Adding audio network adaptor to AudioEncoderOpus.

webrtc/modules/audio_coding/codecs/audio_encoder.h

 /*
  *  Copyright (c) 2014 The WebRTC project authors. All Rights Reserved.
  *
  *  Use of this source code is governed by a BSD-style license
  *  that can be found in the LICENSE file in the root of the source
  *  tree. An additional intellectual property rights grant can be found
  *  in the file PATENTS.  All contributing project authors may
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
@@ skipping 144 matching lines @@
 
   // Causes this encoder to let go of any other encoders it contains, and
   // returns a pointer to an array where they are stored (which is required to
   // live as long as this encoder). Unless the returned array is empty, you may
   // not call any methods on this encoder afterwards, except for the
   // destructor. The default implementation just returns an empty array.
   // NOTE: This method is subject to change. Do not call or override it.
   virtual rtc::ArrayView<std::unique_ptr<AudioEncoder>>
   ReclaimContainedEncoders();
 
+  // Enable/disable audio network adaptor. Return whether setting is succeeded.

[kwiberg-webrtc, 2016/09/27 09:35:54]

Is it supposed to be default enabled or default disabled?

Also, verbs should be third person singular present tense (a.k.a., tack on an
"s" at the end). And the grammar in the second sentence is a bit off. Maybe

  Returns true if successful.

?

[minyue-webrtc, 2016/09/27 16:02:32]

As FEC, the default value is given by VoE. A difference between FEC and audio
network adaptor is that FEC can be parsed from SDP, while audio network adaptor
cannot yet be set through SDP. I am not sure how to handle that. But I think it
is good for now to let VoE decide it, through AudioOption, specifically.

and will improve wording here.

[kwiberg-webrtc, 2016/09/28 08:24:09]

I think it makes more sense for it to be default disabled; someone would then
have to enable it for new encoder objects. This will probably become obvious if
this method starts requiring an ANA configuration when you want to switch ANA
on.

[minyue-webrtc, 2016/09/29 15:34:24]

Now it is default disabled since |audio_network_adaptor_| is initialized as
nullptr.

+  virtual bool SetAudioNetworkAdaptor(bool enable);
+
+  // Provides uplink bandwidth to this encoder to allow it to adapt.
+  virtual void OnReceivedUplinkBandwidth(int uplink_bandwidth_bps);
+
+  // Provides uplink packet loss fraction to this encoder to allow it to adapt.
+  virtual void OnReceivedUplinkPacketLossFraction(
+      float uplink_packet_loss_fraction);
+
+  // Provides target audio bitrate to this encoder to allow it to adapt.
+  virtual void OnReceivedTargetAudioBitrate(int target_audio_bitrate_bps);
+
+  // Provides RTT to this encoder to allow it to adapt.
+  virtual void OnReceivedRtt(int rtt_ms);
+
+  // To allow encoder to adapt its frame length, it must be provided the frame
+  // length range that receives can accept.
+  virtual void SetReceiverFrameLengthRange(int min_frame_length_ms,
+                                           int max_frame_length_ms);

[kwiberg-webrtc, 2016/09/27 09:35:54]

How is the encoder supposed to behave prior to the first call to this function?

Also, "receivers".

[minyue-webrtc, 2016/09/27 16:02:33]

Audio network adaptor is configured so that it only uses initial frame length
because receiver frame length range is given.

+
  protected:
   // Subclasses implement this to perform the actual encoding. Called by
   // Encode().
   virtual EncodedInfo EncodeImpl(uint32_t rtp_timestamp,
                                  rtc::ArrayView<const int16_t> audio,
                                  rtc::Buffer* encoded) = 0;
 };
 }  // namespace webrtc
 #endif  // WEBRTC_MODULES_AUDIO_CODING_CODECS_AUDIO_ENCODER_H_