17315
|
1 ;;; skkdic-utl.el --- Utility functions for handling SKK dictionary
|
17052
|
2
|
|
3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
|
18377
|
4 ;; Licensed to the Free Software Foundation.
|
17052
|
5
|
|
6 ;; Keywords: mule, multilingual, Japanese, SKK
|
|
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
|
17071
|
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.
|
17052
|
24
|
|
25 ;;; Commentary:
|
|
26
|
|
27 ;; SKK is a free Japanese input method running on Mule created by
|
18645
|
28 ;; Masahiko Sato <masahiko@sato.riec.tohoku.ac.jp>. The Emacs Lisp
|
|
29 ;; library kkc.el provides a facility to convert a Japanese kana
|
|
30 ;; string to a kanji-kana-mixed string by using a SKK dictionary.
|
|
31 ;;
|
|
32 ;; This file provides a generic function to look up a SKK dictionary.
|
|
33 ;;
|
|
34 ;; The original SKK dictionary SKK-JISYO.L is converted to skkdic.el.
|
|
35 ;; We get entries of the dictionary in four variables (listed below)
|
|
36 ;; by loadig this file (or byte-compiled version skkdic.elc).
|
17052
|
37
|
|
38 ;;; Code:
|
|
39
|
18645
|
40 ;; The following four variables are set by loading skkdic.el[c].
|
17052
|
41 (defvar skkdic-okuri-ari nil
|
18645
|
42 "Nested alist for OKURI-ARI entries of SKK dictionary.")
|
|
43
|
17052
|
44 (defvar skkdic-postfix nil
|
18645
|
45 "Nested alist for SETSUBIJI (postfix) entries of SKK dictionary.")
|
|
46
|
17052
|
47 (defvar skkdic-prefix nil
|
18645
|
48 "Nested alist SETTOUJI (prefix) entries of SKK dictionary.")
|
|
49
|
17052
|
50 (defvar skkdic-okuri-nasi nil
|
18645
|
51 "Nested alist for OKURI-NASI entries of SKK dictionary.")
|
17052
|
52
|
|
53 (defconst skkdic-okurigana-table
|
|
54 '((?$B$!(B . ?a) (?$B$"(B . ?a) (?$B$#(B . ?i) (?$B$$(B . ?i) (?$B$%(B . ?u)
|
|
55 (?$B$&(B . ?u) (?$B$'(B . ?e) (?$B$((B . ?e) (?$B$)(B . ?o) (?$B$*(B . ?o)
|
|
56 (?$B$+(B . ?k) (?$B$,(B . ?g) (?$B$-(B . ?k) (?$B$.(B . ?g) (?$B$/(B . ?k)
|
|
57 (?$B$0(B . ?g) (?$B$1(B . ?k) (?$B$2(B . ?g) (?$B$3(B . ?k) (?$B$4(B . ?g)
|
|
58 (?$B$5(B . ?s) (?$B$6(B . ?z) (?$B$7(B . ?s) (?$B$8(B . ?j) (?$B$9(B . ?s)
|
|
59 (?$B$:(B . ?z) (?$B$;(B . ?s) (?$B$<(B . ?z) (?$B$=(B . ?s) (?$B$>(B . ?z)
|
|
60 (?$B$?(B . ?t) (?$B$@(B . ?d) (?$B$A(B . ?t) (?$B$B(B . ?d) (?$B$C(B . ?t)
|
|
61 (?$B$D(B . ?t) (?$B$E(B . ?d) (?$B$F(B . ?t) (?$B$G(B . ?d) (?$B$H(B . ?t) (?$B$I(B . ?d)
|
|
62 (?$B$J(B . ?n) (?$B$K(B . ?n) (?$B$L(B . ?n) (?$B$M(B . ?n) (?$B$N(B . ?n)
|
|
63 (?$B$O(B . ?h) (?$B$P(B . ?b) (?$B$Q(B . ?p) (?$B$R(B . ?h) (?$B$S(B . ?b)
|
|
64 (?$B$T(B . ?p) (?$B$U(B . ?h) (?$B$V(B . ?b) (?$B$W(B . ?p) (?$B$X(B . ?h)
|
|
65 (?$B$Y(B . ?b) (?$B$Z(B . ?p) (?$B$[(B . ?h) (?$B$\(B . ?b) (?$B$](B . ?p)
|
|
66 (?$B$^(B . ?m) (?$B$_(B . ?m) (?$B$`(B . ?m) (?$B$a(B . ?m) (?$B$b(B . ?m)
|
|
67 (?$B$c(B . ?y) (?$B$d(B . ?y) (?$B$e(B . ?y) (?$B$f(B . ?y) (?$B$g(B . ?y) (?$B$h(B . ?y)
|
|
68 (?$B$i(B . ?r) (?$B$j(B . ?r) (?$B$k(B . ?r) (?$B$l(B . ?r) (?$B$m(B . ?r)
|
|
69 (?$B$o(B . ?w) (?$B$p(B . ?w) (?$B$q(B . ?w) (?$B$r(B . ?w)
|
|
70 (?$B$s(B . ?n)
|
18645
|
71 )
|
|
72 "Alist of Okuriganas vs trailing ASCII letters in OKURI-ARI entry.")
|
17052
|
73
|
|
74 (defun skkdic-merge-head-and-tail (heads tails postfix)
|
20736
|
75 (let ((min-len 2)
|
17052
|
76 l)
|
|
77 (while heads
|
|
78 (if (or (not postfix)
|
|
79 (>= (length (car heads)) min-len))
|
|
80 (let ((tail tails))
|
|
81 (while tail
|
|
82 (if (or postfix
|
|
83 (>= (length (car tail)) min-len))
|
|
84 (setq l (cons (concat (car heads) (car tail)) l)))
|
|
85 (setq tail (cdr tail)))))
|
|
86 (setq heads (cdr heads)))
|
|
87 l))
|
|
88
|
|
89 (defconst skkdic-jisx0208-hiragana-block (nth 1 (split-char ?$B$"(B)))
|
|
90
|
18645
|
91 (defun skkdic-lookup-key (seq len &optional postfix prefer-noun)
|
17052
|
92 "Return a list of conversion string for sequence SEQ of length LEN.
|
|
93
|
|
94 SEQ is a vector of Kana characters to be converted by SKK dictionary.
|
|
95 If LEN is shorter than the length of KEYSEQ, the first LEN keys in SEQ
|
|
96 are took into account.
|
|
97
|
18645
|
98 Optional 3rd arg POSTFIX non-nil means SETSUBIJI (postfix) are also
|
|
99 considered to find conversion strings.
|
|
100
|
|
101 Optional 4th arg PREFER-NOUN non-nil means that the conversions
|
|
102 without okurigana are placed at the head of the returned list."
|
17052
|
103 (or skkdic-okuri-nasi
|
|
104 (condition-case err
|
|
105 (load-library "skk/skkdic")
|
|
106 (error (ding)
|
|
107 (with-output-to-temp-buffer "*Help*"
|
|
108 (princ "The library `skkdic' can't be loaded.
|
|
109
|
|
110 The most common case is that you have not yet installed the library
|
|
111 included in LEIM (Libraries of Emacs Input Method) which is
|
|
112 distributed separately from Emacs.
|
|
113
|
|
114 LEIM is available from the same ftp directory as Emacs."))
|
|
115 (signal (car err) (cdr err)))))
|
|
116
|
|
117 (let ((vec (make-vector len 0))
|
|
118 (i 0)
|
|
119 entry)
|
|
120 ;; At first, generate vector VEC from SEQ for looking up SKK
|
|
121 ;; alists. Nth element in VEC corresponds to Nth element in SEQ.
|
|
122 ;; The values are decided as follows.
|
|
123 ;; If SEQ[N] is `$B!<(B', VEC[N] is 0,
|
18645
|
124 ;; else if SEQ[N] is a Hiragana character, VEC[N] is:
|
|
125 ;; ((The 2nd position code of SEQ[N]) - 32),
|
|
126 ;; else VEC[N] is 128.
|
17052
|
127 (while (< i len)
|
|
128 (let ((ch (aref seq i))
|
|
129 elts)
|
|
130 (if (= ch ?$B!<(B)
|
|
131 (aset vec i 0)
|
|
132 (setq elts (split-char ch))
|
|
133 (if (and (eq (car elts) 'japanese-jisx0208)
|
|
134 (= (nth 1 elts) skkdic-jisx0208-hiragana-block))
|
|
135 (aset vec i (- (nth 2 elts) 32))
|
|
136 (aset vec i 128))))
|
|
137 (setq i (1+ i)))
|
|
138
|
|
139 ;; Search OKURI-NASI entries.
|
|
140 (setq entry (lookup-nested-alist vec skkdic-okuri-nasi len 0 t))
|
|
141 (if (consp (car entry))
|
|
142 (setq entry (copy-sequence (car entry)))
|
|
143 (setq entry nil))
|
|
144
|
|
145 (if postfix
|
|
146 ;; Search OKURI-NASI entries with postfixes.
|
|
147 (let ((break (max (- len (car skkdic-postfix)) 1))
|
|
148 entry-head entry-postfix entry2)
|
|
149 (while (< break len)
|
|
150 (if (and (setq entry-head
|
|
151 (lookup-nested-alist vec skkdic-okuri-nasi
|
|
152 break 0 t))
|
|
153 (consp (car entry-head))
|
|
154 (setq entry-postfix
|
|
155 (lookup-nested-alist vec skkdic-postfix
|
|
156 len break t))
|
|
157 (consp (car entry-postfix))
|
|
158 (setq entry2 (skkdic-merge-head-and-tail
|
|
159 (car entry-head) (car entry-postfix) t)))
|
|
160 (if entry
|
|
161 (nconc entry entry2)
|
|
162 (setq entry entry2)))
|
|
163 (setq break (1+ break)))))
|
|
164
|
|
165 ;; Search OKURI-NASI entries with prefixes.
|
|
166 (let ((break (min (car skkdic-prefix) (- len 2)))
|
|
167 entry-prefix entry-tail entry2)
|
|
168 (while (> break 0)
|
|
169 (if (and (setq entry-prefix
|
|
170 (lookup-nested-alist vec skkdic-prefix break 0 t))
|
|
171 (consp (car entry-prefix))
|
|
172 (setq entry-tail
|
|
173 (lookup-nested-alist vec skkdic-okuri-nasi len break t))
|
|
174 (consp (car entry-tail))
|
|
175 (setq entry2 (skkdic-merge-head-and-tail
|
|
176 (car entry-prefix) (car entry-tail) nil)))
|
18645
|
177 (progn
|
|
178 (if entry
|
|
179 (nconc entry entry2)
|
|
180 (setq entry entry2))))
|
17052
|
181 (setq break (1- break))))
|
|
182
|
|
183 ;; Search OKURI-ARI entries.
|
|
184 (let ((okurigana (assq (aref seq (1- len)) skkdic-okurigana-table))
|
|
185 orig-element entry2)
|
|
186 (if okurigana
|
|
187 (progn
|
|
188 (setq orig-element (aref vec (1- len)))
|
|
189 (aset vec (1- len) (- (cdr okurigana)))
|
|
190 (if (and (setq entry2 (lookup-nested-alist vec skkdic-okuri-ari
|
|
191 len 0 t))
|
|
192 (consp (car entry2)))
|
|
193 (progn
|
|
194 (setq entry2 (copy-sequence (car entry2)))
|
|
195 (let ((l entry2)
|
|
196 (okuri (char-to-string (aref seq (1- len)))))
|
|
197 (while l
|
|
198 (setcar l (concat (car l) okuri))
|
|
199 (setq l (cdr l)))
|
|
200 (if entry
|
18645
|
201 (if prefer-noun
|
|
202 (nconc entry entry2)
|
|
203 (setq entry2 (nreverse entry2))
|
|
204 (nconc entry2 entry)
|
|
205 (setq entry entry2))
|
|
206 (setq entry (nreverse entry2))))))
|
17052
|
207 (aset vec (1- len) orig-element))))
|
|
208
|
|
209 entry))
|
|
210
|
|
211 ;;
|
|
212 (provide 'skkdic-utl)
|
|
213
|
|
214 ;; skkdic-utl.el ends here
|