Bug 215677 – Move node geometry functions from Range to RenderObject

• Component: Layout and Rendering
• Status: RESOLVED FIXED

[Darin Adler]

Move node geometry functions from Range to RenderObject

[Darin Adler]

Created attachment 406911 [details]
Patch

[Simon Fraser (smfr)]

Are you going to hold off from landing this?

[Darin Adler]

Yes.

[Darin Adler]

Created attachment 406939 [details]
Patch

[Darin Adler]

Created attachment 406962 [details]
Patch

[Darin Adler]

Hooray, all tests passing now, so ready for review. I won’t be landing this until Saturday at the soonest, but would love to get review soon.

[alan]

Comment on attachment 406962 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=406962&action=review

> Source/WebCore/rendering/RenderObject.h:573
> +    WEBCORE_EXPORT static Vector<FloatQuad> absoluteTextQuads(const SimpleRange&, OptionSet<BoundingRectBehavior> = { });
> +    WEBCORE_EXPORT static Vector<IntRect> absoluteTextRects(const SimpleRange&, OptionSet<BoundingRectBehavior> = { });
> +    WEBCORE_EXPORT static Vector<FloatRect> absoluteBorderAndTextRects(const SimpleRange&, OptionSet<BoundingRectBehavior> = { });
> +    static Vector<FloatRect> clientBorderAndTextRects(const SimpleRange&);

It's a bit unfortunate that functions with _text_ and _border_  in their names end up in RenderObject (where we have neither border nor text) but I understand that it's hard to find a better place for a concept (Range) that has no associated renderer.

[Darin Adler]

Comment on attachment 406962 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=406962&action=review

>> Source/WebCore/rendering/RenderObject.h:573
>> +    static Vector<FloatRect> clientBorderAndTextRects(const SimpleRange&);
> 
> It's a bit unfortunate that functions with _text_ and _border_  in their names end up in RenderObject (where we have neither border nor text) but I understand that it's hard to find a better place for a concept (Range) that has no associated renderer.

I am totally willing to follow up and move them somewhere else in the rendering world. They don’t have to stay on RenderObject.

[Darin Adler]

Committed r266028: <https://trac.webkit.org/changeset/266028>

[Radar WebKit Bug Importer]

<rdar://problem/67617601>

[Aakash Jain]

(In reply to Darin Adler from comment #9)
> Committed r266028: <https://trac.webkit.org/changeset/266028>
Seems like it broke one test on ios-wk2: platform/ios/ios/fast/coordinates/range-client-rects.html. Tracked in Bug 215772.

Also, unfortunately ios-wk2 ews was broken for some time late last week (due to Bug 215742), due to which the green ios-wk2 bubble for this bug is misleading (iOS layout-tests weren't actually run).