[Webkit-unassigned] [Bug 260621] AX: aria-controls-with-tabs fails in ITM
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Fri Aug 25 04:55:56 PDT 2023
https://bugs.webkit.org/show_bug.cgi?id=260621
--- Comment #13 from Andres Gonzalez <andresg_22 at apple.com> ---
(In reply to Joshua Hoffman from comment #7)
> (In reply to Tyler Wilcock from comment #5)
> > Comment on attachment 467398 [details]
> > Patch
> >
> > View in context:
> > https://bugs.webkit.org/attachment.cgi?id=467398&action=review
> >
> > > Source/WebCore/accessibility/AXObjectCache.cpp:1645
> > > + // FIXME: optimize with ancestor flags to determine if an element is in a tabpanel
> >
> > I feel like this could be written more clearly for future us...isn't the
> > idea more like "FIXME: Consider creating a new ancestor flag to only do this
> > work when |oldNode| or |newNode| have a tab panel ancestor (as that's the
> > only time it is necessary)"? Or is that not accurate?
> >
>
> Yes, that wording is more clear—I'll update that.
>
> > > Source/WebCore/accessibility/AXObjectCache.cpp:1685
> > > else if (is<HTMLOptionElement>(node))
> > > postNotification(node, AXSelectedStateChanged);
> > > + else if (nodeHasRole(node, "tab"_s))
> > > + postNotification(node, AXSelectedStateChanged);
> >
> > Can we combine this with the previous else if rather than adding a new case?
>
> Yep.
>
> > else if (is<HTMLOptionElement>(node) || nodeHasRole(node, "tab"_s))
> >
> > > Source/WebCore/accessibility/AXObjectCache.h:570
> > > + void handleTabOrControlledTabPanelSelected(Node*);
> >
> > Any time we add a method here, we need to make sure we add a no-op method to
> > the !ENABLE(ACCESSIBILITY) section of this file to avoid breaking the
> > Playstation build. See `handleRoleChanged` as an example.
>
> Got it!
See my comment above. Only do that when needed.
--
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/20230825/2d044921/attachment-0001.htm>
More information about the webkit-unassigned
mailing list