Mpeg2 TS parser: Es parsing using an ES byte queue.

media/formats/mp2t/es_parser.h

Index: media/formats/mp2t/es_parser.h
diff --git a/media/formats/mp2t/es_parser.h b/media/formats/mp2t/es_parser.h
index a1959c20a341ca6c7752c796e62faf561b5bce27..b22fe5d16983c88610c2f6d9459c31c66d08a97b 100644
--- a/media/formats/mp2t/es_parser.h
+++ b/media/formats/mp2t/es_parser.h
@@ -5,6 +5,9 @@
 #ifndef MEDIA_FORMATS_MP2T_ES_PARSER_H_
 #define MEDIA_FORMATS_MP2T_ES_PARSER_H_
 
+#include <list>
+#include <utility>
+
 #include "base/basictypes.h"
 #include "base/callback.h"
 #include "base/memory/ref_counted.h"
@@ -14,6 +17,7 @@
 
 namespace media {
 
+class OffsetByteQueue;
 class StreamParserBuffer;
 
 namespace mp2t {
@@ -22,20 +26,62 @@ class MEDIA_EXPORT EsParser {
  public:
   typedef base::Callback<void(scoped_refptr<StreamParserBuffer>)> EmitBufferCB;
 
-  EsParser() {}
-  virtual ~EsParser() {}
+  EsParser();
+  virtual ~EsParser();
 
   // ES parsing.
   // Should use kNoTimestamp when a timestamp is not valid.
-  virtual bool Parse(const uint8* buf, int size,
-                     base::TimeDelta pts,
-                     DecodeTimestamp dts) = 0;
+  bool Parse(const uint8* buf, int size,
+             base::TimeDelta pts,
+             DecodeTimestamp dts);
 
   // Flush any pending buffer.
   virtual void Flush() = 0;
 
   // Reset the state of the ES parser.
-  virtual void Reset() = 0;
+  void Reset();
+
+ protected:
+  struct TimingDesc {
+    TimingDesc();
+    TimingDesc(DecodeTimestamp dts, base::TimeDelta pts);
+
+    DecodeTimestamp dts;
+    base::TimeDelta pts;
+  };
+
+  // Parse ES data from |es_queue_|.
+  // Return true when successful.
+  virtual bool ParseFromEsQueue() = 0;

[damienv1, 2014/08/27 16:37:21]

Maybe at some point, we should consider making EsParserAdts, EsParserH264 some
delegates instead of using derivation where it seems that composition might be a
better design.
EsParserH264 and EsParserAdts would both derive from EsParser::Delegate. The
delegate would consume data from an EsQueue (which includes both the data and
timing).

Could add a TODO for that purpose ?

[wolenetz, 2014/08/27 18:23:02]

I have no strong feeling here. For now, derivation sgtm.

+
+  // Reset the internal state of the ES parser.
+  virtual void ResetInternal() = 0;
+
+  // Get the timing descriptor with the largest byte count that is less or
+  // equal to |es_byte_count|.
+  // This timing descriptor and all the ones that come before (in stream order)
+  // are removed from list |timing_desc_list_|.
+  // If no timing descriptor is found, then the default TimingDesc is returned.
+  TimingDesc GetTimingDescriptor(int64 es_byte_count);
+
+  // Bytes of the ES stream that have not been emitted yet.
+  scoped_ptr<media::OffsetByteQueue> es_queue_;
+
+ private:
+  // Anchor some timing information into the ES queue.
+  // Here are two examples how this timing info is applied according to
+  // the MPEG-2 TS spec - ISO/IEC 13818:
+  // - "In the case of audio, if a PTS is present in PES packet header it shall
+  // refer to the first access unit commencing in the PES packet. An audio
+  // access unit commences in a PES packet if the first byte of the audio
+  // access unit is present in the PES packet."
+  // - "For AVC video streams conforming to one or more profiles defined
+  // in Annex A of Rec. ITU-T H.264 | ISO/IEC 14496-10 video, if a PTS is
+  // present in the PES packet header, it shall refer to the first AVC access
+  // unit that commences in this PES packet.
+  std::list<std::pair<int64, TimingDesc> > timing_desc_list_;
+
+  DISALLOW_COPY_AND_ASSIGN(EsParser);
 };
 
 }  // namespace mp2t