[BRLTTY] Re: Speech enhancements

[Yannick PLASSIARD]

On Fri, Mar 17, 2006 at 10:42:43AM +0100, Samuel Thibault wrote:
> Hi,
> 
> Dave Mielke, le Fri 17 Mar 2006 03:01:46 -0500, a écrit :
> > [quoted lines by Jason White on 2006/03/17 at 18:11 +1100]
> > 
> > >Perhaps the word identification function should be accessed via a pointer that
> > >can be updated depending on the language of the input.
> > 
> > At the very least, the initial code, which, for expediency, might as well be
> > something simple, should be encapsulated in a function.  That'd make it easy to
> > change when future generalizations are adequately understood.
> 
> Well, ideally enough, such function should be provided by _libc_
> itself...
> 
> >From reading the grep source code about word separation, it seems
> that they just look for contiguity of the isalnum() property of
> characters. This is an idea of an easy fully-localized first
> implementation for brltty.
> 
> The issue of developing better word separation should be developped in
> libc, not in brltty.
	I aggree with Samuel , it depends on the locale and for that, it would be done in conjunction with the locale into the GNU C library. To let us play with "echo by word" features, it seems to be approved to do a "basic word separation" (with blanks (spaces tabs ...)), and see users' feedbacks.

> 
> Regards,
> Samuel
> _______________________________________________
> This message was sent via the BRLTTY mailing list.
> To post a message, send an e-mail to: BRLTTY at mielke.cc
> For general information, go to: http://mielke.cc/mailman/listinfo/brltty