[Webkit-unassigned] [Bug 21546] [GTK] ATK accessibility enhancements

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Mar 31 08:08:46 PDT 2009


https://bugs.webkit.org/show_bug.cgi?id=21546





------- Comment #7 from apinheiro at igalia.com  2009-03-31 08:08 PDT -------
Created an attachment (id=29115)
 --> (https://bugs.webkit.org/attachment.cgi?id=29115&action=view)
Update the first patch

When I tried to use the patch at the comment 1, this doesn't apply directly,
and in the same way, it doesn't compile, so I needed to fix these problems. I
suppose that it was a experimental patch. I also made the minor change
suggested at comment 3. I don't apply the changes at comment 4 as they are
mostly structural, but almost with no additional functionality.

In the same way I made some minor changes:

  * I commented some parts of code, to avoid "not used definition" compile
warnings
  * I change the call to AXObjectCache->get for AXObjectCache->getOrCreate on
some places. The first one was on webkitwebview.cpp, when you try to get the
AccessibleObject you are wrapping. I needed to do that because the call to
->get returned NULL, so no DOM tree was showed on Accerciser. I needed to do
that as well in other places, in order to avoid crashes.
  * As far as I see, the only place where the cache is updated is on
getOrCreate. In the same way, I think that has sense that a cache should work
always in this way, checking if you have the object yet, and if not, create the
object, so getOrCreate should be the most called function. Probably I'm wrong,
but in this case, we should add NULL checks on the ->get return, as can cause
crashes.
  * About the functionality, currently the position is not computed properly.
You can see it easily with the Accerciser, and see that the html objects are
not positionated properly (although it seems that computes properly the size).
The AtkComponent implementation were added with the patch at comment 1, so
probably it needs to be reviewed. I don't know if this affect to Orca or other
screen readers, but should be reviewed.
  * I experiment other crashes, but I can't conclude if was again for the
get/getOrCreate problem yet.


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



More information about the webkit-unassigned mailing list