1 : /**********************************************************************
2 : iso8859_10.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_10_TO_LOWER_CASE(c) EncISO_8859_10_ToLowerCaseTable[c]
33 : #define ENC_IS_ISO_8859_10_CTYPE(code,ctype) \
34 : ((EncISO_8859_10_CtypeTable[code] & ctype) != 0)
35 :
36 : static const UChar EncISO_8859_10_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', '\261', '\262', '\263', '\264', '\265', '\266', '\247',
58 : '\270', '\271', '\272', '\273', '\274', '\255', '\276', '\277',
59 : '\260', '\261', '\262', '\263', '\264', '\265', '\266', '\267',
60 : '\270', '\271', '\272', '\273', '\274', '\275', '\276', '\277',
61 : '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
62 : '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
63 : '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
64 : '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\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_10_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, 0x20a0, 0x21a0, 0x21a0, 0x21a0,
77 : 0x21a0, 0x21a0, 0x21a0, 0x20a0, 0x21a0, 0x21a0, 0x21a0, 0x21a0,
78 : 0x38b0, 0x38b0, 0x38b0, 0x38b0, 0x38b0, 0x38b0, 0x38b0, 0x38b0,
79 : 0x38b0, 0x38b0, 0x21a0, 0x21a0, 0x20a0, 0x20a0, 0x20a0, 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, 0x20a0, 0x31a0,
84 : 0x20a0, 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, 0x20a0, 0x21a0, 0x20a0, 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, 0x00a0,
93 : 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x01a0, 0x14a2, 0x14a2,
94 : 0x00a0, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x01a0,
95 : 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x01a0, 0x10e2, 0x10e2,
96 : 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
97 : 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
98 : 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2,
99 : 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x14a2, 0x10e2,
100 : 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
101 : 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
102 : 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2,
103 : 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2, 0x10e2
104 : };
105 :
106 : static int
107 : iso_8859_10_mbc_to_normalize(OnigAmbigType flag,
108 : const UChar** pp, const UChar* end, UChar* lower)
109 0 : {
110 0 : const UChar* p = *pp;
111 :
112 0 : if (end > p + 1 && (flag & ONIGENC_AMBIGUOUS_MATCH_COMPOUND) != 0) {
113 0 : if ((*p == 's' && *(p+1) == 's') ||
114 : ((flag & ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) != 0 &&
115 : (*p == 'S' && *(p+1) == 'S'))) {
116 0 : *lower = 0xdf;
117 0 : (*pp) += 2;
118 0 : return 1;
119 : }
120 : }
121 :
122 0 : if (((flag & ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) != 0 &&
123 : ONIGENC_IS_MBC_ASCII(p)) ||
124 : ((flag & ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE) != 0 &&
125 : !ONIGENC_IS_MBC_ASCII(p))) {
126 0 : *lower = ENC_ISO_8859_10_TO_LOWER_CASE(*p);
127 : }
128 : else {
129 0 : *lower = *p;
130 : }
131 0 : (*pp)++;
132 0 : return 1; /* return byte length of converted char to lower */
133 : }
134 :
135 : static int
136 : iso_8859_10_is_mbc_ambiguous(OnigAmbigType flag,
137 : const UChar** pp, const UChar* end)
138 0 : {
139 0 : const UChar* p = *pp;
140 :
141 0 : if ((flag & ONIGENC_AMBIGUOUS_MATCH_COMPOUND) != 0) {
142 0 : if (end > p + 1) {
143 0 : if ((*p == 's' && *(p+1) == 's') ||
144 : ((flag & ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) != 0 &&
145 : (*p == 'S' && *(p+1) == 'S'))) {
146 0 : (*pp) += 2;
147 0 : return TRUE;
148 : }
149 : }
150 :
151 0 : if (*p == 0xdf) {
152 0 : (*pp)++;
153 0 : return TRUE;
154 : }
155 : }
156 :
157 0 : (*pp)++;
158 0 : if (((flag & ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) != 0 &&
159 : ONIGENC_IS_MBC_ASCII(p)) ||
160 : ((flag & ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE) != 0 &&
161 : !ONIGENC_IS_MBC_ASCII(p))) {
162 : int v = (EncISO_8859_10_CtypeTable[*p] &
163 0 : (ONIGENC_CTYPE_UPPER | ONIGENC_CTYPE_LOWER));
164 :
165 : if ((v | ONIGENC_CTYPE_LOWER) != 0) {
166 : /* 0xdf is lower case letter, but can't convert. */
167 0 : if (*p == 0xdf)
168 0 : return FALSE;
169 : else
170 0 : return TRUE;
171 : }
172 :
173 : return (v != 0 ? TRUE : FALSE);
174 : }
175 0 : return FALSE;
176 : }
177 :
178 : static int
179 : iso_8859_10_is_code_ctype(OnigCodePoint code, unsigned int ctype)
180 0 : {
181 0 : if (code < 256)
182 0 : return ENC_IS_ISO_8859_10_CTYPE(code, ctype);
183 : else
184 0 : return FALSE;
185 : }
186 :
187 : static int
188 : iso_8859_10_get_all_pair_ambig_codes(OnigAmbigType flag,
189 : const OnigPairAmbigCodes** ccs)
190 0 : {
191 : static const OnigPairAmbigCodes cc[] = {
192 : { 0xa1, 0xb1 },
193 : { 0xa2, 0xb2 },
194 : { 0xa3, 0xb3 },
195 : { 0xa4, 0xb4 },
196 : { 0xa5, 0xb5 },
197 : { 0xa6, 0xb6 },
198 : { 0xa8, 0xb8 },
199 : { 0xa9, 0xb9 },
200 : { 0xaa, 0xba },
201 : { 0xab, 0xbb },
202 : { 0xac, 0xbc },
203 : { 0xae, 0xbe },
204 : { 0xaf, 0xbf },
205 :
206 : { 0xb1, 0xa1 },
207 : { 0xb2, 0xa2 },
208 : { 0xb3, 0xa3 },
209 : { 0xb4, 0xa4 },
210 : { 0xb5, 0xa5 },
211 : { 0xb6, 0xa6 },
212 : { 0xb8, 0xa8 },
213 : { 0xb9, 0xa9 },
214 : { 0xba, 0xaa },
215 : { 0xbb, 0xab },
216 : { 0xbc, 0xac },
217 : { 0xbe, 0xae },
218 : { 0xbf, 0xaf },
219 :
220 : { 0xc0, 0xe0 },
221 : { 0xc1, 0xe1 },
222 : { 0xc2, 0xe2 },
223 : { 0xc3, 0xe3 },
224 : { 0xc4, 0xe4 },
225 : { 0xc5, 0xe5 },
226 : { 0xc6, 0xe6 },
227 : { 0xc7, 0xe7 },
228 : { 0xc8, 0xe8 },
229 : { 0xc9, 0xe9 },
230 : { 0xca, 0xea },
231 : { 0xcb, 0xeb },
232 : { 0xcc, 0xec },
233 : { 0xcd, 0xed },
234 : { 0xce, 0xee },
235 : { 0xcf, 0xef },
236 :
237 : { 0xd0, 0xf0 },
238 : { 0xd1, 0xf1 },
239 : { 0xd2, 0xf2 },
240 : { 0xd3, 0xf3 },
241 : { 0xd4, 0xf4 },
242 : { 0xd5, 0xf5 },
243 : { 0xd6, 0xf6 },
244 : { 0xd7, 0xf7 },
245 : { 0xd8, 0xf8 },
246 : { 0xd9, 0xf9 },
247 : { 0xda, 0xfa },
248 : { 0xdb, 0xfb },
249 : { 0xdc, 0xfc },
250 : { 0xdd, 0xfd },
251 : { 0xde, 0xfe },
252 :
253 : { 0xe0, 0xc0 },
254 : { 0xe1, 0xc1 },
255 : { 0xe2, 0xc2 },
256 : { 0xe3, 0xc3 },
257 : { 0xe4, 0xc4 },
258 : { 0xe5, 0xc5 },
259 : { 0xe6, 0xc6 },
260 : { 0xe7, 0xc7 },
261 : { 0xe8, 0xc8 },
262 : { 0xe9, 0xc9 },
263 : { 0xea, 0xca },
264 : { 0xeb, 0xcb },
265 : { 0xec, 0xcc },
266 : { 0xed, 0xcd },
267 : { 0xee, 0xce },
268 : { 0xef, 0xcf },
269 :
270 : { 0xf0, 0xd0 },
271 : { 0xf1, 0xd1 },
272 : { 0xf2, 0xd2 },
273 : { 0xf3, 0xd3 },
274 : { 0xf4, 0xd4 },
275 : { 0xf5, 0xd5 },
276 : { 0xf6, 0xd6 },
277 : { 0xf7, 0xd7 },
278 : { 0xf8, 0xd8 },
279 : { 0xf9, 0xd9 },
280 : { 0xfa, 0xda },
281 : { 0xfb, 0xdb },
282 : { 0xfc, 0xdc },
283 : { 0xfd, 0xdd },
284 : { 0xfe, 0xde }
285 : };
286 :
287 0 : if (flag == ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE) {
288 0 : *ccs = OnigAsciiPairAmbigCodes;
289 0 : return 52;
290 : }
291 0 : if (flag == ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE) {
292 0 : *ccs = cc;
293 0 : return sizeof(cc) / sizeof(OnigPairAmbigCodes);
294 : }
295 : else
296 0 : return 0;
297 : }
298 :
299 : OnigEncodingType OnigEncodingISO_8859_10 = {
300 : onigenc_single_byte_mbc_enc_len,
301 : "ISO-8859-10", /* name */
302 : 1, /* max enc length */
303 : 1, /* min enc length */
304 : (ONIGENC_AMBIGUOUS_MATCH_ASCII_CASE |
305 : ONIGENC_AMBIGUOUS_MATCH_NONASCII_CASE |
306 : ONIGENC_AMBIGUOUS_MATCH_COMPOUND),
307 : {
308 : (OnigCodePoint )'\\' /* esc */
309 : , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* anychar '.' */
310 : , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* anytime '*' */
311 : , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* zero or one time '?' */
312 : , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* one or more time '+' */
313 : , (OnigCodePoint )ONIG_INEFFECTIVE_META_CHAR /* anychar anytime */
314 : },
315 : onigenc_is_mbc_newline_0x0a,
316 : onigenc_single_byte_mbc_to_code,
317 : onigenc_single_byte_code_to_mbclen,
318 : onigenc_single_byte_code_to_mbc,
319 : iso_8859_10_mbc_to_normalize,
320 : iso_8859_10_is_mbc_ambiguous,
321 : iso_8859_10_get_all_pair_ambig_codes,
322 : onigenc_ess_tsett_get_all_comp_ambig_codes,
323 : iso_8859_10_is_code_ctype,
324 : onigenc_not_support_get_ctype_code_range,
325 : onigenc_single_byte_left_adjust_char_head,
326 : onigenc_always_true_is_allowed_reverse_match
327 : };
|
