Bug 95831 – [Shadow DOM][Refactoring] Element subclasses should have a way to reject author shadows.

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

[Hajime Morrita]

Each element subclasses should have a method to reject author shadows to protect its UA shadow.
Currently we have such checkin ShadowRoot but it should be handled by each element.

[Hajime Morrita]

Created attachment 162202 [details]
Patch

[Hajime Morrita]

Comment on attachment 162202 [details]
Patch

Attached a wrong patch...

[Hajime Morrita]

Created attachment 162206 [details]
Patch

[Build Bot]

Comment on attachment 162206 [details]
Patch

Attachment 162206 [details] did not pass win-ews (win):
Output: http://queues.webkit.org/results/13765174

[Dimitri Glazkov (Google)]

Comment on attachment 162206 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=162206&action=review

The change is good, but needs spelling fixes and a rename to plural.

> Source/WebCore/ChangeLog:9
> +        This chagne moves that responsibility to each Element subclasses to let coming

chagne->change, and "to let coming" -> "to make future"

> Source/WebCore/ChangeLog:10
> +        author shdaow support improvement local to each replaced elements, rather than

shdaow->shadow :)

> Source/WebCore/html/HTMLInputElement.h:295
> +    virtual bool isAuthorShadowsAllowed() const OVERRIDE { return false; }

is->are

[Hajime Morrita]

Created attachment 162385 [details]
Patch

[Hajime Morrita]

Thanks for take a look, Dimitri! I kicked my ashamed misspelling out. Could you PTAL?

[Hajime Morrita]

Dimitri: ping? I want this to mask some unpolished elements.

[Hajime Morrita]

Comment on attachment 162385 [details]
Patch

Thanks ;-)

[WebKit Review Bot]

Comment on attachment 162385 [details]
Patch

Clearing flags on attachment: 162385

Committed r127811: <http://trac.webkit.org/changeset/127811>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[Elliott Sprehn]

Comment on attachment 162385 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=162385&action=review

> Source/WebCore/svg/SVGElement.h:141
> +    virtual bool areAuthorShadowsAllowed() const OVERRIDE { return false; }

Note that all SVG elements inherit from SVGElement. If this was to return true you'd allow attaching shadows to things like <path> and <rect> which doesn't make any sense.

This should only ever return true from SVGSVGElement.h

[Hajime Morrita]

(In reply to comment #12)
> 
> Note that all SVG elements inherit from SVGElement. If this was to return true you'd allow attaching shadows to things like <path> and <rect> which doesn't make any sense.
> 
> This should only ever return true from SVGSVGElement.h
Could you elaborate?

Currently SVG uses ShadowRoot in some tricky way and
we don't make it work with author shadows yet.
This is why we prohibit author shadows here.

It has been like that from the beginning and this patch just clarifies
what we've been doing and has no behavioral change.

Does this make sense?