[XeTeX] Traditional TeX ligatures once again

Evgenie Medvedev medvedev at project7.ru
Thu Jun 28 15:21:16 CEST 2007


I happen to write mostly in Russian and use utf-8 out of sheer
stubbornness everywhere where at all possible, and I like the pdf files
XeTeX produces a lot. I am trying to set up a document workflow based on
XeTeX, in such a way as to keep my LaTeX source files as compatible
across all the various *TeX flavours that can consume utf-8 input as
reasonably possible, so that I could give them away if necessary, maybe
even to people who don't have or use XeTeX. I'm not having much success,
probably because I'm just a few days into this.

The trend around the cyrillic TeX community is to use "<<" and ">>" for
french-style quotes (\guillemotleft and \guillemotright) which are
standard typography in Russian, and I would rather keep this convention.
This is unavoidable, since Russian keyboards don't have these characters
either.

Unfortunately, while these << and >> sequences translate into ligatures
in pre-XeTeX fonts, so it worked before most of the time, they are not
defined in tex-text font mapping as of version 0.996, nor in the
subversion repository, and as far as I can see, there's no good way to
fix that with xunicode or using the macros provided by xunicode. Using
"< and "> as described in babel docs produces symbols which do not
translate to U+00AA and U+00AB in the resulting pdf file, and don't look
right anyway. Maybe it would be a good idea to add these mappings in a
future version? Meanwhile, how do I define these sequences in the
document itself, or nearby, without making my TexLive setup incompatible
with everybody else and without using a preprocessor? What should I do
if I decide to forgo compatibility, after editing tex-text.map to fit?

I've been tearing my hair out for most of yesterday and today and still
no clue.

I've also been struggling with \cyr<letter> not being defined while
required by babel, (strange, but I got no problems with hypenation which
have been reported so often, so far) but found the cyrunicode.tex in
this list's archives, which fixed it -- could it be eventually added to
xunicode, or something? Or is it a question of testing to see if it's
complete and correct? Being a native speaker, I could do that.

-- 
Evgenie Medvedev


More information about the XeTeX mailing list