Motown: Improvements to packet definition

services/media/framework_ffmpeg/ffmpeg_audio_decoder.cc

Index: services/media/framework_ffmpeg/ffmpeg_audio_decoder.cc
diff --git a/services/media/framework_ffmpeg/ffmpeg_audio_decoder.cc b/services/media/framework_ffmpeg/ffmpeg_audio_decoder.cc
index 11e2679aba266dca51bdb7c8729952a46f97a8a6..45e1adb516cbf951776d76202df4987b49d0204d 100644
--- a/services/media/framework_ffmpeg/ffmpeg_audio_decoder.cc
+++ b/services/media/framework_ffmpeg/ffmpeg_audio_decoder.cc
@@ -28,7 +28,7 @@ FfmpegAudioDecoder::~FfmpegAudioDecoder() {}
 
 void FfmpegAudioDecoder::Flush() {
   FfmpegDecoderBase::Flush();
-  next_presentation_time_= Packet::kUnknownPresentationTime;
+  next_pts_= Packet::kUnknownPts;
 }
 
 int FfmpegAudioDecoder::Decode(
@@ -41,11 +41,11 @@ int FfmpegAudioDecoder::Decode(
   DCHECK(context());
   DCHECK(av_frame_ptr);
 
-  if (next_presentation_time_ == Packet::kUnknownPresentationTime) {
+  if (next_pts_ == Packet::kUnknownPts) {
     if (av_packet.pts == AV_NOPTS_VALUE) {
-      next_presentation_time_ = 0;
+      next_pts_ = 0;

[johngro, 2016/03/21 19:47:56]

I'm still a bit confused by this logic; eg, why is the ffmpeg decoder wrapper
making timestamp interpolation decisions instead of just putting the timestamps
produced by ffmpeg onto the payloads during CreateOutputPacket?

No action is being requested here.  Just noting that it is something we need to
sit down and discuss at some point.

[dalesat, 2016/03/21 22:30:59]

Acknowledged.

     } else {
-      next_presentation_time_ = av_packet.pts;
+      next_pts_ = av_packet.pts;
     }
   }
 
@@ -74,10 +74,10 @@ PacketPtr FfmpegAudioDecoder::CreateOutputPacket(
     PayloadAllocator* allocator) {
   DCHECK(allocator);
 
-  int64_t presentation_time = av_frame.pts;
-  if (presentation_time == AV_NOPTS_VALUE) {
-    presentation_time = next_presentation_time_;
-    next_presentation_time_ += av_frame.nb_samples;
+  int64_t pts = av_frame.pts;
+  if (pts == AV_NOPTS_VALUE) {
+    pts = next_pts_;
+    next_pts_ += av_frame.nb_samples;
   }
 
   uint64_t payload_size;
@@ -110,8 +110,7 @@ PacketPtr FfmpegAudioDecoder::CreateOutputPacket(
   }
 
   return Packet::Create(
-      presentation_time,
-      av_frame.nb_samples,
+      pts,
       false, // The base class is responsible for end-of-stream.
       payload_size,
       payload_buffer,
@@ -119,7 +118,7 @@ PacketPtr FfmpegAudioDecoder::CreateOutputPacket(
 }
 
 PacketPtr FfmpegAudioDecoder::CreateOutputEndOfStreamPacket() {
-  return Packet::CreateEndOfStream(next_presentation_time_);
+  return Packet::CreateEndOfStream(next_pts_);
 }
 
 int FfmpegAudioDecoder::AllocateBufferForAvFrame(