Indexing of same files gotten worse

simondedman
I setup Zotero & manually added 70 pdfs of which it indexed ~35
After playing around with my setup for cloud sync reasons, I deleted my whole library & restarted (since it was easier & i didn't have any notes or anything), readded all 70 pdfs, asked them all to index, only 10 have done so successfully. I've not changed anything with the PDF files. Any first thoughts why this might be?
Thanks!
  • simondedman
    For some of them I'm not sure if I changed the file names between the switch, from a variety of styles to a semi-uniform:
    Authorsurname1,Authorsurname2.YYYY.Paper name.pdf

    I can't imagine the lack of spaces makes a difference, right? What info does the metadata retrieval process use?
  • dstillman
    Generate a Debug ID for an indexing attempt (clicking the arrows in the right-hand pane next to "Indexed: No") for one of the unindexed PDFs.
  • simondedman
    Done. The Debug ID is D1253654479. Sincere apologies, however, as I'm using the wrong terminology for what is (i believe) two different things: many of them say that they ARE indexed (and indeed it seems I can manually refresh index them successfully), rather they are no longer retrieving metadata successfully. The debug ID is me successfully manually refresh indexing a few pdfs, then requesting metadata retrieval for ~ 30 which all fail.
    Again, apologies for calling an apple and orange in the first post 7 thread title.
  • dstillman
    Ah, OK.

    So the short answer is that you're hitting Google's request limits for lookup attempts using Google Scholar. If you wait a while and try again (and do fewer at once), you might have better luck.

    The longer answer is that the only reason it's using Google Scholar is that all of the PDFs from this journal have a ligature in the DOI on the title page: "fi" instead of "fi". Zotero sends the DOI as is, and CrossRef has no record of the DOI, since the ligature is a totally different character. This is kind of just a bug in the PDF, but I guess we could correct for it. I'll start a separate thread about this.
Sign In or Register to comment.