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

   1 ;;; japanese.el --- support for Japanese -*- coding: iso-2022-7bit; no-byte-compile: t -*-
   2
   3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
   4 ;;   Licensed to the Free Software Foundation.
   5 ;; Copyright (C) 2003
   6 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
   7 ;;   Registration Number H13PRO009
   8
   9 ;; Keywords: multilingual, Japanese
  10
  11 ;; This file is part of GNU Emacs.
  12
  13 ;; GNU Emacs is free software; you can redistribute it and/or modify
  14 ;; it under the terms of the GNU General Public License as published by
  15 ;; the Free Software Foundation; either version 2, or (at your option)
  16 ;; any later version.
  17
  18 ;; GNU Emacs is distributed in the hope that it will be useful,
  19 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  20 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  21 ;; GNU General Public License for more details.
  22
  23 ;; You should have received a copy of the GNU General Public License
  24 ;; along with GNU Emacs; see the file COPYING.  If not, write to the
  25 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  26 ;; Boston, MA 02111-1307, USA.
  27
  28 ;;; Commentary:
  29
  30 ;; For Japanese, character sets JISX0201, JISX0208, JISX0212 are
  31 ;; supported.
  32
  33 ;;; Code:
  34
  35 ;;; Load translation tables for CP932.
  36 (load "international/cp51932")
  37 (load "international/eucjp-ms")
  38
  39 (define-coding-system 'iso-2022-jp
  40   "ISO 2022 based 7bit encoding for Japanese (MIME:ISO-2022-JP)."
  41   :coding-type 'iso-2022
  42   :mnemonic ?J
  43   :designation [(ascii japanese-jisx0208-1978 japanese-jisx0208
  44                        latin-jisx0201)
  45                 nil nil nil]
  46   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation)
  47   :charset-list '(ascii japanese-jisx0208
  48                         japanese-jisx0208-1978 latin-jisx0201)
  49   :mime-charset 'iso-2022-jp)
  50
  51 (define-coding-system-alias 'junet 'iso-2022-jp)
  52
  53 (define-coding-system 'iso-2022-jp-2
  54   "ISO 2022 based 7bit encoding for CJK, Latin-1, Greek (MIME:ISO-2022-JP-2)."
  55   :coding-type 'iso-2022
  56   :mnemonic ?J
  57   :designation [(ascii japanese-jisx0208-1978 japanese-jisx0208
  58                        latin-jisx0201 japanese-jisx0212
  59                        chinese-gb2312 korean-ksc5601)
  60                 nil
  61                 (nil latin-iso8859-1 greek-iso8859-7)
  62                 nil]
  63   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation single-shift
  64                  init-at-bol)
  65   :charset-list '(ascii japanese-jisx0208 japanese-jisx0212
  66                         latin-jisx0201 japanese-jisx0208-1978
  67                         chinese-gb2312 korean-ksc5601
  68                         latin-iso8859-1 greek-iso8859-7)
  69   :mime-charset 'iso-2022-jp-2)
  70
  71 (let ((map                      ; JIS           vs      CP932
  72        '((#x301C . #xFF5E)      ; WAVE DASH             FULLWIDTH TILDE
  73          (#x2014 . #x2015)      ; EM DASH               HORIZONTAL BAR
  74          (#x2016 . #x2225)      ; DOUBLE VERTICAL LINE  PARALLEL TO
  75          (#x2212 . #xFF0D)      ; MINUS SIGN            FULLWIDTH HYPHEN-MINUS
  76          (#x00A2 . #xFFE0)      ; CENT SIGN             FULLWIDTH CENT SIGN
  77          (#x00A3 . #xFFE1)      ; POUND SIGN            FULLWIDTH POUND SIGN
  78          (#x00AC . #xFFE2)      ; NOT SIGN              FULLWIDTH NOT SIGN
  79          (#x00A6 . #xFFE4)      ; BROKEN LINE           FULLWIDTH BROKEN LINE
  80          )))
  81   (define-translation-table 'japanese-ucs-jis-to-cp932-map map)
  82   (mapc #'(lambda (x) (let ((tmp (car x)))
  83                         (setcar x (cdr x)) (setcdr x tmp)))
  84         map)
  85   (define-translation-table 'japanese-ucs-cp932-to-jis-map map))
  86
  87 ;; U+2014 (EM DASH) vs U+2015 (HORIZONTAL BAR)
  88 (define-translation-table 'japanese-ucs-glibc-to-jis-map '((#x2015 . #x2014)))
  89 (define-translation-table 'japanese-ucs-jis-to-glibc-map '((#x2014 . #x2015)))
  90
  91 (define-coding-system 'japanese-shift-jis
  92   "Shift-JIS 8-bit encoding for Japanese (MIME:SHIFT_JIS)"
  93   :coding-type 'shift-jis
  94   :mnemonic ?S
  95   :charset-list '(ascii katakana-jisx0201 japanese-jisx0208)
  96   :mime-charset 'shift_jis)
  97
  98 (define-coding-system-alias 'shift_jis 'japanese-shift-jis)
  99 (define-coding-system-alias 'sjis 'japanese-shift-jis)
 100 (define-coding-system-alias 'cp932 'japanese-shift-jis)
 101
 102 (define-coding-system 'japanese-cp932
 103   "CP932 (Microsoft shift-jis)"
 104   :coding-type 'charset
 105   :mnemonic ?S
 106   :charset-list '(ascii katakana-sjis cp932-2-byte))
 107
 108 (define-coding-system-alias 'cp932 'japanese-cp932)
 109
 110 (define-coding-system 'japanese-iso-7bit-1978-irv
 111   "ISO 2022 based 7-bit encoding for Japanese JISX0208-1978 and JISX0201-Roman."
 112   :coding-type 'iso-2022
 113   :mnemonic ?j
 114   :designation [(latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
 115                                 japanese-jisx0212 katakana-jisx0201)
 116                 nil nil nil]
 117   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation
 118                  use-roman use-oldjis)
 119   :charset-list '(ascii latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
 120                         japanese-jisx0212))
 121
 122 (define-coding-system-alias 'iso-2022-jp-1978-irv 'japanese-iso-7bit-1978-irv)
 123 (define-coding-system-alias 'old-jis 'japanese-iso-7bit-1978-irv)
 124
 125 (define-coding-system 'japanese-iso-8bit
 126   "ISO 2022 based EUC encoding for Japanese (MIME:EUC-JP)."
 127   :coding-type 'iso-2022
 128   :mnemonic ?E
 129   :designation [ascii japanese-jisx0208 katakana-jisx0201 japanese-jisx0212]
 130   :flags '(short ascii-at-eol ascii-at-cntl single-shift)
 131   :charset-list '(ascii latin-jisx0201 japanese-jisx0208
 132                         katakana-jisx0201 japanese-jisx0212
 133                         japanese-jisx0208-1978)
 134   :mime-charset 'euc-jp)
 135
 136 (define-coding-system-alias 'euc-japan-1990 'japanese-iso-8bit)
 137 (define-coding-system-alias 'euc-japan 'japanese-iso-8bit)
 138 (define-coding-system-alias 'euc-jp 'japanese-iso-8bit)
 139
 140 (define-coding-system 'eucjp-ms
 141   "eucJP-ms (like EUC-JP but with CP932 extension).
 142 eucJP-ms is defined in <http://www.opengroup.or.jp/jvc/cde/appendix.html>."
 143   :coding-type 'iso-2022
 144   :mnemonic ?E
 145   :designation [ascii japanese-jisx0208 katakana-jisx0201 japanese-jisx0212]
 146   :flags '(short ascii-at-eol ascii-at-cntl single-shift)
 147   :charset-list '(ascii latin-jisx0201 japanese-jisx0208
 148                         katakana-jisx0201 japanese-jisx0212)
 149   :decode-translation-table 'eucjp-ms-decode
 150   :encode-translation-table 'eucjp-ms-encode)
 151
 152 (define-coding-system 'iso-2022-jp-2004
 153   "ISO 2022 based 7bit encoding for JIS X 0213:2004 (MIME:ISO-2022-JP-2004)."
 154   :coding-type 'iso-2022
 155   :mnemonic ?J
 156   :designation [(ascii japanese-jisx0208 japanese-jisx0213.2004-1
 157                        japanese-jisx0213-1 japanese-jisx0213-2)
 158                 nil nil nil]
 159   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation)
 160                  ;; init-at-bol)
 161   :charset-list '(ascii japanese-jisx0208 japanese-jisx0213.2004-1
 162                         japanese-jisx0213-1 japanese-jisx0213-2)
 163   :mime-charset 'iso-2022-jp-2004)
 164
 165 (define-coding-system-alias 'iso-2022-jp-3 'iso-2022-jp-2004)
 166
 167 (define-coding-system 'euc-jis-2004
 168   "ISO 2022 based EUC encoding for JIS X 0213 (MIME:EUC-JIS-2004)."
 169   :coding-type 'iso-2022
 170   :mnemonic ?E
 171   :designation [ascii japanese-jisx0213.2004-1 katakana-jisx0201
 172                       japanese-jisx0213-2]
 173   :flags '(short ascii-at-eol ascii-at-cntl single-shift)
 174   :charset-list '(ascii latin-jisx0201 japanese-jisx0213.2004-1
 175                         japanese-jisx0213-1 katakana-jisx0201
 176                         japanese-jisx0213-2)
 177   :mime-charset 'euc-jis-2004)
 178
 179 (define-coding-system-alias 'euc-jisx0213 'euc-jis-2004)
 180
 181 (define-coding-system 'japanese-shift-jis-2004
 182   "Shift_JIS 8-bit encodinf for Japanese (MIME:SHIFT_JIS-2004)"
 183   :coding-type 'shift-jis
 184   :mnemonic ?S
 185   :charset-list '(ascii katakana-jisx0201
 186                         japanese-jisx0213.2004-1 japanese-jisx0213-2))
 187
 188 (define-coding-system-alias 'shift_jis-2004 'japanese-shift-jis-2004)
 189
 190 (set-language-info-alist
 191  "Japanese" '((setup-function . setup-japanese-environment-internal)
 192               (tutorial . "TUTORIAL.ja")
 193               (charset japanese-jisx0208
 194                        japanese-jisx0212 latin-jisx0201 katakana-jisx0201
 195                        japanese-jisx0213.2004-1 japanese-jisx0213-1
 196                        japanese-jisx0213-2 japanese-jisx0208-1978)
 197               (coding-system iso-2022-jp japanese-iso-8bit
 198                              japanese-shift-jis japanese-iso-7bit-1978-irv
 199                              iso-2022-jp-2004 japanese-shift-jis-2004
 200                              euc-jis-2004)
 201               (coding-priority iso-2022-jp japanese-iso-8bit
 202                                japanese-shift-jis
 203                                iso-2022-jp-2004 euc-jis-2004
 204                                japanese-shift-jis-2004
 205                                iso-2022-jp-2)
 206               (input-method . "japanese")
 207               (features japan-util)
 208               (sample-text . "Japanese (\e$BF|K\8l\e(B)   \e$B$3$s$K$A$O\e(B, \e(I:]FAJ\e(B")
 209               (documentation . t)))
 210
 211 (let ((map
 212        ;; JISX0213-1 vs Unicode
 213        '((#x2477 . [#x304B #x309A])
 214          (#x2478 . [#x304D #x309A])
 215          (#x2479 . [#x304F #x309A])
 216          (#x247a . [#x3051 #x309A])
 217          (#x247b . [#x3053 #x309A])
 218          (#x2577 . [#x30AB #x309A])
 219          (#x2578 . [#x30AD #x309A])
 220          (#x2579 . [#x30AF #x309A])
 221          (#x257a . [#x30B1 #x309A])
 222          (#x257b . [#x30B3 #x309A])
 223          (#x257c . [#x30BB #x309A])
 224          (#x257d . [#x30C4 #x309A])
 225          (#x257e . [#x30C8 #x309A])
 226          (#x2678 . [#x31F7 #x309A])
 227          (#x2b44 . [#x00E6 #x0300])
 228          (#x2b48 . [#x0254 #x0300])
 229          (#x2b49 . [#x0254 #x0301])
 230          (#x2b4a . [#x028C #x0300])
 231          (#x2b4b . [#x028C #x0301])
 232          (#x2b4c . [#x0259 #x0300])
 233          (#x2b4d . [#x0259 #x0301])
 234          (#x2b4e . [#x025A #x0300])
 235          (#x2b4f . [#x025A #x0301])
 236          (#x2b65 . [#x02E9 #x02E5])
 237          (#x2b66 . [#x02E5 #x02E9])))
 238       table)
 239   (dolist (elt map)
 240     (setcar elt (decode-char 'japanese-jisx0213-1 (car elt))))
 241   (setq table (make-translation-table-from-alist map))
 242   (define-translation-table 'jisx0213-to-unicode table)
 243   (define-translation-table 'unicode-to-jisx0213
 244     (char-table-extra-slot table 0)))
 245
 246 (provide 'japanese)
 247
 248 ;;; arch-tag: 450f5537-9d53-4d5e-b731-4cf116d8cbc9
 249 ;;; japanese.el ends here