On 1/23/2018 10:46 PM, wm4 wrote:
> On Tue, 23 Jan 2018 17:03:26 -0800
> Richard Shaffer <rshaf...@tunein.com> wrote:
>
>> On Tue, Jan 23, 2018 at 2:32 PM, wm4 <nfx...@googlemail.com> wrote:
>>
>>>
>>> I think this test program is pretty nice, though usually we try to get
>>> by with running the "ffmpeg" utility to test the libs. Having a
>>> relatively big program to test some obscure functionality might have a
>>> maintenance burden. I'm not sure how others feel about it; if nobody
>>> has a strong opinion on this I'll probably apply this in a week or so.
>>>
>>
>> It's definitely not my desire to create a maintenance burden for anyone. I
>> used the ffmpeg utility initially to test my changes, and I'm sure that
>> shell scripting would have been easier than writing a C program.
>>
>> There were pretty good templates for other test programs, but I didn't come
>> across a collection of shell scripts or Makefile targets that seemed to
>> illustrate a clear pattern. I'll look a little harder. If there is a
>> pointer to an example test, though, that would probably be useful.
>
> Most FATE tests use some sort of ffmpeg CLI invocation via shell, but
> it's pretty cryptic. Not sure what's a good example.
ffprobe -show_format is probably best for an id3v2 tag. Add a new
probeformat() function to fate-run.sh like
probeformat(){
run ffprobe${PROGSUF} -show_format -v 0 "$@"
}
Then call it with the sample in question a argument. See how exif.mak
uses probeframes().
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
