gesture recognizer: Expose bounding-box for the points in a gesture-event.

gesture recognizer: Expose bounding-box for the points in a gesture-event.

The bounding box is the smallest axis-aligned rectangle that encloses all the
bounding-boxes of all the touch-points in the gesture.

BUG=none
TEST=existing aura_unittests

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=147975

[tdanderson, 2012-07-23 20:21:02]

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_seque...
File ui/base/gestures/gesture_sequence.cc (right):

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_seque...
ui/base/gestures/gesture_sequence.cc:490: // TODO(sad): Recreating the bounding
box at every touch-event is not very
Does sad == sadrul ?

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_seque...
ui/base/gestures/gesture_sequence.cc:493: for (int i = 0; i < kMaxGesturePoints;
++i) {
Since this loops through all of the gesture points, wouldn't the bounding box
for a two finger tap be useless for fuzzing purposes? The bounding box would
contain the enclosing rects for both fingers, but we only want to fuzz using the
enclosing rect for the first finger that touched the screen.

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_seque...
ui/base/gestures/gesture_sequence.cc:603:
GestureEventDetails(ui::ET_GESTURE_TAP, 0, 0),
Since you are storing the bounding box in the details, it seems there would be
no way for WebKit to access this information. In PlatformGestureEventBuilder, we
need to pull the bounding box dimensions from a WebGestureEvent in order to set
the m_area field for a GestureTap (and soon, a GestureLongPress as well)

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_types.h
File ui/base/gestures/gesture_types.h (right):

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_types...
ui/base/gestures/gesture_types.h:113: // touch-points in the gesture.
Instead shouldn't this comment be "contains all the enclosing rectangles of the
touch-points in the gesture" ?

[sadrul, 2012-07-23 20:34:41]

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_seque...
File ui/base/gestures/gesture_sequence.cc (right):

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_seque...
ui/base/gestures/gesture_sequence.cc:490: // TODO(sad): Recreating the bounding
box at every touch-event is not very
On 2012/07/23 20:21:02, tdanderson wrote:
> Does sad == sadrul ?

Sadly, yes.

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_seque...
ui/base/gestures/gesture_sequence.cc:493: for (int i = 0; i < kMaxGesturePoints;
++i) {
On 2012/07/23 20:21:02, tdanderson wrote:
> Since this loops through all of the gesture points, wouldn't the bounding box
> for a two finger tap be useless for fuzzing purposes? The bounding box would
> contain the enclosing rects for both fingers, but we only want to fuzz using
the
> enclosing rect for the first finger that touched the screen.

We need to expose details about individual fingers at some point:
crbug.com/137679. This will be taken care of from that.

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_seque...
ui/base/gestures/gesture_sequence.cc:603:
GestureEventDetails(ui::ET_GESTURE_TAP, 0, 0),
On 2012/07/23 20:21:02, tdanderson wrote:
> Since you are storing the bounding box in the details, it seems there would be
> no way for WebKit to access this information. In PlatformGestureEventBuilder,
we
> need to pull the bounding box dimensions from a WebGestureEvent in order to
set
> the m_area field for a GestureTap (and soon, a GestureLongPress as well)

Good catch! I have added a workaround for this for TAPs. I think WebGestureEvent
should be updated to include this information.

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_types.h
File ui/base/gestures/gesture_types.h (right):

http://codereview.chromium.org/10808083/diff/1/ui/base/gestures/gesture_types...
ui/base/gestures/gesture_types.h:113: // touch-points in the gesture.
On 2012/07/23 20:21:02, tdanderson wrote:
> Instead shouldn't this comment be "contains all the enclosing rectangles of
the
> touch-points in the gesture" ?

Done.

[sky, 2012-07-23 22:33:47]

LGTM

http://codereview.chromium.org/10808083/diff/9015/content/browser/renderer_ho...
File content/browser/renderer_host/render_widget_host_view_win.cc (right):

http://codereview.chromium.org/10808083/diff/9015/content/browser/renderer_ho...
content/browser/renderer_host/render_widget_host_view_win.cc:1298: 
nit: remove newline.

[tdanderson, 2012-07-23 22:38:07]

lgtm

[sadrul, 2012-07-23 22:40:55]

http://codereview.chromium.org/10808083/diff/9015/content/browser/renderer_ho...
File content/browser/renderer_host/render_widget_host_view_win.cc (right):

http://codereview.chromium.org/10808083/diff/9015/content/browser/renderer_ho...
content/browser/renderer_host/render_widget_host_view_win.cc:1298: 
On 2012/07/23 22:33:47, sky wrote:
> nit: remove newline.

Done.