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, 2009, 2010, 2011
   5 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
   6 ;;   Registration Number H14PRO021
   7
   8 ;; Keywords: i18n, 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 of the License, or
  15 ;; (at your option) 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.  If not, see <http://www.gnu.org/licenses/>.
  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
  92   (cons (decode-char 'japanese-jisx0208 #x2421)
  93         (decode-char 'japanese-jisx0208 #x247E)))
  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 (substitute-command-keys "\
 113 The library `ja-dic' can't be loaded.
 114
 115 The most common case is that you have not yet installed the library
 116 included in LEIM (Libraries of Emacs Input Method) which is
 117 distributed separately from Emacs.
 118
 119 LEIM is available from the same ftp directory as Emacs.")))
 120                (signal (car err) (cdr err)))))
 121
 122   (let ((vec (make-vector len 0))
 123         (i 0)
 124         entry)
 125     ;; At first, generate vector VEC from SEQ for looking up SKK
 126     ;; alists.  Nth element in VEC corresponds to Nth element in SEQ.
 127     ;; The values are decided as follows.
 128     ;;   If SEQ[N] is `\e$B!<\e(B', VEC[N] is 0,
 129     ;;   else if SEQ[N] is a Hiragana character, VEC[N] is:
 130     ;;     ((The 2nd position code of SEQ[N]) - 32),
 131     ;;   else VEC[N] is 128.
 132     (while (< i len)
 133       (let ((ch (aref seq i))
 134             code)
 135         (cond ((= ch ?\e$B!<\e(B)
 136                (aset vec i 0))
 137               ((and (>= ch (car skkdic-jisx0208-hiragana-block))
 138                     (<= ch (cdr skkdic-jisx0208-hiragana-block)))
 139                (setq code (encode-char ch 'japanese-jisx0208))
 140                (if code
 141                    (aset vec i (- (logand code #xFF) 32))
 142                  (aset vec i 128)))
 143               (t
 144                (aset vec i 128))))
 145       (setq i (1+ i)))
 146
 147     ;; Search OKURI-NASI entries.
 148     (setq entry (lookup-nested-alist vec skkdic-okuri-nasi len 0 t))
 149     (if (consp (car entry))
 150         (setq entry (copy-sequence (car entry)))
 151       (setq entry nil))
 152
 153     (if postfix
 154         ;; Search OKURI-NASI entries with postfixes.
 155         (let ((break (max (- len (car skkdic-postfix)) 1))
 156               entry-head entry-postfix entry2)
 157           (while (< break len)
 158             (if (and (setq entry-head
 159                            (lookup-nested-alist vec skkdic-okuri-nasi
 160                                                 break 0 t))
 161                      (consp (car entry-head))
 162                      (setq entry-postfix
 163                            (lookup-nested-alist vec skkdic-postfix
 164                                                 len break t))
 165                      (consp (car entry-postfix))
 166                      (setq entry2 (skkdic-merge-head-and-tail
 167                                    (car entry-head) (car entry-postfix) t)))
 168                 (if entry
 169                     (nconc entry entry2)
 170                   (setq entry entry2)))
 171             (setq break (1+ break)))))
 172
 173     ;; Search OKURI-NASI entries with prefixes.
 174     (let ((break (min (car skkdic-prefix) (- len 2)))
 175           entry-prefix entry-tail entry2)
 176       (while (> break 0)
 177         (if (and (setq entry-prefix
 178                        (lookup-nested-alist vec skkdic-prefix break 0 t))
 179                  (consp (car entry-prefix))
 180                  (setq entry-tail
 181                        (lookup-nested-alist vec skkdic-okuri-nasi len break t))
 182                  (consp (car entry-tail))
 183                  (setq entry2 (skkdic-merge-head-and-tail
 184                                (car entry-prefix) (car entry-tail) nil)))
 185             (progn
 186               (if entry
 187                   (nconc entry entry2)
 188                 (setq entry entry2))))
 189         (setq break (1- break))))
 190
 191     ;; Search OKURI-ARI entries.
 192     (let ((okurigana (assq (aref seq (1- len)) skkdic-okurigana-table))
 193           orig-element entry2)
 194       (if okurigana
 195           (progn
 196             (setq orig-element (aref vec (1- len)))
 197             (aset vec (1- len) (- (cdr okurigana)))
 198             (if (and (setq entry2 (lookup-nested-alist vec skkdic-okuri-ari
 199                                                        len 0 t))
 200                      (consp (car entry2)))
 201                 (progn
 202                   (setq entry2 (copy-sequence (car entry2)))
 203                   (let ((l entry2)
 204                         (okuri (char-to-string (aref seq (1- len)))))
 205                     (while l
 206                       (setcar l (concat (car l) okuri))
 207                       (setq l (cdr l)))
 208                     (if entry
 209                         (if prefer-noun
 210                             (nconc entry entry2)
 211                           (setq entry2 (nreverse entry2))
 212                           (nconc entry2 entry)
 213                           (setq entry entry2))
 214                       (setq entry (nreverse entry2))))))
 215             (aset vec (1- len) orig-element))))
 216
 217     entry))
 218
 219 ;;
 220 (provide 'ja-dic-utl)
 221
 222 ;; Local Variables:
 223 ;; coding: iso-2022-7bit
 224 ;; End:
 225
 226 ;;; ja-dic-utl.el ends here