Added VideoTrackInterface::GetSink method, for use by VideoRtpReceiver. Get rid of FrameInput.

Added VideoTrackInterface::GetSink method, for use by VideoRtpReceiver. This lets us delete FrameInput.

This way, the use of VideoSourceInterface::FrameInput method becomes
an implementation details in VideoTrack. When VideoTrack is changed to
become a VideoSink, the implementation of VideoTrack::GetSink can be
changed to simply return |this|.

Having an extra method which will ultimately just returning |this|
looks silly, but may have some advantage when using VideoTrackProxy
(which I don't fully understand), since we return the underlying
object instead of proxying OnFrame calls.

On top of https://codereview.webrtc.org/1668493002/

BUG=webrtc:5426

[nisse-webrtc, 2016-02-03 13:37:14]

Description was changed from

==========
Added VideoTrackInterface::GetSink method, for use by VideoRtpReceiver.

This way, the use of VideoSourceInterface::FrameInput method becomes
an implementation details in VideoTrack. When VideoTrack is changed to
become a VideoSink, the implementation of VideoTrack::GetSink can be
changed to simply return |this|.

Having an extra method which will ultimately just returning |this|
looks silly, but may have some advantage when using VideoTrackProxy
(which I don't fully understand), since we return the underlying
object instead of proxying OnFrame calls.

On top of https://codereview.webrtc.org/1668493002/

BUG=webrtc:5426
==========

to

==========
Added VideoTrackInterface::GetSink method, for use by VideoRtpReceiver. This
lets us delete FrameInput.

This way, the use of VideoSourceInterface::FrameInput method becomes
an implementation details in VideoTrack. When VideoTrack is changed to
become a VideoSink, the implementation of VideoTrack::GetSink can be
changed to simply return |this|.

Having an extra method which will ultimately just returning |this|
looks silly, but may have some advantage when using VideoTrackProxy
(which I don't fully understand), since we return the underlying
object instead of proxying OnFrame calls.

On top of https://codereview.webrtc.org/1668493002/

BUG=webrtc:5426
==========

[nisse-webrtc, 2016-02-03 13:37:14]

nisse@webrtc.org changed reviewers:
+ perkj@webrtc.org, pthatcher@webrtc.org

[nisse-webrtc, 2016-02-03 13:42:30]

This is my first refactoring on the receiver side. It gets rid of the FrameInput
method and the FrameInputWrapper class.

I had to disable one test, VideoSourceTest.StartStopRemote, since it now lacks a
method to inject frames given only a VideoSource.

Maybe a VideoSource should also be a mux, with an OnFrame method? Even though
except for tests, OnFrame would be called only from within the class itself, not
externally.

[nisse-webrtc, 2016-02-03 13:45:22]

The CQ bit was checked by nisse@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2016-02-03 13:45:26]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1664773002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1664773002/20001

[commit-bot: I haz the power, 2016-02-03 15:46:03]

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

[commit-bot: I haz the power, 2016-02-03 15:46:03]

Dry run: Try jobs failed on following builders:
  android_arm64_rel on tryserver.webrtc (JOB_TIMED_OUT, no build URL)
  linux_baremetal on tryserver.webrtc (JOB_TIMED_OUT, no build URL)
  win_baremetal on tryserver.webrtc (JOB_TIMED_OUT, no build URL)

[pthatcher1, 2016-02-03 16:24:29]

You may consider pulling the renderer => sink part of this CL out into a smaller
one rather than the "add GetSink to track" part.  That might make more sense.

In either case, I like it all, except for the part I keep repeating about making
a VideoTrackInterface implement a VideoSinkInterface.

https://codereview.webrtc.org/1664773002/diff/20001/talk/app/webrtc/mediastre...
File talk/app/webrtc/mediastreaminterface.h (right):

https://codereview.webrtc.org/1664773002/diff/20001/talk/app/webrtc/mediastre...
talk/app/webrtc/mediastreaminterface.h:148: virtual VideoSourceInterface*
GetSource() const = 0;
As mentioned in the other CL, I think we can just have VideoTrackInterface
implement rtc::VideoSinkInterface<cricket::VideoFrame> and then we don't need
GetSink() and we can just use a VideoTrackInterface as a sink.

https://codereview.webrtc.org/1664773002/diff/20001/talk/app/webrtc/mediastre...
File talk/app/webrtc/mediastreamprovider.h (right):

https://codereview.webrtc.org/1664773002/diff/20001/talk/app/webrtc/mediastre...
talk/app/webrtc/mediastreamprovider.h:96:
rtc::VideoSinkInterface<cricket::VideoFrame>* sink) = 0;
You could do this in the smaller CL as well.  There's nothing blocking it, I
don't think.

https://codereview.webrtc.org/1664773002/diff/20001/talk/app/webrtc/rtpreceiv...
File talk/app/webrtc/rtpreceiver.cc (right):

https://codereview.webrtc.org/1664773002/diff/20001/talk/app/webrtc/rtpreceiv...
talk/app/webrtc/rtpreceiver.cc:89: provider_->SetVideoPlayout(ssrc_, true,
track_->GetSink());
As mentioned in the other CL, being able to just pass in track_ here would be
nicer.

https://codereview.webrtc.org/1664773002/diff/20001/talk/app/webrtc/videotrac...
File talk/app/webrtc/videotrack.cc (right):

https://codereview.webrtc.org/1664773002/diff/20001/talk/app/webrtc/videotrac...
talk/app/webrtc/videotrack.cc:63: return &renderers_;
As mentioned in the other CL, you could just implement OnFrame here by calling
renderers_->OnFrame()