[BRLTTY] xbrlapi cannot translate unicode characters defined in text table

[Samuel Thibault]

Hello,

payman shaykhmehdi, on Wed 27 Apr 2016 11:25:46 +0430, wrote:

> and this keysym is different from what
> I have defined in text table. For example, I have defined the following mapping
> in text table for ARABIC COMMA :
> 
> char \u060C (    5   )
> 
> and when I press Dot5 key, xbrlapi says :  couldn't translate keysym "0100060C"
> to keycode.

As Dave explained, this is expected.

> couldn't translate keysym "########" to keycode,

> Is it anything wrong with this char directive or some configuration needs to be
> done with xbrlapi or OS?

Actually it's the OS.  Currently xbrlapi is relatively dumb: it is only
able to translate a keysym to a key press simulation if that involves
only one key press (optionally with a modifier). This means that you
need to have loaded a keymap which allows to type the letter directly or
with shift, alt, altgr. There are ideas to make xbrlapi automatically
dynamically remap an unused keycode before simulating it, it's a bit
ugly, but I guess we'll have to resort to using that so as to properly
emit some letters which really need several keypresses, such as letter
with combining accents.

Samuel