After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 575319 - Problems with GtkAssistant accessibility
Problems with GtkAssistant accessibility
Status: RESOLVED FIXED
Product: gtk+
Classification: Platform
Component: Widget: Other
2.16.x
Other Linux
: Normal normal
: ---
Assigned To: gtk-bugs
gtk-bugs
Depends on:
Blocks:
 
 
Reported: 2009-03-14 02:26 UTC by Michael Terry
Modified: 2009-05-29 17:34 UTC
See Also:
GNOME target: ---
GNOME version: ---


Attachments
vala test case (781 bytes, text/x-vala)
2009-03-14 02:27 UTC, Michael Terry
  Details
C test case (1.43 KB, text/x-csrc)
2009-03-14 02:28 UTC, Michael Terry
  Details
C test case header (341 bytes, text/x-chdr)
2009-03-14 02:28 UTC, Michael Terry
  Details
Cleaned-up C test case (1.23 KB, text/x-csrc)
2009-03-16 14:10 UTC, Michael Terry
  Details
Proposed patch (505 bytes, patch)
2009-05-29 10:59 UTC, Eitan Isaacson
committed Details | Review

Description Michael Terry 2009-03-14 02:26:51 UTC
I may be misunderstanding something.  Please correct me if so.  I can't get GtkAssistant and ldtp to play nice.

Namely, I don't get the expected results when running 'getobjectlist' on an Assistant.

I've attached a test case.  It's in vala and you can run "valac assist.vala --pkg gtk+-2.0" to compile.  I've also attached the generated C code if that's easier.

After running the test case (and optionally pressing 'Apply' in the dialog), I run the following in a different console:

python -c 'import ldtp; print ldtp.getobjectlist("frmAssist")'

I get the following reply:

[u'btnLast', u'frmAssist', u'btnForward', u'btnClose', u'lblSetup1', u'flr0', u'btnCancel', u'btnBack', u'lblSetup', u'btnApply']

I don't know what lblSetup1 is supposed to be.  I'm surprised I don't see 'lblSummary', 'lblsetuplabel', or 'lblsummarylabel'.  Why don't I?  (See the test case to know what I mean.)
Comment 1 Michael Terry 2009-03-14 02:27:26 UTC
Created attachment 130624 [details]
vala test case
Comment 2 Michael Terry 2009-03-14 02:28:13 UTC
Created attachment 130625 [details]
C test case
Comment 3 Michael Terry 2009-03-14 02:28:29 UTC
Created attachment 130626 [details]
C test case header
Comment 4 Nagappan Alagappan 2009-03-16 10:24:03 UTC
(In reply to comment #0)
> I may be misunderstanding something.  Please correct me if so.  I can't get
> GtkAssistant and ldtp to play nice.
> 
> Namely, I don't get the expected results when running 'getobjectlist' on an
> Assistant.
> 
> I've attached a test case.  It's in vala and you can run "valac assist.vala
> --pkg gtk+-2.0" to compile.  I've also attached the generated C code if that's
> easier.

Thanks for the detailed info :)

> 
> After running the test case (and optionally pressing 'Apply' in the dialog), I
> run the following in a different console:
> 
> python -c 'import ldtp; print ldtp.getobjectlist("frmAssist")'
> 
> I get the following reply:
> 
> [u'btnLast', u'frmAssist', u'btnForward', u'btnClose', u'lblSetup1', u'flr0',
> u'btnCancel', u'btnBack', u'lblSetup', u'btnApply']
> 
> I don't know what lblSetup1 is supposed to be.  I'm surprised I don't see
> 'lblSummary', 'lblsetuplabel', or 'lblsummarylabel'.  Why don't I?  (See the
> test case to know what I mean.)
> 

When I tried with accerciser application, I notice the list only generated by LDTP. No additional information. We may need to involve some one from valac team to debug this issue. For sure, this is not a LDTP bug :)

Thanks
Comment 5 Michael Terry 2009-03-16 14:03:46 UTC
Definitely not a vala bug.  I just attached the vala source because it was quicker for me to mock it up that way.  But really, the source is small.  A cleaned-up, C-only test case is coming shortly, to assuage any fears that it has something to do with vala.

I also see the same results with accerciser, so yeah, I guess it's not a ldtp bug either.  I'm switching this bug to GTK+...  Is it possible GtkAssistant just doesn't expose the widgets correctly via AT-SPI?
Comment 6 Michael Terry 2009-03-16 14:08:26 UTC
Definitely not a vala bug.  I just attached the vala source because it was quicker for me to mock it up that way.  But really, the source is small.  A cleaned-up, C-only test case is coming shortly, to assuage any fears that it has something to do with vala.

I also see the same results with accerciser, so yeah, I guess it's not a ldtp bug either.  I'm switching this bug to GTK+...  Is it possible GtkAssistant just doesn't expose the widgets correctly via AT-SPI?
Comment 7 Michael Terry 2009-03-16 14:10:55 UTC
Created attachment 130747 [details]
Cleaned-up C test case

Compile with "gcc `pkg-config --cflags --libs gtk+-2.0` assist.c -o assist"
Comment 8 Michael Terry 2009-03-17 18:08:56 UTC
Switching to 'atk' product; I thought that was under the gtk+ umbrella, but guess not.
Comment 9 Eitan Isaacson 2009-05-29 10:59:12 UTC
Created attachment 135549 [details] [review]
Proposed patch

Hey Terry.

Please try this, let me know if it solves all or some of the symptoms. Also, for the record, could you use accerciser and list the exact issues you encountered? The LDTP diagnosis above is a bit hard to read.
Comment 10 Michael Terry 2009-05-29 13:16:56 UTC
Oh, glorious days; that fixed it!  Thank you!

Noted about accerciser.
Comment 11 Eitan Isaacson 2009-05-29 14:14:17 UTC
Excellent!

Could someone please commit this?
Comment 12 Matthias Clasen 2009-05-29 16:53:41 UTC
Please commit to master and gtk-2-16
Comment 13 Eitan Isaacson 2009-05-29 17:34:25 UTC
Thanks Matthias,

Committed to master and gtk-2-16