Bug 78313 – Element::removeShadowRoot(), setShadowRoot() should be moved into ShadowTree

• Component: DOM
• Status: RESOLVED FIXED

[Shinya Kawanaka]

Patch for Bug 78069 will emulate shadowRoot(), removeShadowRoot(), and setShadowRoot() using shadowRootList().
These API should be removed and replaced using ShadowRootList API.

[Shinya Kawanaka]

Created attachment 128976 [details]
Patch

[Hajime Morrita]

Comment on attachment 128976 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=128976&action=review

Basically looks good. Could you polish a bit?

> Source/WebCore/ChangeLog:9
> +

Could you mention now ShadowTree is on its own heap?

> Source/WebCore/dom/Element.cpp:125
> +    if (hasShadowRoot())

Can be hasShadowTree()?

> Source/WebCore/dom/Element.cpp:126
> +        shadowTree()->removeAllShadowRoots();

It looks we can do removeAllShadowRoots() in the ShadowTree dtor then we can just delete the ShadowTree instance here.

[Shinya Kawanaka]

(In reply to comment #2)
> (From update of attachment 128976 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=128976&action=review
> 
> Basically looks good. Could you polish a bit?
> 
> > Source/WebCore/ChangeLog:9
> > +
> 
> Could you mention now ShadowTree is on its own heap?
> 
> > Source/WebCore/dom/Element.cpp:125
> > +    if (hasShadowRoot())
> 
> Can be hasShadowTree()?
> 
> > Source/WebCore/dom/Element.cpp:126
> > +        shadowTree()->removeAllShadowRoots();
> 
> It looks we can do removeAllShadowRoots() in the ShadowTree dtor then we can just delete the ShadowTree instance here.

Unfortunately removeAllShadowRoots() now requires the element. So what I can do here is just to call rareData()->m_shadowTree.clear().

Currently our code assumes that a shadow root could be removed from an element now, and we have tests for it. However, since we should not have such API, we can remove that assumption. This will reduce the code and make it clear. I think it should be done in another bug though.

[Shinya Kawanaka]

Created attachment 128982 [details]
Patch

[Hajime Morrita]

Comment on attachment 128982 [details]
Patch

>Unfortunately removeAllShadowRoots() now requires the element. So what I can do here is just to call rareData()->m_shadowTree.clear().
>
>Currently our code assumes that a shadow root could be removed from an element now, and we have tests for it. However, since we should not have such API, we can remove that assumption. This will reduce the code and make it clear. I think it should be done in another bug though.

This looks reasonable. When we need the host element, there should be at least one shadow root.
If there is no shadow root. We won't need the reference the the host.

[Shinya Kawanaka]

Created attachment 129189 [details]
Patch for landing

[Early Warning System Bot]

Comment on attachment 129189 [details]
Patch for landing

Attachment 129189 [details] did not pass qt-ews (qt):
Output: http://queues.webkit.org/results/11644419

[Shinya Kawanaka]

Created attachment 129192 [details]
Patch for landing

[Shinya Kawanaka]

Committed r109084: <http://trac.webkit.org/changeset/109084>