[Webkit-unassigned] [Bug 25415] [GTK][ATK] Please implement support for get_text_at_offset

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu May 21 18:42:15 PDT 2009


------- Comment #16 from joanmarie.diggs at gmail.com  2009-05-21 18:42 PDT -------
Xan, given your work on this bug and other recent fixes, I'm thinking we're
approaching the point where it would be worth getting a
skeleton/debugging/testing script for WebKit in place in Orca.
(Up to now, I've done all testing in Accerciser.) Unfortunately, the mere act
of using the GtkLauncher or Epiphany while Orca is running crashes
GtkLauncher/Epiphany 100% of the time with this error:

getTextHelper(GetTextFunctionType, AtkText*, gint, AtkTextBoundary, gint*,
gint*): code should not be reached

Looking at what code in Orca was triggering the crash, it turned out to be
attempting to get the text of a line. 

I realize that you're still working on getting the visible line. The problem is
that while Orca speaks the character, word, etc. based on how the user is
navigating, it ALWAYS brailles the line. Hence the constant crashes.

I'm guessing that the problem is the assertion at the bottom of this block:

+        if (boundaryType == ATK_TEXT_BOUNDARY_WORD_START)
+            predicate = isWordStart;
+        else if (boundaryType == ATK_TEXT_BOUNDARY_WORD_END)
+            predicate = isWordEnd;
+        else if (boundaryType == ATK_TEXT_BOUNDARY_SENTENCE_START)
+            predicate = isSentenceStart;
+        else if (boundaryType == ATK_TEXT_BOUNDARY_SENTENCE_END)
+            predicate = isSentenceEnd;
+        else
+            g_assert_not_reached();

If so, would it be possible to return ('', 0, 0) instead? If you'd like, you
could also display a grumpy error message. ;-) ;-) That way, while you're
working on the line implementation, I could start implementing support in Orca
for things which are not lines.

If this would be more hassle than it's worth, no worries, I'll wait.

Thanks either way!

Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

More information about the webkit-unassigned mailing list