[Webkit-unassigned] [Bug 203264] [Release][ iOS ] Three editing/pasteboard/smart-paste-paragraph tests have been flaky since they landed in r243124

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=203264

--- Comment #15 from Wenson Hsieh <wenson_hsieh at apple.com> ---
(In reply to Wenson Hsieh from comment #14)
> (In reply to Wenson Hsieh from comment #13)
> > (In reply to Wenson Hsieh from comment #12)
> > > (In reply to Alexey Proskuryakov from comment #11)
> > > > This makes me wonder if this has the same root cause as the bug Wenson just
> > > > fixed, with software keyboard appearing unexpectedly.
> > > 
> > > I just ran this test with and without having the test force the software
> > > keyboard (using `UIHelper.setHardwareKeyboardAttached`), and confirmed that
> > > I see the same output in both cases.
> > 
> > (The same test failure output, that is)
> 
> It seems that this line:
> 
>     await
> UIHelper.selectWordByDoubleTapOrClick(document.getElementById('test'));
> 
> …is failing to actually produce a ranged selection.

The test seems to start passing if I perturb the double tap location a bit, like so:

diff --git a/LayoutTests/resources/ui-helper.js b/LayoutTests/resources/ui-helper.js
index 3aaba293a2c..49c3d526023 100644
--- a/LayoutTests/resources/ui-helper.js
+++ b/LayoutTests/resources/ui-helper.js
@@ -201,7 +201,8 @@ window.UIHelper = class UIHelper {
         const y = boundingRect.y + relativeY;
         if (this.isIOSFamily()) {
             await UIHelper.activateAndWaitForInputSessionAt(x, y);
-            await UIHelper.doubleTapAt(x, y);
+            await UIHelper.activateAt(x + 1, y + 1);
+            await UIHelper.activateAt(x + 1, y + 1);
             // This is only here to deal with async/sync copy/paste calls, so
             // once <rdar://problem/16207002> is resolved, should be able to remove for faster tests.
             await new Promise(resolve => testRunner.runUIScript("uiController.uiScriptComplete()", resolve));

…this appears to share the same root cause as https://bugs.webkit.org/show_bug.cgi?id=203392. We should really consider fixing position information request caching in WKContentViewInteraction, so that we don’t end up with a stale value for `_positionInformation.nodeAtPositionIsFocusedElement` after focusing an 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/20191025/63710ef1/attachment.htm>