[FFmpeg-devel] [PATCH] fftools: Avoid build failure on OSX 10.6 with gcc 4.2.1

[Pavel Koshevoy]

On Tue, Jul 18, 2023 at 3:59 PM Andreas Rheinhardt <
andreas.rheinhardt at outlook.com> wrote:

> Pavel Koshevoy:
> > ping ... could someone review this please
> >
> > On Sat, Jul 15, 2023 at 12:33 PM Pavel Koshevoy <pkoshevoy at gmail.com>
> wrote:
> >
> >> ... and make code more readable / human friendly.
> >> ---
> >>  fftools/ffmpeg_mux_init.c | 3 ++-
> >>  1 file changed, 2 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/fftools/ffmpeg_mux_init.c b/fftools/ffmpeg_mux_init.c
> >> index 6458414b5f..bade35896f 100644
> >> --- a/fftools/ffmpeg_mux_init.c
> >> +++ b/fftools/ffmpeg_mux_init.c
> >> @@ -1648,7 +1648,8 @@ read_fail:
> >>
> >>  static int create_streams(Muxer *mux, const OptionsContext *o)
> >>  {
> >> -    static const int (*map_func[])(Muxer *mux, const OptionsContext
> *o) =
> >> {
> >> +    typedef int(*TMapFunc)(Muxer *, const OptionsContext *);
> >> +    static const TMapFunc map_func[] = {
> >>          [AVMEDIA_TYPE_VIDEO]    = map_auto_video,
> >>          [AVMEDIA_TYPE_AUDIO]    = map_auto_audio,
> >>          [AVMEDIA_TYPE_SUBTITLE] = map_auto_subtitle,
> >> --
> >> 2.35.3
> >>
> >>
>
> This issue has already been fixed in
> 153cf85b246a7931fac6344f2189cd268ca9e0aa.
>

I see ... I can check if it compiles on OSX 10.6 over the weekend.
I've got to say that a typedef for the function pointer is arguably a
better solution, as it doesn't require nearly as much mental gymnastics to
parse what is happening on that line.
Code should be easy to read, and C function pointer syntax is very awkward
... combining it all into an array declaration at the same time is just
making it worse.

... but this is not my code to maintain, so feel free to ignore.

Pavel.