Re: Possible change to startup.el

[Kenichi Handa <handa@m17n.org>]

In article <20050323.190146.104245427.wl@gnu.org>, Werner LEMBERG <wl@gnu.org> writes:
> For this Thai text string (in UTF-8 encoding) I just get boxes.  Doing
> `C-u C-x =' I see that the characters are taken from the
> `mule-unicode-0100-24ff' charset; the font I use for it
> (-Efont-Biwidth-Medium-R-Normal--24-240-75-75-P-120-ISO10646-1)
> doesn't have them.  On the other hand, TIS-620 encoding displays just
> fine.  How can I make Emacs use my TIS-620 font for Thai characters if
> the encoding is UTF-8?  Neither emacs.info nor elisp.info give a clue.

Emacs is still not that good at handling Unicode characters.
It is better to use/recommend the charset thai-tis620 for
Thai characters for the moment.

It's possible to display Thai characters in
mule-unicode-0100-24ff by TIS-620 font by the attached code.
But, I don't want to commit such an adhoc code at this
moment.

(define-translation-table 'thai-tis620-encode-table
  ucs-thai-tis620-encode-table)

(define-ccl-program ccl-encode-tis620-font
  '(0
    ((if (r2 > 0)
	 ((r1 = ((r1 << 7) | r2))
	  (r2 = 0)
	  (translate-character thai-tis620-encode-table r0 r1)))
     (r1 |= #x80)))
  "CCL program to encode Thai characters to TIS620 font.")

(add-to-list 'font-ccl-encoder-alist '("tis620" . ccl-encode-tis620-font))

(set-fontset-font
 "fontset-default"
 (cons (decode-char 'ucs #x0E00)
       (decode-char 'ucs #x0E7F))
 '(nil . "tis620"))

Another way is to use setup utf-fragmentation-table property
and set utf-fragment-on-decoding to t so that Emacs decodes
Thai characters in utf-8 not to mule-unicode-0100-24ff but
to thai-tis620.  But, that is also fairly an adhoc way.

---
Ken'ichi HANDA
handa@m17n.org