1#!/usr/local/bin/perl -w
2#
3# Copyright (c) 2002 Victor Ivanov <v0rbiz@yahoo.com>
4# All rights reserved.
5#
6# Redistribution and use in source and binary forms, with or without
7# modification, are permitted provided that the following conditions
8# are met:
9# 1. Redistributions of source code must retain the above copyright
10#    notice, this list of conditions and the following disclaimer.
11# 2. Redistributions in binary form must reproduce the above copyright
12#    notice, this list of conditions and the following disclaimer in the
13#    documentation and/or other materials provided with the distribution.
14#
15# THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
16# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18# ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
19# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
25# SUCH DAMAGE.
26#
27
28use strict;
29use vars qw($VERSION %IRSSI);
30
31$VERSION = '1.0b3';
32%IRSSI = (
33    authors	=> 'Victor Ivanov',
34    contact	=> 'v0rbiz@yahoo.com',
35    name	=> 'romaji',
36    description => 'translates romaji to hiragana or katakana in text enclosed in ^R',
37    license	=> 'BSD 2-clause',
38    url		=> 'http://irssi.org/scripts/'
39);
40
41
42my(%hira) = (
43    "a"   => "あ", "i"   => "い", "u"   => "う", "e"   => "え", "o"   => "お",
44    "ka"  => "か", "ki"  => "き", "ku"  => "く", "ke"  => "け", "ko"  => "こ",
45    "sa"  => "さ", "shi" => "し", "su"  => "す", "se"  => "せ", "so"  => "そ",
46    "ta"  => "た", "chi" => "ち", "tsu" => "つ", "te"  => "て", "to"  => "と",
47    "na"  => "な", "ni"  => "に", "nu"  => "ぬ", "ne"  => "ね", "no"  => "の",
48    "ha"  => "は", "hi"  => "ひ", "hu"  => "ふ", "he"  => "へ", "ho"  => "ほ", "fu"  => "ふ",
49    "ma"  => "ま", "mi"  => "み", "mu"  => "む", "me"  => "め", "mo"  => "も",
50    "ya"  => "や", "yu"  => "ゆ", "yo"  => "よ",
51    "ra"  => "ら", "ri"  => "り", "ru"  => "る", "re"  => "れ", "ro"  => "ろ",
52    "wa"  => "わ", "wi"  => "ゐ", "we"  => "ゑ", "wo"  => "を",
53    "n"   => "ん",
54    "m"   => "ん",
55
56    "ga"  => "が", "gi"  => "ぎ", "gu"  => "ぐ", "ge"  => "げ", "go"  => "ご",
57    "za"  => "ざ", "ji"  => "じ", "zu"  => "ず", "ze"  => "ぜ", "zo"  => "ぞ",
58    "da"  => "だ", "dzi" => "ぢ", "dzu" => "づ", "de"  => "で", "do"  => "ど",
59    "ba"  => "ば", "bi"  => "び", "bu"  => "ぶ", "be"  => "べ", "bo"  => "ぼ",
60    "pa"  => "ぱ", "pi"  => "ぴ", "pu"  => "ぷ", "pe"  => "ぺ", "po"  => "ぽ",
61
62    "fa"  => "ふぁ", "fi"  => "ふぃ", "fe"  => "ふぇ", "fo"  => "ふぉ",
63    "di"  => "でぃ",
64
65    "kya" => "きゃ", "kyu" => "きゅ", "kyo" => "きょ",
66    "sha" => "しゃ", "shu" => "しゅ", "sho" => "しょ",
67    "cha" => "ちゃ", "chu" => "ちゅ", "cho" => "ちょ",
68    "nya" => "にゃ", "nyu" => "にゅ", "nyo" => "にょ",
69    "hya" => "ひゃ", "hyu" => "ひゅ", "hyo" => "ひょ",
70    "mya" => "みゃ", "myu" => "みゅ", "myo" => "みょ",
71    "rya" => "りゃ", "ryu" => "りゅ", "ryo" => "りょ",
72    "gya" => "ぎゃ", "gyu" => "ぎゅ", "gyo" => "ぎょ",
73    "ja"  => "じゃ", "ju"  => "じゅ", "jo"  => "じょ",
74    "jya" => "じゃ", "jyu" => "じゅ", "jyo" => "じょ",
75    "dza" => "ぢゃ", "dju" => "ぢゅ", "dzo" => "ぢょ",
76    "dja" => "ぢゃ",                  "djo" => "ぢょ",
77    "bya" => "びゃ", "byu" => "びゅ", "byo" => "びょ",
78    "pya" => "ぴゃ", "pyu" => "ぴゅ", "pyo" => "ぴょ",
79
80    "TSU" => "っ"
81);
82
83my(%kata) = (
84    "a"   => "ア", "i"   => "イ", "u"   => "ウ", "e"   => "エ", "o"   => "オ",
85    "ka"  => "カ", "ki"  => "キ", "ku"  => "ク", "ke"  => "ケ", "ko"  => "コ",
86    "sa"  => "サ", "shi" => "シ", "su"  => "ス", "se"  => "セ", "so"  => "ソ",
87    "ta"  => "タ", "chi" => "チ", "tsu" => "ツ", "te"  => "テ", "to"  => "ト",
88    "na"  => "ナ", "ni"  => "ニ", "nu"  => "ヌ", "ne"  => "ネ", "no"  => "ノ",
89    "ha"  => "ハ", "hi"  => "ヒ", "hu"  => "フ", "he"  => "ヘ", "ho"  => "ホ", "fu"  => "フ",
90    "ma"  => "マ", "mi"  => "ミ", "mu"  => "ム", "me"  => "メ", "mo"  => "モ",
91    "ya"  => "ヤ", "yu"  => "ユ", "yo"  => "ヨ", "ye"  => "エ",
92    "ra"  => "ラ", "ri"  => "リ", "ru"  => "ル", "re"  => "レ", "ro"  => "ロ",
93    "wa"  => "ワ", "wi"  => "ヰ", "we"  => "ヱ", "wo"  => "ヲ",
94    "n"   => "ン",
95    "m"   => "ン",
96
97    "ga"  => "ガ", "gi"  => "ギ", "gu"  => "グ", "ge"  => "ゲ", "go"  => "ゴ",
98    "za"  => "ザ", "ji"  => "ジ", "zu"  => "ズ", "ze"  => "ゼ", "zo"  => "ゾ",
99    "da"  => "ダ", "dzi" => "ヂ", "dzu" => "ヅ", "de"  => "デ", "do"  => "ド",
100    "ba"  => "バ", "bi"  => "ビ", "bu"  => "ブ", "be"  => "ベ", "bo"  => "ボ",
101    "pa"  => "パ", "pi"  => "ピ", "pu"  => "プ", "pe"  => "ペ", "po"  => "ポ",
102
103    "va"  => "ヴァ", "vi"  => "ヴィ", "vu"  => "ヴ",   "ve"  => "ヴェ", "vo"  => "ヴォ",
104    "fa"  => "ファ", "fi"  => "フィ", "fe"  => "フェ", "fo"  => "フォ",
105    "di"  => "ディ",
106
107    "dje" => "ヂェ", "dze" => "ヂェ",
108
109    "kya" => "キャ", "kyu" => "キュ", "kyo" => "キョ",
110    "sha" => "シャ", "shu" => "シュ", "sho" => "ショ",
111    "cha" => "チャ", "chu" => "チュ", "cho" => "チョ",
112    "nya" => "ニャ", "nyu" => "ニュ", "nyo" => "ニョ",
113    "hya" => "ヒャ", "hyu" => "ヒュ", "hyo" => "ヒョ",
114    "mya" => "ミャ", "myu" => "ミュ", "myo" => "ミョ",
115    "rya" => "リャ", "ryu" => "リュ", "ryo" => "リョ",
116    "gya" => "ギャ", "gyu" => "ギュ", "gyo" => "ギョ",
117    "ja"  => "ジャ", "ju"  => "ジュ", "jo"  => "ジョ",
118    "jya" => "ジャ", "jyu" => "ジュ", "jyo" => "ジョ",
119    "dza" => "ヂャ", "dju" => "ヂュ", "dzo" => "ヂョ",
120    "dja" => "ヂャ",                  "djo" => "ヂョ",
121    "bya" => "ビャ", "byu" => "ビュ", "byo" => "ビョ",
122    "pya" => "ピャ", "pyu" => "ピュ", "pyo" => "ピョ",
123
124    "TSU" => "ッ"
125);
126
127my(%comn) = (
128    "-"   => "ー",
129    "."   => "。",
130    ","   => "、",
131    "!"   => "！",
132    "?"   => "？",
133    "~"   => "〜",
134    "  "  => "　",
135    "["   => "〔", "]"   => "〕",
136    "{"   => "【", "}"   => "】",
137    "("   => "（", ")"   => "）",
138    "0"   => "０", "1"   => "１", "2"   => "２", "3"   => "３", "4"   => "４",
139    "5"   => "５", "6"   => "６", "7"   => "７", "8"   => "８", "9"   => "９",
140    "*"   => "★", # ☆ is uglier :P
141    # where to put ♪ ?
142);
143
144my(@squot) = ( "「", "」" );
145my($squoti) = 0;
146my(@dquot) = ( "『", "』" );
147my($dquoti) = 0;
148
149sub r2hk ($$) {
150    my($str) = "";
151    my($pos) = 0;
152    my($inlen) = length($_[0]);
153    my($last) = "";
154    my($href) = $_[1];
155    my($inp) = lc($_[0]);
156
157    while ($pos < $inlen) {
158	my($len);
159	my($p) = substr($inp, $pos, 3);
160	my($h) = ${$href}{$p};
161
162	# this could be done with another cycle, but this way's faster i guess
163	if ($h) {
164	    $len = 3;
165	} else {
166	    $p = substr($inp, $pos, 2);
167	    $h = ${$href}{$p};
168	    if ($h) {
169		$len = 2;
170	    } else {
171		$p = substr($inp, $pos, 1);
172		$h = ${$href}{$p};
173		if (!$h) {
174		    if ($p eq "'") {
175			$h = $squot[$squoti];
176			$squoti = 1 - $squoti;
177		    } elsif ($p eq "\"") {
178			$h = $dquot[$dquoti];
179			$dquoti = 1 - $dquoti;
180		    } else {
181			$h = $p;
182		    }
183		}
184		$len = 1;
185	    }
186	}
187
188	if ($h ne $p) {
189	    if ($last) {
190		if ($last eq substr($p, 0, 1)) {
191		    $str .= ${$href}{"TSU"};
192		} else {
193		    $str .= $last;
194		}
195		$last = "";
196	    }
197	} else {
198	    $str .= $last;
199	    $last = $p;
200	    $h = "";
201	}
202
203	$str .= $h;
204
205	$pos += $len;
206    }
207
208    $str .= $last;
209
210    return $str;
211}
212
213my($lock_ev) = 0;
214
215sub event1 {
216    my ($line, $server, $witem) = @_;
217
218    return unless ref $witem;
219    if ($lock_ev) { return };
220    $squoti = 0;
221    $dquoti = 0;
222
223    my ($str) = "";
224    my (@p) = split(//, $line);
225    my ($i);
226    my ($inside) = 0;
227    my ($empty) = 0;
228
229    for ($i = 0; $i <= $#p; $i++) {
230	if ($inside) {
231	    if (!$p[$i]) {
232		$empty++;
233	    } else {
234		if ($empty == 0) {
235		    $str .= r2hk($p[$i], \%hira);
236		} else {
237		    $str .= r2hk($p[$i], \%kata);
238		}
239		$empty = 0;
240		$inside = 0;
241	    }
242	} else {
243	    $str .= $p[$i];
244	    $inside = 1;
245	}
246    }
247
248    $lock_ev = 1;
249    Irssi::signal_emit('send command', $str, $server, $witem);
250    Irssi::signal_stop();
251    $lock_ev = 0;
252}
253
254sub cmd_romaji {
255    Irssi::print('%BRomaji (with ひらがな and カタカナ support) version '.$VERSION);
256    Irssi::print('(this is amateur product and comes with %Wno warranty%n, see the source)');
257    Irssi::print('Text enclosed in Ctrl-Rs (like this) will be converted to hiragana.');
258    Irssi::print('If the opening ^R is doubled, it will be converted to katakana.');
259    Irssi::print('Example: genki -> げんき and genki -> ゲンキ');
260}
261
262Irssi::signal_add('send command', "event1");
263Irssi::command_bind('romaji', \&cmd_romaji);
264
265Irssi::print('%B'.$IRSSI{name}.' '.$VERSION.'%n loaded; type /romaji for more info');
266
267# Add the common hash to hiragana and kitakana hashes
268my($k, $v);
269
270while (($k, $v) = each %comn) {
271    $hira{$k} = $v;
272    $kata{$k} = $v;
273}
274