On Sun, 16 Oct 2011 22:11:10 -0700, "Ronald S. Bultje" <[email protected]> wrote: > Hi, > > On Wed, Oct 12, 2011 at 10:12 AM, Anton Khirnov <[email protected]> wrote: > > Manual replacements are done in this commit. > > --- > > avconv.c | 5 ++++- > > libavformat/4xm.c | 2 +- > > libavformat/applehttp.c | 3 ++- > > libavformat/avidec.c | 2 +- > > libavformat/avisynth.c | 4 ++-- > > libavformat/avs.c | 4 ++-- > > libavformat/bink.c | 2 +- > > libavformat/c93.c | 2 +- > > libavformat/flic.c | 2 +- > > libavformat/flvdec.c | 3 ++- > > libavformat/gxf.c | 3 ++- > > libavformat/idroqdec.c | 2 +- > > libavformat/jvdec.c | 2 +- > > libavformat/libnut.c | 2 +- > > libavformat/lmlm4.c | 2 +- > > libavformat/lxfdec.c | 2 +- > > libavformat/mov.c | 3 ++- > > libavformat/mpeg.c | 3 ++- > > libavformat/mpegts.c | 18 ++++++++++++------ > > libavformat/mtv.c | 10 ++++------ > > libavformat/mxfdec.c | 3 ++- > > libavformat/mxg.c | 10 ++++------ > > libavformat/nsvdec.c | 6 ++++-- > > libavformat/nutdec.c | 2 +- > > libavformat/nuv.c | 4 ++-- > > libavformat/oggdec.c | 2 +- > > libavformat/pva.c | 2 +- > > libavformat/r3d.c | 2 +- > > libavformat/rdt.c | 3 ++- > > libavformat/rtsp.c | 2 +- > > libavformat/sapdec.c | 4 ++-- > > libavformat/swfdec.c | 9 ++++++--- > > libavformat/wtv.c | 3 ++- > > libavformat/xmv.c | 2 +- > > libavformat/yop.c | 2 +- > > 35 files changed, 75 insertions(+), 57 deletions(-) > [..] > > index 567baeb..eadb230 100644 > > --- a/libavformat/applehttp.c > > +++ b/libavformat/applehttp.c > > @@ -505,11 +505,12 @@ static int applehttp_read_header(AVFormatContext *s, > > AVFormatParameters *ap) > > snprintf(bitrate_str, sizeof(bitrate_str), "%d", v->bandwidth); > > /* Create new AVStreams for each stream in this variant */ > > for (j = 0; j < v->ctx->nb_streams; j++) { > > - AVStream *st = av_new_stream(s, i); > > + AVStream *st = avformat_new_stream(s, NULL); > > if (!st) { > > ret = AVERROR(ENOMEM); > > goto fail; > > } > > + st->id = i; > > I see this everywhere, is it really a good idea to remove the argument > if we're going to have to add an extra line of code for each use? >
Not really, in most cases it's not set. In many remaining cases it can be removed, I just didn't want to make the patch too complicated. I think id is actually used for something meaningful in about 5-10 places. Out of 180. -- Anton Khirnov _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
