Add read support of RtpStreamId/RepairedRtpStreamId header extensions.

webrtc/common_types.cc

 /*
  *  Copyright (c) 2012 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.
  */
 
 #include "webrtc/common_types.h"
 
+#include <string.h>
 #include <limits>
-#include <string.h>
+#include <type_traits>
 
 #include "webrtc/base/checks.h"
 #include "webrtc/base/stringutils.h"
 
 namespace webrtc {
 
 StreamDataCounters::StreamDataCounters() : first_packet_time_ms(-1) {}
 
+constexpr size_t StreamId::kMaxSize;
+
+void StreamId::Set(const void* data, size_t size) {
+  RTC_DCHECK_LE(size, kMaxSize);

[nisse-webrtc, 2017/04/10 12:25:00]

I'd like an RTC_DCHECK(memchr(data, 0, size) == nullptr) here, to make sure
conversion makes sense. 

Or if you don't want to repeat this as a DCHECK here and a production check in
the parsing functions, you could let this method have a bool return value to
indicate success.

[danilchap, 2017/04/10 13:27:05]

Prefer to find those value acceptible. Added comment instead.

One of the reason - to make tests simpler and more natural: currently StreamId
created there from const char[N] where N would be string length including
terminating zero, but I still treat them as one-char shorter strings.

[nisse-webrtc, 2017/04/11 07:59:24]

Is that that the ArrayView constructor? I think that's a bit counter-intuitive
when it comes to string literals, I would have expected ArrayView<char>("foo")
to construct an array of length 3. 

If that can't be fixed, could tests go via a std::string instead, to get the
intended length?

[danilchap, 2017/04/11 10:05:26]

std::string_view would make more sense for StreamId, but that is available only
in c++17
When it is same complexity, I choose std::string in tests, but when testing
parsing RTPHeader, then comparing StreamId looks cleaner

+  memcpy(value_, data, size);
+  if (size < kMaxSize)
+    value_[size] = 0;
+}
+
+static_assert(std::is_trivially_destructible<StreamId>::value, "");
+
 RTPHeaderExtension::RTPHeaderExtension()
     : hasTransmissionTimeOffset(false),
       transmissionTimeOffset(0),
       hasAbsoluteSendTime(false),
       absoluteSendTime(0),
       hasTransportSequenceNumber(false),
       transportSequenceNumber(0),
       hasAudioLevel(false),
       voiceActivity(false),
       audioLevel(0),
@@ skipping 145 matching lines @@
 // Get the sum of all the temporal layer for a specific spatial layer.
 uint32_t BitrateAllocation::GetSpatialLayerSum(size_t spatial_index) const {
   RTC_CHECK_LT(spatial_index, kMaxSpatialLayers);
   uint32_t sum = 0;
   for (int i = 0; i < kMaxTemporalStreams; ++i)
     sum += bitrates_[spatial_index][i];
   return sum;
 }
 
 }  // namespace webrtc