Bug 40133 – AX: need ListItemRole and PresentationalRole

• Status: RESOLVED
• Resolution: FIXED
• Component: Accessibility

[chris fleizach]

We need both these roles to be exposed directly to help solve https://bugs.webkit.org/show_bug.cgi?id=40132

[chris fleizach]

no functional changes. just prep work

[chris fleizach]

Created attachment 57796 [details]
Patch

[WebKit Review Bot]

Attachment 57796 [details] did not build on chromium:
Build output: http://webkit-commit-queue.appspot.com/results/3102006

[chris fleizach]

Created attachment 58083 [details]
patch

this should hopefully work for chomium too

[chris fleizach]

Created attachment 58084 [details]
patch

now with correct changelogs

[chris fleizach]

Created attachment 58086 [details]
patch

for some reason the last patch didn't take

[chris fleizach]

weird only details is showing up

[Darin Adler]

Comment on attachment 58086 [details]
patch

> +Node* AccessibilityRenderObject::node() const
> +{
> +    if (!m_renderer) 
> +        return 0;
> +    return m_renderer->node();
> +}

What is this function for? It's added in the patch and not used.

The patch otherwise seems great. review- because of the seemingly-unrelated change.

[chris fleizach]

(In reply to comment #8)
> (From update of attachment 58086 [details])
> > +Node* AccessibilityRenderObject::node() const
> > +{
> > +    if (!m_renderer) 
> > +        return 0;
> > +    return m_renderer->node();
> > +}
> 
> What is this function for? It's added in the patch and not used.
> 
> The patch otherwise seems great. review- because of the seemingly-unrelated change.

That was part of the follow up patch. Forgot to take it out of this one

[chris fleizach]

Created attachment 58664 [details]
patch

Patch is same as before, but removes the not necessary node() function from this patch

[Darin Adler]

Comment on attachment 58664 [details]
patch

> +        (WebCore::):

Lines like this should be deleted from the ChangeLog. And we should fix prepare-ChangeLog to not generate them.

> +        (RoleEntry::):

Ditto.

[chris fleizach]

http://trac.webkit.org/changeset/61151

[Eric Seidel (no email)]

Seems to have caused a Gtk API test regression:

ERROR:../../WebKit/gtk/tests/testatkroles.c:98:get_child_and_test_role: assertion failed: (child_role == role)

[Eric Seidel (no email)]

Sheriffbot didn't notice because the buildbot seems to have not been notified of this revision.  Not sure why or how.

[chris fleizach]

(In reply to comment #14)
> Sheriffbot didn't notice because the buildbot seems to have not been notified of this revision.  Not sure why or how.

hopefully i can fix this tonite

[chris fleizach]

looks like
http://trac.webkit.org/changeset/61176 fixed it