Bug 90977 - 'Verify written data' always fails when filenames contain non-ASCII characters
Summary: 'Verify written data' always fails when filenames contain non-ASCII characters
Alias: None
Product: k3b
Classification: Applications
Component: general (show other bugs)
Version: unspecified
Platform: unspecified Linux
: NOR normal
Target Milestone: ---
Assignee: Sebastian Trueg
Depends on:
Reported: 2004-10-08 18:41 UTC by Michał Kosmulski
Modified: 2005-12-10 13:54 UTC (History)
0 users

See Also:
Latest Commit:
Version Fixed In:


Note You need to log in before you can comment on or make changes to this bug.
Description Michał Kosmulski 2004-10-08 18:41:43 UTC
Version:           0.11.17 (using KDE KDE 3.3.0)
Compiler:          gcc 3.3.4 
OS:                Linux

I record my CDs with both Joliet and RockRidge extensions turned on, so I expect that all filenames be written to the disc using Unicode characters. My system uses a UTF-8 locale. When there is a file containing non-ASCII characters (e.g. Polish diacritics) in its name and 'Verify written data' is enabled, verifying the recorded disc always fails, even if a manual md5sum check confirms that the disc was recorded correctly.
Comment 1 Michał Kosmulski 2004-10-12 20:54:22 UTC
Additional info:
I use supermount (supermount-ng.sourceforge.net) for mounting the CD-burner, the line in /etc/fstab is:
none /mnt/cdrom supermount fs=iso9660,dev=/dev/cdrom,--,auto,ro,utf8  0   0
but I also tested with a normal mount:
/dev/cdrom /mnt/cdrom iso9660 defaults,users,noauto,ro,utf8  0  0
and got the same results. Verifying fails, even though I can manually check that all files were recorded correctly, with the message 'Could not find XYZ', where XYZ is the name of one of the files whose names contain non-ASCII characters. I burn with all 'compatibility' options ('allow 31 char names' etc.) disabled.
Comment 2 Helge Hielscher 2004-11-19 23:40:35 UTC
sounds like a dup of bug 83315
Comment 3 Sebastian Trueg 2005-12-10 13:54:34 UTC
SVN commit 487359 by trueg:

Copied iso9660 filename generation code from mkisofs. With it K3b is able
to determine the iso9660 names generated by mkisofs and use them for data
project verification.
The result is: no charset problems anymore.

BUG: 90974
BUG: 84544
BUG: 90977
BUG: 105649
BUG: 113052
BUG: 83315
BUG: 80593

 M  +620 -39   projects/datacd/k3bdatadoc.cpp  
 M  +2 -0      projects/datacd/k3bdatadoc.h  
 M  +11 -0     projects/datacd/k3bdataitem.cpp  
 M  +15 -3     projects/datacd/k3bdataitem.h  
 M  +1 -2      projects/datacd/k3bdataverifyingjob.cpp  
 M  +109 -29   tools/k3biso9660.cpp  
 M  +29 -2     tools/k3biso9660.h