skip last record, since it''s not done
for(var i in records) {
var newItem = new Scholar.Item();
// create new record
var rec = new record();
rec.importBinary(records[i]);
rec.translate(newItem);
closes #78, figure out import/export architecture closes #100, migrate ingester to Scholar.Translate closes #88, migrate scrapers away from RDF closes #9, pull out LC subject heading tags references #87, add fromArray() and toArray() methods to item objects API changes: all translation (import/export/web) now goes through Scholar.Translate all Scholar-specific functions in scrapers start with "Scholar." rather than the jumbled up piggy bank un-namespaced confusion scrapers now longer specify items through RDF (the beginning of an item.fromArray()-like function exists in Scholar.Translate.prototype._itemDone()) scrapers can be any combination of import, export, and web (type is the sum of 1/2/4 respectively) scrapers now contain functions (doImport, doExport, doWeb) rather than loose code scrapers can call functions in other scrapers or just call the function to translate itself export accesses items item-by-item, rather than accepting a huge array of items MARC functions are now in the MARC import translator, and accessed by the web translators new features: import now works rudimentary RDF (unqualified dublin core only), RIS, and MARC import translators are implemented (although they are a little picky with respect to file extensions at the moment) items appear as they are scraped MARC import translator pulls out tags, although this seems to slow things down no icon appears next to a the URL when Scholar hasn't detected metadata, since this seemed somewhat confusing apologizes for the size of this diff. i figured if i was going to re-write the API, i might as well do it all at once and get everything working right.
2006-07-17 04:06:58 +00:00
newItem.complete();
}
} else {
holdOver += text;
}
}
}');
REPLACE INTO "csl" VALUES('http://purl.org/net/xbiblio/csl/styles/apa.csl', '2006-08-12 19:22:00', 'APA',
'<?xml version="1.0" encoding="UTF-8"?>
<?oxygen RNGSchema="file:/Users/darcusb/xbiblio/csl/schema/trunk/csl-alt.rnc" type="compact"?>
<style xmlns="http://purl.org/net/xbiblio/csl" class="author-date" xml:lang="en">
<info>
<title>American Psychological Association</title>
<id>http://purl.org/net/xbiblio/csl/styles/apa.csl</id>
<link>http://purl.org/net/xbiblio/csl/styles/apa.csl</link>
<author>
<name>Bruce DArcus</name>
<email>bdarcus@sourceforge.net</email>
</author>
<contributor>
<name>Simon Kornblith</name>
<email>simon@simonster.com</email>
</contributor>
<updated>2006-08-19T17:12:00-05:00</updated>
</info>
<defaults>
<contributor name-as-sort-order="no">
<name and="symbol" initialize-with="."/>
<label term-set="roles-short" prefix=", " text-transform="capitalize" suffix="."/>
</contributor>
<author name-as-sort-order="all">
<name and="symbol" sort-separator=", " initialize-with="."/>
<label term-set="roles-short" prefix=" (" suffix=".)" text-transform="capitalize"/>
<substitute>
<choose>
<editor/>
<translator/>
<titles/>
</choose>
</substitute>
</author>
<locator>
<number/>
</locator>
<identifier>
<number/>
</identifier>
<titles>
<title/>
</titles>
<date>
<year/>
<month prefix=", "/>
<day prefix=" "/>
</date>
<publisher>
<place suffix=": "/>
<name/>
</publisher>
<access>
<text term-name="retrieved" text-transform="capitalize"/>
<date suffix=", ">
<month/>
<day suffix=", "/>
<year/>
</date>
<text term-name="from"/>
<url/>
<date prefix=", "/>
</access>
</defaults>
<citation prefix="(" suffix=")" delimiter="; ">
<et-al min-authors="6" use-first="6" position="first"/>
<et-al min-authors="6" use-first="1" position="subsequent"/>
<layout>
<item>
<author form="short" suffix=", "/>
<date>
<year/>
</date>
<locator prefix=": "/>
</item>
</layout>
</citation>
<bibliography hanging-indent="true">
<sort algorithm="author-date"/>
<et-al min-authors="4" use-first="3"/>
<layout>
<list>
<heading>
<text term-name="references"/>
</heading>
</list>
<item suffix=".">
<choose>
<type name="book">
<author/>
<date prefix=" (" suffix=").">
<year/>
</date>
<group suffix=".">
<titles font-style="italic" prefix=" "/>
<group prefix=" (" suffix=")" delimiter=", ">
<editor/>
<translator/>
</group>
</group>
<publisher prefix=" "/>
<access prefix=" "/>
</type>
<type name="chapter">
<author/>
<date prefix=" (" suffix=").">
<year/>
</date>
<titles font-style="italic" prefix=" "/>
<group class="container" prefix=" ">
<text term-name="in" text-transform="capitalize"/>
<editor prefix=" " suffix=",">
<name and="symbol" sort-separator=", " initialize-with="."/>
<label term-set="roles-short" prefix=" (" suffix=")" text-transform="capitalize"/>
</editor>
<translator prefix=" " suffix=",">
<name and="symbol" sort-separator=", " initialize-with="."/>
<label term-set="roles-short" prefix=" (" suffix=")" text-transform="capitalize"/>
</translator>
<titles relation="container" font-style="italic" prefix=" " suffix="."/>
<titles relation="collection" prefix=" " suffix="."/>
<publisher prefix=" "/>
<pages prefix=" (" suffix=")">
<label term-set="locators-short" text-transform="capitalize" suffix=". "/>
<number/>
</pages>
</group>
<access prefix=" "/>
</type>
<type name="article">
<author/>
<date prefix=" (" suffix=").">
<year/>
</date>
<group suffix=".">
<titles prefix=" "/>
<group prefix=" (" suffix=")" delimiter=", ">
<editor/>
<translator/>
</group>
</group>
<group class="container" prefix=" " suffix=".">
<titles relation="container" font-style="italic"/>
<volume prefix=", " font-style="italic"/>
<issue prefix="(" suffix=")"/>
<pages prefix=", "/>
</group>
<access prefix=" "/>
</type>
</choose>
</item>
</layout>
</bibliography>
</style>');
REPLACE INTO "csl" VALUES('http://purl.org/net/xbiblio/csl/styles/chicago-note.csl', '2006-08-12 19:22:00', 'Chicago (Footnote)',
'<?xml version="1.0" encoding="UTF-8"?>
<?oxygen RNGSchema="../schema/trunk/csl.rnc" type="compact"?>
<style xmlns="http://purl.org/net/xbiblio/csl" class="note" xml:lang="en">
<info>
<title>Chicago Note Sans Reference List</title>
<id>http://purl.org/net/xbiblio/csl/styles/chicago-note.csl</id>
<author>
<name>Bruce DArcus</name>
<email>bdarcus@sourceforge.net</email>
</author>
<contributor>
<name>Simon Kornblith</name>
<email>simon@simonster.com</email>
</contributor>
<updated>2006-08-19T17:12:00-05:00</updated>
<summary>The note-without-bibliography variant of the Chicago style.</summary>
</info>
<defaults>
<contributor>
<label term-set="roles-short" suffix=". " text-transform="lowercase"/>
<name and="text"/>
</contributor>
<author>
<name and="text"/>
<label term-set="roles-short" prefix=", " suffix="." text-transform="lowercase"/>
<substitute>
<choose>
<editor/>
<translator/>
</choose>
</substitute>
</author>
<locator>
<number/>
</locator>
<titles>
<title/>
</titles>
<date>
<year/>
</date>
<publisher>
<place suffix=": "/>
<name/>
</publisher>
<access>
<url/>
<date prefix=" "/>
</access>
</defaults>
<citation suffix=".">
<et-al min-authors="4" use-first="1"/>
<layout>
<item>
<choose>
<type name="book">
<author suffix=", "/>
<titles font-style="italic"/>
<editor prefix=", "/>
<translator prefix=", "/>
<group prefix=" (" suffix=")" delimiter=", ">
<publisher/>
<date/>
</group>
<pages prefix=", "/>
</type>
<type name="chapter">
<author suffix=", "/>
<titles prefix="&#8220;" suffix=",&#8221; "/>
<group class="container">
<text term-name="in" text-transform="lowercase"/>
<titles relation="container" prefix=" " font-style="italic"/>
<editor prefix=", "/>
<translator prefix=", "/>
<pages prefix=", "/>
<group prefix=" (" suffix=")" delimiter=", ">
<publisher/>
<date/>
</group>
</group>
</type>
<type name="journal-article">
<author suffix=", "/>
<titles prefix="&#8220;" suffix=",&#8221; "/>
<titles relation="container" font-style="italic"/>
<volume prefix=" "/>
<issue prefix=" (" suffix=")"/>
<pages prefix=": "/>
</type>
<type name="article">
<author suffix=", "/>
<titles prefix="&#8220;" suffix=",&#8221; "/>
<titles relation="container" font-style="italic"/>
<date prefix=", ">
<day suffix=" "/>
<month suffix=" " text-transform="capitalize"/>
<year/>
</date>
</type>
</choose>
</item>
<item position="subsequent" ibid="true">
<author/>
<title prefix=", "/>
</item>
</layout>
</citation>
</style>');