Bug 78875 – [v8] v8 doesn't assume to do 'new WebKitShadowRoot(host)'.

• Status: RESOLVED
• Resolution: FIXED
• Component: DOM

[Shinya Kawanaka]

new WebKitShadowRoot(host) and internals.ensureShadowRoot(host) returns a different object now.
These should be the same, however they return different objects, which contain the same ShadowRoot object though.

[Shinya Kawanaka]

Created attachment 127521 [details]
Test

[Kentaro Hara]

Comment on attachment 127521 [details]
Test

View in context: https://bugs.webkit.org/attachment.cgi?id=127521&action=review

> Source/WebCore/bindings/scripts/CodeGeneratorV8.pm:1687
> +    # A DOMObject that is an ActiveDOMObject and also a DOMNode should be treated as an DOMNode here.
> +    # setJSWrapperForDOMNode() will look if node is active and choose correct map to add node to.

The change looks OK, but the comment is confusing to me. Maybe you can just remove the comment.

Why "should" an ActiveDOMObject-and-DOMNode object be treated as DOMNode? In my understanding,
- There is no ActiveDOMObject-and-DOMNode object for now.
- If an ActiveDOMObject-and-DOMNode appeared in the future, we need another logic, like "setJSWrapperForActiveDOMNodeObject()".

[Shinya Kawanaka]

Created attachment 127524 [details]
Patch

[Shinya Kawanaka]

(In reply to comment #2)
> (From update of attachment 127521 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=127521&action=review
> 
> > Source/WebCore/bindings/scripts/CodeGeneratorV8.pm:1687
> > +    # A DOMObject that is an ActiveDOMObject and also a DOMNode should be treated as an DOMNode here.
> > +    # setJSWrapperForDOMNode() will look if node is active and choose correct map to add node to.
> 
> The change looks OK, but the comment is confusing to me. Maybe you can just remove the comment.
> 
> Why "should" an ActiveDOMObject-and-DOMNode object be treated as DOMNode? In my understanding,
> - There is no ActiveDOMObject-and-DOMNode object for now.
> - If an ActiveDOMObject-and-DOMNode appeared in the future, we need another logic, like "setJSWrapperForActiveDOMNodeObject()".

Thanks. I've removed the comment.

[Shinya Kawanaka]

Created attachment 127525 [details]
Patch

[Kentaro Hara]

Comment on attachment 127525 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=127525&action=review

> LayoutTests/ChangeLog:17
> +        * platform/wincairo/Skipped:

platform/wk2/Skipped might be needed...? (I am not sure.)

> LayoutTests/fast/dom/shadow/shadow-root-new.html:9
> +<div id="container"></div>

Nit: This line is not necessary.

[Shinya Kawanaka]

(In reply to comment #6)
> (From update of attachment 127525 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=127525&action=review
> 
> > LayoutTests/ChangeLog:17
> > +        * platform/wincairo/Skipped:
> 
> platform/wk2/Skipped might be needed...? (I am not sure.)

No one has added Skipped line for SHADOW_DOM there, but let's add this time for defensive style :-)

> 
> > LayoutTests/fast/dom/shadow/shadow-root-new.html:9
> > +<div id="container"></div>
> 
> Nit: This line is not necessary.

Thanks.

After fixing them, I'll land this.

[Shinya Kawanaka]

Committed r108035: <http://trac.webkit.org/changeset/108035>