Support slotchange event

Support slotchange event

This CL implements slotchange event,
which is added to the microtask queue whenever its assigned nodes
or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback
calculations in SlotAssignment.
Also note that slotchange event gets fired at most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288
See the spec here: https://dom.spec.whatwg.org/#mutation-algorithms

BUG=595287
Committed: https://crrev.com/723a4b536fa9a2f0071c510d14c36918cfee83a9
Cr-Commit-Position: refs/heads/master@{#389056}

[yuzuchan, 2016-04-18 04:49:00]

Description was changed from

==========
Support SlotChange event

This CL implements slotchange event, which is added to the microtask queue
whenever its assigned nodes or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback
calculations in SlotAssignment. Also note that slotchange event gets fired at
most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288

BUG=595287
==========

to

==========
Support slotchange event

This CL implements slotchange event, which is added to the microtask queue
whenever its assigned nodes or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback
calculations in SlotAssignment. Also note that slotchange event gets fired at
most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288

BUG=595287
==========

[yuzuchan, 2016-04-18 04:49:00]

yuzus@chromium.org changed reviewers:
+ hayato@chromium.org, kochi@chromium.org

[yuzuchan, 2016-04-18 04:51:14]

PTAL :) 

This is how I feel about this patch: http://i.imgur.com/zGusm9A.gif

[hayato, 2016-04-18 07:16:31]

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/dom/Element.h (right):

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/dom/Element.h:874: if
(document().shadowCascadeOrder() == ShadowCascadeOrder::ShadowCascadeV1) {
Given that CSS is no longer only the client of this flag, 
it would be better to rename shadowCascadeOrder to more general name.

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/dom/Node.cpp (right):

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/dom/Node.cpp:573: void Node::recalcAssignment()
We do not want to call recalcAssignment recursively.

Given that we call updateAssignment in each DOM mutation, and the result of
assignments is purely a local effect within one node tree, I would rather call
assignV1() (or something else) for a node tree which would be directly affected,
in each DOM mutation.

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/dom/shadow/SlotAssignment.cpp (right):

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/dom/shadow/SlotAssignment.cpp:74: Name2Slot
name2slot;
resolveDistribution does not need `name2slot`.

[yuzuchan, 2016-04-18 09:43:27]

PTAL again :)

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/dom/Element.h (right):

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/dom/Element.h:874: if
(document().shadowCascadeOrder() == ShadowCascadeOrder::ShadowCascadeV1) {
On 2016/04/18 at 07:16:31, hayato wrote:
> Given that CSS is no longer only the client of this flag, 
> it would be better to rename shadowCascadeOrder to more general name.

Found out that I do not need to use the function here. Will rename it in another
CL.

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/dom/Node.cpp (right):

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/dom/Node.cpp:573: void Node::recalcAssignment()
On 2016/04/18 at 07:16:31, hayato wrote:
> We do not want to call recalcAssignment recursively.
> 
> Given that we call updateAssignment in each DOM mutation, and the result of
assignments is purely a local effect within one node tree, I would rather call
assignV1() (or something else) for a node tree which would be directly affected,
in each DOM mutation.

Thanks for the feedback! Done.

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/dom/shadow/SlotAssignment.cpp (right):

https://codereview.chromium.org/1899653002/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/dom/shadow/SlotAssignment.cpp:74: Name2Slot
name2slot;
On 2016/04/18 at 07:16:31, hayato wrote:
> resolveDistribution does not need `name2slot`.

Done.

[hayato, 2016-04-19 06:39:34]

FYI. I and Anne are updating DOM Standard to define "slotchange" formally.
https://github.com/whatwg/dom/pull/229

There, the timing and condition of slotchange events are strictly defined. You
might want to check it.

The basic idea is the same to this CL. It would be nice if we could try to
implement the current definition to know its feasibility.

Let's chat offline later.

[yuzuchan, 2016-04-21 08:52:03]

On 2016/04/19 at 06:39:34, hayato wrote:
> FYI. I and Anne are updating DOM Standard to define "slotchange" formally.
> https://github.com/whatwg/dom/pull/229
> 
> There, the timing and condition of slotchange events are strictly defined. You
might want to check it.
> 
> The basic idea is the same to this CL. It would be nice if we could try to
implement the current definition to know its feasibility.
> 
> Let's chat offline later.

I made the assignment steps more consistent with the current spec. PTAL :)

[yuzuchan, 2016-04-22 02:07:03]

PTAL again :)

[hayato, 2016-04-22 05:33:22]

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Lay...
File third_party/WebKit/LayoutTests/shadow-dom/slotchange-event.html (right):

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/shadow-dom/slotchange-event.html:47: test.done();
Can we remove "test.done()" from d1_s1?

I guess it is guaranteed that d1_d1_s1's 2nd slotchange event will be fired
after d1_s1's 2nd slotchange event.

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/shadow-dom/slotchange-event.html:77: }, 100);
0 (zero) is good enough, instead of 100.

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/html/HTMLSlotElement.cpp (right):

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/html/HTMLSlotElement.cpp:348: ShadowRoot*
shadowRoot = shadowRootIfV1();
This would be:
  shadowHost.shadowRootIfV1()

[yuzuchan, 2016-04-22 06:59:11]

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Lay...
File third_party/WebKit/LayoutTests/shadow-dom/slotchange-event.html (right):

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/shadow-dom/slotchange-event.html:47: test.done();
On 2016/04/22 at 05:33:22, hayato wrote:
> Can we remove "test.done()" from d1_s1?
> 
> I guess it is guaranteed that d1_d1_s1's 2nd slotchange event will be fired
after d1_s1's 2nd slotchange event.

Because of syntax, this test.done() cannot be removed, even though the order is
guaranteed.

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/shadow-dom/slotchange-event.html:77: }, 100);
On 2016/04/22 at 05:33:22, hayato wrote:
> 0 (zero) is good enough, instead of 100.

Done.

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/html/HTMLSlotElement.cpp (right):

https://codereview.chromium.org/1899653002/diff/200001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/html/HTMLSlotElement.cpp:348: ShadowRoot*
shadowRoot = shadowRootIfV1();
On 2016/04/22 at 05:33:22, hayato wrote:
> This would be:
>   shadowHost.shadowRootIfV1()

Done.

[hayato, 2016-04-22 07:32:13]

LGTM.

Could you update the description so that it does not have a long line?

I think we can land this patch as long as there is no regression in Layout
Tests.
This CL should not cause any performance regression unless a v1 shadow tree is
used.

Let's land this CL. Let me optimize the performance later in another CL.

I am so happy to see this CL in your internship's last day!

[yuzuchan, 2016-04-22 07:33:12]

Description was changed from

==========
Support slotchange event

This CL implements slotchange event, which is added to the microtask queue
whenever its assigned nodes or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback
calculations in SlotAssignment. Also note that slotchange event gets fired at
most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288

BUG=595287
==========

to

==========
Support slotchange event

This CL implements slotchange event, which is added to the microtask queue
whenever its assigned nodes or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback
calculations in SlotAssignment.
Also note that slotchange event gets fired at most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288

BUG=595287
==========

[yuzuchan, 2016-04-22 07:33:47]

Description was changed from

==========
Support slotchange event

This CL implements slotchange event, which is added to the microtask queue
whenever its assigned nodes or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback
calculations in SlotAssignment.
Also note that slotchange event gets fired at most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288

BUG=595287
==========

to

==========
Support slotchange event

This CL implements slotchange event,
which is added to the microtask queue whenever its assigned nodes
or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback 
calculations in SlotAssignment.
Also note that slotchange event gets fired at most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288

BUG=595287
==========

[yuzuchan, 2016-04-22 07:34:28]

The CQ bit was checked by yuzus@chromium.org

[commit-bot: I haz the power, 2016-04-22 07:34:39]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1899653002/220001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1899653002/220001

[yuzuchan, 2016-04-22 07:34:40]

The CQ bit was unchecked by yuzus@chromium.org

[yuzuchan, 2016-04-22 07:34:54]

The CQ bit was checked by yuzus@chromium.org

[commit-bot: I haz the power, 2016-04-22 07:35:09]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1899653002/220001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1899653002/220001

[yuzuchan, 2016-04-22 08:02:19]

Description was changed from

==========
Support slotchange event

This CL implements slotchange event,
which is added to the microtask queue whenever its assigned nodes
or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback 
calculations in SlotAssignment.
Also note that slotchange event gets fired at most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288

BUG=595287
==========

to

==========
Support slotchange event

This CL implements slotchange event,
which is added to the microtask queue whenever its assigned nodes
or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback 
calculations in SlotAssignment.
Also note that slotchange event gets fired at most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288
See the spec here: https://dom.spec.whatwg.org/#mutation-algorithms

BUG=595287
==========

[commit-bot: I haz the power, 2016-04-22 08:33:34]

Committed patchset #12 (id:220001)

[commit-bot: I haz the power, 2016-04-22 19:45:29]

Description was changed from

==========
Support slotchange event

This CL implements slotchange event,
which is added to the microtask queue whenever its assigned nodes
or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback 
calculations in SlotAssignment.
Also note that slotchange event gets fired at most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288
See the spec here: https://dom.spec.whatwg.org/#mutation-algorithms

BUG=595287
==========

to

==========
Support slotchange event

This CL implements slotchange event,
which is added to the microtask queue whenever its assigned nodes
or its fallback nodes change.
In order to achieve it, this CL separates assignment, distribution and fallback
calculations in SlotAssignment.
Also note that slotchange event gets fired at most once per a microtask.

See the spec issue here: https://github.com/w3c/webcomponents/issues/288
See the spec here: https://dom.spec.whatwg.org/#mutation-algorithms

BUG=595287

Committed: https://crrev.com/723a4b536fa9a2f0071c510d14c36918cfee83a9
Cr-Commit-Position: refs/heads/master@{#389056}
==========

[commit-bot: I haz the power, 2016-04-22 19:45:31]

Patchset 12 (id:??) landed as
https://crrev.com/723a4b536fa9a2f0071c510d14c36918cfee83a9
Cr-Commit-Position: refs/heads/master@{#389056}