Explicitly suppress scrolling for wheel events that will trigger zooming

content/browser/web_contents/web_contents_impl.cc

Index: content/browser/web_contents/web_contents_impl.cc
diff --git a/content/browser/web_contents/web_contents_impl.cc b/content/browser/web_contents/web_contents_impl.cc
index e6f8713fdb8985d0631ca19b12868461d106921c..b80f85230d4dfa3d773a6c91c6446a407116f246 100644
--- a/content/browser/web_contents/web_contents_impl.cc
+++ b/content/browser/web_contents/web_contents_impl.cc
@@ -1331,12 +1331,9 @@ bool WebContentsImpl::HandleWheelEvent(
   //   -if a user starts an inertial scroll, let's go, and presses control
   //      (i.e. control+tab) then the OS's buffered scroll events will come in
   //      with control key set which isn't what the user wants
-  if (delegate_ &&
-      event.wheelTicksY &&
-      (event.modifiers & blink::WebInputEvent::ControlKey) &&
+  if (delegate_ && event.wheelTicksY && event.suppressScroll) {

[Rick Byers, 2014/11/27 17:31:32]

I think we should probably keep the ControlKey check here.  I know this
duplicates some logic, but it makes a bit more sense for WebContents to say that
we'll zoom IF Control is pressed AND the platform has chosen to suppress
scrolling for this event.  It's not unreasonable that someone may choose to
suppress scrolling in some other non-Ctrl case, and it would be odd if that
triggered zooming on it's own.

[lanwei, 2014/12/02 06:28:54]

Done.

       // Avoid adjusting the zoom in response to two-finger-scrolling touchpad
       // gestures, which are regrettably easy to trigger accidentally.

[tdresser, 2014/11/27 15:06:08]

Indentation is wrong here, but I think this comment can be removed completely.

[Rick Byers, 2014/11/27 17:31:32]

The comment is still useful but now belongs in MakeWebMouseWheelEvent, so just
move it there.

[lanwei, 2014/12/02 06:28:54]

Done.

[lanwei, 2014/12/02 06:28:54]

Done.

-      !event.hasPreciseScrollingDeltas) {
     delegate_->ContentsZoomChange(event.wheelTicksY > 0);
     return true;
   }