Bug 111986 – [EFL] accessibility/aria-invalid is failing

• Component: WebKit EFL
• Status: RESOLVED FIXED

[Krzysztof Czech]

accessibility/aria-invalid is failing is failing on all EFL platforms.

[Denis Nomiyama (dnomi)]

The implementation of aria-invalid has been added to the GTK port in bug 98350 and bug 121668.
The EFL port can also use this code by implementing a small piece of code at:

Tools/DumpRenderTree/atk/AccessibilityCallbacksAtk.cpp
- axObjectEventListener()
- addAccessibilityNotificationHandler()
- removeAccessibilityNotificationHandler()

and

Tools/WebKitTestRunner/InjectedBundle/atk/AccessibilityNotificationHandlerAtk.cpp
- axObjectEventListener()
- setNotificationFunctionCallback()
- removeAccessibilityNotificationHandler()

Currently there is only implementation for the GTK platform (e.g. #if PLATFORM(GTK)), and the code needs basically to get the JS context.

[Krzysztof Czech]

(In reply to comment #1)
> The implementation of aria-invalid has been added to the GTK port in bug 98350 and bug 121668.
> The EFL port can also use this code by implementing a small piece of code at:
> 

Great, I will take a look at this. Thanks

[Krzysztof Czech]

Created attachment 212339 [details]
Patch

[Mario Sanchez Prada]

Comment on attachment 212339 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=212339&action=review

> Tools/DumpRenderTree/atk/AccessibilityCallbacksAtk.cpp:131
> +#elif PLATFORM(EFL) && HAVE(ACCESSIBILITY)

The whole content of this file is wrapper inside an HAVE(ACCESSIBILITY) guard, so you don't need to add it explicitly here.

> Tools/DumpRenderTree/atk/AccessibilityCallbacksAtk.cpp:232
> +#elif PLATFORM(EFL) && HAVE(ACCESSIBILITY)

Same here

> Tools/DumpRenderTree/atk/AccessibilityCallbacksAtk.cpp:266
> +#elif PLATFORM(EFL) && HAVE(ACCESSIBILITY)

And here :)

> Tools/WebKitTestRunner/InjectedBundle/atk/AccessibilityNotificationHandlerAtk.cpp:116
> +#if PLATFORM(GTK) || (PLATFORM(EFL) && HAVE(ACCESSIBILITY))

In this case is different because this file is not guarded at all as the other ones.

Perhaps we should do it here too? After all, this file does not make much sense at all if you don't have accessibility enabled, does it? :)

[Krzysztof Czech]

 > > Tools/DumpRenderTree/atk/AccessibilityCallbacksAtk.cpp:131
> > +#elif PLATFORM(EFL) && HAVE(ACCESSIBILITY)
> 
> The whole content of this file is wrapper inside an HAVE(ACCESSIBILITY) guard, so you don't need to add it explicitly here.
> 
Indeed, sorry, I did not catch this.
> > Tools/WebKitTestRunner/InjectedBundle/atk/AccessibilityNotificationHandlerAtk.cpp:116
> > +#if PLATFORM(GTK) || (PLATFORM(EFL) && HAVE(ACCESSIBILITY))
> 
> In this case is different because this file is not guarded at all as the other ones.
> 
> Perhaps we should do it here too? After all, this file does not make much sense at all if you don't have accessibility enabled, does it? :)

Yes, you're right, we should have this guard. Those are specific accessibility bits.

[Krzysztof Czech]

Created attachment 212344 [details]
Patch

[Krzysztof Czech]

> In this case is different because this file is not guarded at all as the other ones.
> 
> Perhaps we should do it here too? After all, this file does not make much sense at all if you don't have accessibility enabled, does it? :)

I will propose a new patch with guards

[Mario Sanchez Prada]

Comment on attachment 212344 [details]
Patch

lgtm

[WebKit Commit Bot]

Comment on attachment 212344 [details]
Patch

Clearing flags on attachment: 212344

Committed r156273: <http://trac.webkit.org/changeset/156273>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.