Upgrade StorageExtension doesn't recognize books with Rero ils catalogue
Uclouvain recently change its catalogue to rero ils and my Zotero extension doesn't recognize books or articles as such when i save them in my library. Instead it's saving them as the web page of the catalogue with a snapshot of it. It's a big problem because they're all named "RERO ils : Catalogue" and they're just unusable. I searched on the forum but couldn't find anything related. I'll put an URL to show it as an exemple because it may be a problem only for me idk. https://ils.bib.uclouvain.be/uclouvain/documents/68705
We are currently working on an expose RIS functionality.
The feature will be available in june.
It will be possible to export search results or single document in RIS format and import into Zotero.
It will be also possible to use the Zotero plugin.
I've just one technical question for tech people on this channel.
They are many ways for doing that: unAPI, COinS, RIS, ...
What's the best practice to expose data for Zotero plugin ?
@Lauren-d:
Zotero has two ways of getting metadata from sites:
1. Generic translators, which use structured metadata like unAPI, COins, or Highwire. This is good for sites like journal databases, where there are likely only to be a few item types (journal articles, reports, preprints) and data will be relatively clean and structured. You *could* expose RERO-ILS metadata that way, but it would really fall short with some of the less common item types that library catalogs tend to have - there's no way to represent a map or a piece of artwork with Highwire, for instance. It also doesn't allow for multiples (clicking the Add to Zotero button on a search results page to add multiple items at once).
2. Site- or app-specific translators. These target a specific system and contain code to translate its most accurate catalog data representation directly into a Zotero item. On some sites that's a relatively structured format like MARC that has to be parsed in a site-specific way (e.g. the BiblioCommons translator); on others it's a completely custom format like Primo Normalized XML. Other sites don't give us any structured metadata at all and we have to scrape from the page. In your case, it looks like you're already exposing structured metadata as JSON - that's great for us and we shouldn't need anything more to get a solid translator working. Adding RIS will definitely be helpful for users, but Zotero isn't going to use it if the JSON metadata representation is stable - there's no reason to use data that has to go through two potentially lossy translation steps (RERO-ILS metadata -> RIS -> Zotero metadata) when we can just use one (RERO-ILS metadata -> Zotero metadata). I'll just write a translator to do that conversion and you shouldn't have to make any changes on your end.
Many thanks for your advices and for the new translator.