Changing styles to include new fields from Multilingual Zotero

sdspieg
I am trying to make Word show some newly created fields I generated for an English-language translation (NOT transliteration) of some foreign titles in multilingual Zotero. I have downloaded the Chicago-fullnote csl-file and opened it in Ultraedit and the logic of that xml-file seems easy to understand.
But so could I just copy-paste the <macro name="title"> into a <macro name="title_en">, and then do a search and replace on "title" in that node and replace it there with title "xml:lang="en"" (that being what I see in the rdf-file as being the new tag for the English translation of the title). Would that work? Or is that too easy?

Thanks for your assistance.

-Stephan
  • ajlyon
    Reminder: The multilingual branch is not even beta code-- it's an experiment that may not have an upgrade path to the final form it takes.

    But see the citeproc-js manual for information on how to access the various forms from CSL: http://gsl-nagoya-u.net/http/pub/citeproc-doc.html#id36
  • bdarcus
    But see the citeproc-js manual for information on how to access the various forms from CSL: http://gsl-nagoya-u.net/http/pub/citeproc-doc.html#id36.
    And this itself should be considered "experiemental" from the standpoint of CSL as well (it's not in the CSL spec, so won't work in other implementations).

    But I think the key thing to resolve on both the CSL API and the Zotero end is the data representation. Frank's solution is a clever piggy-backing on existing infrastructure, but I think there are going to need to be some fundamental model changes to support this properly. Namely, I think we're going to need to move towards a system where field values can be tagged with a language, and can be repeated (in those other languages). This will inevitably take a fair bit of time to work out.

    I will add that there' s at least one piece of this puzzle that already supports this: BIBO RDF (since RDF values can have optional language tags already).
  • ajlyon
    Bruce: Within Zotero, the underlying structure of the multilingual branch's data model is pretty much what you describe -- the piggy-backing (concatenating language-tagged values in a single string) is happening just to accommodate encapsulate the data for syncing.

    The CSL situation will of course also be subject to change, but some of it could be covered by the recent changes to locale identifiers; see http://www.alvestrand.no/pipermail/ietf-languages/2010-December/010811.html and subsequent posts in that thread.

    This discussion should probably move to one or more of the zotero-dev, xbiblio or citeproc-js mailing lists, but I'm glad it's happening-- we need to work this out if this is ever going to make it into the spec and into a stable release of Zotero.
  • fbennett
    The multilingual version will definitely move to a cleaner data model for input. It will take awhile to get the processor and the several relevant layers of Zotero adapted to it, but it's coming. I'll certainly float proposals on the CSL list and on the forums here before making further moves in the code, but change is in the cards.
  • sdspieg
    As you folks are working on this - is there anybody out there who would be able to help us to generate a Chicago-compliant multilingual style sheet that could be used to generate both footnotes and a full bibliography? If so, we should be able to reimburse you for your time (within reason of course :)), and to make the result freely available to everybody as an (early) workable solution for multilingual citations. If you're interested, please send me a pm. Thanks!

    -Stephan
Sign In or Register to comment.