1 /**********************************************************************
2 iso8859_1.c - Oniguruma (regular expression library)
3 **********************************************************************/
4 /*-
5 * Copyright (c) 2002-2020 K.Kosako
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 LARGE_S 0x53
33 #define SMALL_S 0x73
34
35 #define ENC_IS_ISO_8859_1_CTYPE(code,ctype) \
36 ((EncISO_8859_1_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
37
38 static const unsigned short EncISO_8859_1_CtypeTable[256] = {
39 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
40 0x4008, 0x420c, 0x4209, 0x4208, 0x4208, 0x4208, 0x4008, 0x4008,
41 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
42 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
43 0x4284, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
44 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
45 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0,
46 0x78b0, 0x78b0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
47 0x41a0, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x74a2,
48 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
49 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
50 0x74a2, 0x74a2, 0x74a2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x51a0,
51 0x41a0, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x70e2,
52 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
53 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
54 0x70e2, 0x70e2, 0x70e2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x4008,
55 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
56 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
57 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
58 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
59 0x0284, 0x01a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
60 0x00a0, 0x00a0, 0x30e2, 0x01a0, 0x00a0, 0x01a0, 0x00a0, 0x00a0,
61 0x00a0, 0x00a0, 0x10a0, 0x10a0, 0x00a0, 0x30e2, 0x00a0, 0x01a0,
62 0x00a0, 0x10a0, 0x30e2, 0x01a0, 0x10a0, 0x10a0, 0x10a0, 0x01a0,
63 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
64 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
65 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x00a0,
66 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x30e2,
67 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
68 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
69 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x00a0,
70 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2
71 };
72
73 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
74 { 0xc0, 0xe0 },
75 { 0xc1, 0xe1 },
76 { 0xc2, 0xe2 },
77 { 0xc3, 0xe3 },
78 { 0xc4, 0xe4 },
79 { 0xc5, 0xe5 },
80 { 0xc6, 0xe6 },
81 { 0xc7, 0xe7 },
82 { 0xc8, 0xe8 },
83 { 0xc9, 0xe9 },
84 { 0xca, 0xea },
85 { 0xcb, 0xeb },
86 { 0xcc, 0xec },
87 { 0xcd, 0xed },
88 { 0xce, 0xee },
89 { 0xcf, 0xef },
90
91 { 0xd0, 0xf0 },
92 { 0xd1, 0xf1 },
93 { 0xd2, 0xf2 },
94 { 0xd3, 0xf3 },
95 { 0xd4, 0xf4 },
96 { 0xd5, 0xf5 },
97 { 0xd6, 0xf6 },
98 { 0xd8, 0xf8 },
99 { 0xd9, 0xf9 },
100 { 0xda, 0xfa },
101 { 0xdb, 0xfb },
102 { 0xdc, 0xfc },
103 { 0xdd, 0xfd },
104 { 0xde, 0xfe }
105 };
106
107 static int
apply_all_case_fold(OnigCaseFoldType flag,OnigApplyAllCaseFoldFunc f,void * arg)108 apply_all_case_fold(OnigCaseFoldType flag,
109 OnigApplyAllCaseFoldFunc f, void* arg)
110 {
111 return onigenc_apply_all_case_fold_with_map(
112 sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 1,
113 flag, f, arg);
114 }
115
116 static int
get_case_fold_codes_by_str(OnigCaseFoldType flag,const OnigUChar * p,const OnigUChar * end,OnigCaseFoldCodeItem items[])117 get_case_fold_codes_by_str(OnigCaseFoldType flag,
118 const OnigUChar* p, const OnigUChar* end,
119 OnigCaseFoldCodeItem items[])
120 {
121 static OnigUChar sa[] = { LARGE_S, SMALL_S };
122 int i, j, n;
123
124 if (0x41 <= *p && *p <= 0x5a) {
125 if (*p == LARGE_S && end > p + 1
126 && (*(p+1) == LARGE_S || *(p+1) == SMALL_S)
127 && CASE_FOLD_IS_NOT_ASCII_ONLY(flag)) { /* SS */
128 ss_combination:
129 items[0].byte_len = 2;
130 items[0].code_len = 1;
131 items[0].code[0] = (OnigCodePoint )0xdf;
132
133 n = 1;
134 for (i = 0; i < 2; i++) {
135 for (j = 0; j < 2; j++) {
136 if (sa[i] == *p && sa[j] == *(p+1))
137 continue;
138
139 items[n].byte_len = 2;
140 items[n].code_len = 2;
141 items[n].code[0] = (OnigCodePoint )sa[i];
142 items[n].code[1] = (OnigCodePoint )sa[j];
143 n++;
144 }
145 }
146 return 4;
147 }
148
149 items[0].byte_len = 1;
150 items[0].code_len = 1;
151 items[0].code[0] = (OnigCodePoint )(*p + 0x20);
152 return 1;
153 }
154 else if (0x61 <= *p && *p <= 0x7a) {
155 if (*p == SMALL_S && end > p + 1
156 && (*(p+1) == SMALL_S || *(p+1) == LARGE_S)
157 && CASE_FOLD_IS_NOT_ASCII_ONLY(flag)) { /* ss */
158 goto ss_combination;
159 }
160
161 items[0].byte_len = 1;
162 items[0].code_len = 1;
163 items[0].code[0] = (OnigCodePoint )(*p - 0x20);
164 return 1;
165 }
166 else if (CASE_FOLD_IS_NOT_ASCII_ONLY(flag)) {
167 if (0xc0 <= *p && *p <= 0xcf) {
168 items[0].byte_len = 1;
169 items[0].code_len = 1;
170 items[0].code[0] = (OnigCodePoint )(*p + 0x20);
171 return 1;
172 }
173 else if (0xd0 <= *p && *p <= 0xdf) {
174 if (*p == 0xdf) {
175 items[0].byte_len = 1;
176 items[0].code_len = 2;
177 items[0].code[0] = (OnigCodePoint )'s';
178 items[0].code[1] = (OnigCodePoint )'s';
179
180 items[1].byte_len = 1;
181 items[1].code_len = 2;
182 items[1].code[0] = (OnigCodePoint )'S';
183 items[1].code[1] = (OnigCodePoint )'S';
184
185 items[2].byte_len = 1;
186 items[2].code_len = 2;
187 items[2].code[0] = (OnigCodePoint )'s';
188 items[2].code[1] = (OnigCodePoint )'S';
189
190 items[3].byte_len = 1;
191 items[3].code_len = 2;
192 items[3].code[0] = (OnigCodePoint )'S';
193 items[3].code[1] = (OnigCodePoint )'s';
194
195 return 4;
196 }
197 else if (*p != 0xd7) {
198 items[0].byte_len = 1;
199 items[0].code_len = 1;
200 items[0].code[0] = (OnigCodePoint )(*p + 0x20);
201 return 1;
202 }
203 }
204 else if (0xe0 <= *p && *p <= 0xef) {
205 items[0].byte_len = 1;
206 items[0].code_len = 1;
207 items[0].code[0] = (OnigCodePoint )(*p - 0x20);
208 return 1;
209 }
210 else if (0xf0 <= *p && *p <= 0xfe) {
211 if (*p != 0xf7) {
212 items[0].byte_len = 1;
213 items[0].code_len = 1;
214 items[0].code[0] = (OnigCodePoint )(*p - 0x20);
215 return 1;
216 }
217 }
218 }
219
220 return 0;
221 }
222
223 static int
mbc_case_fold(OnigCaseFoldType flag,const UChar ** pp,const UChar * end ARG_UNUSED,UChar * lower)224 mbc_case_fold(OnigCaseFoldType flag, const UChar** pp,
225 const UChar* end ARG_UNUSED, UChar* lower)
226 {
227 const UChar* p = *pp;
228
229 if (*p == 0xdf && (flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) != 0) {
230 *lower++ = 's';
231 *lower = 's';
232 (*pp)++;
233 return 2;
234 }
235
236 if (CASE_FOLD_IS_NOT_ASCII_ONLY(flag) || ONIGENC_IS_ASCII_CODE(*p))
237 *lower = ONIGENC_ISO_8859_1_TO_LOWER_CASE(*p);
238 else
239 *lower = *p;
240
241 (*pp)++;
242 return 1;
243 }
244
245 static int
is_code_ctype(OnigCodePoint code,unsigned int ctype)246 is_code_ctype(OnigCodePoint code, unsigned int ctype)
247 {
248 if (code < 256)
249 return ENC_IS_ISO_8859_1_CTYPE(code, ctype);
250 else
251 return FALSE;
252 }
253
254 OnigEncodingType OnigEncodingISO_8859_1 = {
255 onigenc_single_byte_mbc_enc_len,
256 "ISO-8859-1", /* name */
257 1, /* max enc length */
258 1, /* min enc length */
259 onigenc_is_mbc_newline_0x0a,
260 onigenc_single_byte_mbc_to_code,
261 onigenc_single_byte_code_to_mbclen,
262 onigenc_single_byte_code_to_mbc,
263 mbc_case_fold,
264 apply_all_case_fold,
265 get_case_fold_codes_by_str,
266 onigenc_minimum_property_name_to_ctype,
267 is_code_ctype,
268 onigenc_not_support_get_ctype_code_range,
269 onigenc_single_byte_left_adjust_char_head,
270 onigenc_always_true_is_allowed_reverse_match,
271 NULL, /* init */
272 NULL, /* is_initialized */
273 onigenc_always_true_is_valid_mbc_string,
274 ENC_FLAG_ASCII_COMPATIBLE|ENC_FLAG_SKIP_OFFSET_1,
275 0, 0
276 };
277