529621 - When a table has only one column per row and that column happens to be a column header its role is exposed wrong

Status: RESOLVED FIXED

(Core :: Disability Access APIs, defect)

Description

[Peter Vágner]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; sk-SK; rv:1.9.3a1pre) Gecko/20091117 Lightning/1.1a1pre SeaMonkey/2.1a1pre
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; sk-SK; rv:1.9.3a1pre) Gecko/20091117 Lightning/1.1a1pre SeaMonkey/2.1a1pre

I think the summary says it all. A column header in the following table is exposed with role equals to 0 under Windows.
I am testing with NVDA so the steps to reproduce will count on it too.

Here is the example
<table>
<thead>
<tr>
<th>col1</th>
</tr>
</thead>
<tbody>
<tr>
<td>bla</td>
</tbody>
</tr>
</tbody>
</table>

Just to ilustrate the difference here is an example of a table which is exposed correctly:
<table>
<thead>
<tr>
<th>col1</th>
<th>col2</th>
</tr>
</thead>
<tbody>
<tr>
<td>bla</td>
<td>bla</td>
</tr>
</tbody>
</table>

Reproducible: Always

Steps to Reproduce:
1. Open the test case in seamonkey
2. navigate to the first row in a table.
3. arrow up and down to identify the issue, you can even try using object navigation.
Actual Results:  
NVDA will anounce the column header as unknown object

Expected Results:  
NVDA should announce column header

Comment 1

[Peter Vágner]

Attachment: the test case

While adding a test case I've noticed the issue can also be seen while browsing this page. Just examine the attachments table.

Comment 2

[Marco Zehe (:MarcoZ)]

This is definitely confirmed also with Firefox 3.6b3 and 3.7a1pre.

Comment 3

[David Bolter [:davidb] (NeedInfo me for attention)]

Can we get a regression window? (Perhaps near the IATable2 work?)

Comment 4

[Marco Zehe (:MarcoZ)]

It's definitely not a problem in Firefox 3.5.3, which is a build I happen to have lying around here randomly. Checking the tables-related ranges now.

Comment 5

[Peter Vágner]

I've upgraded yesterday evening to the Seamonkey version I've reported against.
Before I was running the following version of Seamonkey and the issue was not present there.
I am not sure if it narrows it somewhat.

Mozilla/5.0 (Windows; U; Windows NT 5.1; sk; rv:1.9.1.5pre) Gecko/20091021 SeaMonkey/2.0.1pre

Comment 6

[Marco Zehe (:MarcoZ)]

This was introduced in this build: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.2a1pre) Gecko/20090514 Minefield/3.6a1pre
The possible range is
http://hg.mozilla.org/mozilla-central/shortlog/6da448e1e340

Comment 7

[alexander :surkov (:asurkov)]

Attachment: patch

Comment 8

[Marco Zehe (:MarcoZ)]

Comment on attachment 413268 [details] [diff] [review]
patch

r=me, thanks!

Comment 9

[David Bolter [:davidb] (NeedInfo me for attention)]

Comment on attachment 413268 [details] [diff] [review]
patch

r=me thanks. I didn't review the tests.

>+
>+  // No elements in siblings what means the table has one column only. Therefore
>+  // it should be column header.
>+  *aRole = nsIAccessibleRole::ROLE_COLUMNHEADER;

nit:  remove "what" to make the comment more grammatical.

Comment 10

[David Bolter [:davidb] (NeedInfo me for attention)]

Comment on attachment 413268 [details] [diff] [review]
patch

192 drivers: this patch is very low risk, has tests and fixes a regression from 3.5.

Comment 11

[alexander :surkov (:asurkov)]

landed on 1.9.3 - http://hg.mozilla.org/mozilla-central/rev/5fc28d369d56

Comment 12

[Marco Zehe (:MarcoZ)]

Comment on attachment 413268 [details] [diff] [review]
patch

Moving request to 1.9.2.1. Regression from 3.5.

Comment 13

[Marco Zehe (:MarcoZ)]

Drivers, any word on this one? Should this be moved to 1.9.2.3 or are you still considering this for 1.9.2.2?

Comment 14

[David Bolter [:davidb] (NeedInfo me for attention)]

Drivers: 1 line, low risk, regression fix.

Comment 15

[Mike Beltzner [:beltzner, not reading bugmail]]

Comment on attachment 413268 [details] [diff] [review]
patch

a=beltzner for 1.9.2.4

Comment 16

[alexander :surkov (:asurkov)]

Attachment: patch1.9.2

mochitests are updated to 1.9.2

Comment 17

[alexander :surkov (:asurkov)]

landed on 1.9.2
http://hg.mozilla.org/releases/mozilla-1.9.2/rev/1f25c58b2fd6
http://hg.mozilla.org/releases/mozilla-1.9.2/rev/2ae9bff42d08

Comment 18

[Al Billings [:abillings - ex-MoCo]]

Marco, can you verify that this is fixed in the nightly 1.9.2 build?

Comment 19

[Marco Zehe (:MarcoZ)]

Verified fixed in Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.2.4pre) Gecko/20100412 Namoroka/3.6.4pre (.NET CLR 3.5.30729)