Bug 89073 – Modify event re-targeting algorithm so that we can tell which distributed node is clicked.

• Status: RESOLVED
• Resolution: FIXED
• Component: UI Events

[Hayato Ito]

Please see the going discussion on the spec side:
https://www.w3.org/Bugs/Public/show_bug.cgi?id=17090

We have to adopt a new algorithm once we agree on the new behavior of event re-targeting on the Shadow DOM spec.

[Hayato Ito]

Created attachment 147520 [details]
WIP. concept of a new re-targeting algorithm

[Dimitri Glazkov (Google)]

Comment on attachment 147520 [details]
WIP. concept of a new re-targeting algorithm

I haven't dug into this deeply, but from the first glance, this looks great!

[Hayato Ito]

Created attachment 148436 [details]
target can now be a distributed node

[Hayato Ito]

Created attachment 148820 [details]
minor update on changelog

[Hayato Ito]

Created attachment 148821 [details]
minor update on changelog

[Hayato Ito]

Since it got r+ for https://bugs.webkit.org/attachment.cgi?id=148820&action=review, let me land this manually.

[Hayato Ito]

Created attachment 148825 [details]
Patch for landing

[Hayato Ito]

Created attachment 148826 [details]
Patch for landing

[WebKit Review Bot]

Comment on attachment 148826 [details]
Patch for landing

Clearing flags on attachment: 148826

Committed r120945: <http://trac.webkit.org/changeset/120945>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[WebKit Review Bot]

Re-opened since this is blocked by 89703

[Kenneth Russell]

*** Bug 89708 has been marked as a duplicate of this bug. ***

[Hayato Ito]

I've investigated the crash of editing/shadow/breaking-editing-boundaries.html.

It seems this re-targeting patch is innocent. This patch just revealed a bogus DOM structure that was caused in editing/shadow/breaking-editing-boundaries.html.

So it seems an editing bug. Let me investigate further.

[Hayato Ito]

It seems parentTreeScope() returns an unexpected node.
And I found this bug. https://bugs.webkit.org/show_bug.cgi?id=89172
That's the root cause.

(In reply to comment #13)
> I've investigated the crash of editing/shadow/breaking-editing-boundaries.html.
> 
> It seems this re-targeting patch is innocent. This patch just revealed a bogus DOM structure that was caused in editing/shadow/breaking-editing-boundaries.html.
> 
> So it seems an editing bug. Let me investigate further.

[Hayato Ito]

I'll re-land this patch once bug 89172 is resolved.

(In reply to comment #14)
> It seems parentTreeScope() returns an unexpected node.
> And I found this bug. https://bugs.webkit.org/show_bug.cgi?id=89172
> That's the root cause.
> 
> (In reply to comment #13)
> > I've investigated the crash of editing/shadow/breaking-editing-boundaries.html.
> > 
> > It seems this re-targeting patch is innocent. This patch just revealed a bogus DOM structure that was caused in editing/shadow/breaking-editing-boundaries.html.
> > 
> > So it seems an editing bug. Let me investigate further.

[Hayato Ito]

Created attachment 149045 [details]
Patch for landing

[Hayato Ito]

Let me re-land this patch since bug 89172 was resolved and confirmed that editing/shadow/breaking-editing-boundaries.html does not crash anymore.

[WebKit Review Bot]

Comment on attachment 149045 [details]
Patch for landing

Clearing flags on attachment: 149045

Committed r121040: <http://trac.webkit.org/changeset/121040>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.