lisp/international/ja-dic-utl.el

   1 ;;; ja-dic-utl.el --- utilities for handling Japanese dictionary (SKK-JISYO.L)
   2
   3 ;; Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   4 ;;   2005, 2006, 2007, 2008
   5 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
   6 ;;   Registration Number H14PRO021
   7
   8 ;; Keywords: mule, multilingual, Japanese
   9
  10 ;; This file is part of GNU Emacs.
  11
  12 ;; GNU Emacs is free software; you can redistribute it and/or modify
  13 ;; it under the terms of the GNU General Public License as published by
  14 ;; the Free Software Foundation; either version 3, or (at your option)
  15 ;; any later version.
  16
  17 ;; GNU Emacs is distributed in the hope that it will be useful,
  18 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  19 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  20 ;; GNU General Public License for more details.
  21
  22 ;; You should have received a copy of the GNU General Public License
  23 ;; along with GNU Emacs; see the file COPYING.  If not, write to the
  24 ;; Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
  25 ;; Boston, MA 02110-1301, USA.
  26
  27 ;;; Commentary:
  28
  29 ;; This file provides a generic function to look up a Japanese
  30 ;; dictionary of SKK format.
  31 ;;
  32 ;; SKK is a free Japanese input method running on Mule created by
  33 ;; Masahiko Sato <masahiko@sato.riec.tohoku.ac.jp>.  The Emacs Lisp
  34 ;; library kkc.el provides a facility to convert a Japanese kana
  35 ;; string to a kanji-kana-mixed string by using SKK's dictionary.
  36 ;;
  37 ;; The original SKK dictionary SKK-JISYO.L is converted to ja-dic.el
  38 ;; by ja-dic-cnv.el.  We get entries of the dictionary in four
  39 ;; variables (listed below) by loading that file (or byte-compiled
  40 ;; version ja-dic.elc).
  41
  42 ;;; Code:
  43
  44 ;; The following four variables are set by loading ja-dic.el[c].
  45 (defvar skkdic-okuri-ari nil
  46   "Nested alist for OKURI-ARI entries of SKK dictionary.")
  47
  48 (defvar skkdic-postfix nil
  49   "Nested alist for SETSUBIJI (postfix) entries of SKK dictionary.")
  50
  51 (defvar skkdic-prefix nil
  52   "Nested alist SETTOUJI (prefix) entries of SKK dictionary.")
  53
  54 (defvar skkdic-okuri-nasi nil
  55   "Nested alist for OKURI-NASI entries of SKK dictionary.")
  56
  57 (defconst skkdic-okurigana-table
  58   '((?\e$B$!\e(B . ?a) (?\e$B$"\e(B . ?a) (?\e$B$#\e(B . ?i) (?\e$B$$\e(B . ?i) (?\e$B$%\e(B . ?u)
  59     (?\e$B$&\e(B . ?u) (?\e$B$'\e(B . ?e) (?\e$B$(\e(B . ?e) (?\e$B$)\e(B . ?o) (?\e$B$*\e(B . ?o)
  60     (?\e$B$+\e(B . ?k) (?\e$B$,\e(B . ?g) (?\e$B$-\e(B . ?k) (?\e$B$.\e(B . ?g) (?\e$B$/\e(B . ?k)
  61     (?\e$B$0\e(B . ?g) (?\e$B$1\e(B . ?k) (?\e$B$2\e(B . ?g) (?\e$B$3\e(B . ?k) (?\e$B$4\e(B . ?g)
  62     (?\e$B$5\e(B . ?s) (?\e$B$6\e(B . ?z) (?\e$B$7\e(B . ?s) (?\e$B$8\e(B . ?j) (?\e$B$9\e(B . ?s)
  63     (?\e$B$:\e(B . ?z) (?\e$B$;\e(B . ?s) (?\e$B$<\e(B . ?z) (?\e$B$=\e(B . ?s) (?\e$B$>\e(B . ?z)
  64     (?\e$B$?\e(B . ?t) (?\e$B$@\e(B . ?d) (?\e$B$A\e(B . ?t) (?\e$B$B\e(B . ?d) (?\e$B$C\e(B . ?t)
  65     (?\e$B$D\e(B . ?t) (?\e$B$E\e(B . ?d) (?\e$B$F\e(B . ?t) (?\e$B$G\e(B . ?d) (?\e$B$H\e(B . ?t) (?\e$B$I\e(B . ?d)
  66     (?\e$B$J\e(B . ?n) (?\e$B$K\e(B . ?n) (?\e$B$L\e(B . ?n) (?\e$B$M\e(B . ?n) (?\e$B$N\e(B . ?n)
  67     (?\e$B$O\e(B . ?h) (?\e$B$P\e(B . ?b) (?\e$B$Q\e(B . ?p) (?\e$B$R\e(B . ?h) (?\e$B$S\e(B . ?b)
  68     (?\e$B$T\e(B . ?p) (?\e$B$U\e(B . ?h) (?\e$B$V\e(B . ?b) (?\e$B$W\e(B . ?p) (?\e$B$X\e(B . ?h)
  69     (?\e$B$Y\e(B . ?b) (?\e$B$Z\e(B . ?p) (?\e$B$[\e(B . ?h) (?\e$B$\\e(B . ?b) (?\e$B$]\e(B . ?p)
  70     (?\e$B$^\e(B . ?m) (?\e$B$_\e(B . ?m) (?\e$B$`\e(B . ?m) (?\e$B$a\e(B . ?m) (?\e$B$b\e(B . ?m)
  71     (?\e$B$c\e(B . ?y) (?\e$B$d\e(B . ?y) (?\e$B$e\e(B . ?y) (?\e$B$f\e(B . ?y) (?\e$B$g\e(B . ?y) (?\e$B$h\e(B . ?y)
  72     (?\e$B$i\e(B . ?r) (?\e$B$j\e(B . ?r) (?\e$B$k\e(B . ?r) (?\e$B$l\e(B . ?r) (?\e$B$m\e(B . ?r)
  73     (?\e$B$o\e(B . ?w) (?\e$B$p\e(B . ?w) (?\e$B$q\e(B . ?w) (?\e$B$r\e(B . ?w)
  74     (?\e$B$s\e(B . ?n)
  75     )
  76   "Alist of Okuriganas vs trailing ASCII letters in OKURI-ARI entry.")
  77
  78 (defun skkdic-merge-head-and-tail (heads tails postfix)
  79   (let ((min-len 2)
  80         l)
  81     (while heads
  82       (if (or (not postfix)
  83               (>= (length (car heads)) min-len))
  84           (let ((tail tails))
  85             (while tail
  86               (if (or postfix
  87                       (>= (length (car tail)) min-len))
  88                   (setq l (cons (concat (car heads) (car tail)) l)))
  89               (setq tail (cdr tail)))))
  90       (setq heads (cdr heads)))
  91     l))
  92
  93 (defconst skkdic-jisx0208-hiragana-block
  94   (cons (decode-char 'japanese-jisx0208 #x2421)
  95         (decode-char 'japanese-jisx0208 #x247E)))
  96
  97 (defun skkdic-lookup-key (seq len &optional postfix prefer-noun)
  98   "Return a list of conversion string for sequence SEQ of length LEN.
  99
 100 SEQ is a vector of Kana characters to be converted by SKK dictionary.
 101 If LEN is shorter than the length of KEYSEQ, the first LEN keys in SEQ
 102 are took into account.
 103
 104 Optional 3rd arg POSTFIX non-nil means SETSUBIJI (postfix) are also
 105 considered to find conversion strings.
 106
 107 Optional 4th arg PREFER-NOUN non-nil means that the conversions
 108 without okurigana are placed at the head of the returned list."
 109   (or skkdic-okuri-nasi
 110       (condition-case err
 111           (load-library "ja-dic/ja-dic")
 112         (error (ding)
 113                (with-output-to-temp-buffer "*Help*"
 114                  (princ "The library `ja-dic' can't be loaded.
 115
 116 The most common case is that you have not yet installed the library
 117 included in LEIM (Libraries of Emacs Input Method) which is
 118 distributed separately from Emacs.
 119
 120 LEIM is available from the same ftp directory as Emacs."))
 121                (signal (car err) (cdr err)))))
 122
 123   (let ((vec (make-vector len 0))
 124         (i 0)
 125         entry)
 126     ;; At first, generate vector VEC from SEQ for looking up SKK
 127     ;; alists.  Nth element in VEC corresponds to Nth element in SEQ.
 128     ;; The values are decided as follows.
 129     ;;   If SEQ[N] is `\e$B!<\e(B', VEC[N] is 0,
 130     ;;   else if SEQ[N] is a Hiragana character, VEC[N] is:
 131     ;;     ((The 2nd position code of SEQ[N]) - 32),
 132     ;;   else VEC[N] is 128.
 133     (while (< i len)
 134       (let ((ch (aref seq i))
 135             code)
 136         (cond ((= ch ?\e$B!<\e(B)
 137                (aset vec i 0))
 138               ((and (>= ch (car skkdic-jisx0208-hiragana-block))
 139                     (<= ch (cdr skkdic-jisx0208-hiragana-block)))
 140                (setq code (encode-char ch 'japanese-jisx0208))
 141                (if code
 142                    (aset vec i (- (logand code #xFF) 32))
 143                  (aset vec i 128)))
 144               (t
 145                (aset vec i 128))))
 146       (setq i (1+ i)))
 147
 148     ;; Search OKURI-NASI entries.
 149     (setq entry (lookup-nested-alist vec skkdic-okuri-nasi len 0 t))
 150     (if (consp (car entry))
 151         (setq entry (copy-sequence (car entry)))
 152       (setq entry nil))
 153
 154     (if postfix
 155         ;; Search OKURI-NASI entries with postfixes.
 156         (let ((break (max (- len (car skkdic-postfix)) 1))
 157               entry-head entry-postfix entry2)
 158           (while (< break len)
 159             (if (and (setq entry-head
 160                            (lookup-nested-alist vec skkdic-okuri-nasi
 161                                                 break 0 t))
 162                      (consp (car entry-head))
 163                      (setq entry-postfix
 164                            (lookup-nested-alist vec skkdic-postfix
 165                                                 len break t))
 166                      (consp (car entry-postfix))
 167                      (setq entry2 (skkdic-merge-head-and-tail
 168                                    (car entry-head) (car entry-postfix) t)))
 169                 (if entry
 170                     (nconc entry entry2)
 171                   (setq entry entry2)))
 172             (setq break (1+ break)))))
 173
 174     ;; Search OKURI-NASI entries with prefixes.
 175     (let ((break (min (car skkdic-prefix) (- len 2)))
 176           entry-prefix entry-tail entry2)
 177       (while (> break 0)
 178         (if (and (setq entry-prefix
 179                        (lookup-nested-alist vec skkdic-prefix break 0 t))
 180                  (consp (car entry-prefix))
 181                  (setq entry-tail
 182                        (lookup-nested-alist vec skkdic-okuri-nasi len break t))
 183                  (consp (car entry-tail))
 184                  (setq entry2 (skkdic-merge-head-and-tail
 185                                (car entry-prefix) (car entry-tail) nil)))
 186             (progn
 187               (if entry
 188                   (nconc entry entry2)
 189                 (setq entry entry2))))
 190         (setq break (1- break))))
 191
 192     ;; Search OKURI-ARI entries.
 193     (let ((okurigana (assq (aref seq (1- len)) skkdic-okurigana-table))
 194           orig-element entry2)
 195       (if okurigana
 196           (progn
 197             (setq orig-element (aref vec (1- len)))
 198             (aset vec (1- len) (- (cdr okurigana)))
 199             (if (and (setq entry2 (lookup-nested-alist vec skkdic-okuri-ari
 200                                                        len 0 t))
 201                      (consp (car entry2)))
 202                 (progn
 203                   (setq entry2 (copy-sequence (car entry2)))
 204                   (let ((l entry2)
 205                         (okuri (char-to-string (aref seq (1- len)))))
 206                     (while l
 207                       (setcar l (concat (car l) okuri))
 208                       (setq l (cdr l)))
 209                     (if entry
 210                         (if prefer-noun
 211                             (nconc entry entry2)
 212                           (setq entry2 (nreverse entry2))
 213                           (nconc entry2 entry)
 214                           (setq entry entry2))
 215                       (setq entry (nreverse entry2))))))
 216             (aset vec (1- len) orig-element))))
 217
 218     entry))
 219
 220 ;;
 221 (provide 'ja-dic-utl)
 222
 223 ;; Local Variables:
 224 ;; coding: iso-2022-7bit
 225 ;; End:
 226
 227 ;; arch-tag: df2218fa-469c-40f6-bace-7f89a053f9c0
 228 ;;; ja-dic-utl.el ends here