[Webkit-unassigned] [Bug 56855] [GTK] Unify accessibility role names in DRT for GTK and Mac platforms

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Fri Mar 25 12:42:02 PDT 2011


https://bugs.webkit.org/show_bug.cgi?id=56855





--- Comment #4 from chris fleizach <cfleizach at apple.com>  2011-03-25 12:42:02 PST ---
(In reply to comment #3)
> (In reply to comment #2)
> > (In reply to comment #1)
> > 
> > You have enumerated part of the problem. There's ultimately logic in the Mac 
> > wrapper and I imagine gtk as well that we want to test in many cases. 
> 
> Yes, we have specific logic in the GTK wrapper to map those WebCore roles to ATK ones, and then we can get an string with the name for those ATK roles through atk_role_get_name(). However, in the mac platform I can see the mapping is more complex, since it not only maps webcore roles to mac roles, but also to subroles and role descriptions which is something we currently lack in the gtk platform because it's not something used in ATK/AT-SPI.
> 

Yea, we'd need another step to convert from the platform name, back to a common name. That would be a mess to maintain though

> > Have a method like isMac() isGTK() so we can conditionalize in the test
> 
> You mean in DRT, right? so we can actually use that isMac() or isGTK() in the Javascript code when calling to the shouldBe() functions?
> 

yea that would mean just one test, maybe different expectations, but maybe not always.

> Still, we wouldn't be able to get rid of the platform-specific expected files, but at least we would need only one test (html) file.
> 

> > Have another mapping method in DRT that took the platform output and 
> > converted to platform neutral role
> 
> That sounds good to me as well, although it would mean having to maintain an extra mapping in Mac and GTK's DRTs, unless we decide using one of the platform rolenames as the neutral ones, in which case I'd obviously choose Mac ones, since they're almost a 1:1 match with WebCore ones.
> 
> The benefit of that, in case it's actually feasible, would be that wouldn't require any change in Mac's code, nor in current layout tests, but just adding a mapping in GTK's DRT (not even in the wrapper, since we don't need that info in ATK/AT-SPI).

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



More information about the webkit-unassigned mailing list