DOI detection in web pages

Greetings,

I am wondering what mechanism causes Zotero to detect the DOI in the following linked pages [1a,b], but *not* recognize the DOI on the second linked page[2]. Can someone help be better understand what Zotero is doing? I'm trying to make web pages with automatic DOI detection in them, so I'm trying to better understand the fails and passes in various use cases.

[1a]: https://catalog.paradisec.org.au/collections/RB5/items/Kainji_Kamuku_wordlists
[1b]: https://catalog.paradisec.org.au/collections/RB5
[2]: https://www.sil.org/resources/archives/52727
  • It detects the DOIs for [2]. It just detects Embedded Metadata too, and that takes priority.

    You can see the available translators by right-clicking on the save button and looking in the Save to Zotero submenu.

    To get multiple results on a generic page (i.e., without a matching site-specific or catalog-specific translator), you'd currently need to use (if I recall) COinS or unAPI. We'll hopefully be adding support for some alternative methods later this year.
  • Thank you that is really helpful.
  • To get multiple results on a generic page (i.e., without a matching site-specific or catalog-specific translator), you'd currently need to use (if I recall) COinS or unAPI.
    Yes, that's right. If you're looking to build a forward-compatible page you want to look at also implementing JSON-LD and schema.org, which are the alternative methods dstillman is talking about and will also be most likely to be supported by other tools.
  • @adamsmith Thank you for the pointer to schema.org I am aware of that project and involved with that community. However, how Zotero interprets that information is still undefined. In some cases it appears that there is not 100% defined overlap between Zotero/CSL item types and item types within schema. So, until there is discussion within the Zotero Dev community on how to support Schema.org it seems that it might be pre-mature to implement it. But it is certainly in view for future work. And I would like to stay in the loop with the Zotero dev team on schema.org implementation choices.
  • So, until there is discussion within the Zotero Dev community on how to support Schema.org it seems that it might be pre-mature to implement it.
    I'd disagree with that. The goal of metadata standards is specifically that you don't need to design with a specific application in mind. An increasing number of tools are reading JSON-LD/schema.org and Zotero is firmly intending to, so there's really no reason not to implement it now.
Sign In or Register to comment.