[XeTeX] XeTex and Japanese kunten and warichū

Tak Yato (ZR) zrbabbler at yahoo.co.jp
Sat Sep 4 12:31:41 CEST 2010


Hello,

I am one of the users on “TeX Forum” (a forum for Japanese TeX users,
http://oku.edu.mie-u.ac.jp/tex/) and argued on this topic with Clemente
Beghi the other day. It is my first post to this ML.

I have read the argument on this thread and inspected gezhu package
and sfkanbun package (a pLaTeX package for writing kunten). Here is my
thought on this issue. I apologize in advance that I do not provide
any actual solution here....

When properly typesetting Japanese and/or Chinese with Xe(La)TeX, there
are two thing that are very intricate but yet indispensable: space
adjustment (around punctuations and at alphabet-CJK boundaries) and
automatic font switch (between non-CJK and CJK). The packages xeCJK
and zhspacing both do those jobs (so they conflict with each other).
(Unfortunately there is no package (at least on CTAN) that is sufficient
for formal Japanese typesetting (in Japan), but I think xeCJK in
‘fullwidth’ pucntuation style will do permissible job.)

Such packages employs \XeTeXinterchartoks intervention in its own way,
and then I had believed that higher level process like kunten and
warichu must be deeply dependent on internals of CJK modules (i.e.
xeCJK and zhspacing). However, it seems that the functioning of gezhu
package is not related with any CJK settings; in fact, gezhu works
along with xeCJK, zhspacing, or 8-bit LaTeX plus CJK. Likewise, I
think the adaptation of sfkanbun package to other CJK-aware LaTeX
environment is not so difficult as I first thought.

Some additional notices:

- Although the whole kunten processing indeed involves some (wierd
kind of) ruby positioning (i.e. okurigana), kunten itself (re-ten,
ichi-ni-ten, etc.) is not in “left-ruby” position, but resides
at the left-bottom of a kanji to be annotated, that is, kunten
goes between kanji.

- As is clearly stated in the documentation of zhspacing, vertical
CJK typesetting in XeLaTeX has major drawback that mixing alphabet
and CJK will produce ugly results because of the different internal
baseline settings between alphabet and CJK fonts. This will not hurt
you when writing classical documents, but will make XeLaTeX less
attractive for modern vertical-writing usage.

- I am also interested in the topic of Polyglossia setup for Japanese.
But I would like to talk on it when I have more free time....

Anyway, I strongly believe it is necessary to give XeLaTeX the
ability to typeset Japanese documents that are acceptable in Japan,
as well as outside, because the time will be inevitably come for
Japanese TeX users to migrate from pTeX to modern Unicode-aware
TeX, i.e. XeTeX and/or LuaTeX.

Best regards,
Tak Yato


--------------------------------------
GyaO! - Anime, Dramas, Movies, and Music videos [FREE]
http://pr.mail.yahoo.co.jp/gyao/


More information about the XeTeX mailing list