GNOME Bugzilla – Bug 486976
[verified]Basic Where Am I in multiline text areas should only present info for current line
Last modified: 2008-07-22 19:32:44 UTC
Steps to reproduce: 1) Run the Application main window demo of gtk-demo 2) Type these lines in the main window: I like monkeys. They swing from trees. They have long tails and knocky knees. 3) Select "swing from trees. They have" leaving the cursor just after the 'e' in 'have'. 4) KP_Enter should read just current line or just the selected text on the current line if text is selected on that line. So, in this case, only "'They have' 'selected'" should be presented. This is the bug. 5) With no text selected on the line, KP_Enter should read the entire line with the caret. This works. 6) Double KP_Enter should speak all selected text. This seems good, too.
See also test/keystrokes/gtk-demo/role_text_multiline.py
Created attachment 99937 [details] [review] proposed patch Tested in gtk-demo, gedit, and OOo Writer. Mike, please test. role_text_multiline.py didn't have an existing test marked with a "bug," so I added some new tests to verify the functionality. In addition, one of the existing tests was failing for me -- off by one character -- independent of this bug. So.... Will, I went ahead and changed the test in question. When I run the attached on my box I get 10 out of 10 successes. Mind giving this a spin on your test box -- and if it fails, letting me know how you would like me to proceed. Already pylinted. Still 10/10. Woo hoo! ;-)
Created attachment 99938 [details] [review] pylint doesn't proofread the comments. D'oh! Corrected a typo in the comment of the test file.
(In reply to comment #3) > Created an attachment (id=99938) [edit] The where_am_I.py patch looks good (please commit that). The differences in the tests are disconcerting. The old way works on my machine and the new changes fail. I looking at this, it seems as though it is occurring after down arrowing from the middle of a line. I wonder if there might be a font difference between your machine and mine. If this is the case, and you are using the default system fonts, then the test cases probably should be changed to allow for the same test situation to happen more reliably. For example, instead of just down arrowing, maybe it needs to be a down arrow, followed by a home, followed by a few right arrows?
Created attachment 100064 [details] [review] The where_am_I.py patch (i.e. sans test assertion changes) Here's what I just committed. I'll look at the testing stuff again in a bit -- after I commit the other where_am_I stuff. Thanks! BTW, I am indeed using the default system fonts, and the default everything else. The only two exceptions are left-handed mouse and the fact that I have a printer installed. I need to set up the Solaris portion of this box as a proper testing environment so I can stop asking you to try things. :-) I'll do that this evening as well. But thanks for giving them a try! :-)
Changing the status to pending as the bug fix has been committed, but adding my name as a reminder that I still need to do something. :-) Shall we adjust the target milestone?
Created attachment 100077 [details] [review] revised test assertions This one works for me in both Ubuntu (hardy) and OpenSolaris (SNV 77 with Vermillion 79) and seems to still test what we want to test. Will please take a look and give it a try on your test box. Thanks!
Works great for me. Thanks!
Thanks. Committed the test assertions patch. This bug is truly "pending" now. :-)
works great
Thanks. Closing as FIXED.