[FFmpeg-devel] [PATCH 4/9] lavf/oggdec: simplify destroying streams with chained audio streams.

[Reimar Döffinger]

On Sat, Sep 15, 2012 at 01:20:43AM +0200, Clément Bœsch wrote:
> nstreams is assumed to be 1 at that point, so the loop is pointless.
> ---
>  libavformat/oggdec.c | 9 +++------
>  1 file changed, 3 insertions(+), 6 deletions(-)
> 
> diff --git a/libavformat/oggdec.c b/libavformat/oggdec.c
> index 9d27393..05aeddd 100644
> --- a/libavformat/oggdec.c
> +++ b/libavformat/oggdec.c
> @@ -256,18 +256,15 @@ static int ogg_read_page(AVFormatContext *s, int *sid)
>      idx = ogg_find_stream (ogg, serial);
>      if (idx < 0){
>          if (ogg->headers) {
> -            int n;
>  
>              if (ogg->nstreams != 1) {
>                  av_log_missing_feature(s, "Changing stream parameters in multistream ogg is", 0);
>                  return idx;
>              }
>  
> -            for (n = 0; n < ogg->nstreams; n++) {
> -                av_freep(&ogg->streams[n].buf);
> -                if (!ogg->state || ogg->state->streams[n].private != ogg->streams[n].private)
> -                    av_freep(&ogg->streams[n].private);
> -            }
> +            av_freep(&ogg->streams[0].buf);
> +            if (!ogg->state || ogg->state->streams[0].private != ogg->streams[0].private)
> +                av_freep(&ogg->streams[0].private);

The thing is that this is supposed to be only temporary until someone
implements the feature.
I don't think this simplifies the code enough to justify making the
code doing something semantically wrong ("free all streams" vs.
"free the first stream"), even if it has the same effect.