Zotero Connector sometimes does not save PDF for certain journals

edited September 6, 2021
For certain journals, Zotero Connector sometimes does not find or extract the PDF of article being saved. There are two cases of this issues. First case is when the journal displays the article in their own custom PDF viewer, see example [1,2]. In this case, PDF is not saved and only the metadata/snapshot is saved. Second case is when viewing the main article page and not the PDF. In this case, Zotero sometimes automatically finds the PDF and saves it, see example [1]. However, other times, Zotero does not find the PDF, see example [3]. Issue test on Zotero v5.0.96.3, Firefox, and Chrome browser.

Example articles:
[1] https://www.sciencedirect.com/science/article/pii/S2214860421003857
[2] https://www.science.org/doi/epdf/10.1126/sciadv.abh0273
[3] https://www.science.org/doi/10.1126/sciadv.abh0273

Update 1:

I reported Science Advances as an example journal with the issue. However I noticed articles in Science also have same issue.
  • edited September 7, 2021
    https://www.sciencedirect.com/science/article/pii/S2214860421003857

    If I go to the main article page for the example article link above, and then click on Zotero Connector, Zotero finds all the metadata properly and also extracts the PDF. This method works as expected. I have submitted a debug report for Zotero Connector with id# D745900980.

    When I go to main article page for the example article link above, I see the "View PDF" button. After clicking the "View PDF" button, the article open in a new browser tab with a Elsevier's custom PDF viewer. I clicked on the Zotero Connector to save article. It is unable to properly grab the metadata and save it as a journal article item. It saves the item as a "Web Page". Furthermore Zotero is unable to extract and save the actual PDF file also. I have submitted the a debug report for Zotero Connector with id# D1434757173 .

    Tested with Firefox.

    Debug report id# D1798051126 for Zotero main app.


  • Yes, you should always save from the main article page when available. We're able to do a translator save from a direct PDF URL on some sites with clear URL schemes, but that's rare and not something you should count on. (More often it will save the PDF directly and then run metadata retrieval, but that's also less reliable than saving from the article page on a supported site.) You shouldn't expect anything to work in a site's embedded PDF viewer.

    @AbeJellinek can look into the problem with the article page on science.org for [3].
  • The Science.org error in [3] is on the site's end - try exporting a citation file (any format) using the site's interface and you should see that it fails there, too. There isn't much we can do in this situation; we need citation metadata from the site in order to import the item. Falling back to the generic Embedded Metadata translator (as it does now) is the correct behavior.
  • edited March 2, 2022
    I am reporting a related issue to the translator for Science. When saving an article on Science Advances using the Zotero Connector, the collected metadata about the article are incomplete. For example, the year field is missing. If I manually add the article using the DOI in Zotero, a more complete metadata is present. Example article: https://www.science.org/doi/10.1126/sciadv.abj8030
  • Cannot save all Quarterly Journal of Economics papers. Example:
    https://academic.oup.com/qje/article/137/2/679/6449024
  • @01baftb: Pushed a fix. (The issue was just that the RIS citation data for that article doesn't include the date, so we have to use a workaround.) Your Zotero Connector should auto-update within 24 hours, or you can update manually by clicking Update Translators in the Advanced pane of the Zotero Connector preferences.
  • @wanghongtao: That article saves fine for me. Try again, and if it's still not working, please submit a Connector Debug ID (not a Zotero Debug ID) for attempting to add one of the articles that's failing.
Sign In or Register to comment.