[pkg-mad-maintainers] Bug #405801 is probably in libid3tag, can you have a look?

[Kurt Roeckx]

On Wed, Jan 10, 2007 at 11:05:57PM +0100, Andreas Henriksson wrote:
> On ons, 2007-01-10 at 22:52 +0100, Kurt Roeckx wrote:
> 
> > The function show is a rather simple function, and I don't see how
> > anything can be wrong with it:
> > id3_length_t id3_ucs4_length(id3_ucs4_t const *ucs4)
> [...]
> > I can only assume that it's being called with an invalid string, like
> > one that isn't encoded in ucs4.  A full backtrace would be nice too.
> > 
> 
> The (only) caller of this function is libid3tag itself if I read the
> code correctly. See id3_compat_fixup in compat.c. The argument passed to
> the function is also constructed here as far as I can see.

The function isn't exported in header files, but can be called in
several ways internally in libid3tag.

Anyway, I can't say wether this is a bug in libid3tag or not from this,
but it's likely.

> I have not reproduced the problem, so unfortunately you can not rely on
> me providing any backtraces (yet). As mentioned in the previous mail
> I've asked the original bug submitter to provide a mp3 as a testcase.
> The original bug report includes instructions on how to create one (if
> you have the right charsets set up?).
> Please see: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=405801

The bug report doesn't contain anything useful for me to test anything,
so I'm waiting for an testcase with an mp3 myself.

Afaik, libid3tag doesn't care about any locale settings, so that's
probably unrelated.

What can be the problem is that the encoding in the file is totaly wrong
and I've already had to fix a few things for that.  We'll see when I get
something I can test it with.


Kurt