# [texhax] how to get from CJK from emacs to UTF-8

Thu Nov 6 15:48:21 CET 2003

(This was posted to Mac OS X TeX, usenet:comp.text.tex and the TUB-Prod
list previously, my apologies to those who're seeing it for the second,
third or even fourth time)

Okay, thanks to Peder Axensten's very generous editng posting of
Olivier Delloye's instructions (on the Mac OS X TeX list), I now have
CJK working in TeXshop.

Unfortunately, I did this so as to typeset a paper composed in emacs,
and something's not quite agreeing betwixt the twain.

The header of the paper in question is (now):
\documentclass[preprint,numbersec]{ltugproc}
\usepackage{array,verbatim}
\usepackage{hgmacs,pinyin}
\usepackage[dvips]{graphics}

\usepackage[encapsulated]{CJK}
\usepackage{ucs}
\usepackage[utf8]{inputenc}

which seems to not quite be compatible.

I now get errors along the lines of:

! Undefined control sequence.
\u-default-960 #1->\textpi

l.254 ...^^c3^^9f^^c2^^a1 ^^c3^^b2^^c2^^a8^^cf^^80

^^c3^^b2^^c5^^93^^e2^^80^...

More seriously, while I've been able to use the Pinyin input scheme to
re-enter those characters for which Pinyin transliterations were
provided, a fair number of them aren't so precisely delineated, so I've
got lots of character runs like:

For example, the stroke
\chnstrok{p} (ò Ú) can have two more variations: \chnstrok{P} (ò®Œò Ú
\py{pi`ngpiev}) (as the top
stroke in ò¶∆) and \chnstrok{q} (ò—®ò Ú \py{shu`piev}) (as the left
most stroke
in òßÕ).

and I'm not seeing an obvious way to get from that to UTF-8.

It's my understanding that when emacs typesets CJK it includes some
directives automagically---anyone know what those are?

William

--