code.delx.au - gnu-emacs/blob - lisp/language/chinese.el

   1 ;;; chinese.el --- support for Chinese -*- coding: iso-2022-7bit; -*-
   2
   3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
   4 ;; Licensed to the Free Software Foundation.
   5
   6 ;; Keywords: 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 ;;; Commentary:
  26
  27 ;; For Chinese, three character sets GB2312, BIG5, and CNS11643 are
  28 ;; supported.
  29
  30 ;;; Code:
  31
  32 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
  33 ;;; Chinese (general)
  34 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
  35
  36 (define-coding-system 'iso-2022-cn
  37  "ISO 2022 based 7bit encoding for Chinese GB and CNS (MIME:ISO-2022-CN)."
  38  :coding-type 'iso-2022
  39  :mnemonic ?C
  40  :charset-list '(ascii chinese-gb2312 chinese-cns11643-1 chinese-cns11643-2)
  41  :designation [ascii
  42                (nil chinese-gb2312 chinese-cns11643-1)
  43                (nil chinese-cns11643-2)
  44                nil]
  45  :flags '(ascii-at-eol ascii-at-cntl 7-bit
  46                        designation locking-shift single-shift init-at-bol)
  47  :mime-charset 'iso-2022-cn)
  48
  49 (define-coding-system-alias 'chinese-iso-7bit 'iso-2022-cn)
  50
  51 (define-coding-system 'iso-2022-cn-ext
  52   "ISO 2022 based 7bit encoding for Chinese GB and CNS (MIME:ISO-2022-CN-EXT)."
  53   :coding-type 'iso-2022
  54   :mnemonic ?C
  55   :charset-list '(ascii
  56                   chinese-gb2312 chinese-cns11643-1
  57                   chinese-cns11643-2 chinese-cns11643-3 chinese-cns11643-4
  58                   chinese-cns11643-5 chinese-cns11643-6 chinese-cns11643-7)
  59   :designation '[ascii
  60                  (nil chinese-gb2312 chinese-cns11643-1)
  61                  (nil chinese-cns11643-2)
  62                  (nil chinese-cns11643-3 chinese-cns11643-4 chinese-cns11643-5
  63                       chinese-cns11643-6 chinese-cns11643-7)]
  64   :flags '(ascii-at-eol ascii-at-cntl 7-bit
  65                         designation locking-shift single-shift init-at-bol)
  66   :mime-charset 'iso-2022-cn-ext)
  67
  68 \f
  69 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
  70 ;;; Chinese GB2312 (simplified)
  71 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
  72
  73 (define-coding-system 'chinese-iso-8bit
  74   "ISO 2022 based EUC encoding for Chinese GB2312 (MIME:CN-GB)."
  75   :coding-type 'iso-2022
  76   :mnemonic ?c
  77   :charset-list '(ascii chinese-gb2312)
  78   :designation [ascii chinese-gb2312 nil nil]
  79   :mime-charset 'cn-gb)
  80
  81 (define-coding-system-alias 'cn-gb-2312 'chinese-iso-8bit)
  82 (define-coding-system-alias 'euc-china 'chinese-iso-8bit)
  83 (define-coding-system-alias 'euc-cn 'chinese-iso-8bit)
  84 (define-coding-system-alias 'cn-gb 'chinese-iso-8bit)
  85 (define-coding-system-alias 'gb2312 'chinese-iso-8bit)
  86
  87 (define-coding-system 'chinese-hz
  88  "Hz/ZW 7-bit encoding for Chinese GB2312 (MIME:HZ-GB-2312)."
  89  :coding-type 'utf-8
  90  :mnemonic ?z
  91  :charset-list '(ascii chinese-gb2312)
  92  :mime-charset 'hz-gb-2312
  93  :post-read-conversion 'post-read-decode-hz
  94  :pre-write-conversion 'pre-write-encode-hz)
  95
  96 (define-coding-system-alias 'hz-gb-2312 'chinese-hz)
  97 (define-coding-system-alias 'hz 'chinese-hz)
  98
  99 (set-language-info-alist
 100  "Chinese-GB" '((charset chinese-gb2312 chinese-sisheng)
 101                 (coding-system chinese-iso-8bit iso-2022-cn chinese-hz)
 102                 (coding-priority chinese-iso-8bit chinese-big5 iso-2022-cn)
 103                 (input-method . "chinese-py-punct")
 104                 (features china-util)
 105                 (sample-text . "Chinese (\e$AVPND\e(B,\e$AFUM(;0\e(B,\e$A::So\e(B)    \e$ADc:C\e(B")
 106                 (documentation . "Support for Chinese GB2312 character set."))
 107  '("Chinese"))
 108
 109 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
 110 ;; Chinese BIG5 (traditional)
 111 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
 112
 113 (define-coding-system 'chinese-big5
 114   "BIG5 8-bit encoding for Chinese (MIME:Big5)"
 115   :coding-type 'charset
 116   :mnemonic ?B
 117   :charset-list '(ascii big5)
 118   :mime-charset 'big5)
 119
 120 (define-coding-system-alias 'big5 'chinese-big5)
 121 (define-coding-system-alias 'cn-big5 'chinese-big5)
 122
 123 (set-language-info-alist
 124  "Chinese-BIG5" '((charset chinese-big5-1 chinese-big5-2)
 125                   (coding-system chinese-big5 chinese-iso-7bit)
 126                   (coding-priority chinese-big5 iso-2022-cn chinese-iso-8bit)
 127                   (input-method . "chinese-py-punct-b5")
 128                   (features china-util)
 129                   (sample-text . "Cantonese (\e$(Gemk#\e(B,\e$(Gl]N)fc\e(B) \e$ATg3?\e(B, \e$ADc:C\e(B")
 130                   (documentation . "Support for Chinese Big5 character set."))
 131  '("Chinese"))
 132
 133 (define-coding-system 'chinese-big5-hkscs
 134   "BIG5-HKSCS 8-bit encoding for Chinese, Hong Kong supplement (MIME:Big5-HKSCS)"
 135   :coding-type 'charset
 136   :mnemonic ?B
 137   :charset-list '(ascii big5-hkscs)
 138   :mime-charset 'big5-hkscs)
 139 (define-coding-system-alias 'big5-hkscs 'chinese-big5-hkscs)
 140 (define-coding-system-alias 'cn-big5-hkscs 'chinese-big5-hkscs)
 141
 142 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
 143 ;; Chinese CNS11643 (traditional)
 144 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
 145
 146 (define-coding-system 'euc-tw
 147   "ISO 2022 based EUC encoding for Chinese CNS11643."
 148   :coding-type 'iso-2022
 149   :mnemonic ?Z
 150   :charset-list '(ascii
 151                   chinese-cns11643-1
 152                   chinese-cns11643-2
 153                   chinese-cns11643-3
 154                   chinese-cns11643-4
 155                   chinese-cns11643-5
 156                   chinese-cns11643-6
 157                   chinese-cns11643-7)
 158   :designation [ascii chinese-cns11643-1 (chinese-cns11643-1
 159                                           chinese-cns11643-2
 160                                           chinese-cns11643-3
 161                                           chinese-cns11643-4
 162                                           chinese-cns11643-5
 163                                           chinese-cns11643-6
 164                                           chinese-cns11643-7) nil]
 165   :mime-charset 'euc-tw)
 166
 167 (define-coding-system-alias 'euc-taiwan 'euc-tw)
 168
 169 (set-language-info-alist
 170  "Chinese-CNS" '((charset chinese-cns11643-1 chinese-cns11643-2
 171                           chinese-cns11643-3 chinese-cns11643-4
 172                           chinese-cns11643-5 chinese-cns11643-6
 173                           chinese-cns11643-7)
 174                  (coding-system iso-2022-cn euc-tw)
 175                  (coding-priority iso-2022-cn euc-tw chinese-big5
 176                                   chinese-iso-8bit)
 177                  (features china-util)
 178                  (input-method . "chinese-cns-quick")
 179                  ;; Fixme: presumably it won't accept big5 now.
 180                  (documentation . "\
 181 Support for Chinese CNS character sets.  Note that EUC-TW coding system
 182 accepts Big5 for input also (which is then converted to CNS)."))
 183  '("Chinese"))
 184
 185 ;;; Chinese GBK
 186
 187 (define-coding-system 'chinese-gbk
 188   "GBK encoding for Chinese (MIME:GBK)."
 189   :coding-type 'charset
 190   :mnemonic ?c
 191   :charset-list '(chinese-gbk)
 192   :mime-charset 'gbk)
 193 (define-coding-system-alias 'gbk 'chinese-gbk)
 194 (define-coding-system-alias 'cp936 'chinese-gbk)
 195 (define-coding-system-alias 'windows-936 'chinese-gbk)
 196
 197 (set-language-info-alist
 198  "Chinese-GBK" '((charset chinese-gbk)
 199                  (coding-system chinese-gbk)
 200                  (coding-priority gbk iso-2022-cn chinese-big5
 201                                   chinese-iso-8bit) ; fixme?
 202                  (input-method . "chinese-py-punct") ; fixme?
 203                  (features china-util)
 204                  (documentation . "Support for Chinese GBK character set."))
 205  '("Chinese"))
 206
 207 ;;; Chinese GB18030
 208
 209 (define-coding-system 'chinese-gb18030
 210   "GB18030 encoding for Chinese (MIME:GB18030)."
 211   :coding-type 'charset
 212   :mnemonic ?c
 213   :charset-list '(gb18030)
 214   :mime-charset 'gb18030
 215   )
 216 (define-coding-system-alias 'gb18030 'chinese-gb18030)
 217
 218 (set-language-info-alist
 219  "Chinese-GB18030" '((charset gb18030)
 220                      (coding-system chinese-gb18030)
 221                      (coding-priority gb18030 gbk iso-2022-cn chinese-big5
 222                                       chinese-iso-8bit) ; fixme?
 223                      (input-method . "chinese-py-punct") ; fixme?
 224                      (features china-util)
 225                      (documentation
 226                       . "Support for Chinese GB18030 character set."))
 227  '("Chinese"))
 228
 229 ;; Fixme: add HKSCS
 230
 231 (provide 'chinese)
 232
 233 ;;; chinese.el ends here