Automatic saving of PDFs

dstillman

This discussion was created from comments split from: How to retrieve full text pdfs from the web for all my references at once automatically.

gerhard.bissels

I agree with pbouts: especially for projects for which larger numbers of papers have to be collected and screened, automatic harvesting would be an essential. I am, however, not sure what Zotero is already supposed to do: you can select in the preferences, general: 'Automatically attach associated pdfs and other files...'. Does this not mean that, when I import a hundred records, I should get a hundred pdfs as well (as long as the university subscribes to all cited journals, or they are Open Access anyway)?

bwiernik

If the site you are importing from provides the PDF and you have access to it, Zotero will generally grab the PDF on import. If that doesn’t happen, feel free to start a new thread here with a specific URL. Note though that some sites don’t provide PDFs (e.g., PubMed, often Google Scholar)—Zotero won’t be able to get PDFs for these. On JSTOR, you need to first manually download one pdf so that you can click their terms of service agreement, then Zotero will be able to grab PDFs.

If you are trying to download hundreds of records at once (e.g., for a systematic review), databases may lock you out. For tips on importing large numbers of records, See “Large-Scale Imports from Databases” here: https://www.zotero.org/support/getting_stuff_into_your_library

gerhard.bissels

I just tried downloading at least a single paper automatically from a few databases. In each case I double-checked that my university had access, and the paper could be downloaded manually without glitches.
- Ovid Medline: Zotero failed to download the pdf
- Ovid Embase: same
- EBSCO CINAHL: same
- embase.com: Zotero doesn't even recognise this is a bibliographic database (no yellow folder in browser bar)
- Scopus: same
In a medical context a reference management tool needs to be able to download large numbers of records, and fetch the related full texts at the same time. But Zotero currently fails to even get hold of the pdf of a single paper. Or am I going wrong somewhere?

dstillman

Can you provide example URLs that don't work?

adamsmith

Most of these don't have PDFs, though, i.e. we wouldn't expect PDFs for Scopus and Ovid -- these are all in the PubMed category mentioned by bwiernik.

EBSCO CINAHL I don't know (we should get PDFs where EBSCO offers them, but that's not the case for all of EBSCO) and embase outside of Ovid I know we're not supporting at all.

I don't think there's any doubt that the originally requested feature (as implemented by Joscha in his PR) would be immensely useful for exactly this reason.

dstillman

(Moving this to a separate thread, since Gerhard's question is about existing save functionality, not the feature request in the other thread.)

Most of these don't have PDFs, though

Gerhard says there are PDFs accessible on these pages, though. But we can look at example URLs.

I don't think there's any doubt that the originally requested feature (as implemented by Joscha in his PR) would be immensely useful for exactly this reason.

(I believe the feature as currently implemented by Joscha only tries to download from original URL, though, so it wouldn't help if there's no PDF available. But as Joscha says, it can be extended to work with other services.)

adamsmith

He doesn't say that there are PDFs, just that his university has access. Functionally hard to tell apart for a user, but very different for Zotero:

E.g. Scopus and Medline (like PubMed) have "Full Text" or "Full text at publisher" buttons that lead to the publisher copy, but that doesn't help Zotero. Those databases themselves definitely don't have PDFs.