Copied to clipboard

Flag this post as spam?

This post will be reported to the moderators as potential spam to be looked at


  • Amir Khan 1287 posts 2744 karma points
    Oct 31, 2012 @ 20:33
    Amir Khan
    0

    Illegal character preventing publishing

    Hi,

    I'm having an issue where I have an illegal character error showing up in my umbracoLog table which I believe is preventing me from publishing pages on my site, how do I find out where this character is so I can actually remove it?

    Thanks!

    Amir

  • Bo Damgaard Mortensen 719 posts 1207 karma points
    Nov 04, 2012 @ 03:20
    Bo Damgaard Mortensen
    0

    Hi Amir,

    I had this issue a few days ago where I copied some text from photoshop and pasted it into the editor in Umbraco. What I did was to simply open notepad, paste in the text and look for any suspecious characters. You could also try the "Paste from Word" option in RTE (if you're using RTE as datatype) to see if that helps.

    - Bo

  • dimi309 245 posts 579 karma points
    Nov 04, 2012 @ 14:32
    dimi309
    0

    Ouch! My colleague actually added a bit of javascript to TinyMCE to remove those. That offered us a permanent solution.

    In any case, the reason why they are invisible is that they are probably control characters. You can use something like Notepad++ if you want to be able to see and remove them manually.

    Cheers,

    Dimitri

  • Amir Khan 1287 posts 2744 karma points
    Nov 05, 2012 @ 22:36
    Amir Khan
    0

    The challenge is that the umbraco.config file doesn't exist and I can't delete any of the pages from the backend so there's really nothing I can open in a text editor to find out where this pesky character is. Aside from going through every page of the site I suppose...

  • dimi309 245 posts 579 karma points
    Nov 05, 2012 @ 22:46
    dimi309
    0

    How about having a look in the database in the umbraco log table? Usually the node ids of the pages that cause a problem are mentioned there...

  • dimi309 245 posts 579 karma points
    Nov 05, 2012 @ 22:49
    dimi309
    0

    Or if you don't find anything there, how about writing an SQL statement that selects all the texts together with the node ids, and then run that through a text editor like notepad plus plus to see where the characters are. It might be a bit too hacky but I suppose it beats going through every page; just an idea...

  • Amir Khan 1287 posts 2744 karma points
    Nov 06, 2012 @ 15:37
    Amir Khan
    0

    Dimitri, both great suggestions. I actually ooked in the log table and saw the errors from the the "illegal character 0x00" or something similar but couldn't seem to find the nod id referenced anywhere?

    Luckily I was able to restore from a backup to before the problem occured and lock it down so hopefully it won't happen again. I'm pretty sure it was a problem related to uBlogsy and comments getting spammy. The client doesn't upload the site very often and don't paste from Word just to avoid formatting issues.

    Thanks for your help!
    Amir

  • dimi309 245 posts 579 karma points
    Nov 06, 2012 @ 15:48
    dimi309
    0

    Thanks Amir! I'm glad you have found a way out of this! About the nodeIds, maybe I was lucky myself. I had found them in the database but it was thanks to the fact that they were crashing Lucene too, besides the Umbraco publishing mechanism, and this second crash was actually reporting the Ids (and flooding the log with an error per second for hours, but that's a separate issue :))

Please Sign in or register to post replies

Write your reply to:

Draft