PDFs saved from JSTOR are corrupt

PDFs saved with Zotero from JSTOR are corrupt. I get this message when I try to open them:
"File does not begin with '%PDF-'."

Files are all 142 bytes in size. I'm using Firefox 2.0, Windows XP, Zotero 1.0.0b2.r2

Thank you for developing Zotero!
  • Can you send one of these files to support at zotero.org? I'm guessing they're probably text files containing an error message, in which case you might also try just opening them in Notepad and pasting the contents here. Thanks.
  • Yes, it's a text file, but not an error message. Here are the contents:

    <html><body marginheight="0" marginwidth="0"><embed type="application/pdf" src="0.pdf" name="plugin" height="100%" width="100%"></body></html>
  • I get the same error message and the supposed pdf file has similar contents:

    <html><body marginheight="0" marginwidth="0"><embed type="application/pdf" src="5.pdf" name="plugin" height="100%" width="100%"></body></html>

    Win XP, FF2, Zotero 1.0.0b2.r2
  • I get the same kind of HTML file masquerading as a PDF when I download PDF files from the Rutgers Optimality Archive (roa.rutgers.edu)



    This looks to me like some kind of 'envelope' for the PDF file rather than the PDF itself. As a matter of curiosity, I manually saved the PDF direct to this location and Zotero then retrieved it perfectly fine - but no way would I be bothered to do that every time (and why should I?).

    Does anyone know a real fix for this please?
  • Saving snapshots of non-HTML documents (PDFs, images, audio files, etc.) doesn't really work in b2.r2. This has been fixed in Beta 3, which should be out in a couple days.
  • edited August 30, 2007
    I have version 1.0.0rc3 (which I think is the most current version)

    I am still having problems with saving pdf's as attachments.

    With chemistry journals (especially on pubs.acs.org) supporting information is saved as a separate pdf (or 2 or 3) which can be accessed from an extended abstract page (example

    I have tried 2 ways of saving pdfs, and both have had issues

    1) navigating to the "attachments" tab of the parent entry, navigating to the pdf in my browser, and selecting add:take snapshot of current page.
    2) navigating to the extended abstract and right clicking on the link to the pdf and selecting "save link as zotero snapshot"

    The problem is intermittent, but approx 50% of the time when I try to access the snapshot I receive the "File does not begin with '%PDF-'." error. This happens regardless of the method of saving the pdf.
  • edited August 30, 2007
    Hi,
    Have you tried saving the files by dragging and dropping rather than by using the snapshot feature?

    For example, in your 1) method, once you have the pdf open in a browser window, can you "pick it up" in the Firefox address bar and then drag down to the Zotero entry where you'd like it to be attached, and drop there? That method works for me with files opened in the Firefox window by plug-ins.

    I think there's just a fundamental difference between snapshots of web pages on the one hand and, on another hand, other documents that can only be viewed in Firefox with a plugin, like pdfs (Adobe reader plugin, etc.) and odts (OpenOffice.org plugin).

    It's kind of confusing since, if the needed plug-ins are installed, all of the types of documents I've mentioned can appear in the browser window, yet it's not possible to take a snapshot unless the source is a web page.

    It doesn't help matters that the Zotero highlighting and annotations toolbar appears whenever Zotero opens a file, even when using a plug-in which means that annotations and highlighting aren't functional (but this is a known issue: see Dan Stillman's comment on April 4th in this thread).
  • Success!!!

    I wasn't aware urls could be dragged like that at all, let alone to be used to add attachment to Zotero. That should solve the problem.

    Thanks for your suggestion.
  • azmanam: Just to explain what's happening, ACS's "[Full text in pdf]" links actually take you to an HTML page that uses either JavaScript or an HTML META Refresh to redirect you to the appropriate PDF file, so there's not really any good way for Zotero to handle such downloads. Sybille's suggestion of loading the PDF and then dragging the address bar icon is a good workaround, though.
  • I hate to be a pain, but it's still happening with pubs.acs.org documents.

    Still with 1.0.0rc3.

    Dragging the address bar to the Zotero library worked when it was suggested. Now, when I drag the address bar, I am again seeing the "file does not start with %pdf" error message. Clicking the link in the right pane works, though.

    Since pubs.acs.org appears to access pdfs differently than other sites (?) is this going to continue to be a problem with pubs.acs.org articles. Is it something Zotero cannot address because of the discrepancies with pubs.acs.org?
  • edited September 18, 2007
    Automatic download of PDFs from ACS should now be working. Please click "Update now" in your Zotero preferences and restart Firefox to begin testing.
Sign In or Register to comment.