GNOME Bugzilla – Bug 542840
We need a Gecko-specific getBrailleContext()
Last modified: 2015-07-24 18:05:35 UTC
The script.py for the Gecko toolkit has its own updateBraille() method which does not call braillegenerator.getBrailleContext() like the default script's updateBraille() does. Presumably doing a blanket call to getBrailleContext() would result in context being added which we don't necessarily want. In addition, the Gecko toolkit and document content is sufficiently different that once we work out when we should and should not getBrailleContext(), the braillegenerator's getBrailleContext() is not likely to do the right thing in all cases. I think we need a Gecko-specific getBrailleContext() -- and need to figure out when to invoke it. :-)
The original purpose of getBrailleContext was to include information for labelled containers in the object's hierarchy (e.g., a panel named "Personal Information:" holding several items such as text entry fields for firstname, lastname, address, etc.). In speech, we try to limit the amount of context spoken by comparing hierarchies from the last object with focus. In braille, we try to show it all all the way up to the application. As with speech, we cheated and added row/col header information in as part of the hierarchical context, and we did so in kind of an ungraceful way. I wondering if it might be possible to try to make a new script method that obtains row/col header information for an object and then call *that* method from the appropriate spots.
Given everything else on the to-do list, non-critical Gecko bugs are going to have to wait until after 3.0, unless someone volunteers to do them.