code.delx.au - gnu-emacs/blob - 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   (setq primary-language "Chinese-GB")
  34
  35   (set-default-coding-systems 'chinese-iso-8bit)
  36   (setq coding-category-iso-8-2 'chinese-iso-8bit)
  37   (setq coding-category-iso-7-else 'chinese-iso-7bit)
  38   (setq coding-category-big5 'chinese-big5)
  39
  40   (set-coding-priority
  41    '(coding-category-iso-7
  42      coding-category-iso-7-else
  43      coding-category-iso-8-2
  44      coding-category-big5
  45      coding-category-iso-8-1
  46      coding-category-emacs-mule
  47      coding-category-iso-8-else))
  48
  49   (setq-default buffer-file-coding-system 'chinese-iso-8bit)
  50   (setq default-terminal-coding-system 'chinese-iso-8bit)
  51   (setq default-keyboard-coding-system 'chinese-iso-8bit)
  52   (setq sendmail-coding-system nil
  53         rmail-file-coding-system 'chinese-iso-7bit)
  54
  55   (setq default-input-method  "chinese-py-punct")
  56   (setq-default default-input-method "chinese-py-punct"))
  57
  58 ;;;###autoload
  59 (defun setup-chinese-big5-environment ()
  60   "Setup multilingual environment (MULE) for Chinese Big5 users."
  61   (interactive)
  62   (setup-english-environment)
  63
  64   (setq primary-language "Chinese-BIG5")
  65
  66   (set-default-coding-systems 'chinese-big5)
  67   (setq coding-category-iso-8-2 'chinese-big5)
  68   (setq coding-category-iso-7-else 'chinese-iso-7bit)
  69   (setq coding-category-big5 'chinese-big5)
  70
  71   (set-coding-priority
  72    '(coding-category-iso-7
  73      coding-category-iso-7-else
  74      coding-category-big5
  75      coding-category-iso-8-2
  76      coding-category-emacs-mule
  77      coding-category-iso-8-else))
  78
  79   (setq-default buffer-file-coding-system 'chinese-big5)
  80   (setq default-terminal-coding-system 'chinese-big5)
  81   (setq default-keyboard-coding-system 'chinese-big5)
  82   (setq sendmail-coding-system nil
  83         rmail-file-coding-system 'chienese-iso-7bit)
  84
  85   (setq default-input-method "chinese-py-punct-b5")
  86   (setq-default default-input-method "chinese-py-punct-b5"))
  87
  88 ;;;###autoload
  89 (defun setup-chinese-cns-environment ()
  90   "Setup multilingual environment (MULE) for Chinese CNS11643 family users."
  91   (interactive)
  92   (setup-english-environment)
  93
  94   (setq primary-language "Chinese-CNS")
  95
  96   (setq coding-category-iso-7-else 'chinese-iso-7bit)
  97   (setq coding-category-big5 'chinese-big5)
  98   (setq coding-category-iso-8-2 'chinese-big5)
  99   (set-default-coding-systems 'chinese-iso-7bit)
 100
 101   (set-coding-priority
 102    '(coding-category-iso-7
 103      coding-category-iso-7-else
 104      coding-category-iso-8-2
 105      coding-category-big5
 106      coding-category-iso-7-else))
 107
 108   (setq-default buffer-file-coding-system 'chinese-iso-7bit)
 109   (setq default-terminal-coding-system 'chinese-iso-7bit)
 110   (setq default-keyboard-coding-system 'chinese-iso-7bit)
 111   (setq sendmail-coding-system nil
 112         rmail-file-coding-system 'chinese-iso-7bit)
 113
 114   (setq default-input-method "chinese-quick-cns")
 115   (setq-default default-input-method "chinese-quick-cns"))
 116
 117 ;; Hz/ZW encoding stuffs
 118
 119 ;; HZ is an encoding method for Chinese character set GB2312 used
 120 ;; widely in Internet.  It is very similar to 7-bit environment of
 121 ;; ISO-2022.  The difference is that HZ uses the sequence "~{" and
 122 ;; "~}" for designating GB2312 and ASCII respectively, hence, it
 123 ;; doesn't uses ESC (0x1B) code.
 124
 125 ;; ZW is another encoding method for Chinese character set GB2312.  It
 126 ;; encodes Chinese characters line by line by starting each line with
 127 ;; the sequence "zW".  It also uses only 7-bit as HZ.
 128
 129 ;; ISO-2022 escape sequence to designate GB2312.
 130 (defvar iso2022-gb-designation "\e$A")
 131 ;; HZ escape sequence to designate GB2312.
 132 (defvar hz-gb-designnation "~{")
 133 ;; ISO-2022 escape sequence to designate ASCII.
 134 (defvar iso2022-ascii-designation "\e(B")
 135 ;; HZ escape sequence to designate ASCII.
 136 (defvar hz-ascii-designnation "~}")
 137 ;; Regexp of ZW sequence to start GB2312.
 138 (defvar zw-start-gb "^zW")
 139 ;; Regexp for start of GB2312 in an encoding mixture of HZ and ZW.
 140 (defvar hz/zw-start-gb (concat hz-gb-designnation "\\|" zw-start-gb))
 141
 142 (defvar decode-hz-line-continuation nil
 143   "Flag to tell if we should care line continuation convention of Hz.")
 144
 145 ;;;###autoload
 146 (defun decode-hz-region (beg end)
 147   "Decode HZ/ZW encoded text in the current region.
 148 Return the length of resulting text."
 149   (interactive "r")
 150   (save-excursion
 151     (save-restriction
 152       (narrow-to-region beg end)
 153
 154       ;; We, at first, convert HZ/ZW to `iso-2022-7bit',
 155       ;; then decode it.
 156
 157       ;; "~\n" -> "\n"
 158       (goto-char (point-min))
 159       (while (search-forward "~" nil t)
 160         (if (= (following-char) ?\n) (delete-char -1))
 161         (if (not (eobp)) (forward-char 1)))
 162
 163       ;; "^zW...\n" -> Chinese GB2312
 164       ;; "~{...~}"  -> Chinese GB2312
 165       (goto-char (point-min))
 166       (let ((chinese-found nil))
 167         (while (re-search-forward hz/zw-start-gb nil t)
 168           (if (= (char-after (match-beginning 0)) ?z)
 169               ;; ZW -> iso-2022-7bit
 170               (progn
 171                 (delete-char -2)
 172                 (insert iso2022-gb-designation)
 173                 (end-of-line)
 174                 (insert iso2022-ascii-designation))
 175             ;; HZ -> iso-2022-7bit
 176             (delete-char -2)
 177             (insert iso2022-gb-designation)
 178             (let ((pos (save-excursion (end-of-line) (point))))
 179               (if (search-forward hz-ascii-designnation pos t)
 180                   (replace-match iso2022-ascii-designation)
 181                 (if (not decode-hz-line-continuation)
 182                     (insert iso2022-ascii-designation)))))
 183           (setq chinese-found t))
 184         (if (or chinese-found
 185                 (let ((enable-multibyte-characters nil))
 186                   ;; Here we check if the text contains EUC (China) codes.
 187                   ;; If any, we had better decode them also.
 188                   (goto-char (point-min))
 189                   (re-search-forward "[\240-\377]" nil t)))
 190             (decode-coding-region (point-min) (point-max) 'euc-china)))
 191
 192       ;; "~~" -> "~"
 193       (goto-char (point-min))
 194       (while (search-forward "~~" nil t) (delete-char -1))
 195       (- (point-max) (point-min)))))
 196
 197 ;;;###autoload
 198 (defun decode-hz-buffer ()
 199   "Decode HZ/ZW encoded text in the current buffer."
 200   (interactive)
 201   (decode-hz-region (point-min) (point-max)))
 202
 203 ;;;###autoload
 204 (defun encode-hz-region (beg end)
 205   "Encode the text in the current region to HZ.
 206 Return the length of resulting text."
 207   (interactive "r")
 208   (save-excursion
 209     (save-restriction
 210       (narrow-to-region beg end)
 211
 212       ;; "~" -> "~~"
 213       (goto-char (point-min))
 214       (while (search-forward "~" nil t) (insert ?~))
 215
 216       ;; Chinese GB2312 -> "~{...~}"
 217       (goto-char (point-min))
 218       (if (re-search-forward "\\cc" nil t)
 219           (let ((enable-multibyte-characters nil)
 220                 pos)
 221             (goto-char (setq pos (match-beginning 0)))
 222             (encode-coding-region pos (point-max) 'iso-2022-7bit)
 223             (goto-char pos)
 224             (while (search-forward iso2022-gb-designation nil t)
 225               (delete-char -3)
 226               (insert hz-gb-designnation))
 227             (goto-char pos)
 228             (while (search-forward iso2022-ascii-designation nil t)
 229               (delete-char -3)
 230               (insert hz-ascii-designnation))))
 231       (- (point-max) (point-min)))))
 232
 233 ;;;###autoload
 234 (defun encode-hz-buffer ()
 235   "Encode the text in the current buffer to HZ."
 236   (interactive)
 237   (encode-hz-region (point-min) (point-max)))
 238
 239 ;;
 240 (provide 'china-util)
 241
 242 ;;; china-util.el ends here