[WebRTC] Implement Outgoing libwebrtc audio source support

[youenn fablet]

[WebRTC] Implement Outgoing LibWebRTCNetwork audio source support

[youenn fablet]

Created attachment 301161 [details]
Patch

[youenn fablet]

Created attachment 301168 [details]
Patch

[youenn fablet]

Created attachment 301176 [details]
Patch

[youenn fablet]

The story about muted/disabled tracks is not clear to me yet.
Should we continue being called by the capture thread?
Should we be notified and stop filling the ring buffer?
Or should we just not send any data to libwebrtc?

[Eric Carlson]

Comment on attachment 301176 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=301176&action=review

> Source/WebCore/ChangeLog:10
> +        Using an AudioSampleDataSource to convert the captured data in libwebrtc expected format.

Nit: "in" -> "to"

> Source/WebCore/ChangeLog:14
> +        Introducing bWebRTCAudioFormat.h to centrialise libwbebrtc expected audio format.

Nit: "bWebRTCAudioFormat.h" -> "WebRTCAudioFormat.h"

> Source/WebCore/platform/audio/mac/AudioSampleDataSource.cpp:289
> +bool AudioSampleDataSource::pullAvalaibleSamplesAsChunks(AudioBufferList& buffer, size_t sampleCountPerChunk, uint64_t timeStamp, Function<void()>&& bufferIsFilled)

Nit: "bufferIsFilled" reads like it should be a bool, something like "consumeSamples", or "processSamples" might be better.

> Source/WebCore/platform/mediastream/mac/RealtimeOutgoingAudioSource.h:70
> +    void audioSamplesAvailable(const MediaTime&,  const PlatformAudioData&, const AudioStreamDescription&, size_t) final;

Nit: extra space.

[youenn fablet]

Created attachment 301195 [details]
Patch for landing

[WebKit Commit Bot]

Comment on attachment 301195 [details]
Patch for landing

Clearing flags on attachment: 301195

Committed r212144: <http://trac.webkit.org/changeset/212144>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.