[webkit-reviews] review granted: [Bug 88066] Make RenderTableSection - nodeAtPoint and paintObject reuse spanning logic : [Attachment 148304] Patch

[bugzilla-daemon at webkit.org]

Julien Chaffraix <jchaffraix at webkit.org> has granted Allan Sandfeld Jensen
<allan.jensen at nokia.com>'s request for review:
Bug 88066: Make RenderTableSection -  nodeAtPoint and paintObject reuse
spanning logic
https://bugs.webkit.org/show_bug.cgi?id=88066

Attachment 148304: Patch
https://bugs.webkit.org/attachment.cgi?id=148304&action=review

------- Additional Comments from Julien Chaffraix <jchaffraix at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=148304&action=review


Sorry for the back and forth.

> Source/WebCore/rendering/RenderTableSection.cpp:1107
> +    if (m_forceSlowPaintPathWithOverflowingCell) {

Note that this should apply also to hit-testing. We just handle it separately
for the moment.

> Source/WebCore/rendering/RenderTableSection.cpp:1129
> +	       --dirtiedColumns.start();

If the alternative is to inline dirtiedRows / dirtiedColumns in paintObject
which is already massive, I prefer it if we don't remove the 2 functions. It
was short sighted from me to ask that as I missed the patching needed to
properly paint the first / last row and column.