Add DesktopFrame rotation functions

webrtc/modules/desktop_capture/desktop_frame_rotator.h

Index: webrtc/modules/desktop_capture/desktop_frame_rotator.h
diff --git a/webrtc/modules/desktop_capture/desktop_frame_rotator.h b/webrtc/modules/desktop_capture/desktop_frame_rotator.h
new file mode 100644
index 0000000000000000000000000000000000000000..3ac31dfa60eac744dac9c8983a765c306fdadb0b
--- /dev/null
+++ b/webrtc/modules/desktop_capture/desktop_frame_rotator.h
@@ -0,0 +1,47 @@
+/*
+ *  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.
+ */
+
+#ifndef WEBRTC_MODULES_DESKTOP_CAPTURE_DESKTOP_FRAME_ROTATOR_H_
+#define WEBRTC_MODULES_DESKTOP_CAPTURE_DESKTOP_FRAME_ROTATOR_H_
+
+#include "webrtc/modules/desktop_capture/desktop_frame.h"
+#include "webrtc/modules/desktop_capture/desktop_geometry.h"
+
+namespace webrtc {
+
+// Represents the rotation of a DesktopFrame.
+enum class Rotation {
+  CLOCK_WISE_0,
+  CLOCK_WISE_90,
+  CLOCK_WISE_180,
+  CLOCK_WISE_270,
+};
+
+// Rotates input DesktopFrame |src|, copies pixel in a rotated rectangle |rect|
+// into another DesktopFrame |target|, and adds |rect| into
+// |target|->mutable_updated_region(). Returns false if |target| is nullptr or
+// it does not cover the |rect|.
+bool CopyRotatedRectTo(const DesktopFrame& src,

[Sergey Ulanov, 2016/11/15 20:50:22]

s/src/source/

[Hzj_jie, 2016/11/16 01:18:33]

Done.

+                       DesktopRect rect,
+                       Rotation rotation,
+                       DesktopFrame* target);
+
+// Rotates input DesktopFrame |src|, copies pixel in a unrotated rectangle

[Sergey Ulanov, 2016/11/15 20:50:22]

I'm not sure I understand the difference between these two functions. Is is
correct interpretation that |rotation| always specifies src->target rotation,
and |rect| is rotated for the first function but not the second one, and that's
the only difference between them? It would be useful to say "|rect| is in |src|
coordinates" or "|rect| is in |target|" coordinates.
Do we need both functions? As alternative we could have one function to rotate a
rect and RotateRect() to help with the reverse operation.

[Hzj_jie, 2016/11/16 01:18:33]

Yes, exactly, |rect| is in |target| coordinate in the first function, but in
|source| coordinate in the second function.
The reason of two functions is because I cannot tell how DXGI APIs return the
updated region, i.e. in |source| coordinate or in |target| coordinate.
I have updated the comments.
It's a trade-off, we need to export two functions, either in current way, or a
CopyRotatedRectTo() and RotateRect(). Personally I prefer current solution,
since users do not need to think about whether a RotateRect() should be applied,
instead they only need to care about whether the |rect| is in |source|
coordinate or |target| coordinate.

+// |rect| into another DesktopFrame |target|, and adds rotated |rect| into
+// |target|->mutable_updated_region(). Returns false if |target| is nullptr or
+// it does not cover the |rect|.
+bool CopyUnrotatedRectTo(const DesktopFrame& src,
+                         DesktopRect rect,
+                         Rotation rotation,
+                         DesktopFrame* target);
+
+}  // namespace webrtc
+
+#endif  // WEBRTC_MODULES_DESKTOP_CAPTURE_DESKTOP_FRAME_ROTATOR_H_