Closed Bug 387925 Opened 17 years ago Closed 17 years ago

STATE_EXPANDABLE isn't exposed correctly

Categories

(Core :: Disability Access APIs, defect)

x86
Linux
defect
Not set
normal

Tracking

()

RESOLVED FIXED

People

(Reporter: jdiggs, Assigned: surkov)

References

(Blocks 1 open bug)

Details

(Keywords: access)

Attachments

(2 files, 1 obsolete file)

Steps to reproduce: 1. Launch Thunderbird and Accerciser 2. Select "Local Folders" (or any other expandable table cell) in Thunderbird 3. Use Accerciser's Interface Viewer to examine the states of the object you chose in step 2. Expected results: "Expandable" would be included in the list of states. Actual results: "Expandable" is not included in the list of states.
Attached patch patch (obsolete) (deleted) — Splinter Review
it's a bit strange but treeitem accessible doesn't call nsAccessible::GetState(). I'm not sure why exactly. It leads we can loose some states like in this bug.
Assignee: aaronleventhal → surkov.alexander
Status: NEW → ASSIGNED
Attachment #272114 - Flags: review?(ginn.chen)
Attached patch patch (deleted) — Splinter Review
correct patch
Attachment #272114 - Attachment is obsolete: true
Attachment #272115 - Flags: review?(ginn.chen)
Attachment #272114 - Flags: review?(ginn.chen)
Attached patch states patch (deleted) — Splinter Review
It looks we never set some extra states because we call nsAccessible::GetState() in begin of nsAccClass::GetState(). So I guess we should set these states after nsAccClass::GetState(). Correct?
Attachment #272122 - Flags: review?(ginn.chen)
Comment on attachment 272115 [details] [diff] [review] patch reduce the number of reviews from Ginn.
Attachment #272115 - Flags: review?(ginn.chen) → review?(aaronleventhal)
Attachment #272122 - Flags: review?(ginn.chen) → review?(aaronleventhal)
Comment on attachment 272115 [details] [diff] [review] patch This patch is okay, the other one isn't.
Attachment #272115 - Flags: review?(aaronleventhal) → review+
Comment on attachment 272122 [details] [diff] [review] states patch We need this code in nsAccessible::GetState() because of states set in ARIA.
Attachment #272122 - Flags: review?(aaronleventhal) → review-
(In reply to comment #6) > (From update of attachment 272122 [details] [diff] [review]) > We need this code in nsAccessible::GetState() because of states set in ARIA. > Sorry? We can't keep this in nsAccessible::GetState() because these states won't be setted if related states are defined in successors of nsAccessible. Why we can't set them after all states are calculated?
Okay, you can move it to GetFinalState() -- does that make sense?
(In reply to comment #8) > Okay, you can move it to GetFinalState() -- does that make sense? > I moved there because we add those states after another states are generated on all accessible objects. I guess it makes more sense than ignore states that will be setted on successors of nsAccessible. Though in this case we can't use them in successors at all. Now we don't use them.
Comment on attachment 272122 [details] [diff] [review] states patch My mistake. What you did is correct, I didn't see the 2nd part of the patch.
Attachment #272122 - Flags: review- → review+
both patches was checked in
Status: ASSIGNED → RESOLVED
Closed: 17 years ago
Resolution: --- → FIXED
Summary: "Expandable" tree table cells should expose STATE_EXPANDABLE → STATE_EXPANDABLE isn't exposed correctly
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: