Mercurial > hg > xemacs-beta
annotate src/regex.h @ 5223:acc4a6c9f5f9
Remove the definition of Q_data from glyphs.c, fixing C++ build.
2010-06-01 Aidan Kehoe <kehoea@parhasard.net>
* glyphs.c (syms_of_glyphs):
Remove the definition of Q_data from this file, now it's in
general-slots.h. Problem picked up by a C++ build.
| author | Aidan Kehoe <kehoea@parhasard.net> |
|---|---|
| date | Tue, 01 Jun 2010 20:31:11 +0100 |
| parents | f283b08ff0c9 |
| children | 308d34e9f07d |
| rev | line source |
|---|---|
| 428 | 1 /* Definitions for data structures and routines for the regular |
| 2 expression library, version 0.12. | |
| 3 | |
| 4 Copyright (C) 1985, 89, 90, 91, 92, 93, 95 Free Software Foundation, Inc. | |
| 5041 | 5 Copyright (C) 2002, 2010 Ben Wing. |
| 428 | 6 |
| 7 This program is free software; you can redistribute it and/or modify | |
| 8 it under the terms of the GNU General Public License as published by | |
| 9 the Free Software Foundation; either version 2, or (at your option) | |
| 10 any later version. | |
| 11 | |
| 12 This program is distributed in the hope that it will be useful, | |
| 13 but WITHOUT ANY WARRANTY; without even the implied warranty of | |
| 14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
| 15 GNU General Public License for more details. | |
| 16 | |
| 17 You should have received a copy of the GNU General Public License | |
| 18 along with this program; see the file COPYING. If not, write to | |
| 19 the Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
| 20 Boston, MA 02111-1307, USA. */ | |
| 21 | |
| 22 /* Synched up with: FSF 19.29. */ | |
| 23 | |
| 440 | 24 #ifndef INCLUDED_regex_h_ |
| 25 #define INCLUDED_regex_h_ | |
| 428 | 26 |
| 446 | 27 #ifdef emacs |
| 28 #define RE_TRANSLATE_TYPE Lisp_Object | |
| 2333 | 29 #define RE_LISP_SHORT_CONTEXT_ARGS_DECL , Lisp_Object UNUSED (lispobj), struct buffer *UNUSED (lispbuf) |
| 826 | 30 #define RE_LISP_SHORT_CONTEXT_ARGS , lispobj, lispbuf |
| 31 #define RE_LISP_CONTEXT_ARGS_DECL , Lisp_Object lispobj, struct buffer *lispbuf, struct syntax_cache *scache | |
| 2333 | 32 #define RE_LISP_CONTEXT_ARGS_MULE_DECL , Lisp_Object lispobj, struct buffer *USED_IF_MULE (lispbuf), struct syntax_cache *scache |
| 826 | 33 #define RE_LISP_CONTEXT_ARGS , lispobj, lispbuf, scache |
| 446 | 34 #else |
| 35 #define RE_TRANSLATE_TYPE char * | |
| 826 | 36 #define RE_LISP_SHORT_CONTEXT_ARGS_DECL |
| 37 #define RE_LISP_SHORT_CONTEXT_ARGS | |
| 38 #define RE_LISP_CONTEXT_ARGS_DECL | |
| 2333 | 39 #define RE_LISP_CONTEXT_ARGS_MULE_DECL |
| 826 | 40 #define RE_LISP_CONTEXT_ARGS |
| 665 | 41 #define Elemcount ssize_t |
| 42 #define Bytecount ssize_t | |
| 446 | 43 #endif /* emacs */ |
| 44 | |
|
5183
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
45 #ifndef emacs |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
46 # ifdef __cplusplus |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
47 # define BEGIN_C_DECLS extern "C" { |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
48 # define END_C_DECLS } |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
49 # else |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
50 # define BEGIN_C_DECLS |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
51 # define END_C_DECLS |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
52 # endif |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
53 #endif /* emacs */ |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
54 |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
55 BEGIN_C_DECLS |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
56 |
| 428 | 57 /* POSIX says that <sys/types.h> must be included (by the caller) before |
| 58 <regex.h>. */ | |
| 59 | |
| 60 | |
| 61 /* The following bits are used to determine the regexp syntax we | |
| 62 recognize. The not-set meaning typically corresponds to the syntax | |
| 63 used by Emacs (the exception is RE_INTERVAL, made for historical | |
| 64 reasons). The bits are given in alphabetical order, and the | |
| 65 definitions shifted by one from the previous bit; thus, when we add or | |
| 66 remove a bit, only one other definition need change. */ | |
| 67 typedef unsigned reg_syntax_t; | |
| 68 | |
| 69 /* If this bit is not set, then \ inside a bracket expression is literal. | |
| 70 If set, then such a \ quotes the following character. */ | |
| 71 #define RE_BACKSLASH_ESCAPE_IN_LISTS (1) | |
| 72 | |
| 73 /* If this bit is not set, then + and ? are operators, and \+ and \? are | |
| 74 literals. | |
| 75 If set, then \+ and \? are operators and + and ? are literals. */ | |
| 76 #define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) | |
| 77 | |
| 78 /* If this bit is set, then character classes are supported. They are: | |
| 79 [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], | |
| 80 [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. | |
| 81 If not set, then character classes are not supported. */ | |
| 82 #define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) | |
| 83 | |
| 84 /* If this bit is set, then ^ and $ are always anchors (outside bracket | |
| 85 expressions, of course). | |
| 86 If this bit is not set, then it depends: | |
| 87 ^ is an anchor if it is at the beginning of a regular | |
| 88 expression or after an open-group or an alternation operator; | |
| 89 $ is an anchor if it is at the end of a regular expression, or | |
| 90 before a close-group or an alternation operator. | |
| 91 | |
| 92 This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because | |
| 93 POSIX draft 11.2 says that * etc. in leading positions is undefined. | |
| 94 We already implemented a previous draft which made those constructs | |
| 95 invalid, though, so we haven't changed the code back. */ | |
| 96 #define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) | |
| 97 | |
| 98 /* If this bit is set, then special characters are always special | |
| 99 regardless of where they are in the pattern. | |
| 100 If this bit is not set, then special characters are special only in | |
| 101 some contexts; otherwise they are ordinary. Specifically, | |
| 102 * + ? and intervals are only special when not after the beginning, | |
| 103 open-group, or alternation operator. */ | |
| 104 #define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) | |
| 105 | |
| 106 /* If this bit is set, then *, +, ?, and { cannot be first in an re or | |
| 107 immediately after an alternation or begin-group operator. */ | |
| 108 #define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) | |
| 109 | |
| 110 /* If this bit is set, then . matches newline. | |
| 111 If not set, then it doesn't. */ | |
| 112 #define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) | |
| 113 | |
| 114 /* If this bit is set, then . doesn't match NUL. | |
| 115 If not set, then it does. */ | |
| 116 #define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) | |
| 117 | |
| 118 /* If this bit is set, nonmatching lists [^...] do not match newline. | |
| 119 If not set, they do. */ | |
| 120 #define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) | |
| 121 | |
| 122 /* If this bit is set, either \{...\} or {...} defines an | |
| 123 interval, depending on RE_NO_BK_BRACES. | |
| 124 If not set, \{, \}, {, and } are literals. */ | |
| 125 #define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) | |
| 126 | |
| 127 /* If this bit is set, +, ? and | aren't recognized as operators. | |
| 128 If not set, they are. */ | |
| 129 #define RE_LIMITED_OPS (RE_INTERVALS << 1) | |
| 130 | |
| 131 /* If this bit is set, newline is an alternation operator. | |
| 132 If not set, newline is literal. */ | |
| 133 #define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) | |
| 134 | |
| 135 /* If this bit is set, then `{...}' defines an interval, and \{ and \} | |
| 136 are literals. | |
| 137 If not set, then `\{...\}' defines an interval. */ | |
| 138 #define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) | |
| 139 | |
| 140 /* If this bit is set, (...) defines a group, and \( and \) are literals. | |
| 141 If not set, \(...\) defines a group, and ( and ) are literals. */ | |
| 142 #define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) | |
| 143 | |
| 144 /* If this bit is set, then \<digit> matches <digit>. | |
| 145 If not set, then \<digit> is a back-reference. */ | |
| 146 #define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) | |
| 147 | |
| 148 /* If this bit is set, then | is an alternation operator, and \| is literal. | |
| 149 If not set, then \| is an alternation operator, and | is literal. */ | |
| 150 #define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) | |
| 151 | |
| 152 /* If this bit is set, then an ending range point collating higher | |
| 153 than the starting range point, as in [z-a], is invalid. | |
| 154 If not set, then when ending range point collates higher than the | |
| 155 starting range point, the range is ignored. */ | |
| 156 #define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) | |
| 157 | |
| 158 /* If this bit is not set, allow minimal matching: | |
| 159 - a*? and a+? and a?? perform shortest-possible matching (compare with a* | |
| 160 and a+ and a?, respectively, which perform longest-possible matching) | |
| 161 - other juxtaposing of * + and ? is rejected. | |
| 162 If this bit is set, consecutive * + and ?'s are collapsed in a logical | |
| 163 manner: | |
| 164 - a*? and a+? are the same as a* | |
| 165 - a?? is the same as a? | |
| 166 */ | |
| 167 #define RE_NO_MINIMAL_MATCHING (RE_NO_EMPTY_RANGES << 1) | |
| 168 | |
| 169 /* If this bit is set, succeed as soon as we match the whole pattern, | |
| 170 without further backtracking. */ | |
| 171 #define RE_NO_POSIX_BACKTRACKING (RE_NO_MINIMAL_MATCHING << 1) | |
| 172 | |
| 173 /* If this bit is not set, (?:re) behaves like (re) (or \(?:re\) behaves like | |
| 174 \(re\)) except that the matched string is not registered. */ | |
| 175 #define RE_NO_SHY_GROUPS (RE_NO_POSIX_BACKTRACKING << 1) | |
| 176 | |
| 177 /* If this bit is set, then an unmatched ) is ordinary. | |
| 178 If not set, then an unmatched ) is invalid. */ | |
| 179 #define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_SHY_GROUPS << 1) | |
| 180 | |
| 502 | 181 /* If this bit is set, then \22 will read as a back reference, |
| 182 provided at least 22 non-shy groups have been seen so far. In all | |
| 183 other cases (bit not set, not 22 non-shy groups seen so far), it | |
| 184 reads as a back reference \2 followed by a digit 2. */ | |
| 185 #define RE_NO_MULTI_DIGIT_BK_REFS (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) | |
| 186 | |
| 428 | 187 /* This global variable defines the particular regexp syntax to use (for |
| 188 some interfaces). When a regexp is compiled, the syntax used is | |
| 189 stored in the pattern buffer, so changing this does not affect | |
| 190 already-compiled regexps. */ | |
| 191 extern reg_syntax_t re_syntax_options; | |
| 192 | |
| 193 /* Define combinations of the above bits for the standard possibilities. | |
| 194 (The [[[ comments delimit what gets put into the Texinfo file, so | |
| 195 don't delete them!) */ | |
| 196 /* [[[begin syntaxes]]] */ | |
| 197 #define RE_SYNTAX_EMACS RE_INTERVALS | |
| 198 | |
| 199 #define RE_SYNTAX_AWK \ | |
| 200 (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ | |
| 201 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | |
| 202 | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ | |
| 203 | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_SHY_GROUPS \ | |
| 502 | 204 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 205 |
| 206 #define RE_SYNTAX_POSIX_AWK \ | |
| 207 (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS) | |
| 208 | |
| 209 #define RE_SYNTAX_GREP \ | |
| 210 (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ | |
| 211 | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ | |
| 212 | RE_NEWLINE_ALT | RE_NO_SHY_GROUPS \ | |
| 502 | 213 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 214 |
| 215 #define RE_SYNTAX_EGREP \ | |
| 216 (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ | |
| 217 | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ | |
| 218 | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ | |
| 219 | RE_NO_BK_VBAR | RE_NO_SHY_GROUPS \ | |
| 502 | 220 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 221 |
| 222 #define RE_SYNTAX_POSIX_EGREP \ | |
| 502 | 223 (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES | \ |
| 224 RE_NO_MULTI_DIGIT_BK_REFS) | |
| 428 | 225 |
| 226 /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ | |
| 227 #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC | |
| 228 | |
| 229 #define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC | |
| 230 | |
| 231 /* Syntax bits common to both basic and extended POSIX regex syntax. */ | |
| 232 #define _RE_SYNTAX_POSIX_COMMON \ | |
| 233 (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ | |
| 234 | RE_INTERVALS | RE_NO_EMPTY_RANGES | RE_NO_SHY_GROUPS \ | |
| 502 | 235 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 236 |
| 237 #define RE_SYNTAX_POSIX_BASIC \ | |
| 238 (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM) | |
| 239 | |
| 240 /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes | |
| 241 RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this | |
| 242 isn't minimal, since other operators, such as \`, aren't disabled. */ | |
| 243 #define RE_SYNTAX_POSIX_MINIMAL_BASIC \ | |
| 244 (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) | |
| 245 | |
| 246 #define RE_SYNTAX_POSIX_EXTENDED \ | |
| 247 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ | |
| 248 | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ | |
| 249 | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ | |
| 250 | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
| 251 | |
| 252 /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS | |
| 253 replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added. */ | |
| 254 #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ | |
| 255 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ | |
| 256 | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ | |
| 257 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | |
| 258 | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
| 259 /* [[[end syntaxes]]] */ | |
| 260 | |
| 261 /* Maximum number of duplicates an interval can allow. Some systems | |
| 262 (erroneously) define this in other header files, but we want our | |
| 263 value, so remove any previous define. */ | |
| 264 #ifdef RE_DUP_MAX | |
| 265 #undef RE_DUP_MAX | |
| 266 #endif | |
| 267 #define RE_DUP_MAX ((1 << 15) - 1) | |
| 268 | |
| 269 | |
| 270 /* POSIX `cflags' bits (i.e., information for `regcomp'). */ | |
| 271 | |
| 272 /* If this bit is set, then use extended regular expression syntax. | |
| 273 If not set, then use basic regular expression syntax. */ | |
| 274 #define REG_EXTENDED 1 | |
| 275 | |
| 276 /* If this bit is set, then ignore case when matching. | |
| 277 If not set, then case is significant. */ | |
| 278 #define REG_ICASE (REG_EXTENDED << 1) | |
| 279 | |
| 280 /* If this bit is set, then anchors do not match at newline | |
| 281 characters in the string. | |
| 282 If not set, then anchors do match at newlines. */ | |
| 283 #define REG_NEWLINE (REG_ICASE << 1) | |
| 284 | |
| 285 /* If this bit is set, then report only success or fail in regexec. | |
| 286 If not set, then returns differ between not matching and errors. */ | |
| 287 #define REG_NOSUB (REG_NEWLINE << 1) | |
| 288 | |
| 289 | |
| 290 /* POSIX `eflags' bits (i.e., information for regexec). */ | |
| 291 | |
| 292 /* If this bit is set, then the beginning-of-line operator doesn't match | |
| 293 the beginning of the string (presumably because it's not the | |
| 294 beginning of a line). | |
| 295 If not set, then the beginning-of-line operator does match the | |
| 296 beginning of the string. */ | |
| 297 #define REG_NOTBOL 1 | |
| 298 | |
| 299 /* Like REG_NOTBOL, except for the end-of-line. */ | |
| 300 #define REG_NOTEOL (1 << 1) | |
| 301 | |
| 302 | |
| 303 /* If any error codes are removed, changed, or added, update the | |
| 304 `re_error_msg' table in regex.c. */ | |
| 305 typedef enum | |
| 306 { | |
| 307 REG_NOERROR = 0, /* Success. */ | |
| 308 REG_NOMATCH, /* Didn't find a match (for regexec). */ | |
| 309 | |
| 310 /* POSIX regcomp return error codes. (In the order listed in the | |
| 311 standard.) */ | |
| 312 REG_BADPAT, /* Invalid pattern. */ | |
| 313 REG_ECOLLATE, /* Not implemented. */ | |
| 314 REG_ECTYPE, /* Invalid character class name. */ | |
| 315 REG_EESCAPE, /* Trailing backslash. */ | |
| 316 REG_ESUBREG, /* Invalid back reference. */ | |
| 317 REG_EBRACK, /* Unmatched left bracket. */ | |
| 318 REG_EPAREN, /* Parenthesis imbalance. */ | |
| 319 REG_EBRACE, /* Unmatched \{. */ | |
| 320 REG_BADBR, /* Invalid contents of \{\}. */ | |
| 321 REG_ERANGE, /* Invalid range end. */ | |
| 322 REG_ESPACE, /* Ran out of memory. */ | |
| 323 REG_BADRPT, /* No preceding re for repetition op. */ | |
| 324 | |
| 325 /* Error codes we've added. */ | |
| 326 REG_EEND, /* Premature end. */ | |
| 327 REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ | |
| 328 REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ | |
| 329 #ifdef emacs | |
| 330 ,REG_ESYNTAX /* Invalid syntax designator. */ | |
| 331 #endif | |
| 332 #ifdef MULE | |
| 333 ,REG_ERANGESPAN /* Ranges may not span charsets. */ | |
| 334 ,REG_ECATEGORY /* Invalid category designator */ | |
| 335 #endif | |
| 336 } reg_errcode_t; | |
| 337 | |
| 338 /* This data structure represents a compiled pattern. Before calling | |
| 339 the pattern compiler, the fields `buffer', `allocated', `fastmap', | |
| 340 `translate', and `no_sub' can be set. After the pattern has been | |
| 341 compiled, the `re_nsub' field is available. All other fields are | |
| 342 private to the regex routines. */ | |
| 343 | |
| 344 struct re_pattern_buffer | |
| 345 { | |
| 346 /* [[[begin pattern_buffer]]] */ | |
| 347 /* Space that holds the compiled pattern. It is declared as | |
| 348 `unsigned char *' because its elements are | |
| 349 sometimes used as array indexes. */ | |
| 350 unsigned char *buffer; | |
| 351 | |
| 352 /* Number of bytes to which `buffer' points. */ | |
| 647 | 353 long allocated; |
| 428 | 354 |
| 355 /* Number of bytes actually used in `buffer'. */ | |
| 647 | 356 long used; |
| 428 | 357 |
| 358 /* Syntax setting with which the pattern was compiled. */ | |
| 359 reg_syntax_t syntax; | |
| 360 | |
| 361 /* Pointer to a fastmap, if any, otherwise zero. re_search uses | |
| 362 the fastmap, if there is one, to skip over impossible | |
| 363 starting points for matches. */ | |
| 364 char *fastmap; | |
| 365 | |
| 366 /* Either a translate table to apply to all characters before | |
| 367 comparing them, or zero for no translation. The translation | |
| 368 is applied to a pattern when it is compiled and to a string | |
| 369 when it is matched. */ | |
| 446 | 370 RE_TRANSLATE_TYPE translate; |
| 428 | 371 |
| 2639 | 372 /* Number of subpatterns (returnable groups) found by the compiler. |
| 373 (This does not count shy groups.) */ | |
| 647 | 374 int re_nsub; |
| 428 | 375 |
| 502 | 376 /* Total number of groups found by the compiler. (Including |
| 377 shy ones.) */ | |
| 378 int re_ngroups; | |
| 379 | |
| 428 | 380 /* Zero if this pattern cannot match the empty string, one else. |
| 381 Well, in truth it's used only in `re_search_2', to see | |
| 382 whether or not we should use the fastmap, so we don't set | |
| 383 this absolutely perfectly; see `re_compile_fastmap' (the | |
| 384 `duplicate' case). */ | |
| 647 | 385 unsigned int can_be_null : 1; |
| 428 | 386 |
| 387 /* If REGS_UNALLOCATED, allocate space in the `regs' structure | |
| 388 for `max (RE_NREGS, re_nsub + 1)' groups. | |
| 389 If REGS_REALLOCATE, reallocate space if necessary. | |
| 390 If REGS_FIXED, use what's there. */ | |
| 391 #define REGS_UNALLOCATED 0 | |
| 392 #define REGS_REALLOCATE 1 | |
| 393 #define REGS_FIXED 2 | |
| 647 | 394 unsigned int regs_allocated : 2; |
| 428 | 395 |
| 396 /* Set to zero when `regex_compile' compiles a pattern; set to one | |
| 397 by `re_compile_fastmap' if it updates the fastmap. */ | |
| 647 | 398 unsigned int fastmap_accurate : 1; |
| 428 | 399 |
| 400 /* If set, `re_match_2' does not return information about | |
| 401 subexpressions. */ | |
| 647 | 402 unsigned int no_sub : 1; |
| 428 | 403 |
| 404 /* If set, a beginning-of-line anchor doesn't match at the | |
| 405 beginning of the string. */ | |
| 647 | 406 unsigned int not_bol : 1; |
| 428 | 407 |
| 408 /* Similarly for an end-of-line anchor. */ | |
| 647 | 409 unsigned int not_eol : 1; |
| 428 | 410 |
| 411 /* If true, an anchor at a newline matches. */ | |
| 647 | 412 unsigned int newline_anchor : 1; |
| 428 | 413 |
| 647 | 414 unsigned int warned_about_incompatible_back_references : 1; |
| 502 | 415 |
| 416 /* Mapping between back references and groups (may not be | |
| 417 equivalent with shy groups). */ | |
| 418 int *external_to_internal_register; | |
| 419 | |
| 420 int external_to_internal_register_size; | |
| 421 | |
| 428 | 422 /* [[[end pattern_buffer]]] */ |
| 423 }; | |
| 424 | |
| 425 typedef struct re_pattern_buffer regex_t; | |
| 426 | |
| 427 /* Type for byte offsets within the string. POSIX mandates this. */ | |
| 428 typedef int regoff_t; | |
| 429 | |
| 430 | |
| 431 /* This is the structure we store register match data in. See | |
| 432 regex.texinfo for a full description of what registers match. */ | |
| 433 struct re_registers | |
| 434 { | |
| 1468 | 435 int num_regs; /* number of registers allocated */ |
| 428 | 436 regoff_t *start; |
| 437 regoff_t *end; | |
| 438 }; | |
| 439 | |
| 440 | |
| 441 /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, | |
| 442 `re_match_2' returns information about at least this many registers | |
| 443 the first time a `regs' structure is passed. */ | |
| 444 #ifndef RE_NREGS | |
| 445 #define RE_NREGS 30 | |
| 446 #endif | |
| 447 | |
| 448 | |
| 449 /* POSIX specification for registers. Aside from the different names than | |
| 450 `re_registers', POSIX uses an array of structures, instead of a | |
| 451 structure of arrays. */ | |
| 452 typedef struct | |
| 453 { | |
| 454 regoff_t rm_so; /* Byte offset from string's start to substring's start. */ | |
| 455 regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ | |
| 456 } regmatch_t; | |
| 457 | |
| 458 /* Declarations for routines. */ | |
| 459 | |
| 460 /* Sets the current default syntax to SYNTAX, and return the old syntax. | |
| 461 You can also simply assign to the `re_syntax_options' variable. */ | |
| 462 reg_syntax_t re_set_syntax (reg_syntax_t syntax); | |
| 463 | |
| 464 /* Compile the regular expression PATTERN, with length LENGTH | |
| 465 and syntax given by the global `re_syntax_options', into the buffer | |
| 466 BUFFER. Return NULL if successful, and an error string if not. */ | |
| 442 | 467 const char *re_compile_pattern (const char *pattern, int length, |
| 428 | 468 struct re_pattern_buffer *buffer); |
| 469 | |
| 470 | |
| 471 /* Compile a fastmap for the compiled pattern in BUFFER; used to | |
| 472 accelerate searches. Return 0 if successful and -2 if was an | |
| 473 internal error. */ | |
| 826 | 474 int re_compile_fastmap (struct re_pattern_buffer *buffer |
| 475 RE_LISP_SHORT_CONTEXT_ARGS_DECL); | |
| 428 | 476 |
| 477 | |
| 478 /* Search in the string STRING (with length LENGTH) for the pattern | |
| 479 compiled into BUFFER. Start searching at position START, for RANGE | |
| 480 characters. Return the starting position of the match, -1 for no | |
| 481 match, or -2 for an internal error. Also return register | |
| 482 information in REGS (if REGS and BUFFER->no_sub are nonzero). */ | |
| 442 | 483 int re_search (struct re_pattern_buffer *buffer, const char *string, |
| 428 | 484 int length, int start, int range, |
| 826 | 485 struct re_registers *regs RE_LISP_CONTEXT_ARGS_DECL); |
| 428 | 486 |
| 487 | |
| 488 /* Like `re_search', but search in the concatenation of STRING1 and | |
| 489 STRING2. Also, stop searching at index START + STOP. */ | |
| 442 | 490 int re_search_2 (struct re_pattern_buffer *buffer, const char *string1, |
| 491 int length1, const char *string2, int length2, int start, | |
| 826 | 492 int range, struct re_registers *regs, int stop |
| 493 RE_LISP_CONTEXT_ARGS_DECL); | |
| 428 | 494 |
| 826 | 495 #ifndef emacs /* never used by XEmacs */ |
| 428 | 496 |
| 497 /* Like `re_search', but return how many characters in STRING the regexp | |
| 498 in BUFFER matched, starting at position START. */ | |
| 442 | 499 int re_match (struct re_pattern_buffer *buffer, const char *string, |
| 826 | 500 int length, int start, struct re_registers *regs |
| 501 RE_LISP_CONTEXT_ARGS_DECL); | |
| 428 | 502 |
| 826 | 503 #endif /* not emacs */ |
| 428 | 504 |
| 505 /* Relates to `re_match' as `re_search_2' relates to `re_search'. */ | |
| 442 | 506 int re_match_2 (struct re_pattern_buffer *buffer, const char *string1, |
| 507 int length1, const char *string2, int length2, | |
| 826 | 508 int start, struct re_registers *regs, int stop |
| 509 RE_LISP_CONTEXT_ARGS_DECL); | |
| 428 | 510 |
| 511 /* Set REGS to hold NUM_REGS registers, storing them in STARTS and | |
| 512 ENDS. Subsequent matches using BUFFER and REGS will use this memory | |
| 513 for recording register information. STARTS and ENDS must be | |
| 514 allocated with malloc, and must each be at least `NUM_REGS * sizeof | |
| 515 (regoff_t)' bytes long. | |
| 516 | |
| 517 If NUM_REGS == 0, then subsequent matches should allocate their own | |
| 518 register data. | |
| 519 | |
| 520 Unless this function is called, the first search or match using | |
| 521 PATTERN_BUFFER will allocate its own register data, without | |
| 522 freeing the old data. */ | |
| 523 void re_set_registers (struct re_pattern_buffer *buffer, | |
| 647 | 524 struct re_registers *regs, int num_regs, |
| 428 | 525 regoff_t *starts, regoff_t *ends); |
| 526 | |
| 527 #ifdef _REGEX_RE_COMP | |
| 528 /* 4.2 bsd compatibility. */ | |
| 442 | 529 char *re_comp (const char *); |
| 530 int re_exec (const char *); | |
| 428 | 531 #endif |
| 532 | |
| 533 /* POSIX compatibility. */ | |
| 442 | 534 int regcomp (regex_t *preg, const char *pattern, int cflags); |
| 535 int regexec (const regex_t *preg, const char *string, size_t nmatch, | |
| 428 | 536 regmatch_t pmatch[], int eflags); |
| 442 | 537 size_t regerror (int errcode, const regex_t *preg, char *errbuf, |
| 428 | 538 size_t errbuf_size); |
| 539 void regfree (regex_t *preg); | |
| 540 | |
| 5041 | 541 enum regex_debug |
| 542 { | |
| 543 RE_DEBUG_COMPILATION = 1 << 0, | |
| 544 RE_DEBUG_FAILURE_POINT = 1 << 1, | |
| 545 RE_DEBUG_MATCHING = 1 << 2, | |
| 546 }; | |
| 547 | |
| 548 extern int debug_regexps; | |
| 549 | |
|
5183
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
550 END_C_DECLS |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
551 |
| 440 | 552 #endif /* INCLUDED_regex_h_ */ |