Add function for getting supported H264 level from max resolution and fps

webrtc/common_video/h264/profile_level_id.h

 /*
  *  Copyright (c) 2016 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 43 matching lines @@
       : profile(profile), level(level) {}
   Profile profile;
   Level level;
 };
 
 // Parse profile level id that is represented as a string of 3 hex bytes.
 // Nothing will be returned if the string is not a recognized H264
 // profile level id.
 rtc::Optional<ProfileLevelId> ParseProfileLevelId(const char* str);
 
+// Given the maximum supported frame size (in pixels) and fps, return the
+// highest H264 level the encoder/decoder is guaranteed to support, or
+// return no level if not even level 1 is supported.

[hta-webrtc, 2016/11/03 14:04:32]

This can be made clearer.
Something like:

Given that a decoder supports up to a given frame size (in pixels) at up to a
given number of frames per second, return the highest H.264 level where it can
guarantee that it will be able to support all valid encoded streams that are
within that level.

For an encoder, I think the logic will be different - you would want the level
that would guarantee that if the other end agreed to handle it, you could drive
the encoder to maximum performance without overrunning the decoder.

Thus, if 1280*720@30 is slightly more than Level 3.1 but less than level 4.1:

SupportedDecoderLevel(1280*720, 30) = Level3.1
SupportedEncoderLevel(1280*720, 30) = Level4.1

[magjed_webrtc, 2016/11/03 14:53:39]

You are right about this and I thought about adding two versions when I
implemented this named LowerBoundLevel/UpperBoundLevel or
SupportedLevel/RequiredLevel. However, I don't think the SupportedEncoderLevel
variant will ever be needed since we should base our offered level solely on
decode level. We offer our decode level in SDP, and we limit the input to the
encoder based on remote decode level.

[magjed_webrtc, 2016/11/03 14:54:52]

I updated the comment according to your suggestion though.

+rtc::Optional<Level> SupportedLevel(int max_frame_pixel_count, int max_fps);
+
 // Returns canonical string representation as three hex bytes of the profile
 // level id, or returns empty string for invalid profile level id.
 std::string ProfileLevelIdToString(const ProfileLevelId& profile_level_id);
 
 }  // namespace H264
 }  // namespace webrtc
 
 #endif  // WEBRTC_COMMON_VIDEO_H264_PROFILE_LEVEL_ID_H_