Digital Commons repository and Zotero

JoshuaLambertLib

My university uses Digital Commons (from Elsevier) as an institutional repository. I thought in the past that saving a citation via the Zotero browser add-on would pull in the PDF file when I was on a Digital Commons metadata page. But, I may be remembering wrong. Now, all I get is a snapshot of that current page rather than the PDF.

https://bearworks.missouristate.edu/theses/4055/

If I actually click on the "Download" button and am viewing the PDF in the browser, I can use the Zotero save icon when it shows as a PDF and it will download the PDF. The problem with that is that it doesn't download the metadata from the repository.

Has anybody saved the PDFs, via a connector, from a Digital Commons repository in the past? If so, is it still working for you? I'm thinking that this was a change in the Digital Commons platform but I can't remember for sure what the behavior used to be.

Thanks.

adamsmith

It's actually just picking up a poor default. Zotero has long talked about improved behavior on generic imports (like here) and I hope that still happen, but you can also manually improve this:
- right-click on the "Save to Zotero" button and then
- select "Save to Zotero (Embedded Metadata)" (the current default is COinS)

That gets you excellent metadata and the PDF attached.

(@AbeJellinek -- what's the status of a merged/improved embedded metadata translator? The current status of this isn't ideal; happy to brainstorm on GH if there are remaining conceptual issues)

JoshuaLambertLib

Thanks adamsmith. That does work the way I wanted. I have used Zotero for 15 years and I either have never right clicked the Zotero add-on icon in the browser bar or I long forgot about it. I'm glad to know it now.

If anyone thinks there used to be different behavior with Digital Commons pages, I would still be interested in your feedback.

adamsmith

I think they may have added COinS on the Digital Commons end, which would change default behavior

JoshuaLambertLib

That would make sense. Digging through the html source shows they are using quite the mix of metadata standards. They include: Open Graph, Schema.org, Elsevier's own standard, COinS, Twitter Cards, and old fashioned HTML meta tags. I guess Zotero would have to choose which one is the priority to pull from.

adamsmith

Zotero throws all of them except COinS together and then prefers COinS. There's a logic to that (sometimes all the other tags are attrocious), but it's not ideal, hence my note for Abe who is the Zotero dev who mainly maintains translators, above.