lisp/language/japanese.el

   1 ;;; japanese.el --- support for Japanese -*- coding: iso-2022-7bit -*-
   2
   3 ;; Copyright (C) 1997, 2001-2016 Free Software Foundation, Inc.
   4 ;; Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   5 ;;   2005, 2006, 2007, 2008, 2009, 2010, 2011
   6 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
   7 ;;   Registration Number H14PRO021
   8 ;; Copyright (C) 2003
   9 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
  10 ;;   Registration Number H13PRO009
  11
  12 ;; Keywords: multilingual, Japanese
  13
  14 ;; This file is part of GNU Emacs.
  15
  16 ;; GNU Emacs is free software: you can redistribute it and/or modify
  17 ;; it under the terms of the GNU General Public License as published by
  18 ;; the Free Software Foundation, either version 3 of the License, or
  19 ;; (at your option) any later version.
  20
  21 ;; GNU Emacs is distributed in the hope that it will be useful,
  22 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  23 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  24 ;; GNU General Public License for more details.
  25
  26 ;; You should have received a copy of the GNU General Public License
  27 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  28
  29 ;;; Commentary:
  30
  31 ;; For Japanese, character sets JISX0201, JISX0208, JISX0212 are
  32 ;; supported.
  33
  34 ;;; Code:
  35
  36 ;;; Load translation tables for CP932.
  37 (require 'cp51932)
  38 (require 'eucjp-ms)
  39
  40 (define-coding-system 'iso-2022-jp
  41   "ISO 2022 based 7bit encoding for Japanese (MIME:ISO-2022-JP)."
  42   :coding-type 'iso-2022
  43   :mnemonic ?J
  44   :designation [(ascii japanese-jisx0208-1978 japanese-jisx0208
  45                        latin-jisx0201)
  46                 nil nil nil]
  47   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation)
  48   :charset-list '(ascii japanese-jisx0208
  49                         japanese-jisx0208-1978 latin-jisx0201)
  50   :mime-charset 'iso-2022-jp
  51   :suitable-for-keyboard t)
  52
  53 (define-coding-system-alias 'junet 'iso-2022-jp)
  54
  55 (define-coding-system 'iso-2022-jp-2
  56   "ISO 2022 based 7bit encoding for CJK, Latin-1, Greek (MIME:ISO-2022-JP-2)."
  57   :coding-type 'iso-2022
  58   :mnemonic ?J
  59   :designation [(ascii japanese-jisx0208-1978 japanese-jisx0208
  60                        latin-jisx0201 japanese-jisx0212
  61                        chinese-gb2312 korean-ksc5601)
  62                 nil
  63                 (nil latin-iso8859-1 greek-iso8859-7)
  64                 nil]
  65   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation single-shift
  66                  init-at-bol)
  67   :charset-list '(ascii japanese-jisx0208 japanese-jisx0212
  68                         latin-jisx0201 japanese-jisx0208-1978
  69                         chinese-gb2312 korean-ksc5601
  70                         latin-iso8859-1 greek-iso8859-7)
  71   :mime-charset 'iso-2022-jp-2
  72   :suitable-for-keyboard t)
  73
  74 (let ((map                      ; JIS           vs      CP932
  75        '((#x301C . #xFF5E)      ; WAVE DASH             FULLWIDTH TILDE
  76          (#x2014 . #x2015)      ; EM DASH               HORIZONTAL BAR
  77          (#x2016 . #x2225)      ; DOUBLE VERTICAL LINE  PARALLEL TO
  78          (#x2212 . #xFF0D)      ; MINUS SIGN            FULLWIDTH HYPHEN-MINUS
  79          (#x00A2 . #xFFE0)      ; CENT SIGN             FULLWIDTH CENT SIGN
  80          (#x00A3 . #xFFE1)      ; POUND SIGN            FULLWIDTH POUND SIGN
  81          (#x00AC . #xFFE2)      ; NOT SIGN              FULLWIDTH NOT SIGN
  82          (#x00A6 . #xFFE4)      ; BROKEN LINE           FULLWIDTH BROKEN LINE
  83          )))
  84   (define-translation-table 'japanese-ucs-jis-to-cp932-map map)
  85   (mapc #'(lambda (x) (let ((tmp (car x)))
  86                         (setcar x (cdr x)) (setcdr x tmp)))
  87         map)
  88   (define-translation-table 'japanese-ucs-cp932-to-jis-map map))
  89
  90 ;; U+2014 (EM DASH) vs U+2015 (HORIZONTAL BAR)
  91 (define-translation-table 'japanese-ucs-glibc-to-jis-map '((#x2015 . #x2014)))
  92 (define-translation-table 'japanese-ucs-jis-to-glibc-map '((#x2014 . #x2015)))
  93
  94 (define-coding-system 'japanese-shift-jis
  95   "Shift-JIS 8-bit encoding for Japanese (MIME:SHIFT_JIS)"
  96   :coding-type 'shift-jis
  97   :mnemonic ?S
  98   :charset-list '(ascii katakana-jisx0201 japanese-jisx0208)
  99   :mime-charset 'shift_jis)
 100
 101 (define-coding-system-alias 'shift_jis 'japanese-shift-jis)
 102 (define-coding-system-alias 'sjis 'japanese-shift-jis)
 103
 104 (define-coding-system 'japanese-cp932
 105   "CP932 (Microsoft shift-jis)"
 106   :coding-type 'charset
 107   :mnemonic ?S
 108   :charset-list '(ascii katakana-sjis cp932-2-byte))
 109
 110 (define-coding-system-alias 'cp932 'japanese-cp932)
 111
 112 (define-coding-system 'japanese-iso-7bit-1978-irv
 113   "ISO 2022 based 7-bit encoding for Japanese JISX0208-1978 and JISX0201-Roman."
 114   :coding-type 'iso-2022
 115   :mnemonic ?j
 116   :designation [(latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
 117                                 japanese-jisx0212 katakana-jisx0201)
 118                 nil nil nil]
 119   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation
 120                  use-roman use-oldjis)
 121   :charset-list '(ascii latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
 122                         japanese-jisx0212))
 123
 124 (define-coding-system-alias 'iso-2022-jp-1978-irv 'japanese-iso-7bit-1978-irv)
 125 (define-coding-system-alias 'old-jis 'japanese-iso-7bit-1978-irv)
 126
 127 (define-coding-system 'japanese-iso-8bit
 128   "ISO 2022 based EUC encoding for Japanese (MIME:EUC-JP)."
 129   :coding-type 'iso-2022
 130   :mnemonic ?E
 131   :designation [ascii japanese-jisx0208 katakana-jisx0201 japanese-jisx0212]
 132   :flags '(short ascii-at-eol ascii-at-cntl single-shift)
 133   :charset-list '(ascii latin-jisx0201 japanese-jisx0208
 134                         katakana-jisx0201 japanese-jisx0212
 135                         japanese-jisx0208-1978)
 136   :mime-charset 'euc-jp)
 137
 138 (define-coding-system-alias 'euc-japan-1990 'japanese-iso-8bit)
 139 (define-coding-system-alias 'euc-japan 'japanese-iso-8bit)
 140 (define-coding-system-alias 'euc-jp 'japanese-iso-8bit)
 141
 142 (define-coding-system 'eucjp-ms
 143   "eucJP-ms (like EUC-JP but with CP932 extension).
 144 eucJP-ms is defined in <http://www.opengroup.or.jp/jvc/cde/appendix.html>."
 145   :coding-type 'iso-2022
 146   :mnemonic ?E
 147   :designation [ascii japanese-jisx0208 katakana-jisx0201 japanese-jisx0212]
 148   :flags '(short ascii-at-eol ascii-at-cntl single-shift)
 149   :charset-list '(ascii latin-jisx0201 japanese-jisx0208
 150                         katakana-jisx0201 japanese-jisx0212)
 151   :decode-translation-table 'eucjp-ms-decode
 152   :encode-translation-table 'eucjp-ms-encode)
 153
 154 (define-coding-system 'iso-2022-jp-2004
 155   "ISO 2022 based 7bit encoding for JIS X 0213:2004 (MIME:ISO-2022-JP-2004)."
 156   :coding-type 'iso-2022
 157   :mnemonic ?J
 158   :designation [(ascii japanese-jisx0208 japanese-jisx0213.2004-1
 159                        japanese-jisx0213-1 japanese-jisx0213-2)
 160                 nil nil nil]
 161   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation)
 162                  ;; init-at-bol)
 163   :charset-list '(ascii japanese-jisx0208 japanese-jisx0213.2004-1
 164                         japanese-jisx0213-1 japanese-jisx0213-2)
 165   :mime-charset 'iso-2022-jp-2004
 166   :suitable-for-keyboard t)
 167
 168 (define-coding-system-alias 'iso-2022-jp-3 'iso-2022-jp-2004)
 169
 170 (define-coding-system 'euc-jis-2004
 171   "ISO 2022 based EUC encoding for JIS X 0213 (MIME:EUC-JIS-2004)."
 172   :coding-type 'iso-2022
 173   :mnemonic ?E
 174   :designation [ascii japanese-jisx0213.2004-1 katakana-jisx0201
 175                       japanese-jisx0213-2]
 176   :flags '(short ascii-at-eol ascii-at-cntl single-shift)
 177   :charset-list '(ascii latin-jisx0201 japanese-jisx0213.2004-1
 178                         japanese-jisx0213-1 katakana-jisx0201
 179                         japanese-jisx0213-2)
 180   :mime-charset 'euc-jis-2004)
 181
 182 (define-coding-system-alias 'euc-jisx0213 'euc-jis-2004)
 183
 184 (define-coding-system 'japanese-shift-jis-2004
 185   "Shift_JIS 8-bit encoding for Japanese (MIME:SHIFT_JIS-2004)"
 186   :coding-type 'shift-jis
 187   :mnemonic ?S
 188   :charset-list '(ascii katakana-jisx0201
 189                         japanese-jisx0213.2004-1 japanese-jisx0213-2))
 190
 191 (define-coding-system-alias 'shift_jis-2004 'japanese-shift-jis-2004)
 192
 193 (set-language-info-alist
 194  "Japanese" '((setup-function . setup-japanese-environment-internal)
 195               (exit-function . use-default-char-width-table)
 196               (iso639-language . ja)
 197               (tutorial . "TUTORIAL.ja")
 198               (charset japanese-jisx0208
 199                        japanese-jisx0212 latin-jisx0201 katakana-jisx0201
 200                        japanese-jisx0213.2004-1 japanese-jisx0213-1
 201                        japanese-jisx0213-2 japanese-jisx0208-1978)
 202               (coding-system iso-2022-jp japanese-iso-8bit
 203                              japanese-shift-jis japanese-iso-7bit-1978-irv
 204                              iso-2022-jp-2004 japanese-shift-jis-2004
 205                              euc-jis-2004)
 206               (coding-priority iso-2022-jp japanese-iso-8bit
 207                                japanese-shift-jis
 208                                iso-2022-jp-2004 euc-jis-2004
 209                                japanese-shift-jis-2004
 210                                iso-2022-jp-2)
 211               (input-method . "japanese")
 212               (features japan-util)
 213               (sample-text . "Japanese (\e$BF|K\8l\e(B)   \e$B$3$s$K$A$O\e(B, \e(I:]FAJ\e(B")
 214               (documentation . t)))
 215
 216 (let ((map
 217        ;; JISX0213-1 vs Unicode
 218        '((#x2477 . [#x304B #x309A])
 219          (#x2478 . [#x304D #x309A])
 220          (#x2479 . [#x304F #x309A])
 221          (#x247a . [#x3051 #x309A])
 222          (#x247b . [#x3053 #x309A])
 223          (#x2577 . [#x30AB #x309A])
 224          (#x2578 . [#x30AD #x309A])
 225          (#x2579 . [#x30AF #x309A])
 226          (#x257a . [#x30B1 #x309A])
 227          (#x257b . [#x30B3 #x309A])
 228          (#x257c . [#x30BB #x309A])
 229          (#x257d . [#x30C4 #x309A])
 230          (#x257e . [#x30C8 #x309A])
 231          (#x2678 . [#x31F7 #x309A])
 232          (#x2b44 . [#x00E6 #x0300])
 233          (#x2b48 . [#x0254 #x0300])
 234          (#x2b49 . [#x0254 #x0301])
 235          (#x2b4a . [#x028C #x0300])
 236          (#x2b4b . [#x028C #x0301])
 237          (#x2b4c . [#x0259 #x0300])
 238          (#x2b4d . [#x0259 #x0301])
 239          (#x2b4e . [#x025A #x0300])
 240          (#x2b4f . [#x025A #x0301])
 241          (#x2b65 . [#x02E9 #x02E5])
 242          (#x2b66 . [#x02E5 #x02E9])))
 243       table)
 244   (dolist (elt map)
 245     (setcar elt (decode-char 'japanese-jisx0213-1 (car elt))))
 246   (setq table (make-translation-table-from-alist map))
 247   (define-translation-table 'jisx0213-to-unicode table)
 248   (define-translation-table 'unicode-to-jisx0213
 249     (char-table-extra-slot table 0)))
 250
 251 (defun compose-gstring-for-variation-glyph (gstring)
 252   "Compose glyph-string GSTRING for graphic display.
 253 GSTRING must have two glyphs; the first is a glyph for a han character,
 254 and the second is a glyph for a variation selector."
 255   (let* ((font (lgstring-font gstring))
 256          (han (lgstring-char gstring 0))
 257          (vs (lgstring-char gstring 1))
 258          (glyphs (font-variation-glyphs font han))
 259          (g0 (lgstring-glyph gstring 0))
 260          (g1 (lgstring-glyph gstring 1)))
 261     (catch 'tag
 262       (dolist (elt glyphs)
 263         (if (= (car elt) vs)
 264             (progn
 265               (lglyph-set-code g0 (cdr elt))
 266               (lglyph-set-from-to g0 (lglyph-from g0) (lglyph-to g1))
 267               (lgstring-set-glyph gstring 1 nil)
 268               (throw 'tag gstring)))))))
 269
 270 (let ((elt '([".." 1 compose-gstring-for-variation-glyph])))
 271   (set-char-table-range composition-function-table '(#xFE00 . #xFE0F) elt)
 272   (set-char-table-range composition-function-table '(#xE0100 . #xE01EF) elt))
 273
 274 (provide 'japanese)
 275
 276 ;;; japanese.el ends here