1 /**********************************************************************
2 koi8_r.c - Oniguruma (regular expression library)
3 **********************************************************************/
4 /*-
5 * Copyright (c) 2002-2006 K.Kosako <sndgk393 AT ybb DOT ne DOT jp>
6 * All rights reserved.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 * SUCH DAMAGE.
28 */
29
30 #include "regenc.h"
31
32 #define ENC_KOI8_R_TO_LOWER_CASE(c) EncKOI8_R_ToLowerCaseTable[c]
33 #define ENC_IS_KOI8_R_CTYPE(code,ctype) \
34 ((EncKOI8_R_CtypeTable[code] & ctype) != 0)
35
36 static const UChar EncKOI8_R_ToLowerCaseTable[256] = {
37 '\000', '\001', '\002', '\003', '\004', '\005', '\006', '\007',
38 '\010', '\011', '\012', '\013', '\014', '\015', '\016', '\017',
39 '\020', '\021', '\022', '\023', '\024', '\025', '\026', '\027',
40 '\030', '\031', '\032', '\033', '\034', '\035', '\036', '\037',
41 '\040', '\041', '\042', '\043', '\044', '\045', '\046', '\047',
42 '\050', '\051', '\052', '\053', '\054', '\055', '\056', '\057',
43 '\060', '\061', '\062', '\063', '\064', '\065', '\066', '\067',
44 '\070', '\071', '\072', '\073', '\074', '\075', '\076', '\077',
45 '\100', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
46 '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
47 '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
48 '\170', '\171', '\172', '\133', '\134', '\135', '\136', '\137',
49 '\140', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
50 '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
51 '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
52 '\170', '\171', '\172', '\173', '\174', '\175', '\176', '\177',
53 '\200', '\201', '\202', '\203', '\204', '\205', '\206', '\207',
54 '\210', '\211', '\212', '\213', '\214', '\215', '\216', '\217',
55 '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227',
56 '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237',
57 '\240', '\241', '\242', '\243', '\244', '\245', '\246', '\247',
58 '\250', '\251', '\252', '\253', '\254', '\255', '\256', '\257',
59 '\260', '\261', '\262', '\243', '\264', '\265', '\266', '\267',
60 '\270', '\271', '\272', '\273', '\274', '\275', '\276', '\277',
61 '\300', '\301', '\302', '\303', '\304', '\305', '\306', '\307',
62 '\310', '\311', '\312', '\313', '\314', '\315', '\316', '\317',
63 '\320', '\321', '\322', '\323', '\324', '\325', '\326', '\327',
64 '\330', '\331', '\332', '\333', '\334', '\335', '\336', '\337',
65 '\300', '\301', '\302', '\303', '\304', '\305', '\306', '\307',
66 '\310', '\311', '\312', '\313', '\314', '\315', '\316', '\317',
67 '\320', '\321', '\322', '\323', '\324', '\325', '\326', '\327',
68 '\330', '\331', '\332', '\333', '\334', '\335', '\336', '\337'
69 };
70
71 static const unsigned short EncKOI8_R_CtypeTable[256] = {
72 0x2008, 0x2008, 0x2008, 0x2008, 0x2008, 0x2008, 0x2008, 0x2008,
73 0x2008, 0x220c, 0x2209, 0x2208, 0x2208, 0x2208, 0x2008, 0x2008,
74 0x2008, 0x2008, 0x2008, 0x2008, 0x2008, 0x2008, 0x2008, 0x2008,
75 0x2008, 0x2008, 0x2008, 0x2008, 0x2008, 0x2008, 0x2008, 0x2008,
76 0x2284, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0,
77 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0,
78 0x38b0, 0x38b0, 0x38b0, 0x38b0, 0x38b0, 0x38b0, 0x38b0, 0x38b0,
79 0x38b0, 0x38b0, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0,
80 0x21a0, 0x3ca2, 0x3ca2, 0x3ca2, 0x3ca2, 0x3ca2, 0x3ca2, 0x34a2,
81 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
82 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
83 0x34a2, 0x34a2, 0x34a2, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x31a0,
84 0x21a0, 0x38e2, 0x38e2, 0x38e2, 0x38e2, 0x38e2, 0x38e2, 0x30e2,
85 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
86 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
87 0x30e2, 0x30e2, 0x30e2, 0x21a0, 0x21a0, 0x21a0, 0x21a0, 0x2008,
88 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
89 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
90 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
91 0x00a0, 0x00a0, 0x0284, 0x00a0, 0x00a0, 0x10a0, 0x01a0, 0x00a0,
92 0x00a0, 0x00a0, 0x00a0, 0x10e2, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
93 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
94 0x00a0, 0x00a0, 0x00a0, 0x14a2, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
95 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
96 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
97 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
98 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
99 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
100 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
101 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
102 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
103 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2
104 };
105
106 static int
koi8_r_mbc_to_normalize(OnigAmbigType flag,const UChar ** pp,const UChar * end,UChar * lower)107 koi8_r_mbc_to_normalize(OnigAmbigType flag,
108 const UChar** pp, const UChar* end, UChar* lower)
109 {
110 const UChar* p = *pp;
111
112 if (((flag & ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) != 0 &&
113 ONIGENC_IS_MBC_ASCII(p)) ||
114 ((flag & ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE) != 0 &&
115 !ONIGENC_IS_MBC_ASCII(p))) {
116 *lower = ENC_KOI8_R_TO_LOWER_CASE(*p);
117 }
118 else {
119 *lower = *p;
120 }
121 (*pp)++;
122 return 1; /* return byte length of converted char to lower */
123 }
124
125 static int
koi8_r_is_mbc_ambiguous(OnigAmbigType flag,const UChar ** pp,const UChar * end)126 koi8_r_is_mbc_ambiguous(OnigAmbigType flag, const UChar** pp, const UChar* end)
127 {
128 const UChar* p = *pp;
129
130 (*pp)++;
131 if (((flag & ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) != 0 &&
132 ONIGENC_IS_MBC_ASCII(p)) ||
133 ((flag & ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE) != 0 &&
134 !ONIGENC_IS_MBC_ASCII(p))) {
135 int v = (EncKOI8_R_CtypeTable[*p] &
136 (ONIGENC_CTYPE_UPPER | ONIGENC_CTYPE_LOWER));
137 return (v != 0 ? TRUE : FALSE);
138 }
139 return FALSE;
140 }
141
142 static int
koi8_r_is_code_ctype(OnigCodePoint code,unsigned int ctype)143 koi8_r_is_code_ctype(OnigCodePoint code, unsigned int ctype)
144 {
145 if (code < 256)
146 return ENC_IS_KOI8_R_CTYPE(code, ctype);
147 else
148 return FALSE;
149 }
150
151 static int
koi8_r_get_all_pair_ambig_codes(OnigAmbigType flag,const OnigPairAmbigCodes ** ccs)152 koi8_r_get_all_pair_ambig_codes(OnigAmbigType flag,
153 const OnigPairAmbigCodes** ccs)
154 {
155 static const OnigPairAmbigCodes cc[] = {
156 { 0xa3, 0xb3 },
157 { 0xb3, 0xa3 },
158
159 { 0xc0, 0xe0 },
160 { 0xc1, 0xe1 },
161 { 0xc2, 0xe2 },
162 { 0xc3, 0xe3 },
163 { 0xc4, 0xe4 },
164 { 0xc5, 0xe5 },
165 { 0xc6, 0xe6 },
166 { 0xc7, 0xe7 },
167 { 0xc8, 0xe8 },
168 { 0xc9, 0xe9 },
169 { 0xca, 0xea },
170 { 0xcb, 0xeb },
171 { 0xcc, 0xec },
172 { 0xcd, 0xed },
173 { 0xce, 0xee },
174 { 0xcf, 0xef },
175
176 { 0xd0, 0xf0 },
177 { 0xd1, 0xf1 },
178 { 0xd2, 0xf2 },
179 { 0xd3, 0xf3 },
180 { 0xd4, 0xf4 },
181 { 0xd5, 0xf5 },
182 { 0xd6, 0xf6 },
183 { 0xd7, 0xf7 },
184 { 0xd8, 0xf8 },
185 { 0xd9, 0xf9 },
186 { 0xda, 0xfa },
187 { 0xdb, 0xfb },
188 { 0xdc, 0xfc },
189 { 0xdd, 0xfd },
190 { 0xde, 0xfe },
191 { 0xdf, 0xff },
192
193 { 0xe0, 0xc0 },
194 { 0xe1, 0xc1 },
195 { 0xe2, 0xc2 },
196 { 0xe3, 0xc3 },
197 { 0xe4, 0xc4 },
198 { 0xe5, 0xc5 },
199 { 0xe6, 0xc6 },
200 { 0xe7, 0xc7 },
201 { 0xe8, 0xc8 },
202 { 0xe9, 0xc9 },
203 { 0xea, 0xca },
204 { 0xeb, 0xcb },
205 { 0xec, 0xcc },
206 { 0xed, 0xcd },
207 { 0xee, 0xce },
208 { 0xef, 0xcf },
209
210 { 0xf0, 0xd0 },
211 { 0xf1, 0xd1 },
212 { 0xf2, 0xd2 },
213 { 0xf3, 0xd3 },
214 { 0xf4, 0xd4 },
215 { 0xf5, 0xd5 },
216 { 0xf6, 0xd6 },
217 { 0xf7, 0xd7 },
218 { 0xf8, 0xd8 },
219 { 0xf9, 0xd9 },
220 { 0xfa, 0xda },
221 { 0xfb, 0xdb },
222 { 0xfc, 0xdc },
223 { 0xfe, 0xde },
224 { 0xff, 0xdf }
225 };
226
227 if (flag == ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) {
228 *ccs = OnigAsciiPairAmbigCodes;
229 return 52;
230 }
231 if (flag == ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE) {
232 *ccs = cc;
233 return sizeof(cc) / sizeof(OnigPairAmbigCodes);
234 }
235 else
236 return 0;
237 }
238
239 OnigEncodingType OnigEncodingKOI8_R = {
240 onigenc_single_byte_mbc_enc_len,
241 "KOI8-R", /* name */
242 1, /* max enc length */
243 1, /* min enc length */
244 (ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE |
245 ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE ),
246 {
247 (OnigCodePoint )'\\' /* esc */
248 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* anychar '.' */
249 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* anytime '*' */
250 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* zero or one time '?' */
251 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* one or more time '+' */
252 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* anychar anytime */
253 },
254 onigenc_is_mbc_newline_0x0a,
255 onigenc_single_byte_mbc_to_code,
256 onigenc_single_byte_code_to_mbclen,
257 onigenc_single_byte_code_to_mbc,
258 koi8_r_mbc_to_normalize,
259 koi8_r_is_mbc_ambiguous,
260 koi8_r_get_all_pair_ambig_codes,
261 onigenc_nothing_get_all_comp_ambig_codes,
262 koi8_r_is_code_ctype,
263 onigenc_not_support_get_ctype_code_range,
264 onigenc_single_byte_left_adjust_char_head,
265 onigenc_always_true_is_allowed_reverse_match
266 };
267
