Error with Zotero.org translator

ajlyon · December 22, 2009

When I try to add the item http://www.zotero.org/groups/mobility_studies_russia/items/85841582 to my library in Zotero, I get a translator error.

Translate: Translation using zotero.org failed: 
thrown exception => item has no title
url => http://www.zotero.org/groups/mobility_studies_russia/items/85841582
downloadAssociatedFiles => true
automaticSnapshots => true

This is debug D1978128722.

I am able to save other items from group libraries on Zotero.org.

ajlyon · February 16, 2010

This error still persists.


(2)(+0000789): Translate: 'repository' field is now 'libraryCatalog'; please fix your code
(2)(+0000000): Translate: Translation using zotero.org failed: 
thrown exception => item has no title
url => http://www.zotero.org/groups/mobility_studies_russia/items/85841582
downloadAssociatedFiles => true
automaticSnapshots => true

I may try again to track this down and patch Zotero.org.js, but wouldn't it be better if Zotero.org used something nice like RDF on its item pages instead of relying on a translator? It seems a little odd that while Zotero urges sites to provide metadata, its own site uses a scraper-style translator.

dstillman · February 17, 2010

That page saves for me. Did you add a title that wasn't there before? It's possible that saving of untitled items will currently fail.

It seems a little odd that while Zotero urges sites to provide metadata, its own site uses a scraper-style translator.

Yes, but see my explanation here: http://forums.zotero.org/discussion/8114/automatically-embed-unapi-in-zoteroorg-public-libraries/#Item_2

ajlyon · February 17, 2010

That page saves for me. Did you add a title that wasn't there before? It's possible that saving of untitled items will currently fail.

That's not my library or item, and it still doesn't save for me (Firefox 3.6, Windows XP, Zotero dev builds). The error is as above, but just for kicks, you can have D137177094 to look at too. Maybe the issue is that the title is too long?

As for embedding nice data, I suppose it might be worth waiting since a scraper works for now. Even so, structured metadata doesn't exist only for the convenience of Zotero, and adding COinS would be pretty trivial-- the code already exists. Of course, this is a different feature request altogether.