Video capture: Refactor GetBestMatchedFormat from Win to OS independent

media/video/capture/video_capture_device.h

Index: media/video/capture/video_capture_device.h
diff --git a/media/video/capture/video_capture_device.h b/media/video/capture/video_capture_device.h
index 187145ffa84596ed5ea6f7eeccb3d790af4dafc2..f2aae06b70e39a8b745d3bc917ec44904d14fd5a 100644
--- a/media/video/capture/video_capture_device.h
+++ b/media/video/capture/video_capture_device.h
@@ -12,6 +12,7 @@
 #ifndef MEDIA_VIDEO_CAPTURE_VIDEO_CAPTURE_DEVICE_H_
 #define MEDIA_VIDEO_CAPTURE_VIDEO_CAPTURE_DEVICE_H_
 
+#include <stdint.h>
 #include <list>
 #include <string>
 
@@ -324,6 +325,19 @@ class MEDIA_EXPORT VideoCaptureDevice {
   static const int kPowerLine60Hz = 60;
 };
 
+// Helper function to compare video capture formats when finding the most
+// closely matched video format given a requested video format.
+// Smaller difference means more closely matched formats.
+// Prioritizes according to this list:
+// 1. Closest matching height.
+// 2. Closest matching width.
+// 3. Closest matching frame rate.
+// 4. Best color format.
+// Nr 4. is not commutative, i.e. the order of the inputs to this
+// function matters.
+uint64_t DiffVideoCaptureFormat(const VideoCaptureFormat& requested,

[mcasas, 2014/10/13 08:18:45]

Why is this in video_capture_device.{h,cc} and not
in video_capture_types.{cc,h} ? AFAIK this is a 
convenience function for manipulating data types
defined in the latter files...

+                                const VideoCaptureFormat& supported);
+
 }  // namespace media
 
 #endif  // MEDIA_VIDEO_CAPTURE_VIDEO_CAPTURE_DEVICE_H_