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