1 /**********************************************************************
2 regposix.c - Oniguruma (regular expression library)
3 **********************************************************************/
4 /*-
5 * Copyright (c) 2002-2019 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 #define regex_t onig_regex_t
31 #include "regint.h"
32 #undef regex_t
33
34 #include "onigposix.h"
35
36 #define ONIG_C(reg) ((onig_regex_t* )((reg)->onig))
37 #define PONIG_C(reg) ((onig_regex_t** )(&(reg)->onig))
38
39 /* #define ENC_STRING_LEN(enc,s,len) len = strlen(s) */
40 #define ENC_STRING_LEN(enc,s,len) do { \
41 if (ONIGENC_MBC_MINLEN(enc) == 1) { \
42 UChar* tmps = (UChar* )(s); \
43 while (*tmps != 0) tmps++; \
44 len = (int )(tmps - (UChar* )(s));\
45 } \
46 else { \
47 len = onigenc_str_bytelen_null(enc, (UChar* )s); \
48 } \
49 } while(0)
50
51 typedef struct {
52 int onig_err;
53 int posix_err;
54 } O2PERR;
55
56 static int
onig2posix_error_code(int code)57 onig2posix_error_code(int code)
58 {
59 static const O2PERR o2p[] = {
60 { ONIG_MISMATCH, REG_NOMATCH },
61 { ONIG_NO_SUPPORT_CONFIG, REG_EONIG_INTERNAL },
62 { ONIG_ABORT, REG_EONIG_INTERNAL },
63 { ONIGERR_MEMORY, REG_ESPACE },
64 { ONIGERR_MATCH_STACK_LIMIT_OVER, REG_EONIG_INTERNAL },
65 { ONIGERR_RETRY_LIMIT_IN_MATCH_OVER, REG_EONIG_INTERNAL },
66 { ONIGERR_TYPE_BUG, REG_EONIG_INTERNAL },
67 { ONIGERR_PARSER_BUG, REG_EONIG_INTERNAL },
68 { ONIGERR_STACK_BUG, REG_EONIG_INTERNAL },
69 { ONIGERR_UNDEFINED_BYTECODE, REG_EONIG_INTERNAL },
70 { ONIGERR_UNEXPECTED_BYTECODE, REG_EONIG_INTERNAL },
71 { ONIGERR_DEFAULT_ENCODING_IS_NOT_SETTED, REG_EONIG_BADARG },
72 { ONIGERR_SPECIFIED_ENCODING_CANT_CONVERT_TO_WIDE_CHAR, REG_EONIG_BADARG },
73 { ONIGERR_FAIL_TO_INITIALIZE, REG_EONIG_INTERNAL },
74 { ONIGERR_INVALID_ARGUMENT, REG_EONIG_BADARG },
75 { ONIGERR_END_PATTERN_AT_LEFT_BRACE, REG_EBRACE },
76 { ONIGERR_END_PATTERN_AT_LEFT_BRACKET, REG_EBRACK },
77 { ONIGERR_EMPTY_CHAR_CLASS, REG_ECTYPE },
78 { ONIGERR_PREMATURE_END_OF_CHAR_CLASS, REG_ECTYPE },
79 { ONIGERR_END_PATTERN_AT_ESCAPE, REG_EESCAPE },
80 { ONIGERR_END_PATTERN_AT_META, REG_EESCAPE },
81 { ONIGERR_END_PATTERN_AT_CONTROL, REG_EESCAPE },
82 { ONIGERR_META_CODE_SYNTAX, REG_BADPAT },
83 { ONIGERR_CONTROL_CODE_SYNTAX, REG_BADPAT },
84 { ONIGERR_CHAR_CLASS_VALUE_AT_END_OF_RANGE, REG_ECTYPE },
85 { ONIGERR_CHAR_CLASS_VALUE_AT_START_OF_RANGE, REG_ECTYPE },
86 { ONIGERR_UNMATCHED_RANGE_SPECIFIER_IN_CHAR_CLASS, REG_ECTYPE },
87 { ONIGERR_TARGET_OF_REPEAT_OPERATOR_NOT_SPECIFIED, REG_BADRPT },
88 { ONIGERR_TARGET_OF_REPEAT_OPERATOR_INVALID, REG_BADRPT },
89 { ONIGERR_NESTED_REPEAT_OPERATOR, REG_BADRPT },
90 { ONIGERR_UNMATCHED_CLOSE_PARENTHESIS, REG_EPAREN },
91 { ONIGERR_END_PATTERN_WITH_UNMATCHED_PARENTHESIS, REG_EPAREN },
92 { ONIGERR_END_PATTERN_IN_GROUP, REG_BADPAT },
93 { ONIGERR_UNDEFINED_GROUP_OPTION, REG_BADPAT },
94 { ONIGERR_INVALID_POSIX_BRACKET_TYPE, REG_BADPAT },
95 { ONIGERR_INVALID_LOOK_BEHIND_PATTERN, REG_BADPAT },
96 { ONIGERR_INVALID_REPEAT_RANGE_PATTERN, REG_BADPAT },
97 { ONIGERR_TOO_BIG_NUMBER, REG_BADPAT },
98 { ONIGERR_TOO_BIG_NUMBER_FOR_REPEAT_RANGE, REG_BADBR },
99 { ONIGERR_UPPER_SMALLER_THAN_LOWER_IN_REPEAT_RANGE, REG_BADBR },
100 { ONIGERR_EMPTY_RANGE_IN_CHAR_CLASS, REG_ECTYPE },
101 { ONIGERR_MISMATCH_CODE_LENGTH_IN_CLASS_RANGE, REG_ECTYPE },
102 { ONIGERR_TOO_MANY_MULTI_BYTE_RANGES, REG_ECTYPE },
103 { ONIGERR_TOO_SHORT_MULTI_BYTE_STRING, REG_BADPAT },
104 { ONIGERR_TOO_BIG_BACKREF_NUMBER, REG_ESUBREG },
105 { ONIGERR_INVALID_BACKREF, REG_ESUBREG },
106 { ONIGERR_NUMBERED_BACKREF_OR_CALL_NOT_ALLOWED, REG_BADPAT },
107 { ONIGERR_TOO_BIG_WIDE_CHAR_VALUE, REG_EONIG_BADWC },
108 { ONIGERR_TOO_LONG_WIDE_CHAR_VALUE, REG_EONIG_BADWC },
109 { ONIGERR_INVALID_CODE_POINT_VALUE, REG_EONIG_BADWC },
110 { ONIGERR_EMPTY_GROUP_NAME, REG_BADPAT },
111 { ONIGERR_INVALID_GROUP_NAME, REG_BADPAT },
112 { ONIGERR_INVALID_CHAR_IN_GROUP_NAME, REG_BADPAT },
113 { ONIGERR_UNDEFINED_NAME_REFERENCE, REG_BADPAT },
114 { ONIGERR_UNDEFINED_GROUP_REFERENCE, REG_BADPAT },
115 { ONIGERR_MULTIPLEX_DEFINED_NAME, REG_BADPAT },
116 { ONIGERR_MULTIPLEX_DEFINITION_NAME_CALL, REG_BADPAT },
117 { ONIGERR_NEVER_ENDING_RECURSION, REG_BADPAT },
118 { ONIGERR_GROUP_NUMBER_OVER_FOR_CAPTURE_HISTORY, REG_BADPAT },
119 { ONIGERR_INVALID_CHAR_PROPERTY_NAME, REG_BADPAT },
120 { ONIGERR_INVALID_IF_ELSE_SYNTAX, REG_BADPAT },
121 { ONIGERR_INVALID_ABSENT_GROUP_PATTERN, REG_BADPAT },
122 { ONIGERR_INVALID_ABSENT_GROUP_GENERATOR_PATTERN, REG_BADPAT },
123 { ONIGERR_INVALID_CALLOUT_PATTERN, REG_BADPAT },
124 { ONIGERR_INVALID_CALLOUT_NAME, REG_BADPAT },
125 { ONIGERR_UNDEFINED_CALLOUT_NAME, REG_BADPAT },
126 { ONIGERR_INVALID_CALLOUT_BODY, REG_BADPAT },
127 { ONIGERR_INVALID_CALLOUT_TAG_NAME, REG_BADPAT },
128 { ONIGERR_INVALID_CALLOUT_ARG, REG_BADPAT },
129 { ONIGERR_NOT_SUPPORTED_ENCODING_COMBINATION, REG_EONIG_BADARG },
130 { ONIGERR_LIBRARY_IS_NOT_INITIALIZED, REG_EONIG_INTERNAL }
131 };
132
133 int i;
134
135 if (code >= 0) return 0;
136
137 for (i = 0; i < (int )(sizeof(o2p) / sizeof(o2p[0])); i++) {
138 if (code == o2p[i].onig_err)
139 return o2p[i].posix_err;
140 }
141
142 return REG_EONIG_INTERNAL; /* but, unknown error code */
143 }
144
145 extern int
regcomp(regex_t * reg,const char * pattern,int posix_options)146 regcomp(regex_t* reg, const char* pattern, int posix_options)
147 {
148 int r, len;
149 OnigSyntaxType* syntax = OnigDefaultSyntax;
150 OnigOptionType options;
151
152 reg->onig = (void* )0;
153
154 if ((posix_options & REG_EXTENDED) == 0)
155 syntax = ONIG_SYNTAX_POSIX_BASIC;
156
157 options = syntax->options;
158 if ((posix_options & REG_ICASE) != 0)
159 ONIG_OPTION_ON(options, ONIG_OPTION_IGNORECASE);
160 if ((posix_options & REG_NEWLINE) != 0) {
161 ONIG_OPTION_ON( options, ONIG_OPTION_NEGATE_SINGLELINE);
162 ONIG_OPTION_OFF(options, ONIG_OPTION_SINGLELINE);
163 }
164
165 reg->comp_options = posix_options;
166
167 ENC_STRING_LEN(OnigEncDefaultCharEncoding, pattern, len);
168 r = onig_new(PONIG_C(reg), (UChar* )pattern, (UChar* )(pattern + len),
169 options, OnigEncDefaultCharEncoding, syntax,
170 (OnigErrorInfo* )NULL);
171 if (r != ONIG_NORMAL) {
172 return onig2posix_error_code(r);
173 }
174
175 reg->re_nsub = ONIG_C(reg)->num_mem;
176 return 0;
177 }
178
179 extern int
regexec(regex_t * reg,const char * str,size_t nmatch,regmatch_t pmatch[],int posix_options)180 regexec(regex_t* reg, const char* str, size_t nmatch,
181 regmatch_t pmatch[], int posix_options)
182 {
183 int r, i, len;
184 UChar* end;
185 regmatch_t* pm;
186 OnigOptionType options;
187
188 options = ONIG_OPTION_POSIX_REGION;
189 if ((posix_options & REG_NOTBOL) != 0) options |= ONIG_OPTION_NOTBOL;
190 if ((posix_options & REG_NOTEOL) != 0) options |= ONIG_OPTION_NOTEOL;
191
192 if (nmatch == 0 || (reg->comp_options & REG_NOSUB) != 0) {
193 pm = (regmatch_t* )NULL;
194 nmatch = 0;
195 }
196 else if ((int )nmatch < ONIG_C(reg)->num_mem + 1) {
197 pm = (regmatch_t* )xmalloc(sizeof(regmatch_t)
198 * (ONIG_C(reg)->num_mem + 1));
199 if (pm == NULL)
200 return REG_ESPACE;
201 }
202 else {
203 pm = pmatch;
204 }
205
206 ENC_STRING_LEN(ONIG_C(reg)->enc, str, len);
207 end = (UChar* )(str + len);
208 r = onig_search(ONIG_C(reg), (UChar* )str, end, (UChar* )str, end,
209 (OnigRegion* )pm, options);
210
211 if (r >= 0) {
212 r = 0; /* Match */
213 if (pm != pmatch && pm != NULL) {
214 xmemcpy(pmatch, pm, sizeof(regmatch_t) * nmatch);
215 }
216 }
217 else if (r == ONIG_MISMATCH) {
218 r = REG_NOMATCH;
219 for (i = 0; i < (int )nmatch; i++)
220 pmatch[i].rm_so = pmatch[i].rm_eo = ONIG_REGION_NOTPOS;
221 }
222 else {
223 r = onig2posix_error_code(r);
224 }
225
226 if (pm != pmatch && pm != NULL)
227 xfree(pm);
228
229 #if 0
230 if (reg->re_nsub > nmatch - 1)
231 reg->re_nsub = (nmatch <= 1 ? 0 : nmatch - 1);
232 #endif
233
234 return r;
235 }
236
237 extern void
regfree(regex_t * reg)238 regfree(regex_t* reg)
239 {
240 onig_free(ONIG_C(reg));
241 reg->onig = (void* )0;
242 }
243
244
245 extern void
reg_set_encoding(int mb_code)246 reg_set_encoding(int mb_code)
247 {
248 OnigEncoding enc;
249
250 switch (mb_code) {
251 case REG_POSIX_ENCODING_ASCII:
252 enc = ONIG_ENCODING_ASCII;
253 break;
254 case REG_POSIX_ENCODING_EUC_JP:
255 enc = ONIG_ENCODING_EUC_JP;
256 break;
257 case REG_POSIX_ENCODING_SJIS:
258 enc = ONIG_ENCODING_SJIS;
259 break;
260 case REG_POSIX_ENCODING_UTF8:
261 enc = ONIG_ENCODING_UTF8;
262 break;
263 case REG_POSIX_ENCODING_UTF16_BE:
264 enc = ONIG_ENCODING_UTF16_BE;
265 break;
266 case REG_POSIX_ENCODING_UTF16_LE:
267 enc = ONIG_ENCODING_UTF16_LE;
268 break;
269
270 default:
271 return ;
272 break;
273 }
274
275 onig_initialize(&enc, 1);
276
277 onigenc_set_default_encoding(enc);
278 }
279
280 extern int
reg_name_to_group_numbers(regex_t * reg,const unsigned char * name,const unsigned char * name_end,int ** nums)281 reg_name_to_group_numbers(regex_t* reg,
282 const unsigned char* name, const unsigned char* name_end, int** nums)
283 {
284 return onig_name_to_group_numbers(ONIG_C(reg), name, name_end, nums);
285 }
286
287 typedef struct {
288 int (*func)(const unsigned char*, const unsigned char*,int,int*,regex_t*,void*);
289 regex_t* reg;
290 void* arg;
291 } i_wrap;
292
293 static int
i_wrapper(const UChar * name,const UChar * name_end,int ng,int * gs,onig_regex_t * reg ARG_UNUSED,void * arg)294 i_wrapper(const UChar* name, const UChar* name_end, int ng, int* gs,
295 onig_regex_t* reg ARG_UNUSED, void* arg)
296 {
297 i_wrap* warg = (i_wrap* )arg;
298
299 return (*warg->func)(name, name_end, ng, gs, warg->reg, warg->arg);
300 }
301
302 extern int
reg_foreach_name(regex_t * reg,int (* func)(const unsigned char *,const unsigned char *,int,int *,regex_t *,void *),void * arg)303 reg_foreach_name(regex_t* reg,
304 int (*func)(const unsigned char*, const unsigned char*,int,int*,regex_t*,void*),
305 void* arg)
306 {
307 i_wrap warg;
308
309 warg.func = func;
310 warg.reg = reg;
311 warg.arg = arg;
312
313 return onig_foreach_name(ONIG_C(reg), i_wrapper, &warg);
314 }
315
316 extern int
reg_number_of_names(regex_t * reg)317 reg_number_of_names(regex_t* reg)
318 {
319 return onig_number_of_names(ONIG_C(reg));
320 }
321
