Don't double correct for discarded codec delay frames.

Don't double correct for discarded codec delay frames.

The existing code was correcting twice for discarded codec delay
frames.  Once by adjusting the initial timestamp and twice by
discarding those frames before generating the output timestamp.

This change also adds a test helper to AudioFileReader so we can
demux audio files in tests without duplicating code.  As a bonus
it provides some additional coverage to AudioFileReader.

I've also reused several FFmpegCommon methods where possible.

BUG=360961
TEST=Adds opus unit tests. Plus a basic playback test for ogg.
NOTRY=true

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=266639

[DaleCurtis, 2014-04-24 23:08:50]

Per discussion.  There's no easy way I can think of to test this since
GetMediaTime() is clamped to zero.  If you have ideas that don't involve boiling
the world, let me know :)

[DaleCurtis, 2014-04-25 17:00:08]

Actually looks like rileya@ has some half baked unit tests in progress that
never got landed. I'll see if I can take that for testing this CL.

https://codereview.chromium.org/134453009/

[DaleCurtis, 2014-04-25 18:23:12]

Now with tests! PTAL.

https://codereview.chromium.org/257563007/diff/30001/media/filters/opus_audio...
File media/filters/opus_audio_decoder.cc (right):

https://codereview.chromium.org/257563007/diff/30001/media/filters/opus_audio...
media/filters/opus_audio_decoder.cc:534:
output_timestamp_helper_->AddFrames(frames_to_output);
This is already fixed in https://codereview.chromium.org/251583003/ so I'll
rebase this once it lands.

[wolenetz, 2014-04-25 23:35:53]

drive-by question: I'm confused a bit. Where were the codec delay frames
previously corrected for, and where are they now corrected for as a result of
this change?

[DaleCurtis, 2014-04-25 23:54:09]

Whoops, I messed up the rebase on top of
https://codereview.chromium.org/251583003/

Thanks for catching, fixed in PS#4.

[acolwell GONE FROM CHROMIUM, 2014-04-26 00:42:15]

lgtm % comment.

https://codereview.chromium.org/257563007/diff/70001/media/filters/opus_audio...
File media/filters/opus_audio_decoder.cc (right):

https://codereview.chromium.org/257563007/diff/70001/media/filters/opus_audio...
media/filters/opus_audio_decoder.cc:292: if (!opus_decoder_)
Why is this needed now?

[DaleCurtis, 2014-04-26 01:04:40]

https://codereview.chromium.org/257563007/diff/70001/media/filters/opus_audio...
File media/filters/opus_audio_decoder.cc (right):

https://codereview.chromium.org/257563007/diff/70001/media/filters/opus_audio...
media/filters/opus_audio_decoder.cc:292: if (!opus_decoder_)
On 2014/04/26 00:42:16, acolwell wrote:
> Why is this needed now?

Pre-existing bug.  If a opus_multistream_decoder_create() fails during a config
change, Stop() can be called later when decoder_ is NULL.

[DaleCurtis, 2014-04-28 16:15:51]

The CQ bit was checked by dalecurtis@chromium.org

[commit-bot: I haz the power, 2014-04-28 16:16:24]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/dalecurtis@chromium.org/257563007/110001

[commit-bot: I haz the power, 2014-04-28 16:56:47]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-04-28 16:56:48]

Try jobs failed on following builders:
  tryserver.chromium on chromium_presubmit

[DaleCurtis, 2014-04-28 17:12:34]

The CQ bit was checked by dalecurtis@chromium.org

[commit-bot: I haz the power, 2014-04-28 17:13:06]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/dalecurtis@chromium.org/257563007/110001

[DaleCurtis, 2014-04-28 19:53:06]

The CQ bit was unchecked by dalecurtis@chromium.org

[DaleCurtis, 2014-04-28 19:53:12]

The CQ bit was checked by dalecurtis@chromium.org

[commit-bot: I haz the power, 2014-04-28 19:53:55]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/dalecurtis@chromium.org/257563007/110001

[commit-bot: I haz the power, 2014-04-28 19:59:39]

Change committed as 266639