Corrupt PDF attachement

jpjacobs
Hi!

I've got a PDF attachement to one of my entries that keeps getting corrupted: from time to time it just start failing to load in acrobat reader (as well in the standalone as the browser plugin).

Now I diff'ed the original file and the one in the zotero storage folder, and it seems that although there are still corresponding bits between the two, the majority of the file is composed of other files mingled together. I found bits of an email I send, some pieces of HTML of a site I visited, some registry entries ...

When I just replace the file with the original, everything works again ... up to the next time it gets corrupted. And it's only _that_ file I'm having trouble with.

Anyone that can shed a light on this bizarre problem?
  • jpjacobs
    And it happened again, now wiping the entire file.

    Anyone any thoughts?
  • mark
    I've never seen anything like this. Certainly it is not the kind of corruption that can occur in high-level applications like Zotero and Acrobat. It sounds more like a file system corruption to me. I would do a thorough disk check and look for corrupt sectors.
  • jpjacobs
    I checked the filesystem, everything seems fine, no errors.
  • mark
    Try deleting the attachment from Zotero, emptying the trash (in Zotero), and then readding it from the original copy. If there are no file system problems and you're not doing anything special with that file (editing? other applications accessing it?), it would be very strange if the problem persists.
  • TP357
    I'm using Zotero 2.1.8, on a Windows 7 machine. It seems that Zotero--during the process of saving a webpage--occasionally writes files with illegal characters or names to the computer's hard drive, which damages/corrupts the Windows NTFS file system. You won't be able to delete the bad file--to get rid of it requires running "chkdsk /r /f" or reformatting the hard drive. This happened to my main hard drive and to two separate external USB drives where I was backing up my Zotero data folder to--each Zotero folder had un-deletable files, one was a jpeg file.
  • ajlyon
    It shouldn't be possible for Zotero to write a file with a name that the file system can't handle, since it's the file system that allows the file to be created in the first place. That said, bugs happen at all levels (Windows, Firefox, Zotero), so it'd be very useful if you could identify the filenames in question and where the attachments came from, so this can be properly diagnosed.
  • TP357
    PS:
    The fact that there was an un-deletable corrupt file in each Zotero data folder on each of the three different hard drives points to Zotero as the cause of the problem. Hopefully, Zotero's programing will be modified to prevent this from happening.
  • ajlyon
    And there's no way to work out how to prevent this unless you can provide more information on where the files came from. That means, probably, what website the snapshot was from, and the filenames in question.
  • TP357
    Here's info about one the corrupt file's and its location (on one of my USB hard drives):

    (E) Zotero > storage > QK6RFTTS > logoafpsmall.jpg

    The jpg file would have come from one of the webpages I saved using Zotero. Again, corrupt file/s were found in three separate Zotero storage folders, on three separate hard drives. In fact, I discovered the corrupt file on one USB hard drive after I tried to delete everything on it--the only un-deletable file was in the Zotero storage folder.
  • TP357
    By the way, I think Zotero is great tool nevertheless.
  • mark
    edited July 25, 2011
    Only if you can reproduce the location of the website you saved the snapshot from, others can try it out on their systems. Three possibilities: (1) it's a problem that is specific to your system and/or to the time you saved this particular snapshot, (2) it's a problem with this specific website, (3) It's a problem with the Zotero translator/snapshot saving code.

    BTW How do you get three separate Zotero storage folders on three separate hard drives? They are copies of each other I presume?
  • ajlyon
    Well, I'd add that #2 implies #3.
  • TP357
    I got the following error message after the operating system failed to delete the corrupt file in the Zotero storage folder:

    An unexpected error is keeping you from deleting the file. If you continue to receive this error, you can use the error code to search for help with this problem.

    Error 0x80070570: The file or directory is corrupted and unreadable.

    logoafpsmall.jpg
    Item type: JPG File
    Rating: Unrated
    Size: 1.30 KB
  • ajlyon
    Can you say what URL the file came from?
  • TP357
    Yes -- same Zotero storage folder on each of the drives. I was backing up the Zotero storage folder on my main hard drive to an external USB hard drive I no longer have--the corrupt file showed up after I tried to delete all the contents on that USB drive. Until recently, I was backing it up to a second USB hard drive (320 G, Seagate)--the corrupt file in the Zotero folder showed up after I tried to delete the entire folder (after deciding to back up the Zotero storage folder to a much larger hard drive). Consequently, to prevent file system corruption from happening to another drive, I'll likely use Zotero's online storage for backup instead.
  • TP357
    No--unfortunately, each time I deleted the Zotero folder, everything in the folder was deleted (URLs, ect) except the corrupt file. There was no URL in the "QK6RFTTS" Zotero storage folder above (Zotero > storage > QK6RFTTS > logoafpsmall.jpg) on the main hard drive, however, based on one of the html files that was in that folder, I think this might be the URL:

    http://www.breitbart.com/article.php?id=CNG.ae71a038e9b3b47af4f0e9eac9598fd8.2b1
Sign In or Register to comment.