Cannot download PDF with Zotero Connector

Hello,

I have recent trouble to download .pdf files from Firefox with the connector.

For example, on arXiv with https://arxiv.org/abs/2004.00313, the item is correctly added to my Zotero library but the .pdf is not downloaded (red cross (X) in popup).

If I view the pdf of the same article (https://arxiv.org/pdf/2004.00313.pdf), it is the same I cannot add the pdf to my library (red cross (X) again).

This problem is common to all the pdf.

I can download in the standard way by clicking the download button of Firefox's pdf viewer

I remember having the problem but it disappeared without doing nothing. Now it is back, I would appreciate any help to solve this issue which is a bit annoying from a workflow point-of-view.

I'm running the last Zotero version (5.0.85) and connector on Windows 10 .
If you require any other information from my side, feel free to ask !

Thank you.

  • Both of those are working for me.

    Can you provide a Debug ID from Zotero (not the Connector) for a save attempt that fails to download the PDFs?
  • edited April 7, 2020
    Here is the Debug ID: D2002377077 (hope I did it correctly).

    I realized the save attempt with https://arxiv.org/pdf/2004.00313.pdf.

    Thank you for replying.
  • edited April 7, 2020
    [JavaScript Error: "{better-bibtex! +0} pathSearch: <Error: Component returned failure code: 0x80520001 (NS_ERROR_FILE_UNRECOGNIZED_PATH) [nsIFile.initWithPath]#
    pathSearch@chrome://zotero-better-bibtex/content/common.js:62436:29
    init@chrome://zotero-better-bibtex/content/common.js:55608:26
    init@chrome://zotero-better-bibtex/content/common.js:55884:15
    init@chrome://zotero-better-bibtex/content/common.js:56754:15
    > "]

    [JavaScript Error: "item.isRegularItem is not a function" {file: "chrome://zotero/content/xpcom/retractions.js" line: 832}]

    [JavaScript Error: "Win error 5 during operation move on file C:\Users\[…]\Zotero\storage\tmp-S2FvHH (Accès refusé.
    )" {file: "(unknown module)" line: 0}]
    You're getting some weird errors.

    Does it work if you disable Better BibTeX? If not, can you provide another Debug ID with that disabled, after restarting Zotero?
  • I disabled BetterBibTeX in Zotero, then restarted both Zotero and Firefox.
    I tried again to download the same document but it failed again.

    So I restarted again and generated the following debug ID after a failed save attempt: D213972880.
  • OK, it's the access-denied error, then. This is usually caused by security software on the system preventing Zotero from accessing the disk. You can also check the your Zotero data directory to make sure your user account has read and write permissions to all files (which it should unless you copied the folder from another computer).
  • Hello,

    So I managed to fix the problem which was not related to a security software but to the read/write permissions.

    The problem it seems I had was a little tricky.

    First, I've checked that Zotero (and other) folder has the right permissions. It appeared that the owner of my folders was the SID of my user account (Security Identifier, ID which is proper to each Windows user account) but not the user name or the FullName (like a pseudo).

    My user account is linked to my Microsoft Account, and for any reasons, my account did not have a FullName (the name Windows displays on the logging screen), only the @outlook.fr address was used as identifying my account.

    I guess that's why all the folders security settings were indicating that their owner was the user account's SID.

    Adding a FullName (https://www.windowscentral.com/how-add-your-full-name-local-account-windows-10) to my user account and adding a name and lastname to my Microsoft account (modify it directly on microsoft's website) fixed the problem (but I guess that only adding a FullName would have solve the problem).

    Hence, I would believe Zotero relies on the FullName to be able to download a pdf in the storage folder. I guess the FullName is, in some ways, used to check whether or not the writting permission are given to Zotero.

    Hope this information would be useful for future debugging.

    Thank you dstillman for pointing me out the problem !
  • The pathSearch error is harmless. I'm testing for the existence of a binary in the PATH, but FileUtils.File errors out if I give it a non-existing path, even if I do intend to later check for .exists() and .isFile. I log the error and proceed.
Sign In or Register to comment.