SABnzbd Forums

Version: (07.20)
OS: Centos 7
Install-type: python source
Skin Default
Firewall Software: firewalld
Are you using IPV6? no
Is the issue reproducible? yes

Hi,
I migrated my system from centos6 to 7 and now i have issues unpacking files that contain utf8 characters in it.
I searched the forum, found only older articles, but 1 seems to match, only it was on a nas. I suggested putting a script CharTranslater.py to rename the files after downloading.
I tried that script but it does not seem to help.

So what happens:
I start a download (from spotweb integration), which is named "Thé Lau - Juliëtte (2014)" so it contains 2 special characters.
In the log i see this:
2015-01-17 15:49:34,983::INFO::[__init__:628] Adding Th_#233_ Lau - Juli_#235_tte (2014).nzb
So something already did some renaming..(is this the rename illegal chars setting in nsabnzb????)
Then the download start, runs ok....
Then the script runs..no errors:
2015-01-17 15:49:40,425::INFO::[postproc:284] Starting PostProcessing on Th_#233_ Lau - Juli_#235_tte (2014) => Repair:True, Unpack:True, Delete:True, Script:CharTranslator.py, Cat:books
2015-01-17 15:49:40,928::INFO::[postproc:574] Par2 check starting on Th_#233_ Lau - Juli_#235_tte (2014)
2015-01-17 15:49:40,928::INFO::[__init__:874] /home/downloads/incomplete/Th_#233_ Lau - Juli_#235_tte (2014).1/__ADMIN__/__verified__ missing
2015-01-17 15:49:40,929::INFO::[postproc:595] Running repair on set Jlttthl.rar
2015-01-17 15:49:40,929::INFO::[newsunpack:838] Quick-check for Jlttthl.rar is OK, skipping repair
2015-01-17 15:49:40,929::INFO::[newsunpack:900] Deleting /home/downloads/incomplete/Th_#233_ Lau - Juli_#235_tte (2014).1/Jlttthl.rar.par2
2015-01-17 15:49:40,930::INFO::[postproc:622] Par2 check finished on Th_#233_ Lau - Juli_#235_tte (2014)
2015-01-17 15:49:40,931::INFO::[misc:811] Creating directories: /home/disk04/downloads_complete/books/Th_#233_ Lau - Juli_#235_tte (2014)
2015-01-17 15:49:40,932::INFO::[postproc:351] Running unpack_magic on Th_#233_ Lau - Juli_#235_tte (2014)
2015-01-17 15:49:40,933::INFO::[newsunpack:215] Unrar starting on /home/downloads/incomplete/Th_#233_ Lau - Juli_#235_tte (2014).1
2015-01-17 15:49:40,933::INFO::[newsunpack:428] Extracting rarfile /home/downloads/incomplete/Th_#233_ Lau - Juli_#235_tte (2014).1/Jlttthl.rar (belonging to Jlttthl) to /home/disk04/downloads_complete/books/_UNPACK_Th_#233_ Lau - Juli_#235_tte (2014)
2015-01-17 15:49:40,953::ERROR::[newsunpack:639] ERROR: write error ( /home/disk04/downloads_complete/books/_UNPACK_Th_#233_ Lau - Juli_#235_tte (2014)/Jlttthl/Lau, Th? - Juliette.epub)
2015-01-17 15:49:40,954::INFO::[newsunpack:219] Unrar finished on /home/downloads/incomplete/Th_#233_ Lau - Juli_#235_tte (2014).1

Ending with the error. Well i don't know if that script is still nescescary, it was an older post but with or without it its not working.
Files with normal names are processed ok, so it is not some permissions issue.

Hope you can help.
Gr,
Ronald

SABnzbd 0.7.x doesn't support Unicode.
Keep those characters out or wait for 0.8.0
Having said that, in principle SABnzbd should be able to deal with it.
Can you email the NZB to bugs@sabnzbd.org ?
Please ZIP (or tar) the file to preserve its original name and add the URL of this message.

The job that you sent, downloads fine with 0.7.20 on Windows.
However, I see that the NZB name doesn't contain problematic characters.
For some reason the "specials" are replaces with #233 and #235, which
will not cause a problem.
The fact that the downloaded RAR contains a file with a "special" name
makes no difference.
So it looks to me that the RAR file that you sent doesn't contain the
filename as you first encountered the problem.

BTW: my email answer bounced.

I think I know which NZB this is. I downloaded it on my Linux without problems: including accented characters.

But what about your: /home/disk04/ ... what kind of filesystem is on it? Can you run:
and post the output here?

EDIT: plus this: ... and post the output here.

@bickel: waiting for you followup ...