]>
Commit | Line | Data |
---|---|---|
2b83a2a4 | 1 | /* Definitions for data structures and routines for the regular |
3b0bdc72 | 2 | expression library. |
d614a753 | 3 | Copyright (C) 1985, 1989-2020 Free Software Foundation, Inc. |
3b0bdc72 | 4 | This file is part of the GNU C Library. |
2b83a2a4 | 5 | |
54d79e99 | 6 | The GNU C Library is free software; you can redistribute it and/or |
41bdb6e2 AJ |
7 | modify it under the terms of the GNU Lesser General Public |
8 | License as published by the Free Software Foundation; either | |
9 | version 2.1 of the License, or (at your option) any later version. | |
2b83a2a4 | 10 | |
54d79e99 UD |
11 | The GNU C Library is distributed in the hope that it will be useful, |
12 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
13 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
41bdb6e2 | 14 | Lesser General Public License for more details. |
2b83a2a4 | 15 | |
41bdb6e2 | 16 | You should have received a copy of the GNU Lesser General Public |
59ba27a6 | 17 | License along with the GNU C Library; if not, see |
eb04c213 | 18 | <https://www.gnu.org/licenses/>. */ |
2b83a2a4 | 19 | |
5107cf1d UD |
20 | #ifndef _REGEX_H |
21 | #define _REGEX_H 1 | |
2b83a2a4 | 22 | |
94c24227 UD |
23 | #include <sys/types.h> |
24 | ||
1fb05e3d UD |
25 | /* Allow the use in C++ code. */ |
26 | #ifdef __cplusplus | |
27 | extern "C" { | |
28 | #endif | |
29 | ||
eb04c213 AZ |
30 | /* Define __USE_GNU to declare GNU extensions that violate the |
31 | POSIX name space rules. */ | |
32 | #ifdef _GNU_SOURCE | |
33 | # define __USE_GNU 1 | |
34 | #endif | |
35 | ||
36 | #ifdef _REGEX_LARGE_OFFSETS | |
37 | ||
38 | /* Use types and values that are wide enough to represent signed and | |
39 | unsigned byte offsets in memory. This currently works only when | |
40 | the regex code is used outside of the GNU C library; it is not yet | |
41 | supported within glibc itself, and glibc users should not define | |
42 | _REGEX_LARGE_OFFSETS. */ | |
43 | ||
44 | /* The type of object sizes. */ | |
45 | typedef size_t __re_size_t; | |
46 | ||
47 | /* The type of object sizes, in places where the traditional code | |
48 | uses unsigned long int. */ | |
49 | typedef size_t __re_long_size_t; | |
50 | ||
51 | #else | |
52 | ||
53 | /* The traditional GNU regex implementation mishandles strings longer | |
54 | than INT_MAX. */ | |
55 | typedef unsigned int __re_size_t; | |
56 | typedef unsigned long int __re_long_size_t; | |
57 | ||
58 | #endif | |
59 | ||
4cca6b86 UD |
60 | /* The following two types have to be signed and unsigned integer type |
61 | wide enough to hold a value of a pointer. For most ANSI compilers | |
62 | ptrdiff_t and size_t should be likely OK. Still size of these two | |
63 | types is 2 for Microsoft C. Ugh... */ | |
64 | typedef long int s_reg_t; | |
65 | typedef unsigned long int active_reg_t; | |
66 | ||
2b83a2a4 RM |
67 | /* The following bits are used to determine the regexp syntax we |
68 | recognize. The set/not-set meanings are chosen so that Emacs syntax | |
69 | remains the value 0. The bits are given in alphabetical order, and | |
70 | the definitions shifted by one from the previous bit; thus, when we | |
71 | add or remove a bit, only one other definition need change. */ | |
4cca6b86 | 72 | typedef unsigned long int reg_syntax_t; |
2b83a2a4 | 73 | |
a53d3f82 | 74 | #ifdef __USE_GNU |
2b83a2a4 RM |
75 | /* If this bit is not set, then \ inside a bracket expression is literal. |
76 | If set, then such a \ quotes the following character. */ | |
a53d3f82 | 77 | # define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1) |
2b83a2a4 RM |
78 | |
79 | /* If this bit is not set, then + and ? are operators, and \+ and \? are | |
54d79e99 | 80 | literals. |
2b83a2a4 | 81 | If set, then \+ and \? are operators and + and ? are literals. */ |
a53d3f82 | 82 | # define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) |
2b83a2a4 RM |
83 | |
84 | /* If this bit is set, then character classes are supported. They are: | |
85 | [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], | |
86 | [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. | |
87 | If not set, then character classes are not supported. */ | |
a53d3f82 | 88 | # define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) |
2b83a2a4 RM |
89 | |
90 | /* If this bit is set, then ^ and $ are always anchors (outside bracket | |
91 | expressions, of course). | |
92 | If this bit is not set, then it depends: | |
a4b89fd8 AR |
93 | ^ is an anchor if it is at the beginning of a regular |
94 | expression or after an open-group or an alternation operator; | |
95 | $ is an anchor if it is at the end of a regular expression, or | |
96 | before a close-group or an alternation operator. | |
2b83a2a4 RM |
97 | |
98 | This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because | |
99 | POSIX draft 11.2 says that * etc. in leading positions is undefined. | |
100 | We already implemented a previous draft which made those constructs | |
101 | invalid, though, so we haven't changed the code back. */ | |
a53d3f82 | 102 | # define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) |
2b83a2a4 RM |
103 | |
104 | /* If this bit is set, then special characters are always special | |
105 | regardless of where they are in the pattern. | |
106 | If this bit is not set, then special characters are special only in | |
54d79e99 | 107 | some contexts; otherwise they are ordinary. Specifically, |
2b83a2a4 RM |
108 | * + ? and intervals are only special when not after the beginning, |
109 | open-group, or alternation operator. */ | |
a53d3f82 | 110 | # define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) |
2b83a2a4 RM |
111 | |
112 | /* If this bit is set, then *, +, ?, and { cannot be first in an re or | |
113 | immediately after an alternation or begin-group operator. */ | |
a53d3f82 | 114 | # define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) |
2b83a2a4 RM |
115 | |
116 | /* If this bit is set, then . matches newline. | |
117 | If not set, then it doesn't. */ | |
a53d3f82 | 118 | # define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) |
2b83a2a4 RM |
119 | |
120 | /* If this bit is set, then . doesn't match NUL. | |
121 | If not set, then it does. */ | |
a53d3f82 | 122 | # define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) |
2b83a2a4 RM |
123 | |
124 | /* If this bit is set, nonmatching lists [^...] do not match newline. | |
125 | If not set, they do. */ | |
a53d3f82 | 126 | # define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) |
2b83a2a4 RM |
127 | |
128 | /* If this bit is set, either \{...\} or {...} defines an | |
54d79e99 | 129 | interval, depending on RE_NO_BK_BRACES. |
2b83a2a4 | 130 | If not set, \{, \}, {, and } are literals. */ |
a53d3f82 | 131 | # define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) |
2b83a2a4 RM |
132 | |
133 | /* If this bit is set, +, ? and | aren't recognized as operators. | |
134 | If not set, they are. */ | |
a53d3f82 | 135 | # define RE_LIMITED_OPS (RE_INTERVALS << 1) |
2b83a2a4 RM |
136 | |
137 | /* If this bit is set, newline is an alternation operator. | |
138 | If not set, newline is literal. */ | |
a53d3f82 | 139 | # define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) |
2b83a2a4 | 140 | |
eb04c213 | 141 | /* If this bit is set, then '{...}' defines an interval, and \{ and \} |
2b83a2a4 | 142 | are literals. |
eb04c213 | 143 | If not set, then '\{...\}' defines an interval. */ |
a53d3f82 | 144 | # define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) |
2b83a2a4 RM |
145 | |
146 | /* If this bit is set, (...) defines a group, and \( and \) are literals. | |
147 | If not set, \(...\) defines a group, and ( and ) are literals. */ | |
a53d3f82 | 148 | # define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) |
2b83a2a4 RM |
149 | |
150 | /* If this bit is set, then \<digit> matches <digit>. | |
151 | If not set, then \<digit> is a back-reference. */ | |
a53d3f82 | 152 | # define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) |
2b83a2a4 | 153 | |
54d79e99 | 154 | /* If this bit is set, then | is an alternation operator, and \| is literal. |
2b83a2a4 | 155 | If not set, then \| is an alternation operator, and | is literal. */ |
a53d3f82 | 156 | # define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) |
2b83a2a4 RM |
157 | |
158 | /* If this bit is set, then an ending range point collating higher | |
159 | than the starting range point, as in [z-a], is invalid. | |
160 | If not set, then when ending range point collates higher than the | |
161 | starting range point, the range is ignored. */ | |
a53d3f82 | 162 | # define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) |
2b83a2a4 RM |
163 | |
164 | /* If this bit is set, then an unmatched ) is ordinary. | |
165 | If not set, then an unmatched ) is invalid. */ | |
a53d3f82 | 166 | # define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) |
2b83a2a4 RM |
167 | |
168 | /* If this bit is set, succeed as soon as we match the whole pattern, | |
169 | without further backtracking. */ | |
a53d3f82 | 170 | # define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) |
2b83a2a4 | 171 | |
4cca6b86 UD |
172 | /* If this bit is set, do not process the GNU regex operators. |
173 | If not set, then the GNU regex operators are recognized. */ | |
a53d3f82 | 174 | # define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1) |
4cca6b86 | 175 | |
51702635 UD |
176 | /* If this bit is set, turn on internal regex debugging. |
177 | If not set, and debugging was on, turn it off. | |
178 | This only works if regex.c is compiled -DDEBUG. | |
179 | We define this bit always, so that all that's needed to turn on | |
180 | debugging is to recompile regex.c; the calling code can always have | |
181 | this bit set, and it won't affect anything in the normal case. */ | |
a53d3f82 | 182 | # define RE_DEBUG (RE_NO_GNU_OPS << 1) |
51702635 | 183 | |
0a45b76c UD |
184 | /* If this bit is set, a syntactically invalid interval is treated as |
185 | a string of ordinary characters. For example, the ERE 'a{1' is | |
186 | treated as 'a\{1'. */ | |
a53d3f82 | 187 | # define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1) |
0a45b76c | 188 | |
3b0bdc72 UD |
189 | /* If this bit is set, then ignore case when matching. |
190 | If not set, then case is significant. */ | |
a53d3f82 | 191 | # define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1) |
3b0bdc72 | 192 | |
134abcb5 UD |
193 | /* This bit is used internally like RE_CONTEXT_INDEP_ANCHORS but only |
194 | for ^, because it is difficult to scan the regex backwards to find | |
195 | whether ^ should be special. */ | |
a53d3f82 | 196 | # define RE_CARET_ANCHORS_HERE (RE_ICASE << 1) |
134abcb5 | 197 | |
eb04c213 AZ |
198 | /* If this bit is set, then \{ cannot be first in a regex or |
199 | immediately after an alternation, open-group or \} operator. */ | |
a53d3f82 | 200 | # define RE_CONTEXT_INVALID_DUP (RE_CARET_ANCHORS_HERE << 1) |
06e8303a | 201 | |
c06a6956 UD |
202 | /* If this bit is set, then no_sub will be set to 1 during |
203 | re_compile_pattern. */ | |
a53d3f82 UD |
204 | # define RE_NO_SUB (RE_CONTEXT_INVALID_DUP << 1) |
205 | #endif | |
c06a6956 | 206 | |
2b83a2a4 RM |
207 | /* This global variable defines the particular regexp syntax to use (for |
208 | some interfaces). When a regexp is compiled, the syntax used is | |
209 | stored in the pattern buffer, so changing this does not affect | |
210 | already-compiled regexps. */ | |
211 | extern reg_syntax_t re_syntax_options; | |
212 | \f | |
a53d3f82 | 213 | #ifdef __USE_GNU |
2b83a2a4 RM |
214 | /* Define combinations of the above bits for the standard possibilities. |
215 | (The [[[ comments delimit what gets put into the Texinfo file, so | |
54d79e99 | 216 | don't delete them!) */ |
2b83a2a4 | 217 | /* [[[begin syntaxes]]] */ |
eb04c213 | 218 | # define RE_SYNTAX_EMACS 0 |
2b83a2a4 | 219 | |
eb04c213 | 220 | # define RE_SYNTAX_AWK \ |
4cca6b86 UD |
221 | (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ |
222 | | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | |
223 | | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ | |
51702635 | 224 | | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \ |
a4b89fd8 | 225 | | RE_CHAR_CLASSES \ |
4cca6b86 UD |
226 | | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS) |
227 | ||
eb04c213 | 228 | # define RE_SYNTAX_GNU_AWK \ |
a4b89fd8 AR |
229 | ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ |
230 | | RE_INVALID_INTERVAL_ORD) \ | |
231 | & ~(RE_DOT_NOT_NULL | RE_CONTEXT_INDEP_OPS \ | |
232 | | RE_CONTEXT_INVALID_OPS )) | |
2b83a2a4 | 233 | |
eb04c213 | 234 | # define RE_SYNTAX_POSIX_AWK \ |
51702635 | 235 | (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ |
a4b89fd8 AR |
236 | | RE_INTERVALS | RE_NO_GNU_OPS \ |
237 | | RE_INVALID_INTERVAL_ORD) | |
2b83a2a4 | 238 | |
eb04c213 AZ |
239 | # define RE_SYNTAX_GREP \ |
240 | ((RE_SYNTAX_POSIX_BASIC | RE_NEWLINE_ALT) \ | |
241 | & ~(RE_CONTEXT_INVALID_DUP | RE_DOT_NOT_NULL)) | |
2b83a2a4 | 242 | |
eb04c213 AZ |
243 | # define RE_SYNTAX_EGREP \ |
244 | ((RE_SYNTAX_POSIX_EXTENDED | RE_INVALID_INTERVAL_ORD | RE_NEWLINE_ALT) \ | |
245 | & ~(RE_CONTEXT_INVALID_OPS | RE_DOT_NOT_NULL)) | |
2b83a2a4 | 246 | |
eb04c213 AZ |
247 | /* POSIX grep -E behavior is no longer incompatible with GNU. */ |
248 | # define RE_SYNTAX_POSIX_EGREP \ | |
249 | RE_SYNTAX_EGREP | |
2b83a2a4 RM |
250 | |
251 | /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ | |
eb04c213 | 252 | # define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC |
2b83a2a4 | 253 | |
eb04c213 | 254 | # define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC |
2b83a2a4 RM |
255 | |
256 | /* Syntax bits common to both basic and extended POSIX regex syntax. */ | |
eb04c213 | 257 | # define _RE_SYNTAX_POSIX_COMMON \ |
2b83a2a4 RM |
258 | (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ |
259 | | RE_INTERVALS | RE_NO_EMPTY_RANGES) | |
260 | ||
eb04c213 | 261 | # define RE_SYNTAX_POSIX_BASIC \ |
06e8303a | 262 | (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM | RE_CONTEXT_INVALID_DUP) |
2b83a2a4 RM |
263 | |
264 | /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes | |
265 | RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this | |
266 | isn't minimal, since other operators, such as \`, aren't disabled. */ | |
eb04c213 | 267 | # define RE_SYNTAX_POSIX_MINIMAL_BASIC \ |
2b83a2a4 RM |
268 | (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) |
269 | ||
eb04c213 | 270 | # define RE_SYNTAX_POSIX_EXTENDED \ |
9281f45d UD |
271 | (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ |
272 | | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ | |
273 | | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ | |
274 | | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
2b83a2a4 | 275 | |
0e179ea6 UD |
276 | /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is |
277 | removed and RE_NO_BK_REFS is added. */ | |
eb04c213 | 278 | # define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ |
2b83a2a4 RM |
279 | (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ |
280 | | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ | |
281 | | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | |
282 | | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
283 | /* [[[end syntaxes]]] */ | |
eb04c213 AZ |
284 | |
285 | /* Maximum number of duplicates an interval can allow. POSIX-conforming | |
286 | systems might define this in <limits.h>, but we want our | |
2b83a2a4 | 287 | value, so remove any previous define. */ |
eb04c213 AZ |
288 | # ifdef _REGEX_INCLUDE_LIMITS_H |
289 | # include <limits.h> | |
290 | # endif | |
a53d3f82 UD |
291 | # ifdef RE_DUP_MAX |
292 | # undef RE_DUP_MAX | |
293 | # endif | |
eb04c213 AZ |
294 | |
295 | /* RE_DUP_MAX is 2**15 - 1 because an earlier implementation stored | |
296 | the counter as a 2-byte signed integer. This is no longer true, so | |
297 | RE_DUP_MAX could be increased to (INT_MAX / 10 - 1), or to | |
298 | ((SIZE_MAX - 9) / 10) if _REGEX_LARGE_OFFSETS is defined. | |
299 | However, there would be a huge performance problem if someone | |
300 | actually used a pattern like a\{214748363\}, so RE_DUP_MAX retains | |
301 | its historical value. */ | |
a53d3f82 UD |
302 | # define RE_DUP_MAX (0x7fff) |
303 | #endif | |
2b83a2a4 RM |
304 | |
305 | ||
eb04c213 | 306 | /* POSIX 'cflags' bits (i.e., information for 'regcomp'). */ |
2b83a2a4 RM |
307 | |
308 | /* If this bit is set, then use extended regular expression syntax. | |
309 | If not set, then use basic regular expression syntax. */ | |
310 | #define REG_EXTENDED 1 | |
311 | ||
312 | /* If this bit is set, then ignore case when matching. | |
313 | If not set, then case is significant. */ | |
eb04c213 | 314 | #define REG_ICASE (1 << 1) |
54d79e99 | 315 | |
2b83a2a4 RM |
316 | /* If this bit is set, then anchors do not match at newline |
317 | characters in the string. | |
318 | If not set, then anchors do match at newlines. */ | |
eb04c213 | 319 | #define REG_NEWLINE (1 << 2) |
2b83a2a4 RM |
320 | |
321 | /* If this bit is set, then report only success or fail in regexec. | |
322 | If not set, then returns differ between not matching and errors. */ | |
eb04c213 | 323 | #define REG_NOSUB (1 << 3) |
2b83a2a4 RM |
324 | |
325 | ||
eb04c213 | 326 | /* POSIX 'eflags' bits (i.e., information for regexec). */ |
2b83a2a4 RM |
327 | |
328 | /* If this bit is set, then the beginning-of-line operator doesn't match | |
329 | the beginning of the string (presumably because it's not the | |
330 | beginning of a line). | |
331 | If not set, then the beginning-of-line operator does match the | |
332 | beginning of the string. */ | |
333 | #define REG_NOTBOL 1 | |
334 | ||
335 | /* Like REG_NOTBOL, except for the end-of-line. */ | |
336 | #define REG_NOTEOL (1 << 1) | |
337 | ||
6fefb4e0 UD |
338 | /* Use PMATCH[0] to delimit the start and end of the search in the |
339 | buffer. */ | |
340 | #define REG_STARTEND (1 << 2) | |
341 | ||
2b83a2a4 RM |
342 | |
343 | /* If any error codes are removed, changed, or added, update the | |
eb04c213 AZ |
344 | '__re_error_msgid' table in regcomp.c. */ |
345 | ||
2b83a2a4 RM |
346 | typedef enum |
347 | { | |
eb04c213 AZ |
348 | _REG_ENOSYS = -1, /* This will never happen for this implementation. */ |
349 | _REG_NOERROR = 0, /* Success. */ | |
350 | _REG_NOMATCH, /* Didn't find a match (for regexec). */ | |
2b83a2a4 RM |
351 | |
352 | /* POSIX regcomp return error codes. (In the order listed in the | |
353 | standard.) */ | |
eb04c213 AZ |
354 | _REG_BADPAT, /* Invalid pattern. */ |
355 | _REG_ECOLLATE, /* Invalid collating element. */ | |
356 | _REG_ECTYPE, /* Invalid character class name. */ | |
357 | _REG_EESCAPE, /* Trailing backslash. */ | |
358 | _REG_ESUBREG, /* Invalid back reference. */ | |
359 | _REG_EBRACK, /* Unmatched left bracket. */ | |
360 | _REG_EPAREN, /* Parenthesis imbalance. */ | |
361 | _REG_EBRACE, /* Unmatched \{. */ | |
362 | _REG_BADBR, /* Invalid contents of \{\}. */ | |
363 | _REG_ERANGE, /* Invalid range end. */ | |
364 | _REG_ESPACE, /* Ran out of memory. */ | |
365 | _REG_BADRPT, /* No preceding re for repetition op. */ | |
2b83a2a4 RM |
366 | |
367 | /* Error codes we've added. */ | |
eb04c213 AZ |
368 | _REG_EEND, /* Premature end. */ |
369 | _REG_ESIZE, /* Too large (e.g., repeat count too large). */ | |
370 | _REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ | |
2b83a2a4 | 371 | } reg_errcode_t; |
eb04c213 AZ |
372 | |
373 | #if defined _XOPEN_SOURCE || defined __USE_XOPEN2K | |
374 | # define REG_ENOSYS _REG_ENOSYS | |
375 | #endif | |
376 | #define REG_NOERROR _REG_NOERROR | |
377 | #define REG_NOMATCH _REG_NOMATCH | |
378 | #define REG_BADPAT _REG_BADPAT | |
379 | #define REG_ECOLLATE _REG_ECOLLATE | |
380 | #define REG_ECTYPE _REG_ECTYPE | |
381 | #define REG_EESCAPE _REG_EESCAPE | |
382 | #define REG_ESUBREG _REG_ESUBREG | |
383 | #define REG_EBRACK _REG_EBRACK | |
384 | #define REG_EPAREN _REG_EPAREN | |
385 | #define REG_EBRACE _REG_EBRACE | |
386 | #define REG_BADBR _REG_BADBR | |
387 | #define REG_ERANGE _REG_ERANGE | |
388 | #define REG_ESPACE _REG_ESPACE | |
389 | #define REG_BADRPT _REG_BADRPT | |
390 | #define REG_EEND _REG_EEND | |
391 | #define REG_ESIZE _REG_ESIZE | |
392 | #define REG_ERPAREN _REG_ERPAREN | |
2b83a2a4 RM |
393 | \f |
394 | /* This data structure represents a compiled pattern. Before calling | |
eb04c213 AZ |
395 | the pattern compiler, the fields 'buffer', 'allocated', 'fastmap', |
396 | and 'translate' can be set. After the pattern has been compiled, | |
397 | the fields 're_nsub', 'not_bol' and 'not_eol' are available. All | |
78e64fdc | 398 | other fields are private to the regex routines. */ |
2b83a2a4 | 399 | |
54d79e99 | 400 | #ifndef RE_TRANSLATE_TYPE |
a53d3f82 UD |
401 | # define __RE_TRANSLATE_TYPE unsigned char * |
402 | # ifdef __USE_GNU | |
403 | # define RE_TRANSLATE_TYPE __RE_TRANSLATE_TYPE | |
404 | # endif | |
405 | #endif | |
406 | ||
407 | #ifdef __USE_GNU | |
408 | # define __REPB_PREFIX(name) name | |
409 | #else | |
410 | # define __REPB_PREFIX(name) __##name | |
03a75825 RM |
411 | #endif |
412 | ||
2b83a2a4 RM |
413 | struct re_pattern_buffer |
414 | { | |
eb04c213 AZ |
415 | /* Space that holds the compiled pattern. The type |
416 | 'struct re_dfa_t' is private and is not declared here. */ | |
417 | struct re_dfa_t *__REPB_PREFIX(buffer); | |
2b83a2a4 | 418 | |
eb04c213 AZ |
419 | /* Number of bytes to which 'buffer' points. */ |
420 | __re_long_size_t __REPB_PREFIX(allocated); | |
2b83a2a4 | 421 | |
eb04c213 AZ |
422 | /* Number of bytes actually used in 'buffer'. */ |
423 | __re_long_size_t __REPB_PREFIX(used); | |
2b83a2a4 | 424 | |
2ff89ea4 | 425 | /* Syntax setting with which the pattern was compiled. */ |
a53d3f82 | 426 | reg_syntax_t __REPB_PREFIX(syntax); |
2b83a2a4 | 427 | |
2ff89ea4 UD |
428 | /* Pointer to a fastmap, if any, otherwise zero. re_search uses the |
429 | fastmap, if there is one, to skip over impossible starting points | |
430 | for matches. */ | |
a53d3f82 | 431 | char *__REPB_PREFIX(fastmap); |
2b83a2a4 | 432 | |
2ff89ea4 UD |
433 | /* Either a translate table to apply to all characters before |
434 | comparing them, or zero for no translation. The translation is | |
435 | applied to a pattern when it is compiled and to a string when it | |
436 | is matched. */ | |
a53d3f82 | 437 | __RE_TRANSLATE_TYPE __REPB_PREFIX(translate); |
2b83a2a4 | 438 | |
2ff89ea4 | 439 | /* Number of subexpressions found by the compiler. */ |
2b83a2a4 RM |
440 | size_t re_nsub; |
441 | ||
2ff89ea4 | 442 | /* Zero if this pattern cannot match the empty string, one else. |
eb04c213 | 443 | Well, in truth it's used only in 're_search_2', to see whether or |
2ff89ea4 | 444 | not we should use the fastmap, so we don't set this absolutely |
eb04c213 | 445 | perfectly; see 're_compile_fastmap' (the "duplicate" case). */ |
a53d3f82 | 446 | unsigned __REPB_PREFIX(can_be_null) : 1; |
2b83a2a4 | 447 | |
eb04c213 AZ |
448 | /* If REGS_UNALLOCATED, allocate space in the 'regs' structure |
449 | for 'max (RE_NREGS, re_nsub + 1)' groups. | |
2ff89ea4 UD |
450 | If REGS_REALLOCATE, reallocate space if necessary. |
451 | If REGS_FIXED, use what's there. */ | |
a53d3f82 UD |
452 | #ifdef __USE_GNU |
453 | # define REGS_UNALLOCATED 0 | |
454 | # define REGS_REALLOCATE 1 | |
455 | # define REGS_FIXED 2 | |
456 | #endif | |
457 | unsigned __REPB_PREFIX(regs_allocated) : 2; | |
2b83a2a4 | 458 | |
eb04c213 AZ |
459 | /* Set to zero when 're_compile_pattern' compiles a pattern; set to |
460 | one by 're_compile_fastmap' if it updates the fastmap. */ | |
a53d3f82 | 461 | unsigned __REPB_PREFIX(fastmap_accurate) : 1; |
2b83a2a4 | 462 | |
eb04c213 | 463 | /* If set, 're_match_2' does not return information about |
2ff89ea4 | 464 | subexpressions. */ |
a53d3f82 | 465 | unsigned __REPB_PREFIX(no_sub) : 1; |
2b83a2a4 | 466 | |
2ff89ea4 UD |
467 | /* If set, a beginning-of-line anchor doesn't match at the beginning |
468 | of the string. */ | |
a53d3f82 | 469 | unsigned __REPB_PREFIX(not_bol) : 1; |
2b83a2a4 | 470 | |
2ff89ea4 | 471 | /* Similarly for an end-of-line anchor. */ |
a53d3f82 | 472 | unsigned __REPB_PREFIX(not_eol) : 1; |
2b83a2a4 | 473 | |
2ff89ea4 | 474 | /* If true, an anchor at a newline matches. */ |
a53d3f82 | 475 | unsigned __REPB_PREFIX(newline_anchor) : 1; |
2b83a2a4 RM |
476 | }; |
477 | ||
478 | typedef struct re_pattern_buffer regex_t; | |
479 | \f | |
480 | /* Type for byte offsets within the string. POSIX mandates this. */ | |
eb04c213 AZ |
481 | #ifdef _REGEX_LARGE_OFFSETS |
482 | /* POSIX 1003.1-2008 requires that regoff_t be at least as wide as | |
483 | ptrdiff_t and ssize_t. We don't know of any hosts where ptrdiff_t | |
484 | is wider than ssize_t, so ssize_t is safe. ptrdiff_t is not | |
485 | visible here, so use ssize_t. */ | |
486 | typedef ssize_t regoff_t; | |
487 | #else | |
488 | /* The traditional GNU regex implementation mishandles strings longer | |
489 | than INT_MAX. */ | |
2b83a2a4 | 490 | typedef int regoff_t; |
eb04c213 | 491 | #endif |
2b83a2a4 RM |
492 | |
493 | ||
a53d3f82 | 494 | #ifdef __USE_GNU |
2b83a2a4 RM |
495 | /* This is the structure we store register match data in. See |
496 | regex.texinfo for a full description of what registers match. */ | |
497 | struct re_registers | |
498 | { | |
eb04c213 | 499 | __re_size_t num_regs; |
2b83a2a4 RM |
500 | regoff_t *start; |
501 | regoff_t *end; | |
502 | }; | |
503 | ||
504 | ||
eb04c213 AZ |
505 | /* If 'regs_allocated' is REGS_UNALLOCATED in the pattern buffer, |
506 | 're_match_2' returns information about at least this many registers | |
507 | the first time a 'regs' structure is passed. */ | |
a53d3f82 UD |
508 | # ifndef RE_NREGS |
509 | # define RE_NREGS 30 | |
510 | # endif | |
2b83a2a4 RM |
511 | #endif |
512 | ||
513 | ||
514 | /* POSIX specification for registers. Aside from the different names than | |
eb04c213 | 515 | 're_registers', POSIX uses an array of structures, instead of a |
2b83a2a4 RM |
516 | structure of arrays. */ |
517 | typedef struct | |
518 | { | |
519 | regoff_t rm_so; /* Byte offset from string's start to substring's start. */ | |
520 | regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ | |
521 | } regmatch_t; | |
522 | \f | |
523 | /* Declarations for routines. */ | |
524 | ||
a53d3f82 | 525 | #ifdef __USE_GNU |
2b83a2a4 | 526 | /* Sets the current default syntax to SYNTAX, and return the old syntax. |
eb04c213 | 527 | You can also simply assign to the 're_syntax_options' variable. */ |
2ff89ea4 | 528 | extern reg_syntax_t re_set_syntax (reg_syntax_t __syntax); |
2b83a2a4 RM |
529 | |
530 | /* Compile the regular expression PATTERN, with length LENGTH | |
eb04c213 | 531 | and syntax given by the global 're_syntax_options', into the buffer |
78e64fdc RT |
532 | BUFFER. Return NULL if successful, and an error string if not. |
533 | ||
eb04c213 AZ |
534 | To free the allocated storage, you must call 'regfree' on BUFFER. |
535 | Note that the translate table must either have been initialized by | |
536 | 'regcomp', with a malloc'ed value, or set to NULL before calling | |
537 | 'regfree'. */ | |
2ff89ea4 UD |
538 | extern const char *re_compile_pattern (const char *__pattern, size_t __length, |
539 | struct re_pattern_buffer *__buffer); | |
2b83a2a4 RM |
540 | |
541 | ||
542 | /* Compile a fastmap for the compiled pattern in BUFFER; used to | |
543 | accelerate searches. Return 0 if successful and -2 if was an | |
544 | internal error. */ | |
2ff89ea4 | 545 | extern int re_compile_fastmap (struct re_pattern_buffer *__buffer); |
2b83a2a4 RM |
546 | |
547 | ||
548 | /* Search in the string STRING (with length LENGTH) for the pattern | |
549 | compiled into BUFFER. Start searching at position START, for RANGE | |
550 | characters. Return the starting position of the match, -1 for no | |
551 | match, or -2 for an internal error. Also return register | |
552 | information in REGS (if REGS and BUFFER->no_sub are nonzero). */ | |
eb04c213 AZ |
553 | extern regoff_t re_search (struct re_pattern_buffer *__buffer, |
554 | const char *__String, regoff_t __length, | |
555 | regoff_t __start, regoff_t __range, | |
556 | struct re_registers *__regs); | |
2b83a2a4 RM |
557 | |
558 | ||
eb04c213 | 559 | /* Like 're_search', but search in the concatenation of STRING1 and |
2b83a2a4 | 560 | STRING2. Also, stop searching at index START + STOP. */ |
eb04c213 AZ |
561 | extern regoff_t re_search_2 (struct re_pattern_buffer *__buffer, |
562 | const char *__string1, regoff_t __length1, | |
563 | const char *__string2, regoff_t __length2, | |
564 | regoff_t __start, regoff_t __range, | |
565 | struct re_registers *__regs, | |
566 | regoff_t __stop); | |
2b83a2a4 RM |
567 | |
568 | ||
eb04c213 | 569 | /* Like 're_search', but return how many characters in STRING the regexp |
2b83a2a4 | 570 | in BUFFER matched, starting at position START. */ |
eb04c213 AZ |
571 | extern regoff_t re_match (struct re_pattern_buffer *__buffer, |
572 | const char *__String, regoff_t __length, | |
573 | regoff_t __start, struct re_registers *__regs); | |
2b83a2a4 RM |
574 | |
575 | ||
eb04c213 AZ |
576 | /* Relates to 're_match' as 're_search_2' relates to 're_search'. */ |
577 | extern regoff_t re_match_2 (struct re_pattern_buffer *__buffer, | |
578 | const char *__string1, regoff_t __length1, | |
579 | const char *__string2, regoff_t __length2, | |
580 | regoff_t __start, struct re_registers *__regs, | |
581 | regoff_t __stop); | |
2b83a2a4 RM |
582 | |
583 | ||
584 | /* Set REGS to hold NUM_REGS registers, storing them in STARTS and | |
585 | ENDS. Subsequent matches using BUFFER and REGS will use this memory | |
586 | for recording register information. STARTS and ENDS must be | |
eb04c213 | 587 | allocated with malloc, and must each be at least 'NUM_REGS * sizeof |
2b83a2a4 RM |
588 | (regoff_t)' bytes long. |
589 | ||
590 | If NUM_REGS == 0, then subsequent matches should allocate their own | |
591 | register data. | |
592 | ||
593 | Unless this function is called, the first search or match using | |
eb04c213 | 594 | BUFFER will allocate its own register data, without |
2b83a2a4 | 595 | freeing the old data. */ |
2ff89ea4 UD |
596 | extern void re_set_registers (struct re_pattern_buffer *__buffer, |
597 | struct re_registers *__regs, | |
eb04c213 | 598 | __re_size_t __num_regs, |
2ff89ea4 | 599 | regoff_t *__starts, regoff_t *__ends); |
a53d3f82 | 600 | #endif /* Use GNU */ |
2b83a2a4 | 601 | |
498afc54 | 602 | #if defined _REGEX_RE_COMP || (defined _LIBC && defined __USE_MISC) |
6973fc01 | 603 | # ifndef _CRAY |
2b83a2a4 | 604 | /* 4.2 bsd compatibility. */ |
2d87db5b UD |
605 | extern char *re_comp (const char *); |
606 | extern int re_exec (const char *); | |
6973fc01 | 607 | # endif |
4cca6b86 | 608 | #endif |
2b83a2a4 | 609 | |
eb04c213 AZ |
610 | /* For plain 'restrict', use glibc's __restrict if defined. |
611 | Otherwise, GCC 2.95 and later have "__restrict"; C99 compilers have | |
612 | "restrict", and "configure" may have defined "restrict". | |
613 | Other compilers use __restrict, __restrict__, and _Restrict, and | |
614 | 'configure' might #define 'restrict' to those words, so pick a | |
615 | different name. */ | |
616 | #ifndef _Restrict_ | |
617 | # if defined __restrict || 2 < __GNUC__ + (95 <= __GNUC_MINOR__) | |
618 | # define _Restrict_ __restrict | |
619 | # elif 199901L <= __STDC_VERSION__ || defined restrict | |
620 | # define _Restrict_ restrict | |
621 | # else | |
622 | # define _Restrict_ | |
7ca404ad | 623 | # endif |
da2a3ca6 | 624 | #endif |
eb04c213 AZ |
625 | /* For [restrict], use glibc's __restrict_arr if available. |
626 | Otherwise, GCC 3.1 (not in C++ mode) and C99 support [restrict]. */ | |
627 | #ifndef _Restrict_arr_ | |
628 | # ifdef __restrict_arr | |
629 | # define _Restrict_arr_ __restrict_arr | |
630 | # elif ((199901L <= __STDC_VERSION__ || 3 < __GNUC__ + (1 <= __GNUC_MINOR__)) \ | |
631 | && !defined __GNUG__) | |
632 | # define _Restrict_arr_ _Restrict_ | |
8343eaee | 633 | # else |
eb04c213 | 634 | # define _Restrict_arr_ |
8343eaee UD |
635 | # endif |
636 | #endif | |
7ca404ad | 637 | |
2b83a2a4 | 638 | /* POSIX compatibility. */ |
eb04c213 AZ |
639 | extern int regcomp (regex_t *_Restrict_ __preg, |
640 | const char *_Restrict_ __pattern, | |
2d87db5b | 641 | int __cflags); |
30baa360 | 642 | |
eb04c213 AZ |
643 | extern int regexec (const regex_t *_Restrict_ __preg, |
644 | const char *_Restrict_ __String, size_t __nmatch, | |
645 | regmatch_t __pmatch[_Restrict_arr_], | |
2d87db5b | 646 | int __eflags); |
30baa360 | 647 | |
eb04c213 AZ |
648 | extern size_t regerror (int __errcode, const regex_t *_Restrict_ __preg, |
649 | char *_Restrict_ __errbuf, size_t __errbuf_size); | |
30baa360 | 650 | |
2d87db5b | 651 | extern void regfree (regex_t *__preg); |
2b83a2a4 | 652 | |
1fb05e3d UD |
653 | |
654 | #ifdef __cplusplus | |
655 | } | |
656 | #endif /* C++ */ | |
657 | ||
5107cf1d | 658 | #endif /* regex.h */ |