[FFmpeg-devel] Would like to work on lavf-based concatenation tool for GSoC qualification

Baptiste Coudurier baptiste.coudurier
Thu Apr 2 19:54:22 CEST 2009


On 4/2/2009 10:48 AM, Michael Niedermayer wrote:
> On Thu, Apr 02, 2009 at 11:01:01AM -0400, Geza Kovacs wrote:
>> On Thu, 2009-04-02 at 14:17 +0200, Michael Niedermayer wrote:
>>> On Thu, Apr 02, 2009 at 05:09:16AM -0400, Geza Kovacs wrote:
>>>> On Wed, 2009-04-01 at 13:50 -0700, Baptiste Coudurier wrote:
>>>>> Yes, please enhance ffmpeg binary. You can put code in a
>>>>> separate file, but the goal is to enhance ffplay and ffmpeg
>>>>> binaries, so factorizing the code is likely needed.
>>>>> 
>>>> I've attached a patch that adds said functionality as an
>>>> ffmpeg command-line option. I assume there's a more clean way
>>>> to do this than by introducing a conditional at the beginning
>>>> of av_encode(); I assume I would probably want to modify
>>>> opt_format and opt_output_file in the final version instead,
>>>> but for the sake of time this seemed to work best with the code
>>>> I had already written for now.
>>> Id like to clarify that the goal of the qualification tasks is
>>> to get code into ffmpeg svn that is usefull for ffmpeg
>>> 
>>> hacks will not reach svn
>>> 
>> Thanks for the response, that was exactly what I was concerned
>> about. What, then, is considered a "non-hack"-ish way of
>> accomplishing this? In this case, one has to override the output
>> format settings based on information from the input format
>> settings, so I don't believe this could be done optimally within
>> opt_format or opt_output_file, or whether this should be done, in a
>> "non-hack"-ish way, within opt_format() and opt_output_file() or
>> instead within av_encode().
> 
> Well, its your qualification task, you picked it, also iam not the
> one who added this to the small tasks page. also forget adding any 
> if(concatenation) do_something_entirely_different() anywhere. chances
> are this is buggy broken, wrong and code duplication
> 
> it could be done in ffmpeg.c & ffplay.c, it could be done as a new
> demuxer and there might be other ways.

Indeed,
I suggest you have a look at eof_reached handling in av_encode in
ffmpeg.c and analyze wether it would be interesting to handle next file
around this.

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



More information about the ffmpeg-devel mailing list