Re: character-spacing & cursively-connected characters

[Owen Taylor <otaylor redhat com>]

On Thu, 2003-02-13 at 20:34, Andrew Dunbar wrote:
>  --- Damon Chaplin <damon karuna uklinux net> wrote: >
> On Thu, 2003-02-13 at 00:09, Matthias Clasen wrote:
> > > On Thu, 2003-02-13 at 01:53, Damon Chaplin wrote:
> > > > 
> > > > Is it possible to determine if 2 clusters are
> > > > cursively-connected? (i.e. the glyphs are
> > > > joined up in some way.)
> > > > 
> > > > I think in some scripts like Arabic nearly all
> > > > clusters are connected, and you can even get
> > > > cursive Latin fonts.
> > > > 
> > > > I'm thinking of using character spacing in my
> > > > justification routines, maybe just as a
> > > > fallback for when word-spacing fails. But I
> > > > can't really do that for cursively-connected
> > > > clusters.
> > > > 
> > > 
> > > How can word-spacing "fail" ?
> > 
> > If only one word fits on the line there may be no
> > spaces to expand.
> > 
> > The TeX algorithm also tries very hard not to
> > stretch spaces above a certain ratio. So this is
> > another sort of failure case.
> 
> Also, some languages including Thai discourage spaces
> between words, and other languages including Khmar
> completely forbid spaces between words.
> I'm not sure if there is a method of justification
> used in print in these languages.

If I'm not mistaken, for Thai, you insert space between
the words when justifying, though you don't for
text set at the natural width.

Regards,
                                     Owen