lisp/international/ja-dic-utl.el

   1 ;;; ja-dic-utl.el --- utilities for handling Japanese dictionary (SKK-JISYO.L)
   2
   3 ;; Copyright (C) 1995, 2000 Electrotechnical Laboratory, JAPAN.
   4 ;; Licensed to the Free Software Foundation.
   5
   6 ;; Keywords: mule, multilingual, Japanese
   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 ;; This file provides a generic function to look up a Japanese
  28 ;; dictionary of SKK format.
  29 ;;
  30 ;; SKK is a free Japanese input method running on Mule created by
  31 ;; Masahiko Sato <masahiko@sato.riec.tohoku.ac.jp>.  The Emacs Lisp
  32 ;; library kkc.el provides a facility to convert a Japanese kana
  33 ;; string to a kanji-kana-mixed string by using SKK's dictionary.
  34 ;;
  35 ;; The original SKK dictionary SKK-JISYO.L is converted to ja-dic.el
  36 ;; by ja-dic-cnv.el.  We get entries of the dictionary in four
  37 ;; variables (listed below) by loading that file (or byte-compiled
  38 ;; version ja-dic.elc).
  39
  40 ;;; Code:
  41
  42 ;; The following four variables are set by loading ja-dic.el[c].
  43 (defvar skkdic-okuri-ari nil
  44   "Nested alist for OKURI-ARI entries of SKK dictionary.")
  45
  46 (defvar skkdic-postfix nil
  47   "Nested alist for SETSUBIJI (postfix) entries of SKK dictionary.")
  48
  49 (defvar skkdic-prefix nil
  50   "Nested alist SETTOUJI (prefix) entries of SKK dictionary.")
  51
  52 (defvar skkdic-okuri-nasi nil
  53   "Nested alist for OKURI-NASI entries of SKK dictionary.")
  54
  55 (defconst skkdic-okurigana-table
  56   '((?\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)
  57     (?\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)
  58     (?\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)
  59     (?\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$4\e(B . ?g)
  60     (?\e$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$9\e(B . ?s)
  61     (?\e$B$:\e(B . ?z) (?\e$B$;\e(B . ?s) (?\e$B$<\e(B . ?z) (?\e$B$=\e(B . ?s) (?\e$B$>\e(B . ?z)
  62     (?\e$B$?\e(B . ?t) (?\e$B$@\e(B . ?d) (?\e$B$A\e(B . ?t) (?\e$B$B\e(B . ?d) (?\e$B$C\e(B . ?t)
  63     (?\e$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$I\e(B . ?d)
  64     (?\e$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$N\e(B . ?n)
  65     (?\e$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$S\e(B . ?b)
  66     (?\e$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$X\e(B . ?h)
  67     (?\e$B$Y\e(B . ?b) (?\e$B$Z\e(B . ?p) (?\e$B$[\e(B . ?h) (?\e$B$\\e(B . ?b) (?\e$B$]\e(B . ?p)
  68     (?\e$B$^\e(B . ?m) (?\e$B$_\e(B . ?m) (?\e$B$`\e(B . ?m) (?\e$B$a\e(B . ?m) (?\e$B$b\e(B . ?m)
  69     (?\e$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$h\e(B . ?y)
  70     (?\e$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$m\e(B . ?r)
  71     (?\e$B$o\e(B . ?w) (?\e$B$p\e(B . ?w) (?\e$B$q\e(B . ?w) (?\e$B$r\e(B . ?w)
  72     (?\e$B$s\e(B . ?n)
  73     )
  74   "Alist of Okuriganas vs trailing ASCII letters in OKURI-ARI entry.")
  75
  76 (defun skkdic-merge-head-and-tail (heads tails postfix)
  77   (let ((min-len 2)
  78         l)
  79     (while heads
  80       (if (or (not postfix)
  81               (>= (length (car heads)) min-len))
  82           (let ((tail tails))
  83             (while tail
  84               (if (or postfix
  85                       (>= (length (car tail)) min-len))
  86                   (setq l (cons (concat (car heads) (car tail)) l)))
  87               (setq tail (cdr tail)))))
  88       (setq heads (cdr heads)))
  89     l))
  90
  91 (defconst skkdic-jisx0208-hiragana-block (nth 1 (split-char ?\e$B$"\e(B)))
  92
  93 (defun skkdic-lookup-key (seq len &optional postfix prefer-noun)
  94   "Return a list of conversion string for sequence SEQ of length LEN.
  95
  96 SEQ is a vector of Kana characters to be converted by SKK dictionary.
  97 If LEN is shorter than the length of KEYSEQ, the first LEN keys in SEQ
  98 are took into account.
  99
 100 Optional 3rd arg POSTFIX non-nil means SETSUBIJI (postfix) are also
 101 considered to find conversion strings.
 102
 103 Optional 4th arg PREFER-NOUN non-nil means that the conversions
 104 without okurigana are placed at the head of the returned list."
 105   (or skkdic-okuri-nasi
 106       (condition-case err
 107           (load-library "ja-dic/ja-dic")
 108         (error (ding)
 109                (with-output-to-temp-buffer "*Help*"
 110                  (princ "The library `ja-dic' can't be loaded.
 111
 112 The most common case is that you have not yet installed the library
 113 included in LEIM (Libraries of Emacs Input Method) which is
 114 distributed separately from Emacs.
 115
 116 LEIM is available from the same ftp directory as Emacs."))
 117                (signal (car err) (cdr err)))))
 118
 119   (let ((vec (make-vector len 0))
 120         (i 0)
 121         entry)
 122     ;; At first, generate vector VEC from SEQ for looking up SKK
 123     ;; alists.  Nth element in VEC corresponds to Nth element in SEQ.
 124     ;; The values are decided as follows.
 125     ;;   If SEQ[N] is `\e$B!<\e(B', VEC[N] is 0,
 126     ;;   else if SEQ[N] is a Hiragana character, VEC[N] is:
 127     ;;     ((The 2nd position code of SEQ[N]) - 32),
 128     ;;   else VEC[N] is 128.
 129     (while (< i len)
 130       (let ((ch (aref seq i))
 131             elts)
 132         (if (= ch ?\e$B!<\e(B)
 133             (aset vec i 0)
 134           (setq elts (split-char ch))
 135           (if (and (eq (car elts) 'japanese-jisx0208)
 136                    (= (nth 1 elts) skkdic-jisx0208-hiragana-block))
 137               (aset vec i (- (nth 2 elts) 32))
 138             (aset vec i 128))))
 139       (setq i (1+ i)))
 140
 141     ;; Search OKURI-NASI entries.
 142     (setq entry (lookup-nested-alist vec skkdic-okuri-nasi len 0 t))
 143     (if (consp (car entry))
 144         (setq entry (copy-sequence (car entry)))
 145       (setq entry nil))
 146
 147     (if postfix
 148         ;; Search OKURI-NASI entries with postfixes.
 149         (let ((break (max (- len (car skkdic-postfix)) 1))
 150               entry-head entry-postfix entry2)
 151           (while (< break len)
 152             (if (and (setq entry-head
 153                            (lookup-nested-alist vec skkdic-okuri-nasi
 154                                                 break 0 t))
 155                      (consp (car entry-head))
 156                      (setq entry-postfix
 157                            (lookup-nested-alist vec skkdic-postfix
 158                                                 len break t))
 159                      (consp (car entry-postfix))
 160                      (setq entry2 (skkdic-merge-head-and-tail
 161                                    (car entry-head) (car entry-postfix) t)))
 162                 (if entry
 163                     (nconc entry entry2)
 164                   (setq entry entry2)))
 165             (setq break (1+ break)))))
 166
 167     ;; Search OKURI-NASI entries with prefixes.
 168     (let ((break (min (car skkdic-prefix) (- len 2)))
 169           entry-prefix entry-tail entry2)
 170       (while (> break 0)
 171         (if (and (setq entry-prefix
 172                        (lookup-nested-alist vec skkdic-prefix break 0 t))
 173                  (consp (car entry-prefix))
 174                  (setq entry-tail
 175                        (lookup-nested-alist vec skkdic-okuri-nasi len break t))
 176                  (consp (car entry-tail))
 177                  (setq entry2 (skkdic-merge-head-and-tail
 178                                (car entry-prefix) (car entry-tail) nil)))
 179             (progn
 180               (if entry
 181                   (nconc entry entry2)
 182                 (setq entry entry2))))
 183         (setq break (1- break))))
 184
 185     ;; Search OKURI-ARI entries.
 186     (let ((okurigana (assq (aref seq (1- len)) skkdic-okurigana-table))
 187           orig-element entry2)
 188       (if okurigana
 189           (progn
 190             (setq orig-element (aref vec (1- len)))
 191             (aset vec (1- len) (- (cdr okurigana)))
 192             (if (and (setq entry2 (lookup-nested-alist vec skkdic-okuri-ari
 193                                                        len 0 t))
 194                      (consp (car entry2)))
 195                 (progn
 196                   (setq entry2 (copy-sequence (car entry2)))
 197                   (let ((l entry2)
 198                         (okuri (char-to-string (aref seq (1- len)))))
 199                     (while l
 200                       (setcar l (concat (car l) okuri))
 201                       (setq l (cdr l)))
 202                     (if entry
 203                         (if prefer-noun
 204                             (nconc entry entry2)
 205                           (setq entry2 (nreverse entry2))
 206                           (nconc entry2 entry)
 207                           (setq entry entry2))
 208                       (setq entry (nreverse entry2))))))
 209             (aset vec (1- len) orig-element))))
 210
 211     entry))
 212
 213 ;;
 214 (provide 'ja-dic-utl)
 215
 216 ;; Local Variables:
 217 ;; coding: iso-2022-7bit
 218 ;; End:
 219
 220 ;;; ja-dic-utl.el ends here