Tobias Burnus wrote:
Hi,
Xiao Jianfeng wrote:
What would be needed to get UTF-8 input running with Chinese?
If you use vim to edit your tex file, maybe you can try "set encoding=utf8", then save and compile. As far as I know, GBK is compatible with unicode.
No, that does not work - that is the reason I started this mail thread. You get the wrong characters and you may get some TeX errors. (And that is the reason Lutz wrote a UTF-8 to GBK converted.)
Hmm, I suspect that some remix between my old (deprecated) Libertine in ConTeXt recipe and the ttf2tfm automatic unicode splitting would have some positive effects. (I would discourage using that recipe for alphabetic (incl Roman) Unicode fonts because it blows away any kerning that would happen between unicode blocks. Is there less kerning among CJK fonts? I would expect so.) Thinking aloud, you'd probably want to include some language-switching commands, to mediate between the calling of unicode fonts for un-named CJK glyphs (just raw conversion from Unicode to font switch + glyph number) to named roman (and other alphabetic) glyphs (conversion from UTF-8 to named glyphs to font+glyph, which retains kerning where it can). I know it's sketchy and vague, but have a look inside font-uni. It's not the most complicated file in the distro. adam -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Adam T. Lindsay, Computing Dept. atl@comp.lancs.ac.uk Lancaster University, InfoLab21 +44(0)1524/510.514 Lancaster, LA1 4WA, UK Fax:+44(0)1524/510.492 -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-