lisp/language/china-util.el

   1 ;;; china-util.el --- utilities for Chinese
   2
   3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
   4 ;; Licensed to the Free Software Foundation.
   5
   6 ;; Keywords: mule, multilingual, Chinese
   7
   8 ;; This file is part of GNU Emacs.
   9
  10 ;; GNU Emacs is free software; you can redistribute it and/or modify
  11 ;; it under the terms of the GNU General Public License as published by
  12 ;; the Free Software Foundation; either version 2, or (at your option)
  13 ;; any later version.
  14
  15 ;; GNU Emacs is distributed in the hope that it will be useful,
  16 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 ;; GNU General Public License for more details.
  19
  20 ;; You should have received a copy of the GNU General Public License
  21 ;; along with GNU Emacs; see the file COPYING.  If not, write to the
  22 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  23 ;; Boston, MA 02111-1307, USA.
  24
  25 ;;; Code:
  26
  27 ;;;###autoload
  28 (defun setup-chinese-gb-environment ()
  29   "Setup multilingual environment (MULE) for Chinese GB2312 users."
  30   (interactive)
  31   (setup-english-environment)
  32
  33   (set-default-coding-systems 'chinese-iso-8bit)
  34   (setq coding-category-iso-8-2 'chinese-iso-8bit)
  35   (setq coding-category-iso-7-else 'chinese-iso-7bit)
  36   (setq coding-category-big5 'chinese-big5)
  37
  38   (set-coding-priority
  39    '(coding-category-iso-7
  40      coding-category-iso-7-else
  41      coding-category-iso-8-2
  42      coding-category-big5
  43      coding-category-iso-8-1
  44      coding-category-emacs-mule
  45      coding-category-iso-8-else))
  46
  47   (setq-default buffer-file-coding-system 'chinese-iso-8bit)
  48   (setq default-terminal-coding-system 'chinese-iso-8bit)
  49   (setq default-keyboard-coding-system 'chinese-iso-8bit)
  50
  51   (setq default-input-method  "chinese-py-punct"))
  52
  53 ;;;###autoload
  54 (defun setup-chinese-big5-environment ()
  55   "Setup multilingual environment (MULE) for Chinese Big5 users."
  56   (interactive)
  57   (setup-english-environment)
  58
  59   (set-default-coding-systems 'chinese-big5)
  60   (setq coding-category-iso-8-2 'chinese-big5)
  61   (setq coding-category-iso-7-else 'chinese-iso-7bit)
  62   (setq coding-category-big5 'chinese-big5)
  63
  64   (set-coding-priority
  65    '(coding-category-iso-7
  66      coding-category-iso-7-else
  67      coding-category-big5
  68      coding-category-iso-8-2
  69      coding-category-emacs-mule
  70      coding-category-iso-8-else))
  71
  72   (setq-default buffer-file-coding-system 'chinese-big5)
  73   (setq default-terminal-coding-system 'chinese-big5)
  74   (setq default-keyboard-coding-system 'chinese-big5)
  75
  76   (setq default-input-method "chinese-py-punct-b5"))
  77
  78 ;;;###autoload
  79 (defun setup-chinese-cns-environment ()
  80   "Setup multilingual environment (MULE) for Chinese CNS11643 family users."
  81   (interactive)
  82   (setup-english-environment)
  83
  84   (setq coding-category-iso-7-else 'chinese-iso-7bit)
  85   (setq coding-category-big5 'chinese-big5)
  86   (setq coding-category-iso-8-2 'chinese-big5)
  87   (set-default-coding-systems 'chinese-iso-7bit)
  88
  89   (set-coding-priority
  90    '(coding-category-iso-7
  91      coding-category-iso-7-else
  92      coding-category-iso-8-2
  93      coding-category-big5
  94      coding-category-iso-7-else))
  95
  96   (setq-default buffer-file-coding-system 'chinese-iso-7bit)
  97   (setq default-terminal-coding-system 'chinese-iso-7bit)
  98   (setq default-keyboard-coding-system 'chinese-iso-7bit)
  99
 100   (setq default-input-method "chinese-quick-cns"))
 101
 102 ;; Hz/ZW encoding stuffs
 103
 104 ;; HZ is an encoding method for Chinese character set GB2312 used
 105 ;; widely in Internet.  It is very similar to 7-bit environment of
 106 ;; ISO-2022.  The difference is that HZ uses the sequence "~{" and
 107 ;; "~}" for designating GB2312 and ASCII respectively, hence, it
 108 ;; doesn't uses ESC (0x1B) code.
 109
 110 ;; ZW is another encoding method for Chinese character set GB2312.  It
 111 ;; encodes Chinese characters line by line by starting each line with
 112 ;; the sequence "zW".  It also uses only 7-bit as HZ.
 113
 114 ;; ISO-2022 escape sequence to designate GB2312.
 115 (defvar iso2022-gb-designation "\e$A")
 116 ;; HZ escape sequence to designate GB2312.
 117 (defvar hz-gb-designnation "~{")
 118 ;; ISO-2022 escape sequence to designate ASCII.
 119 (defvar iso2022-ascii-designation "\e(B")
 120 ;; HZ escape sequence to designate ASCII.
 121 (defvar hz-ascii-designnation "~}")
 122 ;; Regexp of ZW sequence to start GB2312.
 123 (defvar zw-start-gb "^zW")
 124 ;; Regexp for start of GB2312 in an encoding mixture of HZ and ZW.
 125 (defvar hz/zw-start-gb (concat hz-gb-designnation "\\|" zw-start-gb))
 126
 127 (defvar decode-hz-line-continuation nil
 128   "Flag to tell if we should care line continuation convention of Hz.")
 129
 130 ;;;###autoload
 131 (defun decode-hz-region (beg end)
 132   "Decode HZ/ZW encoded text in the current region.
 133 Return the length of resulting text."
 134   (interactive "r")
 135   (save-excursion
 136     (save-restriction
 137       (narrow-to-region beg end)
 138
 139       ;; We, at first, convert HZ/ZW to `iso-2022-7bit',
 140       ;; then decode it.
 141
 142       ;; "~\n" -> "\n"
 143       (goto-char (point-min))
 144       (while (search-forward "~" nil t)
 145         (if (= (following-char) ?\n) (delete-char -1))
 146         (if (not (eobp)) (forward-char 1)))
 147
 148       ;; "^zW...\n" -> Chinese GB2312
 149       ;; "~{...~}"  -> Chinese GB2312
 150       (goto-char (point-min))
 151       (let ((chinese-found nil))
 152         (while (re-search-forward hz/zw-start-gb nil t)
 153           (if (= (char-after (match-beginning 0)) ?z)
 154               ;; ZW -> iso-2022-7bit
 155               (progn
 156                 (delete-char -2)
 157                 (insert iso2022-gb-designation)
 158                 (end-of-line)
 159                 (insert iso2022-ascii-designation))
 160             ;; HZ -> iso-2022-7bit
 161             (delete-char -2)
 162             (insert iso2022-gb-designation)
 163             (let ((pos (save-excursion (end-of-line) (point))))
 164               (if (search-forward hz-ascii-designnation pos t)
 165                   (replace-match iso2022-ascii-designation)
 166                 (if (not decode-hz-line-continuation)
 167                     (insert iso2022-ascii-designation)))))
 168           (setq chinese-found t))
 169         (if (or chinese-found
 170                 (let ((enable-multibyte-characters nil))
 171                   ;; Here we check if the text contains EUC (China) codes.
 172                   ;; If any, we had better decode them also.
 173                   (goto-char (point-min))
 174                   (re-search-forward "[\240-\377]" nil t)))
 175             (decode-coding-region (point-min) (point-max) 'euc-china)))
 176
 177       ;; "~~" -> "~"
 178       (goto-char (point-min))
 179       (while (search-forward "~~" nil t) (delete-char -1))
 180       (- (point-max) (point-min)))))
 181
 182 ;;;###autoload
 183 (defun decode-hz-buffer ()
 184   "Decode HZ/ZW encoded text in the current buffer."
 185   (interactive)
 186   (decode-hz-region (point-min) (point-max)))
 187
 188 ;;;###autoload
 189 (defun encode-hz-region (beg end)
 190   "Encode the text in the current region to HZ.
 191 Return the length of resulting text."
 192   (interactive "r")
 193   (save-excursion
 194     (save-restriction
 195       (narrow-to-region beg end)
 196
 197       ;; "~" -> "~~"
 198       (goto-char (point-min))
 199       (while (search-forward "~" nil t) (insert ?~))
 200
 201       ;; Chinese GB2312 -> "~{...~}"
 202       (goto-char (point-min))
 203       (if (re-search-forward "\\cc" nil t)
 204           (let ((enable-multibyte-characters nil)
 205                 pos)
 206             (goto-char (setq pos (match-beginning 0)))
 207             (encode-coding-region pos (point-max) 'iso-2022-7bit)
 208             (goto-char pos)
 209             (while (search-forward iso2022-gb-designation nil t)
 210               (delete-char -3)
 211               (insert hz-gb-designnation))
 212             (goto-char pos)
 213             (while (search-forward iso2022-ascii-designation nil t)
 214               (delete-char -3)
 215               (insert hz-ascii-designnation))))
 216       (- (point-max) (point-min)))))
 217
 218 ;;;###autoload
 219 (defun encode-hz-buffer ()
 220   "Encode the text in the current buffer to HZ."
 221   (interactive)
 222   (encode-hz-region (point-min) (point-max)))
 223
 224 ;;
 225 (provide 'china-util)
 226
 227 ;;; china-util.el ends here