Re: [orca-list] Likely bug in braille word wrapping

[Joanmarie Diggs <jdiggs igalia com>]

Hey again.

Hopefully I've successfully fixed the creeping backward when panning to 
the left problem as well.

Given that I really need to focus on some other things, I admittedly 
didn't do much testing of either change. It would be super awesome if 
you could do so.

Thanks again!
--joanie

On 4/28/20 11:06, Joanmarie Diggs wrote:
> Hey Jason.
>
> This bug is now (hopefully) partially fixed. Namely the panning to the 
> right. Your suspicions were correct that the full display size was being 
> added without taking ranges into account. That's the part that is fixed.
>
> The creeping slowly back when moving to the left is still there at the 
> present time.
>
> Keep those suspicions coming! :)
> --joanie
>
> On 4/27/20 19:13, Jason White via orca-list wrote:
> > Adding this to Joanie's ever-growing queue.
> >
> > Note: I'm using a 40-cell braille display here.
> >
> > Steps to reproduce:
> >
> > 1. With UEB contracted braille and word wrap enabled, display the 
> > following line (taken from my emial this morning):
> >
> > git clone https://gitlab.gnome.org/GNOME/orca.git
> >
> > 2. Move the cursor to the beginning of the line. The braille display 
> > shows "git clone" (contracted, of course) - fine so far.
> >
> > 3. Pan the braille display to the right, then back to the left.
> >
> > Expected result: the display should show a continuation of the string 
> > ("http" etc. Upon panning back to the left, it should restore the 
> > original text (i.e., "git clone").
> >
> > Actual result: everything from "http" up to "NOME/orca" is skipped 
> > upon panning to the right, and panning to the left moves gradually 
> > back through the string instead of straight back to the beginning.
> >
> > I'll send Joanie full debug output separately.
> >
> > Having looked at the debug output and the code, I have suspicions 
> > about the "range" computation (with "git clone" displayed, the range 
> > is 48-57, but after panning, it becomes 88-128 (i.e., adding the full 
> > display size to starting offset). The range isn't adjusted because the 
> > remainder of the line won't fit on the display and cannot be 
> > subdivided as there are no spaces. It seems to me that the revised 
> > offset should perhaps still be 57, though, and this difference could 
> > account for the text being skipped. However, I didn't succeed in 
> > working out exactly what was happening by inspecting the source.
> >
> >
> > _______________________________________________
> > orca-list mailing list
> > orca-list gnome org
> > https://mail.gnome.org/mailman/listinfo/orca-list
> > Orca wiki: https://wiki.gnome.org/Projects/Orca
> > Orca documentation: https://help.gnome.org/users/orca/stable/
> > GNOME Universal Access guide: 
> > https://help.gnome.org/users/gnome-help/stable/a11y.html
>
> _______________________________________________
> orca-list mailing list
> orca-list gnome org
> https://mail.gnome.org/mailman/listinfo/orca-list
> Orca wiki: https://wiki.gnome.org/Projects/Orca
> Orca documentation: https://help.gnome.org/users/orca/stable/
> GNOME Universal Access guide: 
> https://help.gnome.org/users/gnome-help/stable/a11y.html