Timeline on website searching PDFs?

Hi, I was hoping to be able to use the Zotero web interface to search my library through the attached PDFs, but I've seen several discussion threads from several years ago saying this is planned for the future. I was wondering if there is a timeline in place for this functionality? Thanks!
  • It's been possible to search PDF content on the website for many years. You have to have "Sync full-text content" enabled in the Sync pane of the Zotero preferences, and you have to select either "Full Text" or "Everything" from the drop-down menu in the website search field, depending on which version of the web library you're using.
  • Thank you for prompt response! I'm still having trouble. I already had the "Sync full-text content" enaabled, and I'm searching using "Full Text" on the website. I'm unable to find library entries when I search for keywords (within Full Text) that would appear in the attached PDFs. In other words, I can find the entries locally when the search entry is in the attached PDF, but I'm unable to do so through the web library. My apologies, am I missing a key step?
  • Can you provide the URL (or just the 8-character item key) when viewing an attachment in your web library that isn't matching a search?
  • Thank you dstillman
    here's the search: mdx/D71-78
    https://www.zotero.org/metabolon/items/q/mdx/D71-78/qmode/everything
    and the item key: 4L8SW969
  • Hi, sorry to be a pest, is there any indication we can solve this problem? If I can search attached PDFs through the web interface effectively, I'd like to upgrade to more storage. Thanks!
  • We're working on it. It turns out that full-text content synced in recent months hasn't been properly indexed. We've fixed the immediate problem, so new (or reindexed) content should now be searchable, but we need to process the backlog of items, which will take a few days.
  • Thank you dstillman, much appreciated. Would it be too much to ask for a comment letting me know the backlog is complete? Or can I reindex on my own to confirm?
  • I'll post here when the backlog is resolved. You can also reindex individual items from the right-hand pane and then sync, which should make them available for searching online.
  • Thanks dstillman. I wanted to comment on my end what I've seen. I went ahead and upgraded storage and then synched my entire local library, which now lives in the web interface, including all attachments. But the "everything" option still doesn't search the pdf attachments; I get the same result with "Title, Creator, Year" as I do with "everything" when if I do the same search locally, I get many many more when I search everything.
  • edited February 6, 2020
    A storage subscription has no effect on this. As long as you had already synced your library data, items added recently won't show up for full-text searches online until we process the backlog. If you reindex an individual attachment locally from the right-hand pane and sync, it should show up within Everything searches within a few seconds, because new full-text content is again being processed correctly.
  • Any updates on the backlog processing?
  • I went ahead and tested reindexing manually a few attachments locally rom the right-hand pane, and I still can't search the PDF attachments in my online library with "Everything" function. I appreciate your help @distillman, I would very much like to resolve this issue and use Zotero to its full functionality, now that I have the storage to do so. Thanks
  • Did you sync after performing the indexing?

    If so, can you right-click on an affected attachment, choose Show File, and provide the 8-character folder name?
  • Yes synced after indexing.
    UXEPUBY2
  • What's a word or phrase that should match that? That item appears to be indexed online.
  • "PD60 hearts" would only be found in the PDF. I can search locally and identify that document, but not on the web interface.
  • That should be matching that item. Just to confirm, that's in Everything mode on the website?
  • Correct, Everything mode
  • Oh, are you searching in a collection? It looks like that wasn't working for full-text searches, but try again now.
  • Dstillman, this works! But only for files (pdfs) I have manually indexed. Is there a way to index a large number of files easily without having to do so manually for each document (using the right-hand toolbar)?
  • If you can construct a search (in the search bar or in a saved search) that matches just the attachments without the parent items, you can do a Select All (Ctrl-A or Cmd-A), right-click, and choose Reindex Items.
  • @metabolon: Well, or you can just rebuild the full-text index completely from the Search pane of the Zotero preferences, and then sync. But if you do it via a search, you could limit it to items added since the summer and not have to reindex and resync the full-text content for all items in your library.

    (And again, we're still planning to reindex the recent ones online, so this is only necessary if you don't want to wait.)
Sign In or Register to comment.