Re-land: bindings: Use MessageEventInit for MessageEvent constructor

Re-land: bindings: Use MessageEventInit for MessageEvent constructor

The previous CL was reverted because it broke browser tests.
We should make sure that MessageEvent.port is non-null when it is created
via WebDOMMessageEvent::initMessageEvent() because content layer uses it
to implement postMessage().

Original description:

This CL also includes some behavior changes to improve spec conformance.
- MessageEvent.ports becomes nullable. If a given MessageEventInit doesn't
  have a value for |ports|, MessageEvent.ports is initialized with null.
  Note that when Blink creates MessageEvent as a result of postMessage(),
  |ports| should be an empty array when there is no ports.
- MessageEvent.data is initialized to null when no |data| is specified.

BUG=433179, 492798
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=196118

[bashi, 2015-05-29 02:05:41]

bashi@chromium.org changed reviewers:
+ haraken@chromium.org

[bashi, 2015-05-29 02:05:42]

PTAL?

https://codereview.chromium.org/1150183007/diff/1/Source/web/WebDOMMessageEve...
File Source/web/WebDOMMessageEvent.cpp (right):

https://codereview.chromium.org/1150183007/diff/1/Source/web/WebDOMMessageEve...
Source/web/WebDOMMessageEvent.cpp:60: if (!ports.get())
This is the only diff from the previous CL.

[haraken, 2015-05-29 02:12:06]

LGTM

[haraken, 2015-05-29 02:12:57]

https://codereview.chromium.org/1150183007/diff/1/Source/web/WebDOMMessageEve...
File Source/web/WebDOMMessageEvent.cpp (right):

https://codereview.chromium.org/1150183007/diff/1/Source/web/WebDOMMessageEve...
Source/web/WebDOMMessageEvent.cpp:60: if (!ports.get())
On 2015/05/29 02:05:42, bashi1 wrote:
> This is the only diff from the previous CL.

.get() won't be needed.

[bashi, 2015-05-29 02:40:30]

https://codereview.chromium.org/1150183007/diff/1/Source/web/WebDOMMessageEve...
File Source/web/WebDOMMessageEvent.cpp (right):

https://codereview.chromium.org/1150183007/diff/1/Source/web/WebDOMMessageEve...
Source/web/WebDOMMessageEvent.cpp:60: if (!ports.get())
On 2015/05/29 02:12:57, haraken wrote:
> On 2015/05/29 02:05:42, bashi1 wrote:
> > This is the only diff from the previous CL.
> 
> .get() won't be needed.

Done.

[bashi, 2015-05-29 02:40:35]

The CQ bit was checked by bashi@chromium.org

[bashi, 2015-05-29 02:40:35]

The patchset sent to the CQ was uploaded after l-g-t-m from haraken@chromium.org
Link to the patchset: https://codereview.chromium.org/1150183007/#ps20001
(title: " ")

[commit-bot: I haz the power, 2015-05-29 02:40:44]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1150183007/20001

[commit-bot: I haz the power, 2015-05-29 03:53:00]

Committed patchset #2 (id:20001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=196118