Mercurial > hg > xemacs-beta
annotate src/regex.h @ 5437:002cb5224e4f
Merge with 21.5 trunk.
| author | Mats Lidell <matsl@xemacs.org> |
|---|---|
| date | Mon, 15 Nov 2010 22:33:52 +0100 |
| parents | 308d34e9f07d |
| children | 1d9f603e9125 |
| 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 |
|
5402
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
7 This file is part of XEmacs. |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
8 |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
9 XEmacs is free software: you can redistribute it and/or modify it |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
10 under the terms of the GNU General Public License as published by the |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
11 Free Software Foundation, either version 3 of the License, or (at your |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
12 option) any later version. |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
13 |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
14 XEmacs is distributed in the hope that it will be useful, but WITHOUT |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
15 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
16 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
17 for more details. |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
18 |
| 428 | 19 You should have received a copy of the GNU General Public License |
|
5402
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5183
diff
changeset
|
20 along with XEmacs. If not, see <http://www.gnu.org/licenses/>. */ |
| 428 | 21 /* Synched up with: FSF 19.29. */ |
| 22 | |
| 440 | 23 #ifndef INCLUDED_regex_h_ |
| 24 #define INCLUDED_regex_h_ | |
| 428 | 25 |
| 446 | 26 #ifdef emacs |
| 27 #define RE_TRANSLATE_TYPE Lisp_Object | |
| 2333 | 28 #define RE_LISP_SHORT_CONTEXT_ARGS_DECL , Lisp_Object UNUSED (lispobj), struct buffer *UNUSED (lispbuf) |
| 826 | 29 #define RE_LISP_SHORT_CONTEXT_ARGS , lispobj, lispbuf |
| 30 #define RE_LISP_CONTEXT_ARGS_DECL , Lisp_Object lispobj, struct buffer *lispbuf, struct syntax_cache *scache | |
| 2333 | 31 #define RE_LISP_CONTEXT_ARGS_MULE_DECL , Lisp_Object lispobj, struct buffer *USED_IF_MULE (lispbuf), struct syntax_cache *scache |
| 826 | 32 #define RE_LISP_CONTEXT_ARGS , lispobj, lispbuf, scache |
| 446 | 33 #else |
| 34 #define RE_TRANSLATE_TYPE char * | |
| 826 | 35 #define RE_LISP_SHORT_CONTEXT_ARGS_DECL |
| 36 #define RE_LISP_SHORT_CONTEXT_ARGS | |
| 37 #define RE_LISP_CONTEXT_ARGS_DECL | |
| 2333 | 38 #define RE_LISP_CONTEXT_ARGS_MULE_DECL |
| 826 | 39 #define RE_LISP_CONTEXT_ARGS |
| 665 | 40 #define Elemcount ssize_t |
| 41 #define Bytecount ssize_t | |
| 446 | 42 #endif /* emacs */ |
| 43 | |
|
5183
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
44 #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
|
45 # 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
|
46 # 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
|
47 # 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
|
48 # else |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
49 # 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
|
50 # 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
|
51 # endif |
|
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 /* emacs */ |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
53 |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
54 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
|
55 |
| 428 | 56 /* POSIX says that <sys/types.h> must be included (by the caller) before |
| 57 <regex.h>. */ | |
| 58 | |
| 59 | |
| 60 /* The following bits are used to determine the regexp syntax we | |
| 61 recognize. The not-set meaning typically corresponds to the syntax | |
| 62 used by Emacs (the exception is RE_INTERVAL, made for historical | |
| 63 reasons). The bits are given in alphabetical order, and the | |
| 64 definitions shifted by one from the previous bit; thus, when we add or | |
| 65 remove a bit, only one other definition need change. */ | |
| 66 typedef unsigned reg_syntax_t; | |
| 67 | |
| 68 /* If this bit is not set, then \ inside a bracket expression is literal. | |
| 69 If set, then such a \ quotes the following character. */ | |
| 70 #define RE_BACKSLASH_ESCAPE_IN_LISTS (1) | |
| 71 | |
| 72 /* If this bit is not set, then + and ? are operators, and \+ and \? are | |
| 73 literals. | |
| 74 If set, then \+ and \? are operators and + and ? are literals. */ | |
| 75 #define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) | |
| 76 | |
| 77 /* If this bit is set, then character classes are supported. They are: | |
| 78 [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], | |
| 79 [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. | |
| 80 If not set, then character classes are not supported. */ | |
| 81 #define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) | |
| 82 | |
| 83 /* If this bit is set, then ^ and $ are always anchors (outside bracket | |
| 84 expressions, of course). | |
| 85 If this bit is not set, then it depends: | |
| 86 ^ is an anchor if it is at the beginning of a regular | |
| 87 expression or after an open-group or an alternation operator; | |
| 88 $ is an anchor if it is at the end of a regular expression, or | |
| 89 before a close-group or an alternation operator. | |
| 90 | |
| 91 This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because | |
| 92 POSIX draft 11.2 says that * etc. in leading positions is undefined. | |
| 93 We already implemented a previous draft which made those constructs | |
| 94 invalid, though, so we haven't changed the code back. */ | |
| 95 #define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) | |
| 96 | |
| 97 /* If this bit is set, then special characters are always special | |
| 98 regardless of where they are in the pattern. | |
| 99 If this bit is not set, then special characters are special only in | |
| 100 some contexts; otherwise they are ordinary. Specifically, | |
| 101 * + ? and intervals are only special when not after the beginning, | |
| 102 open-group, or alternation operator. */ | |
| 103 #define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) | |
| 104 | |
| 105 /* If this bit is set, then *, +, ?, and { cannot be first in an re or | |
| 106 immediately after an alternation or begin-group operator. */ | |
| 107 #define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) | |
| 108 | |
| 109 /* If this bit is set, then . matches newline. | |
| 110 If not set, then it doesn't. */ | |
| 111 #define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) | |
| 112 | |
| 113 /* If this bit is set, then . doesn't match NUL. | |
| 114 If not set, then it does. */ | |
| 115 #define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) | |
| 116 | |
| 117 /* If this bit is set, nonmatching lists [^...] do not match newline. | |
| 118 If not set, they do. */ | |
| 119 #define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) | |
| 120 | |
| 121 /* If this bit is set, either \{...\} or {...} defines an | |
| 122 interval, depending on RE_NO_BK_BRACES. | |
| 123 If not set, \{, \}, {, and } are literals. */ | |
| 124 #define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) | |
| 125 | |
| 126 /* If this bit is set, +, ? and | aren't recognized as operators. | |
| 127 If not set, they are. */ | |
| 128 #define RE_LIMITED_OPS (RE_INTERVALS << 1) | |
| 129 | |
| 130 /* If this bit is set, newline is an alternation operator. | |
| 131 If not set, newline is literal. */ | |
| 132 #define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) | |
| 133 | |
| 134 /* If this bit is set, then `{...}' defines an interval, and \{ and \} | |
| 135 are literals. | |
| 136 If not set, then `\{...\}' defines an interval. */ | |
| 137 #define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) | |
| 138 | |
| 139 /* If this bit is set, (...) defines a group, and \( and \) are literals. | |
| 140 If not set, \(...\) defines a group, and ( and ) are literals. */ | |
| 141 #define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) | |
| 142 | |
| 143 /* If this bit is set, then \<digit> matches <digit>. | |
| 144 If not set, then \<digit> is a back-reference. */ | |
| 145 #define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) | |
| 146 | |
| 147 /* If this bit is set, then | is an alternation operator, and \| is literal. | |
| 148 If not set, then \| is an alternation operator, and | is literal. */ | |
| 149 #define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) | |
| 150 | |
| 151 /* If this bit is set, then an ending range point collating higher | |
| 152 than the starting range point, as in [z-a], is invalid. | |
| 153 If not set, then when ending range point collates higher than the | |
| 154 starting range point, the range is ignored. */ | |
| 155 #define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) | |
| 156 | |
| 157 /* If this bit is not set, allow minimal matching: | |
| 158 - a*? and a+? and a?? perform shortest-possible matching (compare with a* | |
| 159 and a+ and a?, respectively, which perform longest-possible matching) | |
| 160 - other juxtaposing of * + and ? is rejected. | |
| 161 If this bit is set, consecutive * + and ?'s are collapsed in a logical | |
| 162 manner: | |
| 163 - a*? and a+? are the same as a* | |
| 164 - a?? is the same as a? | |
| 165 */ | |
| 166 #define RE_NO_MINIMAL_MATCHING (RE_NO_EMPTY_RANGES << 1) | |
| 167 | |
| 168 /* If this bit is set, succeed as soon as we match the whole pattern, | |
| 169 without further backtracking. */ | |
| 170 #define RE_NO_POSIX_BACKTRACKING (RE_NO_MINIMAL_MATCHING << 1) | |
| 171 | |
| 172 /* If this bit is not set, (?:re) behaves like (re) (or \(?:re\) behaves like | |
| 173 \(re\)) except that the matched string is not registered. */ | |
| 174 #define RE_NO_SHY_GROUPS (RE_NO_POSIX_BACKTRACKING << 1) | |
| 175 | |
| 176 /* If this bit is set, then an unmatched ) is ordinary. | |
| 177 If not set, then an unmatched ) is invalid. */ | |
| 178 #define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_SHY_GROUPS << 1) | |
| 179 | |
| 502 | 180 /* If this bit is set, then \22 will read as a back reference, |
| 181 provided at least 22 non-shy groups have been seen so far. In all | |
| 182 other cases (bit not set, not 22 non-shy groups seen so far), it | |
| 183 reads as a back reference \2 followed by a digit 2. */ | |
| 184 #define RE_NO_MULTI_DIGIT_BK_REFS (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) | |
| 185 | |
| 428 | 186 /* This global variable defines the particular regexp syntax to use (for |
| 187 some interfaces). When a regexp is compiled, the syntax used is | |
| 188 stored in the pattern buffer, so changing this does not affect | |
| 189 already-compiled regexps. */ | |
| 190 extern reg_syntax_t re_syntax_options; | |
| 191 | |
| 192 /* Define combinations of the above bits for the standard possibilities. | |
| 193 (The [[[ comments delimit what gets put into the Texinfo file, so | |
| 194 don't delete them!) */ | |
| 195 /* [[[begin syntaxes]]] */ | |
| 196 #define RE_SYNTAX_EMACS RE_INTERVALS | |
| 197 | |
| 198 #define RE_SYNTAX_AWK \ | |
| 199 (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ | |
| 200 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | |
| 201 | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ | |
| 202 | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_SHY_GROUPS \ | |
| 502 | 203 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 204 |
| 205 #define RE_SYNTAX_POSIX_AWK \ | |
| 206 (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS) | |
| 207 | |
| 208 #define RE_SYNTAX_GREP \ | |
| 209 (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ | |
| 210 | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ | |
| 211 | RE_NEWLINE_ALT | RE_NO_SHY_GROUPS \ | |
| 502 | 212 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 213 |
| 214 #define RE_SYNTAX_EGREP \ | |
| 215 (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ | |
| 216 | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ | |
| 217 | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ | |
| 218 | RE_NO_BK_VBAR | RE_NO_SHY_GROUPS \ | |
| 502 | 219 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 220 |
| 221 #define RE_SYNTAX_POSIX_EGREP \ | |
| 502 | 222 (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES | \ |
| 223 RE_NO_MULTI_DIGIT_BK_REFS) | |
| 428 | 224 |
| 225 /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ | |
| 226 #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC | |
| 227 | |
| 228 #define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC | |
| 229 | |
| 230 /* Syntax bits common to both basic and extended POSIX regex syntax. */ | |
| 231 #define _RE_SYNTAX_POSIX_COMMON \ | |
| 232 (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ | |
| 233 | RE_INTERVALS | RE_NO_EMPTY_RANGES | RE_NO_SHY_GROUPS \ | |
| 502 | 234 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 235 |
| 236 #define RE_SYNTAX_POSIX_BASIC \ | |
| 237 (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM) | |
| 238 | |
| 239 /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes | |
| 240 RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this | |
| 241 isn't minimal, since other operators, such as \`, aren't disabled. */ | |
| 242 #define RE_SYNTAX_POSIX_MINIMAL_BASIC \ | |
| 243 (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) | |
| 244 | |
| 245 #define RE_SYNTAX_POSIX_EXTENDED \ | |
| 246 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ | |
| 247 | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ | |
| 248 | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ | |
| 249 | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
| 250 | |
| 251 /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS | |
| 252 replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added. */ | |
| 253 #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ | |
| 254 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ | |
| 255 | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ | |
| 256 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | |
| 257 | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
| 258 /* [[[end syntaxes]]] */ | |
| 259 | |
| 260 /* Maximum number of duplicates an interval can allow. Some systems | |
| 261 (erroneously) define this in other header files, but we want our | |
| 262 value, so remove any previous define. */ | |
| 263 #ifdef RE_DUP_MAX | |
| 264 #undef RE_DUP_MAX | |
| 265 #endif | |
| 266 #define RE_DUP_MAX ((1 << 15) - 1) | |
| 267 | |
| 268 | |
| 269 /* POSIX `cflags' bits (i.e., information for `regcomp'). */ | |
| 270 | |
| 271 /* If this bit is set, then use extended regular expression syntax. | |
| 272 If not set, then use basic regular expression syntax. */ | |
| 273 #define REG_EXTENDED 1 | |
| 274 | |
| 275 /* If this bit is set, then ignore case when matching. | |
| 276 If not set, then case is significant. */ | |
| 277 #define REG_ICASE (REG_EXTENDED << 1) | |
| 278 | |
| 279 /* If this bit is set, then anchors do not match at newline | |
| 280 characters in the string. | |
| 281 If not set, then anchors do match at newlines. */ | |
| 282 #define REG_NEWLINE (REG_ICASE << 1) | |
| 283 | |
| 284 /* If this bit is set, then report only success or fail in regexec. | |
| 285 If not set, then returns differ between not matching and errors. */ | |
| 286 #define REG_NOSUB (REG_NEWLINE << 1) | |
| 287 | |
| 288 | |
| 289 /* POSIX `eflags' bits (i.e., information for regexec). */ | |
| 290 | |
| 291 /* If this bit is set, then the beginning-of-line operator doesn't match | |
| 292 the beginning of the string (presumably because it's not the | |
| 293 beginning of a line). | |
| 294 If not set, then the beginning-of-line operator does match the | |
| 295 beginning of the string. */ | |
| 296 #define REG_NOTBOL 1 | |
| 297 | |
| 298 /* Like REG_NOTBOL, except for the end-of-line. */ | |
| 299 #define REG_NOTEOL (1 << 1) | |
| 300 | |
| 301 | |
| 302 /* If any error codes are removed, changed, or added, update the | |
| 303 `re_error_msg' table in regex.c. */ | |
| 304 typedef enum | |
| 305 { | |
| 306 REG_NOERROR = 0, /* Success. */ | |
| 307 REG_NOMATCH, /* Didn't find a match (for regexec). */ | |
| 308 | |
| 309 /* POSIX regcomp return error codes. (In the order listed in the | |
| 310 standard.) */ | |
| 311 REG_BADPAT, /* Invalid pattern. */ | |
| 312 REG_ECOLLATE, /* Not implemented. */ | |
| 313 REG_ECTYPE, /* Invalid character class name. */ | |
| 314 REG_EESCAPE, /* Trailing backslash. */ | |
| 315 REG_ESUBREG, /* Invalid back reference. */ | |
| 316 REG_EBRACK, /* Unmatched left bracket. */ | |
| 317 REG_EPAREN, /* Parenthesis imbalance. */ | |
| 318 REG_EBRACE, /* Unmatched \{. */ | |
| 319 REG_BADBR, /* Invalid contents of \{\}. */ | |
| 320 REG_ERANGE, /* Invalid range end. */ | |
| 321 REG_ESPACE, /* Ran out of memory. */ | |
| 322 REG_BADRPT, /* No preceding re for repetition op. */ | |
| 323 | |
| 324 /* Error codes we've added. */ | |
| 325 REG_EEND, /* Premature end. */ | |
| 326 REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ | |
| 327 REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ | |
| 328 #ifdef emacs | |
| 329 ,REG_ESYNTAX /* Invalid syntax designator. */ | |
| 330 #endif | |
| 331 #ifdef MULE | |
| 332 ,REG_ERANGESPAN /* Ranges may not span charsets. */ | |
| 333 ,REG_ECATEGORY /* Invalid category designator */ | |
| 334 #endif | |
| 335 } reg_errcode_t; | |
| 336 | |
| 337 /* This data structure represents a compiled pattern. Before calling | |
| 338 the pattern compiler, the fields `buffer', `allocated', `fastmap', | |
| 339 `translate', and `no_sub' can be set. After the pattern has been | |
| 340 compiled, the `re_nsub' field is available. All other fields are | |
| 341 private to the regex routines. */ | |
| 342 | |
| 343 struct re_pattern_buffer | |
| 344 { | |
| 345 /* [[[begin pattern_buffer]]] */ | |
| 346 /* Space that holds the compiled pattern. It is declared as | |
| 347 `unsigned char *' because its elements are | |
| 348 sometimes used as array indexes. */ | |
| 349 unsigned char *buffer; | |
| 350 | |
| 351 /* Number of bytes to which `buffer' points. */ | |
| 647 | 352 long allocated; |
| 428 | 353 |
| 354 /* Number of bytes actually used in `buffer'. */ | |
| 647 | 355 long used; |
| 428 | 356 |
| 357 /* Syntax setting with which the pattern was compiled. */ | |
| 358 reg_syntax_t syntax; | |
| 359 | |
| 360 /* Pointer to a fastmap, if any, otherwise zero. re_search uses | |
| 361 the fastmap, if there is one, to skip over impossible | |
| 362 starting points for matches. */ | |
| 363 char *fastmap; | |
| 364 | |
| 365 /* Either a translate table to apply to all characters before | |
| 366 comparing them, or zero for no translation. The translation | |
| 367 is applied to a pattern when it is compiled and to a string | |
| 368 when it is matched. */ | |
| 446 | 369 RE_TRANSLATE_TYPE translate; |
| 428 | 370 |
| 2639 | 371 /* Number of subpatterns (returnable groups) found by the compiler. |
| 372 (This does not count shy groups.) */ | |
| 647 | 373 int re_nsub; |
| 428 | 374 |
| 502 | 375 /* Total number of groups found by the compiler. (Including |
| 376 shy ones.) */ | |
| 377 int re_ngroups; | |
| 378 | |
| 428 | 379 /* Zero if this pattern cannot match the empty string, one else. |
| 380 Well, in truth it's used only in `re_search_2', to see | |
| 381 whether or not we should use the fastmap, so we don't set | |
| 382 this absolutely perfectly; see `re_compile_fastmap' (the | |
| 383 `duplicate' case). */ | |
| 647 | 384 unsigned int can_be_null : 1; |
| 428 | 385 |
| 386 /* If REGS_UNALLOCATED, allocate space in the `regs' structure | |
| 387 for `max (RE_NREGS, re_nsub + 1)' groups. | |
| 388 If REGS_REALLOCATE, reallocate space if necessary. | |
| 389 If REGS_FIXED, use what's there. */ | |
| 390 #define REGS_UNALLOCATED 0 | |
| 391 #define REGS_REALLOCATE 1 | |
| 392 #define REGS_FIXED 2 | |
| 647 | 393 unsigned int regs_allocated : 2; |
| 428 | 394 |
| 395 /* Set to zero when `regex_compile' compiles a pattern; set to one | |
| 396 by `re_compile_fastmap' if it updates the fastmap. */ | |
| 647 | 397 unsigned int fastmap_accurate : 1; |
| 428 | 398 |
| 399 /* If set, `re_match_2' does not return information about | |
| 400 subexpressions. */ | |
| 647 | 401 unsigned int no_sub : 1; |
| 428 | 402 |
| 403 /* If set, a beginning-of-line anchor doesn't match at the | |
| 404 beginning of the string. */ | |
| 647 | 405 unsigned int not_bol : 1; |
| 428 | 406 |
| 407 /* Similarly for an end-of-line anchor. */ | |
| 647 | 408 unsigned int not_eol : 1; |
| 428 | 409 |
| 410 /* If true, an anchor at a newline matches. */ | |
| 647 | 411 unsigned int newline_anchor : 1; |
| 428 | 412 |
| 647 | 413 unsigned int warned_about_incompatible_back_references : 1; |
| 502 | 414 |
| 415 /* Mapping between back references and groups (may not be | |
| 416 equivalent with shy groups). */ | |
| 417 int *external_to_internal_register; | |
| 418 | |
| 419 int external_to_internal_register_size; | |
| 420 | |
| 428 | 421 /* [[[end pattern_buffer]]] */ |
| 422 }; | |
| 423 | |
| 424 typedef struct re_pattern_buffer regex_t; | |
| 425 | |
| 426 /* Type for byte offsets within the string. POSIX mandates this. */ | |
| 427 typedef int regoff_t; | |
| 428 | |
| 429 | |
| 430 /* This is the structure we store register match data in. See | |
| 431 regex.texinfo for a full description of what registers match. */ | |
| 432 struct re_registers | |
| 433 { | |
| 1468 | 434 int num_regs; /* number of registers allocated */ |
| 428 | 435 regoff_t *start; |
| 436 regoff_t *end; | |
| 437 }; | |
| 438 | |
| 439 | |
| 440 /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, | |
| 441 `re_match_2' returns information about at least this many registers | |
| 442 the first time a `regs' structure is passed. */ | |
| 443 #ifndef RE_NREGS | |
| 444 #define RE_NREGS 30 | |
| 445 #endif | |
| 446 | |
| 447 | |
| 448 /* POSIX specification for registers. Aside from the different names than | |
| 449 `re_registers', POSIX uses an array of structures, instead of a | |
| 450 structure of arrays. */ | |
| 451 typedef struct | |
| 452 { | |
| 453 regoff_t rm_so; /* Byte offset from string's start to substring's start. */ | |
| 454 regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ | |
| 455 } regmatch_t; | |
| 456 | |
| 457 /* Declarations for routines. */ | |
| 458 | |
| 459 /* Sets the current default syntax to SYNTAX, and return the old syntax. | |
| 460 You can also simply assign to the `re_syntax_options' variable. */ | |
| 461 reg_syntax_t re_set_syntax (reg_syntax_t syntax); | |
| 462 | |
| 463 /* Compile the regular expression PATTERN, with length LENGTH | |
| 464 and syntax given by the global `re_syntax_options', into the buffer | |
| 465 BUFFER. Return NULL if successful, and an error string if not. */ | |
| 442 | 466 const char *re_compile_pattern (const char *pattern, int length, |
| 428 | 467 struct re_pattern_buffer *buffer); |
| 468 | |
| 469 | |
| 470 /* Compile a fastmap for the compiled pattern in BUFFER; used to | |
| 471 accelerate searches. Return 0 if successful and -2 if was an | |
| 472 internal error. */ | |
| 826 | 473 int re_compile_fastmap (struct re_pattern_buffer *buffer |
| 474 RE_LISP_SHORT_CONTEXT_ARGS_DECL); | |
| 428 | 475 |
| 476 | |
| 477 /* Search in the string STRING (with length LENGTH) for the pattern | |
| 478 compiled into BUFFER. Start searching at position START, for RANGE | |
| 479 characters. Return the starting position of the match, -1 for no | |
| 480 match, or -2 for an internal error. Also return register | |
| 481 information in REGS (if REGS and BUFFER->no_sub are nonzero). */ | |
| 442 | 482 int re_search (struct re_pattern_buffer *buffer, const char *string, |
| 428 | 483 int length, int start, int range, |
| 826 | 484 struct re_registers *regs RE_LISP_CONTEXT_ARGS_DECL); |
| 428 | 485 |
| 486 | |
| 487 /* Like `re_search', but search in the concatenation of STRING1 and | |
| 488 STRING2. Also, stop searching at index START + STOP. */ | |
| 442 | 489 int re_search_2 (struct re_pattern_buffer *buffer, const char *string1, |
| 490 int length1, const char *string2, int length2, int start, | |
| 826 | 491 int range, struct re_registers *regs, int stop |
| 492 RE_LISP_CONTEXT_ARGS_DECL); | |
| 428 | 493 |
| 826 | 494 #ifndef emacs /* never used by XEmacs */ |
| 428 | 495 |
| 496 /* Like `re_search', but return how many characters in STRING the regexp | |
| 497 in BUFFER matched, starting at position START. */ | |
| 442 | 498 int re_match (struct re_pattern_buffer *buffer, const char *string, |
| 826 | 499 int length, int start, struct re_registers *regs |
| 500 RE_LISP_CONTEXT_ARGS_DECL); | |
| 428 | 501 |
| 826 | 502 #endif /* not emacs */ |
| 428 | 503 |
| 504 /* Relates to `re_match' as `re_search_2' relates to `re_search'. */ | |
| 442 | 505 int re_match_2 (struct re_pattern_buffer *buffer, const char *string1, |
| 506 int length1, const char *string2, int length2, | |
| 826 | 507 int start, struct re_registers *regs, int stop |
| 508 RE_LISP_CONTEXT_ARGS_DECL); | |
| 428 | 509 |
| 510 /* Set REGS to hold NUM_REGS registers, storing them in STARTS and | |
| 511 ENDS. Subsequent matches using BUFFER and REGS will use this memory | |
| 512 for recording register information. STARTS and ENDS must be | |
| 513 allocated with malloc, and must each be at least `NUM_REGS * sizeof | |
| 514 (regoff_t)' bytes long. | |
| 515 | |
| 516 If NUM_REGS == 0, then subsequent matches should allocate their own | |
| 517 register data. | |
| 518 | |
| 519 Unless this function is called, the first search or match using | |
| 520 PATTERN_BUFFER will allocate its own register data, without | |
| 521 freeing the old data. */ | |
| 522 void re_set_registers (struct re_pattern_buffer *buffer, | |
| 647 | 523 struct re_registers *regs, int num_regs, |
| 428 | 524 regoff_t *starts, regoff_t *ends); |
| 525 | |
| 526 #ifdef _REGEX_RE_COMP | |
| 527 /* 4.2 bsd compatibility. */ | |
| 442 | 528 char *re_comp (const char *); |
| 529 int re_exec (const char *); | |
| 428 | 530 #endif |
| 531 | |
| 532 /* POSIX compatibility. */ | |
| 442 | 533 int regcomp (regex_t *preg, const char *pattern, int cflags); |
| 534 int regexec (const regex_t *preg, const char *string, size_t nmatch, | |
| 428 | 535 regmatch_t pmatch[], int eflags); |
| 442 | 536 size_t regerror (int errcode, const regex_t *preg, char *errbuf, |
| 428 | 537 size_t errbuf_size); |
| 538 void regfree (regex_t *preg); | |
| 539 | |
| 5041 | 540 enum regex_debug |
| 541 { | |
| 542 RE_DEBUG_COMPILATION = 1 << 0, | |
| 543 RE_DEBUG_FAILURE_POINT = 1 << 1, | |
| 544 RE_DEBUG_MATCHING = 1 << 2, | |
| 545 }; | |
| 546 | |
| 547 extern int debug_regexps; | |
| 548 | |
|
5183
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
549 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
|
550 |
| 440 | 551 #endif /* INCLUDED_regex_h_ */ |