[Webkit-unassigned] [Bug 94936] Incomplete rect-based hit-test result when hit-test region overlaps layers with different z-order.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=94936





--- Comment #8 from Antonio Gomes <tonikitoo at webkit.org>  2012-08-27 08:28:31 PST ---
(In reply to comment #7)
> That is consistent with what I'm seeing for the example.  The <iframe>s are detected but do not provide enough information for touch adjustment since smaller elements inside the <iframe>s are not picked up as candidates.
> 
> One solution may be to repeat the hit test on each of the <iframe>s from the first test.  Alternatively, could repeat the hit-test just for the <iframe> that contains the midpoint of the touch.

Right. this is what we do for blackberry port.

FatFingers.cpp:
...
335         bool isElement = curNode->isElementNode();
336         if (isElement && isValidFrameOwner(toElement(curNode))) {
337 
338             HTMLFrameOwnerElement* owner = static_cast<HTMLFrameOwnerElement*>(curNode);
339             Document* childDocument = owner && owner->contentFrame() ? owner->contentFrame()->document() : 0;
340             if (!childDocument)
341                 continue;
342 
343             ASSERT(childDocument->frame()->view());
344 
345             foundOne |= findIntersectingRegions(childDocument, intersectingRegions, remainingFingerRegion);

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.