1 /**********************************************************************
2 iso8859_5.c - Oniguruma (regular expression library)
3 **********************************************************************/
4 /*-
5 * Copyright (c) 2002-2005 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_ISO_8859_5_TO_LOWER_CASE(c) EncISO_8859_5_ToLowerCaseTable[c]
33 #define ENC_IS_ISO_8859_5_CTYPE(code,ctype) \
34 ((EncISO_8859_5_CtypeTable[code] & ctype) != 0)
35
36 static const UChar EncISO_8859_5_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', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
58 '\370', '\371', '\372', '\373', '\374', '\255', '\376', '\377',
59 '\320', '\321', '\322', '\323', '\324', '\325', '\326', '\327',
60 '\330', '\331', '\332', '\333', '\334', '\335', '\336', '\337',
61 '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
62 '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
63 '\320', '\321', '\322', '\323', '\324', '\325', '\326', '\327',
64 '\330', '\331', '\332', '\333', '\334', '\335', '\336', '\337',
65 '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
66 '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
67 '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
68 '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\377'
69 };
70
71 static const unsigned short EncISO_8859_5_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 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
89 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
90 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
91 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
92 0x0284, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
93 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x01a0, 0x14a2, 0x14a2,
94 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
95 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
96 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
97 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
98 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
99 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
100 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
101 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
102 0x00a0, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
103 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x00a0, 0x10e2, 0x10e2
104 };
105
106 static int
iso_8859_5_mbc_to_normalize(OnigAmbigType flag,const UChar ** pp,const UChar * end,UChar * lower)107 iso_8859_5_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_ISO_8859_5_TO_LOWER_CASE(*p);
117 }
118 else {
119 *lower = *p;
120 }
121
122 (*pp)++;
123 return 1; /* return byte length of converted char to lower */
124 }
125
126 static int
iso_8859_5_is_mbc_ambiguous(OnigAmbigType flag,const UChar ** pp,const UChar * end)127 iso_8859_5_is_mbc_ambiguous(OnigAmbigType flag,
128 const UChar** pp, const UChar* end)
129 {
130 const UChar* p = *pp;
131
132 (*pp)++;
133 if (((flag & ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) != 0 &&
134 ONIGENC_IS_MBC_ASCII(p)) ||
135 ((flag & ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE) != 0 &&
136 !ONIGENC_IS_MBC_ASCII(p))) {
137 int v = (EncISO_8859_5_CtypeTable[*p] &
138 (ONIGENC_CTYPE_UPPER | ONIGENC_CTYPE_LOWER));
139 return (v != 0 ? TRUE : FALSE);
140 }
141 return FALSE;
142 }
143
144 static int
iso_8859_5_is_code_ctype(OnigCodePoint code,unsigned int ctype)145 iso_8859_5_is_code_ctype(OnigCodePoint code, unsigned int ctype)
146 {
147 if (code < 256)
148 return ENC_IS_ISO_8859_5_CTYPE(code, ctype);
149 else
150 return FALSE;
151 }
152
153 static int
iso_8859_5_get_all_pair_ambig_codes(OnigAmbigType flag,const OnigPairAmbigCodes ** ccs)154 iso_8859_5_get_all_pair_ambig_codes(OnigAmbigType flag,
155 const OnigPairAmbigCodes** ccs)
156 {
157 static const OnigPairAmbigCodes cc[] = {
158 { 0xa1, 0xf1 },
159 { 0xa2, 0xf2 },
160 { 0xa3, 0xf3 },
161 { 0xa4, 0xf4 },
162 { 0xa5, 0xf5 },
163 { 0xa6, 0xf6 },
164 { 0xa7, 0xf7 },
165 { 0xa8, 0xf8 },
166 { 0xa9, 0xf9 },
167 { 0xaa, 0xfa },
168 { 0xab, 0xfb },
169 { 0xac, 0xfc },
170 { 0xae, 0xfe },
171 { 0xaf, 0xff },
172
173 { 0xb0, 0xd0 },
174 { 0xb1, 0xd1 },
175 { 0xb2, 0xd2 },
176 { 0xb3, 0xd3 },
177 { 0xb4, 0xd4 },
178 { 0xb5, 0xd5 },
179 { 0xb6, 0xd6 },
180 { 0xb7, 0xd7 },
181 { 0xb8, 0xd8 },
182 { 0xb9, 0xd9 },
183 { 0xba, 0xda },
184 { 0xbb, 0xdb },
185 { 0xbc, 0xdc },
186 { 0xbd, 0xdd },
187 { 0xbe, 0xdf },
188 { 0xbf, 0xdf },
189
190 { 0xc0, 0xe0 },
191 { 0xc1, 0xe1 },
192 { 0xc2, 0xe2 },
193 { 0xc3, 0xe3 },
194 { 0xc4, 0xe4 },
195 { 0xc5, 0xe5 },
196 { 0xc6, 0xe6 },
197 { 0xc7, 0xe7 },
198 { 0xc8, 0xe8 },
199 { 0xc9, 0xe9 },
200 { 0xca, 0xea },
201 { 0xcb, 0xeb },
202 { 0xcc, 0xec },
203 { 0xcd, 0xed },
204 { 0xce, 0xee },
205 { 0xcf, 0xef },
206
207 { 0xd0, 0xb0 },
208 { 0xd1, 0xb1 },
209 { 0xd2, 0xb2 },
210 { 0xd3, 0xb3 },
211 { 0xd4, 0xb4 },
212 { 0xd5, 0xb5 },
213 { 0xd6, 0xb6 },
214 { 0xd7, 0xb7 },
215 { 0xd8, 0xb8 },
216 { 0xd9, 0xb9 },
217 { 0xda, 0xba },
218 { 0xdb, 0xbb },
219 { 0xdc, 0xbc },
220 { 0xdd, 0xbd },
221 { 0xde, 0xbe },
222 { 0xdf, 0xbf },
223
224 { 0xe0, 0xc0 },
225 { 0xe1, 0xc1 },
226 { 0xe2, 0xc2 },
227 { 0xe3, 0xc3 },
228 { 0xe4, 0xc4 },
229 { 0xe5, 0xc5 },
230 { 0xe6, 0xc6 },
231 { 0xe7, 0xc7 },
232 { 0xe8, 0xc8 },
233 { 0xe9, 0xc9 },
234 { 0xea, 0xca },
235 { 0xeb, 0xcb },
236 { 0xec, 0xcc },
237 { 0xed, 0xcd },
238 { 0xee, 0xce },
239 { 0xef, 0xcf },
240
241 { 0xf1, 0xa1 },
242 { 0xf2, 0xa2 },
243 { 0xf3, 0xa3 },
244 { 0xf4, 0xa4 },
245 { 0xf5, 0xa5 },
246 { 0xf6, 0xa6 },
247 { 0xf7, 0xa7 },
248 { 0xf8, 0xa8 },
249 { 0xf9, 0xa9 },
250 { 0xfa, 0xaa },
251 { 0xfb, 0xab },
252 { 0xfc, 0xac },
253 { 0xfe, 0xae },
254 { 0xff, 0xaf }
255 };
256
257 if (flag == ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) {
258 *ccs = OnigAsciiPairAmbigCodes;
259 return 52;
260 }
261 if (flag == ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE) {
262 *ccs = cc;
263 return sizeof(cc) / sizeof(OnigPairAmbigCodes);
264 }
265 else
266 return 0;
267 }
268
269 OnigEncodingType OnigEncodingISO_8859_5 = {
270 onigenc_single_byte_mbc_enc_len,
271 "ISO-8859-5", /* name */
272 1, /* max enc length */
273 1, /* min enc length */
274 (ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE |
275 ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE ),
276 {
277 (OnigCodePoint )'\\' /* esc */
278 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* anychar '.' */
279 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* anytime '*' */
280 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* zero or one time '?' */
281 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* one or more time '+' */
282 , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* anychar anytime */
283 },
284 onigenc_is_mbc_newline_0x0a,
285 onigenc_single_byte_mbc_to_code,
286 onigenc_single_byte_code_to_mbclen,
287 onigenc_single_byte_code_to_mbc,
288 iso_8859_5_mbc_to_normalize,
289 iso_8859_5_is_mbc_ambiguous,
290 iso_8859_5_get_all_pair_ambig_codes,
291 onigenc_nothing_get_all_comp_ambig_codes,
292 iso_8859_5_is_code_ctype,
293 onigenc_not_support_get_ctype_code_range,
294 onigenc_single_byte_left_adjust_char_head,
295 onigenc_always_true_is_allowed_reverse_match
296 };
297
