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$s\e(B . ?n)
  75     )
  76   "Alist of Okuriganas vs trailing ASCII letters in OKURI-ARI entry.")
  77
  78 (defun skkdic-merge-head-and-tail (heads tails postfix)
  79   (let ((min-len 2)
  80         l)
  81     (while heads
  82       (if (or (not postfix)
  83               (>= (length (car heads)) min-len))
  84           (let ((tail tails))
  85             (while tail
  86               (if (or postfix
  87                       (>= (length (car tail)) min-len))
  88                   (setq l (cons (concat (car heads) (car tail)) l)))
  89               (setq tail (cdr tail)))))
  90       (setq heads (cdr heads)))
  91     l))
  92
  93 (defconst skkdic-jisx0208-hiragana-block (nth 1 (split-char ?\e$B$"\e(B)))
  94
  95 (defun skkdic-lookup-key (seq len &optional postfix prefer-noun)
  96   "Return a list of conversion string for sequence SEQ of length LEN.
  97
  98 SEQ is a vector of Kana characters to be converted by SKK dictionary.
  99 If LEN is shorter than the length of KEYSEQ, the first LEN keys in SEQ
 100 are took into account.
 101
 102 Optional 3rd arg POSTFIX non-nil means SETSUBIJI (postfix) are also
 103 considered to find conversion strings.
 104
 105 Optional 4th arg PREFER-NOUN non-nil means that the conversions
 106 without okurigana are placed at the head of the returned list."
 107   (or skkdic-okuri-nasi
 108       (condition-case err
 109           (load-library "ja-dic/ja-dic")
 110         (error (ding)
 111                (with-output-to-temp-buffer "*Help*"
 112                  (princ "The library `ja-dic' can't be loaded.
 113
 114 The most common case is that you have not yet installed the library
 115 included in LEIM (Libraries of Emacs Input Method) which is
 116 distributed separately from Emacs.
 117
 118 LEIM is available from the same ftp directory as Emacs."))
 119                (signal (car err) (cdr err)))))
 120
 121   (let ((vec (make-vector len 0))
 122         (i 0)
 123         entry)
 124     ;; At first, generate vector VEC from SEQ for looking up SKK
 125     ;; alists.  Nth element in VEC corresponds to Nth element in SEQ.
 126     ;; The values are decided as follows.
 127     ;;   If SEQ[N] is `\e$B!<\e(B', VEC[N] is 0,
 128     ;;   else if SEQ[N] is a Hiragana character, VEC[N] is:
 129     ;;     ((The 2nd position code of SEQ[N]) - 32),
 130     ;;   else VEC[N] is 128.
 131     (while (< i len)
 132       (let ((ch (aref seq i))
 133             elts)
 134         (if (= ch ?\e$B!<\e(B)
 135             (aset vec i 0)
 136           (setq elts (split-char ch))
 137           (if (and (eq (car elts) 'japanese-jisx0208)
 138                    (= (nth 1 elts) skkdic-jisx0208-hiragana-block))
 139               (aset vec i (- (nth 2 elts) 32))
 140             (aset vec i 128))))
 141       (setq i (1+ i)))
 142
 143     ;; Search OKURI-NASI entries.
 144     (setq entry (lookup-nested-alist vec skkdic-okuri-nasi len 0 t))
 145     (if (consp (car entry))
 146         (setq entry (copy-sequence (car entry)))
 147       (setq entry nil))
 148
 149     (if postfix
 150         ;; Search OKURI-NASI entries with postfixes.
 151         (let ((break (max (- len (car skkdic-postfix)) 1))
 152               entry-head entry-postfix entry2)
 153           (while (< break len)
 154             (if (and (setq entry-head
 155                            (lookup-nested-alist vec skkdic-okuri-nasi
 156                                                 break 0 t))
 157                      (consp (car entry-head))
 158                      (setq entry-postfix
 159                            (lookup-nested-alist vec skkdic-postfix
 160                                                 len break t))
 161                      (consp (car entry-postfix))
 162                      (setq entry2 (skkdic-merge-head-and-tail
 163                                    (car entry-head) (car entry-postfix) t)))
 164                 (if entry
 165                     (nconc entry entry2)
 166                   (setq entry entry2)))
 167             (setq break (1+ break)))))
 168
 169     ;; Search OKURI-NASI entries with prefixes.
 170     (let ((break (min (car skkdic-prefix) (- len 2)))
 171           entry-prefix entry-tail entry2)
 172       (while (> break 0)
 173         (if (and (setq entry-prefix
 174                        (lookup-nested-alist vec skkdic-prefix break 0 t))
 175                  (consp (car entry-prefix))
 176                  (setq entry-tail
 177                        (lookup-nested-alist vec skkdic-okuri-nasi len break t))
 178                  (consp (car entry-tail))
 179                  (setq entry2 (skkdic-merge-head-and-tail
 180                                (car entry-prefix) (car entry-tail) nil)))
 181             (progn
 182               (if entry
 183                   (nconc entry entry2)
 184                 (setq entry entry2))))
 185         (setq break (1- break))))
 186
 187     ;; Search OKURI-ARI entries.
 188     (let ((okurigana (assq (aref seq (1- len)) skkdic-okurigana-table))
 189           orig-element entry2)
 190       (if okurigana
 191           (progn
 192             (setq orig-element (aref vec (1- len)))
 193             (aset vec (1- len) (- (cdr okurigana)))
 194             (if (and (setq entry2 (lookup-nested-alist vec skkdic-okuri-ari
 195                                                        len 0 t))
 196                      (consp (car entry2)))
 197                 (progn
 198                   (setq entry2 (copy-sequence (car entry2)))
 199                   (let ((l entry2)
 200                         (okuri (char-to-string (aref seq (1- len)))))
 201                     (while l
 202                       (setcar l (concat (car l) okuri))
 203                       (setq l (cdr l)))
 204                     (if entry
 205                         (if prefer-noun
 206                             (nconc entry entry2)
 207                           (setq entry2 (nreverse entry2))
 208                           (nconc entry2 entry)
 209                           (setq entry entry2))
 210                       (setq entry (nreverse entry2))))))
 211             (aset vec (1- len) orig-element))))
 212
 213     entry))
 214
 215 ;;
 216 (provide 'ja-dic-utl)
 217
 218 ;; Local Variables:
 219 ;; coding: iso-2022-7bit
 220 ;; End:
 221
 222 ;;; arch-tag: df2218fa-469c-40f6-bace-7f89a053f9c0
 223 ;;; ja-dic-utl.el ends here