[Groff] Re: What's missing for Unicode support of groff?

[Werner LEMBERG]

> > I'll work with gpreconv, looking into your and Bernd Haible's
> > solution to make a water-proof UTF-8 -> groff entity conversion.
> 
> Two modifications of gpreconv are attached.
> They both output \[uXXXX] instead of UTF-8.
> One version relies exclusively on iconv for conversion,
> while the other retains some built-in coversion
> routines from original gpreconv.

Thanks.  Note that I don't like libutf-8 (besides licensing issues see
http://jamesthornton.com/linux/HOWTO/Unicode-HOWTO-6.html).  Instead,
I want handling of UTF-8 completely in gpreconv.

> But there are "not implemented yet" comments
> in roff-related parts of the code.

This comment is incorrect -- Tomohiro has forgotten to remove it.

> Please, take a look at
> http://www.thessalonica.org.ru/en/fonts.html.

Nice.

> And what exactly do you need?  A complete set of base PS fonts with
> additional glyphs?

Yes.  This is what the Cyrillic extension of the URW fonts actually
do.  It's a pity that those fonts are so buggy.

> As a partial workaround, there are some tables floating around the
> Net, that describe ways to compose some extedned Latin glyphs from
> glyphs in base PS fonts.

My main concern isn't Latin but Cyrillic and (modern) Greek.  On the
other hand, it is not that important because groff's main usage is the
display of man pages, and most people do that on a TTY.  It should be
straightforward to create groff support packages for fonts which can't
be distributed with groff directly -- in the TeX world this has
already been done for many fonts.

I'll now concentrate on integrating gpreconv.


    Werner