[Webkit-unassigned] [Bug 182300] New: Click event node approximation does not correctly click targets in subframes

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Jan 30 11:23:35 PST 2018


https://bugs.webkit.org/show_bug.cgi?id=182300

            Bug ID: 182300
           Summary: Click event node approximation does not correctly
                    click targets in subframes
           Product: WebKit
           Version: WebKit Nightly Build
          Hardware: iPhone / iPad
                OS: All
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: Event Handling
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: aestes at apple.com
                CC: benjamin at webkit.org, dino at apple.com,
                    simon.fraser at apple.com, thorton at apple.com,
                    wenson_hsieh at apple.com, youennf at gmail.com,
                    zalan at apple.com

Created attachment 332674

  --> https://bugs.webkit.org/attachment.cgi?id=332674&action=review

test case

See the attached test case.

In Frame::nodeRespondingToClickEvents(), we sample points in a 30px radius around the touch location to find a clickable element to receive synthetic mouse events. We try to find the clickable element that has the largest intersection with the 30x30 sample rectangle.

We do this by taking each node's absoluteBoundingBoxRect(), intersecting that with the the sample rectangle, and using the node with the largest intersection. However, absoluteBoundingBoxRect() returns frame-relative coordinates, so if a candidate node is in a different frame than the sample rectangle, comparing their bounding boxes is bogus.

The result of this tends to be that subframes containing clickable elements are penalized in nodeRespondingToClickEvents() and you are more likely to click a same-frame element even if its intersection is smaller than the subframe element.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20180130/08e759ac/attachment-0001.html>


More information about the webkit-unassigned mailing list