Send lostpointercapture on touch capturing

Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches
by setting the implicit pointer capture right after the touch
starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.

More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553

BUG=596505
Committed: https://crrev.com/268b54a5f857d2d8a89199e3ada14b0afed2075e
Cr-Commit-Position: refs/heads/master@{#385219}

[Navid Zolghadr, 2016-03-29 19:44:17]

Description was changed from

==========
Send lostpointercapture on touch capturing

Sending lostpointercapture on touch released or cancel
rightaway after pointerup/cancel as we will never get
another pointer event with the same id to trigger
sending the lostpointercapture.

BUG=596505
==========

to

==========
Send lostpointercapture on touch capturing

Also sending lostpointercapture on touch released or
cancel rightaway after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.
Setting the pointer capture right after the touch starts.

More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

[Navid Zolghadr, 2016-03-29 19:44:42]

nzolghadr@chromium.org changed reviewers:
+ mustaq@chromium.org

[mustaq, 2016-03-30 21:21:55]

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/LayoutTe...
File
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html
(right):

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:42:
var eventList = ["pointerenter", "pointerleave", "pointerover", "pointerout",
"pointermove", "pointerdown", "pointerup", "pointercancel",
Please add touchstart/end/move in this list.

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:69:
debug("    **** Move to grey box & jiggle ***** ");
The "grey" touch activity doesn't add any value I think. Remove?

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:102:

Please add a multitouch case: touch+drag on G & B, cancel/release 1st/2nd touch
separately.

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/input/PointerEventManager.cpp (right):

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/input/PointerEventManager.cpp:318:
WebInputEventResult PointerEventManager::sendTouchPointerEvent(
Forgot to include the caller?

[Navid Zolghadr, 2016-03-31 14:41:47]

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/input/PointerEventManager.cpp (right):

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/input/PointerEventManager.cpp:318:
WebInputEventResult PointerEventManager::sendTouchPointerEvent(
On 2016/03/30 21:21:54, mustaq wrote:
> Forgot to include the caller?

I'm not quite sure what you mean here. Are you talking about EventHandler that
calls this function? What should I do with that?

[mustaq, 2016-03-31 15:24:44]

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/input/PointerEventManager.cpp (right):

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/input/PointerEventManager.cpp:318:
WebInputEventResult PointerEventManager::sendTouchPointerEvent(
On 2016/03/31 14:41:47, Navid Zolghadr wrote:
> On 2016/03/30 21:21:54, mustaq wrote:
> > Forgot to include the caller?
> 
> I'm not quite sure what you mean here. Are you talking about EventHandler that
> calls this function? What should I do with that? 

EventHandler::handleTouchEvent::freshTouchEvents would trigger setPointerCapture
here, but just realized that bool is true only when all TouchPoints are
touchstarts.

But wait, the same idea works: you can call setPointerCapture when
state==PlatformTouchPoint::TouchPressed, right? Not passing |isNew| to |create|
looks cleaner to me.

[Navid Zolghadr, 2016-03-31 16:22:26]

ptal

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/LayoutTe...
File
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html
(right):

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:42:
var eventList = ["pointerenter", "pointerleave", "pointerover", "pointerout",
"pointermove", "pointerdown", "pointerup", "pointercancel",
On 2016/03/30 21:21:54, mustaq wrote:
> Please add touchstart/end/move in this list.

Done.

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:69:
debug("    **** Move to grey box & jiggle ***** ");
On 2016/03/30 21:21:54, mustaq wrote:
> The "grey" touch activity doesn't add any value I think. Remove?

Done.

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/LayoutTe...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:102:

On 2016/03/30 21:21:54, mustaq wrote:
> Please add a multitouch case: touch+drag on G & B, cancel/release 1st/2nd
touch
> separately.

Is this what you meant?

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/input/PointerEventManager.cpp (right):

https://codereview.chromium.org/1838973003/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/input/PointerEventManager.cpp:318:
WebInputEventResult PointerEventManager::sendTouchPointerEvent(
On 2016/03/31 15:24:44, mustaq wrote:
> On 2016/03/31 14:41:47, Navid Zolghadr wrote:
> > On 2016/03/30 21:21:54, mustaq wrote:
> > > Forgot to include the caller?
> > 
> > I'm not quite sure what you mean here. Are you talking about EventHandler
that
> > calls this function? What should I do with that? 
> 
> EventHandler::handleTouchEvent::freshTouchEvents would trigger
setPointerCapture
> here, but just realized that bool is true only when all TouchPoints are
> touchstarts.
> 
> But wait, the same idea works: you can call setPointerCapture when
> state==PlatformTouchPoint::TouchPressed, right? Not passing |isNew| to
|create|
> looks cleaner to me.

That definitely sounds cleaner. I haven't thought of that.

[mustaq, 2016-04-04 15:37:26]

This is much cleaner, thanks. Looks good, needs a slightly deeper multitouch
test.

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html
(right):

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:32:
var x2 = rect.left + 5;
Unused: x2, y2.

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:108:
eventSender.cancelTouchPoint(1);
I meant separate end/cancel for 0 & 1, to assert that the pointer events sent
right before the two lost-pointer-capture events are sane.

Here is a possibility, also checks per-pointer capture:
- Touch with finger #0 on G and #1 on B.
- Move both of them across: #0 -> B, #1 -> G.
- Release/cancel #0.
- Jiggle #1.
- Release/cancel #1.

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/dom/Element.cpp (right):

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/dom/Element.cpp:2654: // TODO(crbug.com/579553):
This next "else if" is a hack to suppress capturing for touches
...hack to notify JS that we don't (yet) support explicit set/release of touch
pointers (which are implicitly captured for performance reasons).

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/dom/Element.cpp:2667: // TODO(crbug.com/579553):
This next "else if" is a hack to suppress capturing for touches
Ditto.

[Navid Zolghadr, 2016-04-04 16:26:58]

ptal

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html
(right):

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:32:
var x2 = rect.left + 5;
On 2016/04/04 15:37:26, mustaq wrote:
> Unused: x2, y2.

Done.

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:108:
eventSender.cancelTouchPoint(1);
On 2016/04/04 15:37:26, mustaq wrote:
> I meant separate end/cancel for 0 & 1, to assert that the pointer events sent
> right before the two lost-pointer-capture events are sane.
> 
> Here is a possibility, also checks per-pointer capture:
> - Touch with finger #0 on G and #1 on B.
> - Move both of them across: #0 -> B, #1 -> G.
> - Release/cancel #0.
> - Jiggle #1.
> - Release/cancel #1.

Does this sound better? I had to send the moves and starts separately as well
since the ordering of sending touchmove/start/end will be random when sending to
different elements and sometime we send the green one first and sometime we send
the blue one first. Just didn't want to address that issue in this CL.

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/dom/Element.cpp (right):

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/dom/Element.cpp:2654: // TODO(crbug.com/579553):
This next "else if" is a hack to suppress capturing for touches
On 2016/04/04 15:37:26, mustaq wrote:
> ...hack to notify JS that we don't (yet) support explicit set/release of touch
> pointers (which are implicitly captured for performance reasons).

Done.

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/dom/Element.cpp:2667: // TODO(crbug.com/579553):
This next "else if" is a hack to suppress capturing for touches
On 2016/04/04 15:37:26, mustaq wrote:
> Ditto.

Done.

[mustaq, 2016-04-05 15:46:18]

LGTM with nits on test clarity & comment.

Please update the CL description: "we are sending both got/lost... events", then
mention what you have now.

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html
(right):

https://codereview.chromium.org/1838973003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:108:
eventSender.cancelTouchPoint(1);
On 2016/04/04 16:26:57, Navid Zolghadr wrote:
> On 2016/04/04 15:37:26, mustaq wrote:
> > I meant separate end/cancel for 0 & 1, to assert that the pointer events
sent
> > right before the two lost-pointer-capture events are sane.
> > 
> > Here is a possibility, also checks per-pointer capture:
> > - Touch with finger #0 on G and #1 on B.
> > - Move both of them across: #0 -> B, #1 -> G.
> > - Release/cancel #0.
> > - Jiggle #1.
> > - Release/cancel #1.
> 
> Does this sound better? I had to send the moves and starts separately as well
> since the ordering of sending touchmove/start/end will be random when sending
to
> different elements and sometime we send the green one first and sometime we
send
> the blue one first. Just didn't want to address that issue in this CL.

I don't think we need to worry about the apparent random order you mentioned.
Sending touch "activities" separately looks fine.

https://codereview.chromium.org/1838973003/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html
(right):

https://codereview.chromium.org/1838973003/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:114:
eventSender.updateTouchPoint(0, x1+1, y1+1);
*** Jiggle second touch on green ***

https://codereview.chromium.org/1838973003/diff/40001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/input/PointerEventManager.cpp (right):

https://codereview.chromium.org/1838973003/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/input/PointerEventManager.cpp:346: // Sending the
leave/out events as well as lostpointercapture if required
Replace "...if required" with something like "...because the next touchevent
will have a different id, so can't trigger...".

[Navid Zolghadr, 2016-04-05 15:58:32]

Description was changed from

==========
Send lostpointercapture on touch capturing

Also sending lostpointercapture on touch released or
cancel rightaway after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.
Setting the pointer capture right after the touch starts.

More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

to

==========
Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches by setting the
implicit pointer capture right after the touch starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.


More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

[Navid Zolghadr, 2016-04-05 16:23:51]

nzolghadr@chromium.org changed reviewers:
+ bokan@chromium.org, dtapuska@chromium.org

[Navid Zolghadr, 2016-04-05 16:23:52]

ptal.

https://codereview.chromium.org/1838973003/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html
(right):

https://codereview.chromium.org/1838973003/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/virtual/pointerevent/fast/events/pointerevents/touch-capture.html:114:
eventSender.updateTouchPoint(0, x1+1, y1+1);
On 2016/04/05 15:46:17, mustaq wrote:
> *** Jiggle second touch on green ***

Done.

https://codereview.chromium.org/1838973003/diff/40001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/input/PointerEventManager.cpp (right):

https://codereview.chromium.org/1838973003/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/input/PointerEventManager.cpp:346: // Sending the
leave/out events as well as lostpointercapture if required
On 2016/04/05 15:46:18, mustaq wrote:
> Replace "...if required" with something like "...because the next touchevent
> will have a different id, so can't trigger...".

Done.

[mustaq, 2016-04-05 16:39:26]

LGTM.

[Navid Zolghadr, 2016-04-05 16:39:57]

Description was changed from

==========
Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches by setting the
implicit pointer capture right after the touch starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.


More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

to

==========
Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches
by setting the implicit pointer capture right after the touch starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.


More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

[Navid Zolghadr, 2016-04-05 16:40:14]

Description was changed from

==========
Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches
by setting the implicit pointer capture right after the touch starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.


More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

to

==========
Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches
by setting the implicit pointer capture right after the touch
starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.


More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

[bokan, 2016-04-05 17:08:31]

rs lgtm

[Navid Zolghadr, 2016-04-05 17:56:00]

The CQ bit was checked by nzolghadr@chromium.org

[Navid Zolghadr, 2016-04-05 17:56:00]

The patchset sent to the CQ was uploaded after l-g-t-m from mustaq@chromium.org
Link to the patchset: https://codereview.chromium.org/1838973003/#ps80001
(title: " ")

[commit-bot: I haz the power, 2016-04-05 17:56:22]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1838973003/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1838973003/80001

[commit-bot: I haz the power, 2016-04-05 18:03:12]

Description was changed from

==========
Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches
by setting the implicit pointer capture right after the touch
starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.


More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

to

==========
Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches
by setting the implicit pointer capture right after the touch
starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.


More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

[commit-bot: I haz the power, 2016-04-05 18:03:13]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2016-04-05 18:04:27]

Description was changed from

==========
Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches
by setting the implicit pointer capture right after the touch
starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.


More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553


BUG=596505
==========

to

==========
Send lostpointercapture on touch capturing

We are sending both got/lostpointercapture events on touches
by setting the implicit pointer capture right after the touch
starts.

Also sending lostpointercapture on touch released or
cancel right away after pointerup/cancel as we will never
get another pointer event with the same id to trigger
sending the lostpointercapture.

More on the interaction of pointer event capturing and
touch capturing will be addressed in crbug.com/579553

BUG=596505

Committed: https://crrev.com/268b54a5f857d2d8a89199e3ada14b0afed2075e
Cr-Commit-Position: refs/heads/master@{#385219}
==========

[commit-bot: I haz the power, 2016-04-05 18:04:28]

Patchset 5 (id:??) landed as
https://crrev.com/268b54a5f857d2d8a89199e3ada14b0afed2075e
Cr-Commit-Position: refs/heads/master@{#385219}