How do Zotero get reference data from these URL:s

beogl
Could someone please tell me how Zotero gets the reference data from URL:s like

http://www.springerlink.com/content/qh290kr305158620/

I would also be glad to know how a link on pubmed is handled, this for example

http://www.ncbi.nlm.nih.gov/pubmed/17501969
  • adamsmith
    I'm not clear what you mean by "how Zotero gets the reference".
    For me getting the citations using the URL-bar item (i.e. the standard method) works well.
  • beogl
    Hi Adam,

    Sorry, I was not clear enough. I meant what Zotero looks for on those pages to get the information.
  • ajlyon
    That depends on the translator. For some sites, we download an RIS or other bibliographic data file. For others, we try to extract the info from the page content or embedded metadata. If you're curious, the best and only way to really see what's going on is to look at specific translators in the translators directory of your Zotero data directory.
  • adamsmith
    For SpringerLink Zotero uses the RIS export data from Springer
    For Pubmed, if I understand the translator correctly, it uses the xml provided by pubmed eutils and cleans it up a little for Zotero purposes.
    You can look at the translator code for each translator (which is in the translator folder in the Zotero data folder http://www.zotero.org/support/zotero_data#locating_your_zotero_library )
    with any text editor.
  • beogl
    Thanks, I have found it. (But I can't figure out how it works. Just replacing the link with "/export.mpx?code=CODE" does not work for me.)
  • ajlyon
    In short, it's not always a straightforward process, but what adamsmith and I have written is about all I can tell you without diving into the code and explicating the whole thing.

    Is there a specific issue you are trying to address here?
  • beogl
    Thanks ajlyon, I understand.

    I am trying to get a quick way to get a bibliography without starting FF which I am the moment is not using very much due to some problem I have had.

    Looking at the Zotero site I see that a standalone version of Zotero is on its way. Sounds very nice. It would be good if that could be used also from the command line and if it could just generate a bibliography and put it on the clipboard (without storing it in the database or just temporary storing there - which would avoid cluttering the db).
  • adamsmith
    that's very unlikely to ever happen - db "cluttering" seems an odd concern for a reference manager storing references.
    Many sites actually allow you to directly export a citation in a standard style. You could also look at tools like citeUlike and Mendeley - the former very light, web-based and less powerful, the latter pretty similar to Zotero, but it's standalone and import works through a bookmarklet.
  • beogl
    Sorry, with cluttering I just meant getting a second copy in the wrong place in the db. And sometimes have just downloaded a report as pdf and just need a bibliography. (But perhaps this means I just should learn more about Zotero.)

    Thanks for the pointers. I really realize I have to learn more about Zotero.

    Another way to solve my problem would perhaps be to from outside of FF/Zotero tell Zotero to open a go to a stored document that is identified by for example DOI. Is that possible?

    (And that makes me wonder if several stored documents in Zotero could have the same DOI.)
  • adamsmith
    DOIs are unique by design.
    For accessing Zotero database outside Zotero, see the gnotero plugin (not available for Mac OS)
    http://www.cogsci.nl/software/gnotero
  • beogl
    Yes, of course DOIs are unique, but only as document identifiers on the net. Is it a unique identifier in Zotero too?

    Thanks, gnotero looks useful, but unfortunately it does not seem to work well yet on MS Windows. I will try to file a bug report for that. (I could not access Zotero from inside FF if it was accessed by gnotero.)
  • adamsmith
    since you can edit the DOI field it's of course possible to have two items with the same DOI in Zotero - I just don't see a case in which that would happen (unless it's the same document).
    Zotero does have internal unique identifiers - there are a whole range of threads on that on the forum.
  • beogl
    I just happened to add the same page two times to Zotero - and got two entries in the Zotero db. The only difference I can see right now is that the dates for "Accessed", "Date Added" and "Modified" are different.

    Should not that be just one entry in the Zotero db? (This is with Zotero 2.0.9)
Sign In or Register to comment.