Change to RIS Export Field Mappings?

mjt772

For years, I've successfully exported RIS files from Zotero and used them in an application that's heavily reliant on standard RIS specs. Recently, though, I noticed that Zotero is mapping fields differently. Title is now mapping to TI (formerly T1), Abstract is now mapping to AB (formerly N2), and Author is now mapping to AU (formerly A1). All of those changes are fine and in accordance with RIS specs.

However, there's one change that isn't in accordance with the specs: The Date field is getting parsed and mapped to two separate fields, so that only the year (YYYY) gets mapped to PY. The full YYYY/MM/DD/etc info now gets mapped to DA (a field which doesn't appear in the RIS specs, afaik).

Can anyone please shed some light on why this change occurred, and whether it can be fixed/undone? This is causing problems. Thanks!

adamsmith

The RIS specs as published by RefMan have relatively recently undergone a long-overdue update, that for the most part reflects already common practice in programs like Endnote. Zotero's RIS translator was update to reflect these changes about one month ago.

The specs are here:
http://www.refman.com/support/risformat_intro.asp
(there is a link to a Zip file with a PDF and an Excel spreadsheet at the bottom of the page).

If you check these specs you'll see that the current PY and DA behavior is exactly according to specs.
If you know some javascript it wouldn't be very hard for you to revert that change in the translator and save it as a separate, export only translator with a different translator ID, so that it wouldn't get overwritten by updates in the future, but I don't really see what we can do differently from the Zotero side here.

mjt772

Ah, okay, thank you. I had been looking at the specs listed elsewhere on that site (http://www.refman.com/support/risformat_tags_03.asp), which are obviously not up-to-date. I'll make the necessary changes to my code, then. Thank you for the quick reply!

adamsmith

out of curiosity - what is this for more specifically? Wouldn't it make sense to use a better specified standard like MODS?

mjt772

I'm working with the Conflict Information Consortium at CU (in a joint project with S-CAR at GMU, actually), exporting RIS files from Zotero and then importing them into a Drupal 6 installation using the Biblio module (something that, judging from the module forums, a lot of other people are doing as well). It seems, though, that the Biblio module maintainers have dropped the ball here. The module doesn't yet conform to the new RIS specs, and it notably doesn't process the DA field at all.

A more robust export format would actually work against us in this case, as we need very streamlined formatting to enable proofreading and editing of the exported file. (We also have to work with a format that is common to both Zotero and Biblio, and that limits us to BibTex and RIS.)

I'm currently opening a ticket with the Biblio module coders, hoping that they'll upgrade their RIS functionality. In the meantime, though, I'll probably just make changes on the fly on my end (I'm the gatekeeper for RIS files that we get from numerous contributors), in order to "regress" the field designations to versions that Biblio can understand.

adamsmith

Makes sense - the big problem with RIS is that it really doesn't have a properly maintained "standard" - that RefMan can just put out with a document that changes half of what's on their web-page without any announcement, while keeping that - now outdated - information online is really quite telling. I wouldn't even really say that Biblio "dropped the ball" here - they're just victims of the terrible state of this standard.

So wherever possible I'd recommend a properly maintained standard (such as MODS, which is painstakenly maintained and documented by LoC), but I see why that doesn't work in your case.

sjgold

I did a test on exporting a MODS file from my current Zotero for Firefox/Windows Vista library (metadata, no files or notes) but MIcrosoft XML viewer complained about invalid hex characters. Is there a better XML viewer out there? Our research project wants to keep the base data (for an (urban) ethnographic artefacts database)in Zotero but we will need to validate the export file ourselves before it disappears off into the college server.

Thanks, Stephen Goldborough for Jane Fernandez at avondale edu au.

adamsmith

Stephen,
please start a new thread - we prefer to keep things one topic at a time to maintain some sense of order. But if you want to validate XML I suggest either using an online validator:
http://validator.nu/
or using Oxygen or emacs in nxml mode. Any follow-ups, please start a new thread.