374712 - Expose posinset/setsize/level for all trees

Status: RESOLVED DUPLICATE of bug 371594

(Core :: Disability Access APIs, defect)

Description

[parente]

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9a3pre) Gecko/20070316 Minefield/3.0a3pre

Object attributes are missing from referenced example.

Comment 1

[Aaron Leventhal]

It's a pretty easy fix:
1) Move the GetDescription() code that calculates positional info out of the MSAA wrap classes and into the GetAttributes() for each widget. Don't do it in a wrap class, because those positional attributes are important on ATK too.
2) Make MSAA's nsAccessibleWrap::get_accDescription() just use those object attributes, when on one of those widgets where it makes sense to provide positional info.

Comment 2

[Aaron Leventhal]

Related to bug 371594 (maybe fixed by that patch, I need to look).

Comment 3

[alexander :surkov (:asurkov)]

(In reply to comment #2)
> Related to bug 371594 (maybe fixed by that patch, I need to look).
> 

Yea, related. The patch exposes those ARIA attributes for those accessibles that should have them. I just didn't get why we should move something from acc_getDescription() to GetDescription. Does ATK have not something similar with IA2 groupPosition or getAttributes?

Comment 4

[Aaron Leventhal]

> I just didn't get why we should move something from
> acc_getDescription() to GetDescription. 
It's the other way around. We want to do it for get_accDescription() so that it only affects MSAA usage on Windows, but not nsIAccessible usage on Windows. get_accDescription() was never really standard, but a few people used it because there was no other way. We will need keep it for a while until AT's move to the new correct IA2 groupPosition method for a couple years. The get_accDescription() way of getting this info will still be supported for now but it's deprecated once we have groupPosition and getAttribute()

> Does ATK have not something similar with IA2 groupPosition or getAttribute.
ATK didn't have anything so we are using getAttribute() there.

Comment 5

[alexander :surkov (:asurkov)]

I guess dupe of bug 371594?