Initial revision
[bpt/emacs.git] / lisp / international / skkdic-utl.el
CommitLineData
4ed46869
KH
1;; skkdic-utl.el -- utility functions for handling skkdic.el
2
3;; Copyright (C) 1995 Free Software Foundation, Inc.
4;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
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
21;; along with GNU Emacs; see the file COPYING. If not, write to
22;; the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
23
24;;; Commentary:
25
26;; SKK is a free Japanese input method running on Mule created by
27;; Masahiko Sato <masahiko@sato.riec.tohoku.ac.jp>. A dictionary of
28;; SKK can be converted by `skkdic-convert' (skkdic-conv.el) to a file
29;; "skkdic.el" in which the dictionary entries are defined in the
30;; format which can be handled by the following functions.
31
32;;; Code:
33
34(defvar skkdic-okuri-ari nil
35 "OKURI-ARI entries of SKK dictionary.")
36(defvar skkdic-postfix nil
37 "POSTFIX entries of SKK dictionary.")
38(defvar skkdic-prefix nil
39 "PREFIX entries of SKK dictionary.")
40(defvar skkdic-okuri-nasi nil
41 "OKURI-NASI entries of SKK dictionary.")
42
43;; Alist of Okuriganas vs trailing ASCII letters in OKURI-ARI entry.
44(defconst skkdic-okurigana-table
45 '((?\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)
46 (?\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)
47 (?\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)
48 (?\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$4\e(B . ?g)
49 (?\e$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$9\e(B . ?s)
50 (?\e$B$:\e(B . ?z) (?\e$B$;\e(B . ?s) (?\e$B$<\e(B . ?z) (?\e$B$=\e(B . ?s) (?\e$B$>\e(B . ?z)
51 (?\e$B$?\e(B . ?t) (?\e$B$@\e(B . ?d) (?\e$B$A\e(B . ?t) (?\e$B$B\e(B . ?d) (?\e$B$C\e(B . ?t)
52 (?\e$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$I\e(B . ?d)
53 (?\e$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$N\e(B . ?n)
54 (?\e$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$S\e(B . ?b)
55 (?\e$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$X\e(B . ?h)
56 (?\e$B$Y\e(B . ?b) (?\e$B$Z\e(B . ?p) (?\e$B$[\e(B . ?h) (?\e$B$\\e(B . ?b) (?\e$B$]\e(B . ?p)
57 (?\e$B$^\e(B . ?m) (?\e$B$_\e(B . ?m) (?\e$B$`\e(B . ?m) (?\e$B$a\e(B . ?m) (?\e$B$b\e(B . ?m)
58 (?\e$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$h\e(B . ?y)
59 (?\e$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$m\e(B . ?r)
60 (?\e$B$o\e(B . ?w) (?\e$B$p\e(B . ?w) (?\e$B$q\e(B . ?w) (?\e$B$r\e(B . ?w)
61 (?\e$B$s\e(B . ?n)
62 ))
63
64(defconst skkdic-jbytes
65 (charset-bytes 'japanese-jisx0208))
66
67(defun skkdic-merge-head-and-tail (heads tails postfix)
68 (let ((min-len (* skkdic-jbytes 2))
69 l)
70 (while heads
71 (if (or (not postfix)
72 (>= (length (car heads)) min-len))
73 (let ((tail tails))
74 (while tail
75 (if (or postfix
76 (>= (length (car tail)) min-len))
77 (setq l (cons (concat (car heads) (car tail)) l)))
78 (setq tail (cdr tail)))))
79 (setq heads (cdr heads)))
80 l))
81
82(defconst skkdic-jisx0208-hiragana-block (nth 1 (split-char ?\e$B$"\e(B)))
83
84(defun skkdic-lookup-key (seq len &optional postfix)
85 "Return a list of conversion string for sequence SEQ of length LEN.
86
87SEQ is a vector of Kana characters to be converted by SKK dictionary.
88If LEN is shorter than the length of KEYSEQ, the first LEN keys in SEQ
89are took into account.
90
91Postfixes are handled only if the optional argument POSTFIX is non-nil."
92 (or skkdic-okuri-nasi
93 (condition-case err
94 (load-library "skk/skkdic")
95 (error (ding)
96 (with-output-to-temp-buffer "*Help*"
97 (princ "The library `skkdic' can't be loaded.
98
99The most common case is that you have not yet installed the library
100included in LEIM (Libraries of Emacs Input Method) which is
101distributed separately from Emacs.
102
103LEIM is available from the same ftp directory as Emacs."))
104 (signal (car err) (cdr err)))))
105
106 (let ((vec (make-vector len 0))
107 (i 0)
108 entry)
109 ;; At first, generate vector VEC from SEQ for looking up SKK
110 ;; alists. Nth element in VEC corresponds to Nth element in SEQ.
111 ;; The values are decided as follows.
112 ;; If SEQ[N] is `\e$B!<\e(B', VEC[N] is 0,
113 ;; Else if SEQ[N] is a Hiragana character, VEC[N] is:
114 ;; ((The 2nd position code o SEQ[N]) - 32),
115 ;; ELse VEC[N] is 128.
116 (while (< i len)
117 (let ((ch (aref seq i))
118 elts)
119 (if (= ch ?\e$B!<\e(B)
120 (aset vec i 0)
121 (setq elts (split-char ch))
122 (if (and (eq (car elts) 'japanese-jisx0208)
123 (= (nth 1 elts) skkdic-jisx0208-hiragana-block))
124 (aset vec i (- (nth 2 elts) 32))
125 (aset vec i 128))))
126 (setq i (1+ i)))
127
128 ;; Search OKURI-NASI entries.
129 (setq entry (lookup-nested-alist vec skkdic-okuri-nasi len 0 t))
130 (if (consp (car entry))
131 (setq entry (copy-sequence (car entry)))
132 (setq entry nil))
133
134 (if postfix
135 ;; Search OKURI-NASI entries with postfixes.
136 (let ((break (max (- len (car skkdic-postfix)) 1))
137 entry-head entry-postfix entry2)
138 (while (< break len)
139 (if (and (setq entry-head
140 (lookup-nested-alist vec skkdic-okuri-nasi
141 break 0 t))
142 (consp (car entry-head))
143 (setq entry-postfix
144 (lookup-nested-alist vec skkdic-postfix
145 len break t))
146 (consp (car entry-postfix))
147 (setq entry2 (skkdic-merge-head-and-tail
148 (car entry-head) (car entry-postfix) t)))
149 (if entry
150 (nconc entry entry2)
151 (setq entry entry2)))
152 (setq break (1+ break)))))
153
154 ;; Search OKURI-NASI entries with prefixes.
155 (let ((break (min (car skkdic-prefix) (- len 2)))
156 entry-prefix entry-tail entry2)
157 (while (> break 0)
158 (if (and (setq entry-prefix
159 (lookup-nested-alist vec skkdic-prefix break 0 t))
160 (consp (car entry-prefix))
161 (setq entry-tail
162 (lookup-nested-alist vec skkdic-okuri-nasi len break t))
163 (consp (car entry-tail))
164 (setq entry2 (skkdic-merge-head-and-tail
165 (car entry-prefix) (car entry-tail) nil)))
166 (if entry
167 (nconc entry entry2)
168 (setq entry entry2)))
169 (setq break (1- break))))
170
171 ;; Search OKURI-ARI entries.
172 (let ((okurigana (assq (aref seq (1- len)) skkdic-okurigana-table))
173 orig-element entry2)
174 (if okurigana
175 (progn
176 (setq orig-element (aref vec (1- len)))
177 (aset vec (1- len) (- (cdr okurigana)))
178 (if (and (setq entry2 (lookup-nested-alist vec skkdic-okuri-ari
179 len 0 t))
180 (consp (car entry2)))
181 (progn
182 (setq entry2 (copy-sequence (car entry2)))
183 (let ((l entry2)
184 (okuri (char-to-string (aref seq (1- len)))))
185 (while l
186 (setcar l (concat (car l) okuri))
187 (setq l (cdr l)))
188 (if entry
189 (nconc entry entry2)
190 (setq entry entry2)))))
191 (aset vec (1- len) orig-element))))
192
193 entry))
194
195;;
196(provide 'skkdic-utl)
197
198;; skkdic-utl.el ends here