Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

BUG=webrtc:6402, webrtc:6337
Committed: https://crrev.com/1e45cc6ee0b368f7cb8b90acc7fe4c1699bb4288
Cr-Commit-Position: refs/heads/master@{#14826}

[magjed_webrtc, 2016-10-26 14:39:44]

Patchset #1 (id:1) has been deleted

[magjed_webrtc, 2016-10-26 14:39:57]

Description was changed from

==========
WIP: Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

BUG=webrtc:6402
==========

to

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

BUG=webrtc:6402
==========

[magjed_webrtc, 2016-10-26 14:40:05]

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

[commit-bot: I haz the power, 2016-10-26 14:40:07]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[magjed_webrtc, 2016-10-26 14:40:35]

The CQ bit was unchecked by magjed@webrtc.org

[magjed_webrtc, 2016-10-26 14:52:07]

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

[magjed_webrtc, 2016-10-26 14:52:14]

Patchset #1 (id:20001) has been deleted

[commit-bot: I haz the power, 2016-10-26 14:52:17]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[magjed_webrtc, 2016-10-26 15:08:15]

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

[commit-bot: I haz the power, 2016-10-26 15:08:24]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[magjed_webrtc, 2016-10-26 15:08:56]

Patchset #1 (id:40001) has been deleted

[magjed_webrtc, 2016-10-26 15:26:26]

Description was changed from

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

BUG=webrtc:6402
==========

to

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
Once all external clients have updated their code, the plan is to remove
the old functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

Another change in this CL is that webrtcvideoengine2.cc is updated to
compare cricket::VideoCodecs using
'bool VideoCodec::Matches(const VideoCodec& codec) const' instead of
just comparing codec names. The purpose is to prepare for H264 profile
matching in VideoCodec::Matches so that webrtcvideoengine2.cc
distinguishs different H264 profiles.

BUG=webrtc:6402
==========

[magjed_webrtc, 2016-10-26 15:26:27]

magjed@webrtc.org changed reviewers:
+ hta@webrtc.org, kthelgason@webrtc.org

[magjed_webrtc, 2016-10-26 15:28:19]

Description was changed from

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
Once all external clients have updated their code, the plan is to remove
the old functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

Another change in this CL is that webrtcvideoengine2.cc is updated to
compare cricket::VideoCodecs using
'bool VideoCodec::Matches(const VideoCodec& codec) const' instead of
just comparing codec names. The purpose is to prepare for H264 profile
matching in VideoCodec::Matches so that webrtcvideoengine2.cc
distinguishs different H264 profiles.

BUG=webrtc:6402
==========

to

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

Another change in this CL is that webrtcvideoengine2.cc is updated to
compare cricket::VideoCodecs using
'bool VideoCodec::Matches(const VideoCodec& codec) const' instead of
just comparing codec names. The purpose is to prepare for H264 profile
matching in VideoCodec::Matches so that webrtcvideoengine2.cc
distinguishs different H264 profiles.

BUG=webrtc:6402
==========

[magjed_webrtc, 2016-10-26 15:29:23]

Description was changed from

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

Another change in this CL is that webrtcvideoengine2.cc is updated to
compare cricket::VideoCodecs using
'bool VideoCodec::Matches(const VideoCodec& codec) const' instead of
just comparing codec names. The purpose is to prepare for H264 profile
matching in VideoCodec::Matches so that webrtcvideoengine2.cc
distinguishs different H264 profiles.

BUG=webrtc:6402
==========

to

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

Another change in this CL is that webrtcvideoengine2.cc is updated to
compare cricket::VideoCodecs using
'bool VideoCodec::Matches(const VideoCodec& codec) const' instead of
just comparing codec names. The purpose is to prepare for H264 profile
matching in VideoCodec::Matches so that webrtcvideoengine2.cc can
distinguishs between different H264 profiles.

BUG=webrtc:6402
==========

[magjed_webrtc, 2016-10-26 15:29:32]

Please take a look.

[commit-bot: I haz the power, 2016-10-26 15:35:39]

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

[commit-bot: I haz the power, 2016-10-26 15:35:40]

Dry run: This issue passed the CQ dry run.

[kthelgason, 2016-10-28 08:26:54]

some small suggestions, otherwise lgtm

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
File webrtc/media/engine/webrtcvideoengine2.cc (right):

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
webrtc/media/engine/webrtcvideoengine2.cc:658: }
This code exists in a few places. We could use std::any_of here to avoid
manually writing the loop, and to make the intent more clear.

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
webrtc/media/engine/webrtcvideoengine2.cc:730: }
The inner loop here is identical to a previous one in this class. Perhaps we can
extract that out into a function, perhaps "bool IsCodecSupported(VideoCodec
codec)" and reuse that here. Then this whole thing could be a single
std::copy_if call.

[magjed_webrtc, 2016-10-28 10:06:50]

Patchset #2 (id:80001) has been deleted

[magjed_webrtc, 2016-10-28 10:27:31]

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
File webrtc/media/engine/webrtcvideoengine2.cc (right):

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
webrtc/media/engine/webrtcvideoengine2.cc:658: }
On 2016/10/28 08:26:54, kthelgason wrote:
> This code exists in a few places. We could use std::any_of here to avoid
> manually writing the loop, and to make the intent more clear.

Thanks! Much better to extract it to a function. I went with std::any_of, but it
might not be more clear than a regular for-loop. It's this:
return std::any_of(supported_codecs.begin(), supported_codecs.end(),
                   [&codec](const VideoCodec& supported_codec) -> bool {
                     return supported_codec.Matches(codec);
                   });
vs this:
for (const VideoCodec& supported_codec : supported_codecs) {
  if (supported_codec.Matches(codec))
    return true;
}
return false;

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
webrtc/media/engine/webrtcvideoengine2.cc:730: }
On 2016/10/28 08:26:54, kthelgason wrote:
> The inner loop here is identical to a previous one in this class. Perhaps we
can
> extract that out into a function, perhaps "bool IsCodecSupported(VideoCodec
> codec)" and reuse that here. Then this whole thing could be a single
> std::copy_if call.

Done, the inner loop is now extracted out into a IsCodecSupported helper
function. I didn't go with std::copy_if, because I think it adds more noise than
clarity. It's this:
  std::copy_if(
      mapped_codecs.begin(), mapped_codecs.end(),
      std::back_inserter(filtered_codecs),
      [&supported_codecs](const VideoCodecSettings& mapped_codec) -> bool {
        return IsCodecSupported(supported_codecs, mapped_codec.codec);
      });
vs
  for (const VideoCodecSettings& mapped_codec : mapped_codecs) {
    if (IsCodecSupported(supported_codecs, mapped_codec.codec))
      filtered_codecs.push_back(mapped_codec);
  }

[hta-webrtc, 2016-10-28 11:13:11]

lgtm

I assume that all tests pass and that it works with Chrome... we should
eventually delete all code that has payload type within the codec, because we
should be capable of supporting scenarios with more than one payload type
pointing to the same codec (it's a map, not an equality) - but that's not part
of this CL.

[magjed_webrtc, 2016-10-28 12:54:28]

Description was changed from

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

Another change in this CL is that webrtcvideoengine2.cc is updated to
compare cricket::VideoCodecs using
'bool VideoCodec::Matches(const VideoCodec& codec) const' instead of
just comparing codec names. The purpose is to prepare for H264 profile
matching in VideoCodec::Matches so that webrtcvideoengine2.cc can
distinguishs between different H264 profiles.

BUG=webrtc:6402
==========

to

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

Another change in this CL is that webrtcvideoengine2.cc is updated to
compare cricket::VideoCodecs using
'bool VideoCodec::Matches(const VideoCodec& codec) const' instead of
just comparing codec names. The purpose is to prepare for H264 profile
matching in VideoCodec::Matches so that webrtcvideoengine2.cc can
distinguishs between different H264 profiles.

BUG=webrtc:6402,webrtc:6337
==========

[kthelgason, 2016-10-28 13:19:45]

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
File webrtc/media/engine/webrtcvideoengine2.cc (right):

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
webrtc/media/engine/webrtcvideoengine2.cc:730: }
On 2016/10/28 10:27:31, magjed_webrtc wrote:
> On 2016/10/28 08:26:54, kthelgason wrote:
> > The inner loop here is identical to a previous one in this class. Perhaps we
> can
> > extract that out into a function, perhaps "bool IsCodecSupported(VideoCodec
> > codec)" and reuse that here. Then this whole thing could be a single
> > std::copy_if call.
> 
> Done, the inner loop is now extracted out into a IsCodecSupported helper
> function. I didn't go with std::copy_if, because I think it adds more noise
than
> clarity.

I agree with your assessment that this adds more noise than clarity. I had
imagined something like:
std::copy_if(mapped_codecs.begin(), mapped_codecs.end(),
    std::back_inserter(filtered_codecs), &IsCodecSupported)

but this is not javascript, and that obviously does not work :)

[magjed_webrtc, 2016-10-28 13:23:18]

Description was changed from

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

Another change in this CL is that webrtcvideoengine2.cc is updated to
compare cricket::VideoCodecs using
'bool VideoCodec::Matches(const VideoCodec& codec) const' instead of
just comparing codec names. The purpose is to prepare for H264 profile
matching in VideoCodec::Matches so that webrtcvideoengine2.cc can
distinguishs between different H264 profiles.

BUG=webrtc:6402,webrtc:6337
==========

to

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

BUG=webrtc:6402,webrtc:6337
==========

[magjed_webrtc, 2016-10-28 13:30:17]

I had to revert comparing codecs with VideoCodec::Matches and compare using
codec names for now. The problem is that VideoCodec::Matches is looking at the
payload_id which complicates things. I need to fix this in a future CL so that
different H264 profiles can be added.

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
File webrtc/media/engine/webrtcvideoengine2.cc (right):

https://codereview.webrtc.org/2449993003/diff/60001/webrtc/media/engine/webrt...
webrtc/media/engine/webrtcvideoengine2.cc:730: }
On 2016/10/28 13:19:45, kthelgason wrote:
> On 2016/10/28 10:27:31, magjed_webrtc wrote:
> > On 2016/10/28 08:26:54, kthelgason wrote:
> > > The inner loop here is identical to a previous one in this class. Perhaps
we
> > can
> > > extract that out into a function, perhaps "bool
IsCodecSupported(VideoCodec
> > > codec)" and reuse that here. Then this whole thing could be a single
> > > std::copy_if call.
> > 
> > Done, the inner loop is now extracted out into a IsCodecSupported helper
> > function. I didn't go with std::copy_if, because I think it adds more noise
> than
> > clarity.
> 
> I agree with your assessment that this adds more noise than clarity. I had
> imagined something like:
> std::copy_if(mapped_codecs.begin(), mapped_codecs.end(),
>     std::back_inserter(filtered_codecs), &IsCodecSupported)
> 
> but this is not javascript, and that obviously does not work :)
> 
> 

Yeah, it's close to working with:
std::copy_if(mapped_codecs.begin(), mapped_codecs.end(),
    std::back_inserter(filtered_codecs),
    std::bind(&IsCodecSupported, supported_codecs));
Not quite though, because:
 1. std::bind is not yet an allowed C++11 feature in WebRTC.
 2. mapped_codec is of type VideoCodecSettings and not VideoCodec, so I need to
do mapped_codec.codec which essentially requires a lambda in C++.
Haskell has much nicer and cleaner glue functions :)

[magjed_webrtc, 2016-10-28 14:42:07]

The CQ bit was checked by magjed@webrtc.org

[magjed_webrtc, 2016-10-28 14:42:07]

The patchset sent to the CQ was uploaded after l-g-t-m from
kthelgason@webrtc.org, hta@webrtc.org
Link to the patchset: https://codereview.webrtc.org/2449993003/#ps120001 (title:
"Compare only codec names, not VideoCodec::Matches")

[commit-bot: I haz the power, 2016-10-28 14:42:14]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[commit-bot: I haz the power, 2016-10-28 14:43:49]

Description was changed from

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

BUG=webrtc:6402,webrtc:6337
==========

to

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

BUG=webrtc:6402,webrtc:6337
==========

[commit-bot: I haz the power, 2016-10-28 14:43:51]

Committed patchset #3 (id:120001)

[commit-bot: I haz the power, 2016-10-28 14:44:00]

Description was changed from

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

BUG=webrtc:6402,webrtc:6337
==========

to

==========
Replace WebRtcVideoEncoderFactory::VideoCodec with cricket::VideoCodec

This CL introduces two new functions to the WebRtcVideoEncoderFactory
interface based on cricket::VideoFormat instead of
WebRtcVideoEncoderFactory::VideoCodec. The functions are:
WebRtcVideoEncoderFactory::CreateVideoEncoder() and
WebRtcVideoEncoderFactory::supported_codecs(). In order to make a smooth
transition to the new interface, the old functions are kept, and default
implementations are provided for both the old and new functions so that
external clients can switch from the old to the new functions in peace.
The default implementations will just convert between
cricket::VideoFormat and WebRtcVideoEncoderFactory::VideoCodec. Once all
external clients have updated their code, the plan is to remove the old
functions and all default implementations to make
WebRtcVideoEncoderFactory a pure interface again.

BUG=webrtc:6402,webrtc:6337

Committed: https://crrev.com/1e45cc6ee0b368f7cb8b90acc7fe4c1699bb4288
Cr-Commit-Position: refs/heads/master@{#14826}
==========

[commit-bot: I haz the power, 2016-10-28 14:44:01]

Patchset 3 (id:??) landed as
https://crrev.com/1e45cc6ee0b368f7cb8b90acc7fe4c1699bb4288
Cr-Commit-Position: refs/heads/master@{#14826}