[MLZ] Bulk import/edit collection with multilingual fields

Hi there,

I am finally trying to get my Zotero collection into MLZ-shape. This means adding multilingual fields to the collection and I am wondering what the most efficient way to do this would be.

Since the Zotero interface for editing items is rather clunky, I'd prefer to export the collection, do the changes and then re-import. However, after a few experiments it seems that the built-in export formats all ignore the MLZ fields. So what is the recommended way for this kind of task?

I looked around for solution, but did not find any solution so far.

Any help appreciated,

  • Frank would have to confirm, but I don't believe MLZ has a full export format so there really is not alternative to working with the interface.
  • Bibliontology RDF is currently the only format that supports multilingual export and import. Scripted editing of the RDF objects outside of Zotero would be daunting, I think.

    You may be able to do something with Firefox Scratchpad (working with a copy of the library in an alternative Firefox profile for safety); Scratchpad can be used to access Zotero functions programmatically, which may be useful, depending on what your source for multilingual variants is like.
  • Thanks for your suggestion, I will investigate.

    Meanwhile, trying to explore other possibilities (and establish the path of least resistance), I was wondering if going through the API would be a possible venue. I would expect that I have access to all information here, no?
  • Scratchpad would be going through the client-side javascript API and that would likely be the better option anyway.
    Since the server API wasn't meant to be used with MLZ, you'd have to see how well that works and, especially, where/if the multilingual fields are returned.
  • Ah, I see. Thank you. It seems I have to swallow the bullet then...
  • Re the API, you could go that route also. The Zotero servers do not support multilingual directly, but if a correctly coded data block is placed in the Extra field, it will sync down. There are strict rules for the construction of the block (if it is not recognized, it will sync down as a string, instead of populating the desired fields). There is no formal documentation on the requirements yet; if you decide to look into it, let me know.
  • Well, in that case, I'd like to give it a try. I understand that I have to go through the Extra field and would appreciate instructions on how to make sure it gets understood. As far as I understand, the Extra field is quite a mixed bag of things and the way MLZ expects its stuff is not entirely clear to me.

    I also looked at the output of CSL-JSON export, which seems to contain the MLZ stuff as well. Would it be possible to massage that file and feed it back to ML-Zotero?
  • There are some support functions in MLZ that can be used or adapted to handle it. I'll post a description tomorrow - have to run for an all-day obligation today, and will be tied down through the morning tomorrow.
Sign In or Register to comment.