[FFmpeg-devel] [PATCH] mov.c: factorize ff_mov_read_stsd_entries()

[Baptiste Coudurier]

On 3/9/2009 3:41 PM, Ronald S. Bultje wrote:
> Hi Baptiste,
> 
> On Mon, Mar 9, 2009 at 5:57 PM, Baptiste Coudurier
> <baptiste.coudurier at gmail.com> wrote:
>> Well I would prefer if the whole function would get changed to parse
>> these structures and populate a MOVStsd struct. You could then use these
>> structs. Besides I believe it would be easier to then create separate
>> tracks and handle multiple data reference.
> 
> I think that was my initial intention, but then it wasn't possible
> because the palette reading code halfway that function directly
> touches stuff in MOVContext. I didn't look too deeply, so it may in
> fact be possible to move that stuff away from MOVContext into MOVStsd
> and then this would indeed be a good approach, at least much better
> than initializing the complete QT demuxer just to read one atom...
> 
> Are you familiar with the palette reading code? Can it be moved (can I
> move it) into MOVStsd without problems? Any reason it isn't there
> already?

I think Palette must be in MOVStsd anyway, and palette stuff is
deprecated ...
Demuxer will gather palette from MOVStsd if it needs too, and update
MOVContext if needed.

How does this sound ?

-- 
Baptiste COUDURIER                              GnuPG Key Id: 0x5C1ABAAA
Key fingerprint                 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA
checking for life_signs in -lkenny... no
FFmpeg maintainer                                  http://www.ffmpeg.org