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

closes #272, problems with Library of Congress ingest

Browse source
This commit is contained in:
Simon Kornblith 2006-09-05 03:06:22 +00:00
parent 05f56aa489
commit cec35d7566
1 changed files with 20 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

23
scrapers.sql
View file

@ -1,4 +1,4 @@
-- 78 -- 79
-- Set the following timestamp to the most recent scraper update date -- Set the following timestamp to the most recent scraper update date
REPLACE INTO "version" VALUES ('repository', STRFTIME('%s', '2006-08-31 22:44:00')); REPLACE INTO "version" VALUES ('repository', STRFTIME('%s', '2006-08-31 22:44:00'));
@ -5722,7 +5722,22 @@ record.prototype.importBinary = function(record) {
var baseAddress = parseInt(this.leader.substr(12, 5), 10); var baseAddress = parseInt(this.leader.substr(12, 5), 10);
// get record data // get record data
this.content = record.substr(baseAddress); var contentTmp = record.substr(baseAddress);
// MARC wants one-byte characters, so when we have multi-byte UTF-8
// sequences, add null characters so that the directory shows up right. we
// can strip the nulls later.
this.content = "";
for(i=0; i<contentTmp.length; i++) {
this.content += contentTmp[i];
if(contentTmp.charCodeAt(i) > 0x00FFFF) {
this.content += "\x00\x00\x00";
} else if(contentTmp.charCodeAt(i) > 0x0007FF) {
this.content += "\x00\x00";
} else if(contentTmp.charCodeAt(i) > 0x00007F) {
this.content += "\x00";
}
}
// read directory // read directory
for(var i=0; i<directory.length; i+=12) { for(var i=0; i<directory.length; i+=12) {
@ -5775,10 +5790,10 @@ record.prototype.getField = function(field) {
for(var i in this.directory[field]) { for(var i in this.directory[field]) {
var location = this.directory[field][i]; var location = this.directory[field][i];
// add to array // add to array, replacing null characters
fields.push([this.content.substr(location[0], this.indicatorLength), fields.push([this.content.substr(location[0], this.indicatorLength),
this.content.substr(location[0]+this.indicatorLength, this.content.substr(location[0]+this.indicatorLength,
location[1]-this.indicatorLength-1)]); location[1]-this.indicatorLength-1).replace(/\x00/g, "")]);
} }
return fields; return fields;