[mpg123-devel] [ mpg123-Bugs-2834636 ] NULL treated as single byte, even for UTF-16

[SourceForge.net <no...@so...>]

Bugs item #2834636, was opened at 2009-08-10 00:34
Message generated for change (Settings changed) made by sobukus
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=733194&aid=2834636&group_id=135704

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: mpglib
Group: svn
Status: Closed
Resolution: Fixed
Priority: 5
Private: No
Submitted By: James Le Cuirot (jerseychewi)
Assigned to: Nobody/Anonymous (nobody)
Summary: NULL treated as single byte, even for UTF-16

Initial Comment:
I reported a bug to the MusicBrainz Picard project because it seemed to be breaking my MP3s. mpg123 was throwing up warnings and XMMS2 wasn't playing them at all. I haven't figured out the problem with XMMS2 yet but it seems that this may have been a coincidence because there is a bug in mpg123. This is what, Lukáš Lalinský, the author of Picard had to say. I also suspected it might be something like this.

--------

So, it turns out I was using an ancient version of mpg123. I checked out the source code from SVN and it indeed did print the warnings. With enough -v options, you can see that it's happening only for TXXX frames. So I had a look at the source code and they have a bug in function next_text in src/libmpg123/id3.c. That function is supposed to find an encoded null character, but it doesn't take into account only byte sequences at aligned positions. So if you have UTF-16 LE text, such as:

41 00 4c 00 00 00 42 00 55 00 4d 00 ('A' 'L' '\0' 'B' 'U' 'M')

The function should return a pointer to 42 00 55 00 4d 00 (ie, read 41 00 4c 00 and find a null character), but instead it returns a pointer to 00 42 00 55 00 4d 00 because it looks for the null character on any position, not just aligned ones. You can use this mail to submit a bug
report to mpg123. :) I'm not sure about XMMS2, but I'm fairly sure this isn't a Picard bug.

----------------------------------------------------------------------

Comment By: James Le Cuirot (jerseychewi)
Date: 2009-08-10 12:10

Message:
That did the trick. Thanks a lot. :)

----------------------------------------------------------------------

Comment By: Thomas Orgis (sobukus)
Date: 2009-08-10 03:02

Message:
Man, Sourceforge is driving me nuts -- I just submitted that comment, and
where is it now? Lost in Lala-land...

Anyhow, here again: I worked on the poor function, would you like to test
the current http://mpg123.org/snapshot if it works now with your tags?


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=733194&aid=2834636&group_id=135704