Page 1 of 1

Invalid Characters after unpacking

Posted: July 26th, 2010, 1:18 pm
by Tscherno
Hi there,

i have a problem. I have running sabnzbd on my QNAP. Sometime after unpacking there are some files which are unaccessible via SMB. When i look via SSH they have strange characters in the filenames in it (lika a CR) where for example german umlauts should be (ä,ü,ö). I guess the unrar choose the wrong encoding. Is there a possibility to set an option like for the folders to remove them? Any possibility to control the encoding for the uncompression? Or at last ressort to put an user script in which removes the characters?

Re: Invalid Characters after unpacking

Posted: July 27th, 2010, 5:31 am
by shypike
Unrar always uses UTF-8 encoding, so your file system should support that
and Samba (the service that simulates Windows shares) should be told about it.
An alternative would be to run the a user script that calls
a very popular Perl script for this (available as a package for most systems).
Unfortunately, I cannot remember its name right now.

Re: Invalid Characters after unpacking

Posted: July 27th, 2010, 5:50 am
by Tscherno
I just searched a little bit and maybe found the tool you mentioned: fslint?

http://en.flossmanuals.net/FSlint/BadNames

I would be more happy with a solution on the unraring-side but if there is none i will try it that way... Is it possible to use the logic which is already build into sabnzbd for the directory cleanup?

Re: Invalid Characters after unpacking

Posted: July 27th, 2010, 5:57 am
by shypike
The best way is to configure your system to support UTF-8.
I cannot influence what rarlabs doesn with unrar.
fslint isn't the one I used, but if it does the job, why not.

We will not build a "solution" into SABnzbd, this is way too complex
to do it right on all platforms in all circumstances.
Even then it would require people to enter encoding info about their system into SABnzbd.
It would be a support nightmare.
Anyway: our opinion is that systems should just support UTF-8
instead of mucking about with obsolete code page methods.

Re: Invalid Characters after unpacking

Posted: July 27th, 2010, 12:04 pm
by Tscherno
shypike wrote: The best way is to configure your system to support UTF-8.
Hmh how to do that?

Re: Invalid Characters after unpacking

Posted: July 27th, 2010, 12:52 pm
by Tscherno
OK i think i found the script you mentioned:

http://j3e.de/linux/convmv/man/

I put that into a user-script which looks like that:

#!/bin/sh
convmv -r -f iso-8859-1 -t utf8 --notest $1

Lets see if this is working...

Re: Invalid Characters after unpacking

Posted: July 27th, 2010, 4:34 pm
by shypike
That's the one.