[FFmpeg-devel] [PATCH] lavfi: unify asink_buffer and vsink_buffer API

[Michael Niedermayer]

On Sat, Aug 27, 2011 at 04:36:44PM +0200, Stefano Sabatini wrote:
> On date Saturday 2011-08-27 01:19:36 +0200, Michael Niedermayer encoded:
> > On Thu, Aug 25, 2011 at 03:09:45PM +0200, Stefano Sabatini wrote:
> > > On date Saturday 2011-08-20 19:44:59 +0200, Stefano Sabatini encoded:
> > > > More generic, less code duplication.
> > > > ---
> > > >  avconv.c                    |    5 +-
> > > >  ffmpeg.c                    |    5 +-
> > > >  ffplay.c                    |    5 +-
> > > >  libavdevice/lavfi.c         |   11 ++--
> > > >  libavfilter/Makefile        |    2 +-
> > > >  libavfilter/asink_abuffer.c |   97 ------------------------------
> > > >  libavfilter/asink_abuffer.h |   47 ---------------
> > > >  libavfilter/vsink_buffer.c  |  139 ++++++++++++++++++++++++++++++++++---------
> > > >  libavfilter/vsink_buffer.h  |   38 +++++++++---
> > > >  9 files changed, 155 insertions(+), 194 deletions(-)
> > > >  delete mode 100644 libavfilter/asink_abuffer.c
> > > >  delete mode 100644 libavfilter/asink_abuffer.h
> > > 
> > > Breaks ABI, adding a backward compatibility layers seems no easy
> > > (unless we add another buffer2 sink), so I suggest to bump major and
> > > get done with it.
> > 
> > ok but i suggest we use this opertunity to make the code not require
> > ABI bumps on extensions where its avoidable
> > so we dont need to bump weekly
> 
> Please check attachment.
> 
> Bikeshed:
> av_sink_buffer_get_buffer_ref
> or
> avfilter_sink_buffer_get_buffer_ref
> ?
> 
> If this is fine I'm also going to rename:
> vsink_buffer.{hc} -> sink_buffer.{hc}
> 
> in a separate commit.
> 
> I'm also attaching the lavfi indev audio support patch for reference.

[...]
> +/**
> + * Get an audio/video buffer data from buffer_sink and put it in bufref.
> + *
> + * This function works with both audio and video buffer sinks.
> + *
> + * @param buffer_sink pointer to a buffersink or abuffersink context

maybe the function could free the parameters too ?
would simplify the calling code if its possible

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Breaking DRM is a little like attempting to break through a door even
though the window is wide open and the only thing in the house is a bunch
of things you dont want and which you would get tomorrow for free anyway
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20110827/44e83df9/attachment.asc>