mirrors/zotero
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages 1 Wiki Activity Actions

Quick update of recognizePDF from Q to Bluebird

Browse source 
This should be refactored further, but this is a start at swapping in
Bluebird. Unfortunately the process doesn't work correctly because of
issues in the translation framework, which will need to be addressed
separately -- right now a parent item is successfully created but the UI
still shows an error and the PDF isn't placed under the parent.
This commit is contained in:
Dan Stillman 2015-12-07 17:23:10 -05:00
parent fb46d6d216
commit bad92a1d22
1 changed files with 144 additions and 116 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

260
chrome/content/zotero/recognizePDF.js
View file

@ -69,62 +69,70 @@ var Zotero_RecognizePDF = new function() {
* process is to be interrupted
* @return {Promise} A promise resolved when PDF metadata has been retrieved
*/
this.recognize = function(file, libraryID, stopCheckCallback) {
this.recognize = Zotero.Promise.coroutine(function* (file, libraryID, stopCheckCallback) {
const MAX_PAGES = 15;
var me = this;
return _extractText(file, MAX_PAGES).then(function(lines) {
// Look for DOI - Use only first 80 lines to avoid catching article references
var allText = lines.join("\n"),
firstChunk = lines.slice(0,80).join('\n'),
doi = Zotero.Utilities.cleanDOI(firstChunk),
promise;
Zotero.debug(allText);
if(!doi) {
// Look for a JSTOR stable URL, which can be converted to a DOI by prepending 10.2307
doi = firstChunk.match(/www.\jstor\.org\/stable\/(\S+)/i);
if(doi) {
doi = Zotero.Utilities.cleanDOI(
doi[1].indexOf('10.') == 0 ? doi[1] : '10.2307/' + doi[1]
);
}
}
var lines = yield _extractText(file, MAX_PAGES);
// Look for DOI - Use only first 80 lines to avoid catching article references
var allText = lines.join("\n"),
firstChunk = lines.slice(0,80).join('\n'),
doi = Zotero.Utilities.cleanDOI(firstChunk),
promise;
Zotero.debug(allText);
if(!doi) {
// Look for a JSTOR stable URL, which can be converted to a DOI by prepending 10.2307
doi = firstChunk.match(/www.\jstor\.org\/stable\/(\S+)/i);
if(doi) {
// Look up DOI
Zotero.debug("RecognizePDF: Found DOI: "+doi);
var translateDOI = new Zotero.Translate.Search();
translateDOI.setTranslator("11645bd1-0420-45c1-badb-53fb41eeb753");
translateDOI.setSearch({"itemType":"journalArticle", "DOI":doi});
promise = _promiseTranslate(translateDOI, libraryID);
} else {
promise = Zotero.Promise.reject("No DOI found in text");
doi = Zotero.Utilities.cleanDOI(
doi[1].indexOf('10.') == 0 ? doi[1] : '10.2307/' + doi[1]
);
}
}
var newItem;
if (doi) {
// Look up DOI
Zotero.debug("RecognizePDF: Found DOI: "+doi);
return promise
// Look for ISBNs if no DOI
.catch(function(error) {
Zotero.debug("RecognizePDF: " + error);
var isbns = _findISBNs(allText);
if (isbns.length) {
Zotero.debug("RecognizePDF: Found ISBNs: " + isbns);
var translate = new Zotero.Translate.Search();
translate.setSearch({"itemType":"book", "ISBN":isbns[0]});
return _promiseTranslate(translate, libraryID);
} else {
return Zotero.Promise.reject("No ISBN found in text.");
}
})
var translateDOI = new Zotero.Translate.Search();
translateDOI.setTranslator("11645bd1-0420-45c1-badb-53fb41eeb753");
translateDOI.setSearch({"itemType":"journalArticle", "DOI":doi});
try {
newItem = yield _promiseTranslate(translateDOI, libraryID);
return newItem;
}
catch (e) {
Zotero.debug("RecognizePDF: " + e);
}
}
else {
Zotero.debug("RecognizePDF: No DOI found in text");
}
// Look for ISBNs if no DOI
var isbns = _findISBNs(allText);
if (isbns.length) {
Zotero.debug("RecognizePDF: Found ISBNs: " + isbns);
var translate = new Zotero.Translate.Search();
translate.setSearch({"itemType":"book", "ISBN":isbns[0]});
try {
newItem = yield _promiseTranslate(translate, libraryID);
return newItem;
}
catch (e) {
// If no DOI or ISBN, query Google Scholar
.catch(function(error) {
Zotero.debug("RecognizePDF: " + error);
return me.GSFullTextSearch.findItem(lines, libraryID, stopCheckCallback);
});
});
}
Zotero.debug("RecognizePDF: " + e);
}
}
else {
Zotero.debug("RecognizePDF: No ISBN found in text");
}
return this.GSFullTextSearch.findItem(lines, libraryID, stopCheckCallback);
});
/**
* Get text from a PDF
@ -183,8 +191,7 @@ var Zotero_RecognizePDF = new function() {
* Attach appropriate handlers to a Zotero.Translate instance and begin translation
* @return {Promise}
*/
function _promiseTranslate(translate, libraryID) {
var deferred = Zotero.Promise.defer();
var _promiseTranslate = Zotero.Promise.coroutine(function* (translate, libraryID) {
translate.setHandler("select", function(translate, items, callback) {
for(var i in items) {
var obj = {};
@ -193,7 +200,7 @@ var Zotero_RecognizePDF = new function() {
return;
}
});
translate.setHandler("done", function(translate, success) {
/*translate.setHandler("done", function(translate, success) {
if(success && translate.newItems.length) {
deferred.resolve(translate.newItems[0]);
} else {
@ -202,10 +209,15 @@ var Zotero_RecognizePDF = new function() {
: "Could not find a translator for given search item"
);
}
});*/
var newItems = yield translate.translate({
libraryID
});
translate.translate(libraryID, false);
return deferred.promise;
}
if (newItems.length) {
return newItems[0];
}
throw new Error("No items found");
});
/**
* Search ISBNs in text
@ -333,14 +345,14 @@ var Zotero_RecognizePDF = new function() {
this._progressWindow.addEventListener("keypress", this._keypressCancelHandler);
this._progressWindow.addEventListener("close", this._cancelHandler, false);
Zotero_RecognizePDF.GSFullTextSearch.resetQueryLimit();
this._recognizeItem();
return this._recognizeItem();
},
/**
* Shifts an item off of this._items and recognizes it, then calls itself again if there are more
* @private
*/
"_recognizeItem": function() {
"_recognizeItem": Zotero.Promise.coroutine(function* () {
const SUCCESS_IMAGE = "chrome://zotero/skin/tick.png";
const FAILURE_IMAGE = "chrome://zotero/skin/cross.png";
const LOADING_IMAGE = "chrome://global/skin/icons/loading_16.png";
@ -367,51 +379,66 @@ var Zotero_RecognizePDF = new function() {
var file = item.getFile(), me = this;
(file
? Zotero_RecognizePDF.recognize(file, item.libraryID, function() { return me._stopped; })
: Q.reject(new Zotero.Exception.Alert("recognizePDF.fileNotFound")))
.then(function(newItem) {
// If already stopped, delete
if(me._stopped) {
Zotero.Items.erase(newItem.id);
throw new Zotero.Exception.Alert('recognizePDF.stopped');
try {
if (file) {
let newItem = yield Zotero_RecognizePDF.recognize(
file,
item.libraryID,
() => this._stopped
);
// If already stopped, delete
if (this._stopped) {
yield Zotero.Items.eraseTx(newItem.id);
throw new Zotero.Exception.Alert('recognizePDF.stopped');
}
// put new item in same collections as the old one
let itemCollections = item.getCollections();
for (let i = 0; i < itemCollections.length; i++) {
let collection = yield Zotero.Collections.getAsync(itemCollections[i]);
yield collection.addItem(newItem.id);
}
// put old item as a child of the new item
item.parentID = newItem.id;
yield item.saveTx();
itemTitle.setAttribute("label", newItem.getField("title"));
itemIcon.setAttribute("src", SUCCESS_IMAGE);
this._rowIDs[rowNumber] = newItem.id;
return this._recognizeItem();
}
// put new item in same collections as the old one
var itemCollections = item.getCollections();
for(var j=0; j<itemCollections.length; j++) {
var collection = Zotero.Collections.get(itemCollections[j]);
collection.addItem(newItem.id);
else {
throw new Zotero.Exception.Alert("recognizePDF.fileNotFound");
}
// put old item as a child of the new item
item.parentID = newItem.id;
item.save();
itemTitle.setAttribute("label", newItem.getField("title"));
itemIcon.setAttribute("src", SUCCESS_IMAGE);
me._rowIDs[rowNumber] = newItem.id;
me._recognizeItem();
})
.catch(function(error) {
Zotero.debug(error);
Zotero.logError(error);
}
catch (e) {
Zotero.logError(e);
itemTitle.setAttribute("label", error instanceof Zotero.Exception.Alert ? error.message : Zotero.getString("recognizePDF.error"));
itemTitle.setAttribute(
"label",
e instanceof Zotero.Exception.Alert
? e.message
: Zotero.getString("recognizePDF.error")
);
itemIcon.setAttribute("src", FAILURE_IMAGE);
// Don't show "completed" label if stopped on last item
if(me._stopped && !me._items.length) {
me._done(true);
if (this._stopped && !this._items.length) {
this._done(true);
} else {
me._recognizeItem();
return this._recognizeItem();
}
}).finally(function() {
}
finally {
// scroll to this item
me._progressWindow.document.getElementById("tree").treeBoxObject.scrollToRow(Math.max(0, me._itemTotal-me._items.length-4));
}).done();
},
this._progressWindow.document.getElementById("tree").treeBoxObject.scrollToRow(
Math.max(0, this._itemTotal - this._items.length - 4)
);
}
}),
/**
* Cleans up after items are recognized, disabling the cancel button and
@ -523,10 +550,10 @@ var Zotero_RecognizePDF = new function() {
this.findItem = function(lines, libraryID, stopCheckCallback) {
if(!inProgress && queryLimitReached) {
// There's no queue, so we can reject immediately
return Q.reject(new Zotero.Exception.Alert("recognizePDF.limit"));
return Zotero.Promise.reject(new Zotero.Exception.Alert("recognizePDF.limit"));
}
var deferred = Q.defer();
var deferred = Zotero.Promise.defer();
queue.push({
deferred: deferred,
lines: lines,
@ -571,7 +598,7 @@ var Zotero_RecognizePDF = new function() {
}
item.deferred.resolve(
Q.try(getGoodLines, item.lines)
Zotero.Promise.try(getGoodLines, item.lines)
.then(function(lines) {
return queryGoogle(lines, item.libraryID, 3); // Try querying 3 times
})
@ -634,7 +661,7 @@ var Zotero_RecognizePDF = new function() {
* @param {Integer} tries Number of queries to attempt before giving up
* @return {Promise} A promise resolved when PDF metadata has been retrieved
*/
function queryGoogle(goodLines, libraryID, tries) {
var queryGoogle = Zotero.Promise.coroutine(function* (goodLines, libraryID, tries) {
if(tries <= 0) throw new Zotero.Exception.Alert("recognizePDF.noMatches");
// Take the relevant parts of some lines (exclude hyphenated word)
@ -671,20 +698,21 @@ var Zotero_RecognizePDF = new function() {
delay = GOOGLE_SCHOLAR_QUERY_DELAY - (Date.now() - Zotero.HTTP.lastGoogleScholarQueryTime);
// Delay
return (delay > 0 ? Q.delay(delay) : Q())
.then(function() {
Zotero.HTTP.lastGoogleScholarQueryTime = Date.now();
return Zotero.HTTP.promise("GET", url, {"responseType":"document"})
})
.then(function(xmlhttp) {
return _checkCaptchaOK(xmlhttp, 3);
},
function(e) {
return _checkCaptchaError(e, 3);
})
.then(function(xmlhttp) {
var doc = xmlhttp.response,
deferred = Q.defer(),
if (delay > 0) {
yield Zotero.Promise.delay(delay);
}
Zotero.HTTP.lastGoogleScholarQueryTime = Date.now();
try {
try {
let xmlhttp = yield Zotero.HTTP.promise("GET", url, {"responseType":"document"});
}
catch (e) {
yield _checkCaptchaError(e, 3);
}
xmlhttp = yield _checkCaptchaOK(xmlhttp, 3);
let doc = xmlhttp.response,
deferred = Zotero.Promise.defer(),
translate = new Zotero.Translate.Web();
translate.setTranslator("57a00950-f0d1-4b41-b6ba-44ff0fc30289");
@ -693,7 +721,7 @@ var Zotero_RecognizePDF = new function() {
if(detected.length) {
deferred.resolve(_promiseTranslate(translate, libraryID));
} else {
deferred.resolve(Q.try(function() {
deferred.resolve(Zotero.Promise.try(function() {
return queryGoogle(goodLines, libraryID, tries-1);
}));
}
@ -701,14 +729,14 @@ var Zotero_RecognizePDF = new function() {
translate.getTranslators();
return deferred.promise;
})
.catch(function(e) {
}
catch (e) {
if(e.name == "recognizePDF.limit") {
queryLimitReached = true;
}
throw e;
});
}
}
});
/**
* Check for CAPTCHA on a page with HTTP 200 status