Mercurial > hg > octave-shane > gnulib-hg
annotate lib/regex.h @ 4554:5a1e7c84e9cd
Use three spaces, rather than tab, after '#'
in shell-script copyright notices.
Suggested by Bruno Haible.
| author | Paul Eggert <eggert@cs.ucla.edu> |
|---|---|
| date | Fri, 15 Aug 2003 18:04:34 +0000 |
| parents | 2d87db0b5530 |
| children | 02efc69861ae |
| rev | line source |
|---|---|
| 14 | 1 /* Definitions for data structures and routines for the regular |
| 4020 | 2 expression library. |
| 4304 | 3 Copyright (C) 1985,1989-93,1995-98,2000,2001,2002,2003 |
| 4020 | 4 Free Software Foundation, Inc. |
| 5 This file is part of the GNU C Library. | |
| 1283 | 6 |
| 4020 | 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. | |
| 14 | 11 |
| 4020 | 12 This program is distributed in the hope that it will be useful, |
| 1283 | 13 but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 4020 | 14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 15 GNU General Public License for more details. | |
| 14 | 16 |
| 4020 | 17 You should have received a copy of the GNU General Public License along |
| 18 with this program; if not, write to the Free Software Foundation, | |
| 19 Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ | |
| 14 | 20 |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
21 #ifndef _REGEX_H |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
22 #define _REGEX_H 1 |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
23 |
| 4304 | 24 #include <sys/types.h> |
| 25 | |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
26 /* Allow the use in C++ code. */ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
27 #ifdef __cplusplus |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
28 extern "C" { |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
29 #endif |
| 14 | 30 |
| 389 | 31 /* POSIX says that <sys/types.h> must be included (by the caller) before |
| 32 <regex.h>. */ | |
| 33 | |
| 1283 | 34 #if !defined _POSIX_C_SOURCE && !defined _POSIX_SOURCE && defined VMS |
| 389 | 35 /* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it |
| 36 should be there. */ | |
| 1283 | 37 # include <stddef.h> |
| 389 | 38 #endif |
| 39 | |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
40 /* The following two types have to be signed and unsigned integer type |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
41 wide enough to hold a value of a pointer. For most ANSI compilers |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
42 ptrdiff_t and size_t should be likely OK. Still size of these two |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
43 types is 2 for Microsoft C. Ugh... */ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
44 typedef long int s_reg_t; |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
45 typedef unsigned long int active_reg_t; |
| 14 | 46 |
| 47 /* The following bits are used to determine the regexp syntax we | |
| 48 recognize. The set/not-set meanings are chosen so that Emacs syntax | |
| 49 remains the value 0. The bits are given in alphabetical order, and | |
| 50 the definitions shifted by one from the previous bit; thus, when we | |
| 51 add or remove a bit, only one other definition need change. */ | |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
52 typedef unsigned long int reg_syntax_t; |
| 14 | 53 |
| 54 /* If this bit is not set, then \ inside a bracket expression is literal. | |
| 55 If set, then such a \ quotes the following character. */ | |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
56 #define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1) |
| 14 | 57 |
| 58 /* If this bit is not set, then + and ? are operators, and \+ and \? are | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
59 literals. |
| 14 | 60 If set, then \+ and \? are operators and + and ? are literals. */ |
| 61 #define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) | |
| 62 | |
| 63 /* If this bit is set, then character classes are supported. They are: | |
| 64 [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], | |
| 65 [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. | |
| 66 If not set, then character classes are not supported. */ | |
| 67 #define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) | |
| 68 | |
| 69 /* If this bit is set, then ^ and $ are always anchors (outside bracket | |
| 70 expressions, of course). | |
| 71 If this bit is not set, then it depends: | |
| 72 ^ is an anchor if it is at the beginning of a regular | |
| 73 expression or after an open-group or an alternation operator; | |
| 74 $ is an anchor if it is at the end of a regular expression, or | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
75 before a close-group or an alternation operator. |
| 14 | 76 |
| 77 This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because | |
| 78 POSIX draft 11.2 says that * etc. in leading positions is undefined. | |
| 79 We already implemented a previous draft which made those constructs | |
| 80 invalid, though, so we haven't changed the code back. */ | |
| 81 #define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) | |
| 82 | |
| 83 /* If this bit is set, then special characters are always special | |
| 84 regardless of where they are in the pattern. | |
| 85 If this bit is not set, then special characters are special only in | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
86 some contexts; otherwise they are ordinary. Specifically, |
| 14 | 87 * + ? and intervals are only special when not after the beginning, |
| 88 open-group, or alternation operator. */ | |
| 89 #define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) | |
| 90 | |
| 91 /* If this bit is set, then *, +, ?, and { cannot be first in an re or | |
| 92 immediately after an alternation or begin-group operator. */ | |
| 93 #define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) | |
| 94 | |
| 95 /* If this bit is set, then . matches newline. | |
| 96 If not set, then it doesn't. */ | |
| 97 #define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) | |
| 98 | |
| 99 /* If this bit is set, then . doesn't match NUL. | |
| 100 If not set, then it does. */ | |
| 101 #define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) | |
| 102 | |
| 103 /* If this bit is set, nonmatching lists [^...] do not match newline. | |
| 104 If not set, they do. */ | |
| 105 #define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) | |
| 106 | |
| 107 /* If this bit is set, either \{...\} or {...} defines an | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
108 interval, depending on RE_NO_BK_BRACES. |
| 14 | 109 If not set, \{, \}, {, and } are literals. */ |
| 110 #define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) | |
| 111 | |
| 112 /* If this bit is set, +, ? and | aren't recognized as operators. | |
| 113 If not set, they are. */ | |
| 114 #define RE_LIMITED_OPS (RE_INTERVALS << 1) | |
| 115 | |
| 116 /* If this bit is set, newline is an alternation operator. | |
| 117 If not set, newline is literal. */ | |
| 118 #define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) | |
| 119 | |
| 120 /* If this bit is set, then `{...}' defines an interval, and \{ and \} | |
| 121 are literals. | |
| 122 If not set, then `\{...\}' defines an interval. */ | |
| 123 #define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) | |
| 124 | |
| 125 /* If this bit is set, (...) defines a group, and \( and \) are literals. | |
| 126 If not set, \(...\) defines a group, and ( and ) are literals. */ | |
| 127 #define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) | |
| 128 | |
| 129 /* If this bit is set, then \<digit> matches <digit>. | |
| 130 If not set, then \<digit> is a back-reference. */ | |
| 131 #define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) | |
| 132 | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
133 /* If this bit is set, then | is an alternation operator, and \| is literal. |
| 14 | 134 If not set, then \| is an alternation operator, and | is literal. */ |
| 135 #define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) | |
| 136 | |
| 137 /* If this bit is set, then an ending range point collating higher | |
| 138 than the starting range point, as in [z-a], is invalid. | |
| 139 If not set, then when ending range point collates higher than the | |
| 140 starting range point, the range is ignored. */ | |
| 141 #define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) | |
| 142 | |
| 143 /* If this bit is set, then an unmatched ) is ordinary. | |
| 144 If not set, then an unmatched ) is invalid. */ | |
| 145 #define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) | |
| 146 | |
| 389 | 147 /* If this bit is set, succeed as soon as we match the whole pattern, |
| 148 without further backtracking. */ | |
| 149 #define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) | |
| 150 | |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
151 /* If this bit is set, do not process the GNU regex operators. |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
152 If not set, then the GNU regex operators are recognized. */ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
153 #define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1) |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
154 |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
155 /* If this bit is set, turn on internal regex debugging. |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
156 If not set, and debugging was on, turn it off. |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
157 This only works if regex.c is compiled -DDEBUG. |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
158 We define this bit always, so that all that's needed to turn on |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
159 debugging is to recompile regex.c; the calling code can always have |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
160 this bit set, and it won't affect anything in the normal case. */ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
161 #define RE_DEBUG (RE_NO_GNU_OPS << 1) |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
162 |
| 3227 | 163 /* If this bit is set, a syntactically invalid interval is treated as |
| 164 a string of ordinary characters. For example, the ERE 'a{1' is | |
| 165 treated as 'a\{1'. */ | |
| 166 #define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1) | |
| 167 | |
| 4020 | 168 /* If this bit is set, then ignore case when matching. |
| 169 If not set, then case is significant. */ | |
| 170 #define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1) | |
| 171 | |
| 14 | 172 /* This global variable defines the particular regexp syntax to use (for |
| 173 some interfaces). When a regexp is compiled, the syntax used is | |
| 174 stored in the pattern buffer, so changing this does not affect | |
| 175 already-compiled regexps. */ | |
| 176 extern reg_syntax_t re_syntax_options; | |
| 177 | |
| 178 /* Define combinations of the above bits for the standard possibilities. | |
| 179 (The [[[ comments delimit what gets put into the Texinfo file, so | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
180 don't delete them!) */ |
| 14 | 181 /* [[[begin syntaxes]]] */ |
| 182 #define RE_SYNTAX_EMACS 0 | |
| 183 | |
| 184 #define RE_SYNTAX_AWK \ | |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
185 (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
186 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
187 | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
188 | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
189 | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS) |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
190 |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
191 #define RE_SYNTAX_GNU_AWK \ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
192 ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \ |
| 4020 | 193 & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS \ |
| 194 | RE_CONTEXT_INVALID_OPS )) | |
| 14 | 195 |
|
4529
2d87db0b5530
White space fixes from coreutils.
Paul Eggert <eggert@cs.ucla.edu>
parents:
4521
diff
changeset
|
196 #define RE_SYNTAX_POSIX_AWK \ |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
197 (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
198 | RE_INTERVALS | RE_NO_GNU_OPS) |
| 14 | 199 |
| 200 #define RE_SYNTAX_GREP \ | |
| 201 (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ | |
| 202 | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ | |
| 203 | RE_NEWLINE_ALT) | |
| 204 | |
| 205 #define RE_SYNTAX_EGREP \ | |
| 206 (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ | |
| 207 | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ | |
| 208 | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ | |
| 209 | RE_NO_BK_VBAR) | |
| 210 | |
| 211 #define RE_SYNTAX_POSIX_EGREP \ | |
| 3227 | 212 (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES \ |
| 213 | RE_INVALID_INTERVAL_ORD) | |
| 14 | 214 |
| 389 | 215 /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ |
| 216 #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC | |
| 217 | |
| 14 | 218 #define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC |
| 219 | |
| 220 /* Syntax bits common to both basic and extended POSIX regex syntax. */ | |
| 221 #define _RE_SYNTAX_POSIX_COMMON \ | |
| 222 (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ | |
| 223 | RE_INTERVALS | RE_NO_EMPTY_RANGES) | |
| 224 | |
| 225 #define RE_SYNTAX_POSIX_BASIC \ | |
| 226 (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM) | |
| 227 | |
| 228 /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes | |
| 229 RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this | |
| 230 isn't minimal, since other operators, such as \`, aren't disabled. */ | |
| 231 #define RE_SYNTAX_POSIX_MINIMAL_BASIC \ | |
| 232 (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) | |
| 233 | |
| 234 #define RE_SYNTAX_POSIX_EXTENDED \ | |
| 2472 | 235 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ |
| 236 | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ | |
| 237 | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ | |
| 238 | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
| 14 | 239 |
| 2472 | 240 /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is |
| 241 removed and RE_NO_BK_REFS is added. */ | |
| 14 | 242 #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ |
| 243 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ | |
| 244 | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ | |
| 245 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | |
| 246 | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
| 247 /* [[[end syntaxes]]] */ | |
| 248 | |
| 249 /* Maximum number of duplicates an interval can allow. Some systems | |
| 250 (erroneously) define this in other header files, but we want our | |
| 251 value, so remove any previous define. */ | |
| 252 #ifdef RE_DUP_MAX | |
| 1283 | 253 # undef RE_DUP_MAX |
| 14 | 254 #endif |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
255 /* If sizeof(int) == 2, then ((1 << 15) - 1) overflows. */ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
256 #define RE_DUP_MAX (0x7fff) |
| 14 | 257 |
| 258 | |
| 259 /* POSIX `cflags' bits (i.e., information for `regcomp'). */ | |
| 260 | |
| 261 /* If this bit is set, then use extended regular expression syntax. | |
| 262 If not set, then use basic regular expression syntax. */ | |
| 263 #define REG_EXTENDED 1 | |
| 264 | |
| 265 /* If this bit is set, then ignore case when matching. | |
| 266 If not set, then case is significant. */ | |
| 267 #define REG_ICASE (REG_EXTENDED << 1) | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
268 |
| 14 | 269 /* If this bit is set, then anchors do not match at newline |
| 270 characters in the string. | |
| 271 If not set, then anchors do match at newlines. */ | |
| 272 #define REG_NEWLINE (REG_ICASE << 1) | |
| 273 | |
| 274 /* If this bit is set, then report only success or fail in regexec. | |
| 275 If not set, then returns differ between not matching and errors. */ | |
| 276 #define REG_NOSUB (REG_NEWLINE << 1) | |
| 277 | |
| 278 | |
| 279 /* POSIX `eflags' bits (i.e., information for regexec). */ | |
| 280 | |
| 281 /* If this bit is set, then the beginning-of-line operator doesn't match | |
| 282 the beginning of the string (presumably because it's not the | |
| 283 beginning of a line). | |
| 284 If not set, then the beginning-of-line operator does match the | |
| 285 beginning of the string. */ | |
| 286 #define REG_NOTBOL 1 | |
| 287 | |
| 288 /* Like REG_NOTBOL, except for the end-of-line. */ | |
| 289 #define REG_NOTEOL (1 << 1) | |
| 290 | |
| 291 | |
| 292 /* If any error codes are removed, changed, or added, update the | |
| 293 `re_error_msg' table in regex.c. */ | |
| 294 typedef enum | |
| 295 { | |
| 1639 | 296 #ifdef _XOPEN_SOURCE |
| 297 REG_ENOSYS = -1, /* This will never happen for this implementation. */ | |
| 1283 | 298 #endif |
| 299 | |
| 14 | 300 REG_NOERROR = 0, /* Success. */ |
| 301 REG_NOMATCH, /* Didn't find a match (for regexec). */ | |
| 302 | |
| 303 /* POSIX regcomp return error codes. (In the order listed in the | |
| 304 standard.) */ | |
| 305 REG_BADPAT, /* Invalid pattern. */ | |
| 306 REG_ECOLLATE, /* Not implemented. */ | |
| 307 REG_ECTYPE, /* Invalid character class name. */ | |
| 308 REG_EESCAPE, /* Trailing backslash. */ | |
| 309 REG_ESUBREG, /* Invalid back reference. */ | |
| 310 REG_EBRACK, /* Unmatched left bracket. */ | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
311 REG_EPAREN, /* Parenthesis imbalance. */ |
| 14 | 312 REG_EBRACE, /* Unmatched \{. */ |
| 313 REG_BADBR, /* Invalid contents of \{\}. */ | |
| 314 REG_ERANGE, /* Invalid range end. */ | |
| 315 REG_ESPACE, /* Ran out of memory. */ | |
| 316 REG_BADRPT, /* No preceding re for repetition op. */ | |
| 317 | |
| 318 /* Error codes we've added. */ | |
| 319 REG_EEND, /* Premature end. */ | |
| 320 REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ | |
| 321 REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ | |
| 322 } reg_errcode_t; | |
| 323 | |
| 324 /* This data structure represents a compiled pattern. Before calling | |
| 325 the pattern compiler, the fields `buffer', `allocated', `fastmap', | |
| 326 `translate', and `no_sub' can be set. After the pattern has been | |
| 327 compiled, the `re_nsub' field is available. All other fields are | |
| 328 private to the regex routines. */ | |
| 329 | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
330 #ifndef RE_TRANSLATE_TYPE |
| 1283 | 331 # define RE_TRANSLATE_TYPE char * |
| 502 | 332 #endif |
| 333 | |
| 14 | 334 struct re_pattern_buffer |
| 335 { | |
| 336 /* [[[begin pattern_buffer]]] */ | |
| 337 /* Space that holds the compiled pattern. It is declared as | |
| 338 `unsigned char *' because its elements are | |
| 339 sometimes used as array indexes. */ | |
| 340 unsigned char *buffer; | |
| 341 | |
| 342 /* Number of bytes to which `buffer' points. */ | |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
343 unsigned long int allocated; |
| 14 | 344 |
| 345 /* Number of bytes actually used in `buffer'. */ | |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
346 unsigned long int used; |
| 14 | 347 |
| 348 /* Syntax setting with which the pattern was compiled. */ | |
| 349 reg_syntax_t syntax; | |
| 350 | |
| 351 /* Pointer to a fastmap, if any, otherwise zero. re_search uses | |
| 352 the fastmap, if there is one, to skip over impossible | |
| 353 starting points for matches. */ | |
| 354 char *fastmap; | |
| 355 | |
| 356 /* Either a translate table to apply to all characters before | |
| 357 comparing them, or zero for no translation. The translation | |
| 358 is applied to a pattern when it is compiled and to a string | |
| 359 when it is matched. */ | |
| 502 | 360 RE_TRANSLATE_TYPE translate; |
| 14 | 361 |
| 362 /* Number of subexpressions found by the compiler. */ | |
| 363 size_t re_nsub; | |
| 364 | |
| 365 /* Zero if this pattern cannot match the empty string, one else. | |
| 366 Well, in truth it's used only in `re_search_2', to see | |
| 367 whether or not we should use the fastmap, so we don't set | |
| 368 this absolutely perfectly; see `re_compile_fastmap' (the | |
| 369 `duplicate' case). */ | |
| 370 unsigned can_be_null : 1; | |
| 371 | |
| 372 /* If REGS_UNALLOCATED, allocate space in the `regs' structure | |
| 373 for `max (RE_NREGS, re_nsub + 1)' groups. | |
| 374 If REGS_REALLOCATE, reallocate space if necessary. | |
| 375 If REGS_FIXED, use what's there. */ | |
| 376 #define REGS_UNALLOCATED 0 | |
| 377 #define REGS_REALLOCATE 1 | |
| 378 #define REGS_FIXED 2 | |
| 379 unsigned regs_allocated : 2; | |
| 380 | |
| 389 | 381 /* Set to zero when `regex_compile' compiles a pattern; set to one |
| 382 by `re_compile_fastmap' if it updates the fastmap. */ | |
| 14 | 383 unsigned fastmap_accurate : 1; |
| 384 | |
| 389 | 385 /* If set, `re_match_2' does not return information about |
| 386 subexpressions. */ | |
| 14 | 387 unsigned no_sub : 1; |
| 388 | |
| 389 /* If set, a beginning-of-line anchor doesn't match at the | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
390 beginning of the string. */ |
| 14 | 391 unsigned not_bol : 1; |
| 392 | |
| 393 /* Similarly for an end-of-line anchor. */ | |
| 394 unsigned not_eol : 1; | |
| 395 | |
| 396 /* If true, an anchor at a newline matches. */ | |
| 397 unsigned newline_anchor : 1; | |
| 398 | |
| 399 /* [[[end pattern_buffer]]] */ | |
| 400 }; | |
| 401 | |
| 402 typedef struct re_pattern_buffer regex_t; | |
| 403 | |
| 404 /* Type for byte offsets within the string. POSIX mandates this. */ | |
| 405 typedef int regoff_t; | |
| 406 | |
| 407 | |
| 408 /* This is the structure we store register match data in. See | |
| 409 regex.texinfo for a full description of what registers match. */ | |
| 410 struct re_registers | |
| 411 { | |
| 412 unsigned num_regs; | |
| 413 regoff_t *start; | |
| 414 regoff_t *end; | |
| 415 }; | |
| 416 | |
| 417 | |
| 418 /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, | |
| 419 `re_match_2' returns information about at least this many registers | |
| 420 the first time a `regs' structure is passed. */ | |
| 421 #ifndef RE_NREGS | |
| 1283 | 422 # define RE_NREGS 30 |
| 14 | 423 #endif |
| 424 | |
| 425 | |
| 426 /* POSIX specification for registers. Aside from the different names than | |
| 427 `re_registers', POSIX uses an array of structures, instead of a | |
| 428 structure of arrays. */ | |
| 429 typedef struct | |
| 430 { | |
| 431 regoff_t rm_so; /* Byte offset from string's start to substring's start. */ | |
| 432 regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ | |
| 433 } regmatch_t; | |
| 434 | |
| 435 /* Declarations for routines. */ | |
| 436 | |
| 437 /* To avoid duplicating every routine declaration -- once with a | |
| 438 prototype (if we are ANSI), and once without (if we aren't) -- we | |
| 439 use the following macro to declare argument types. This | |
| 440 unfortunately clutters up the declarations a bit, but I think it's | |
| 389 | 441 worth it. */ |
| 14 | 442 |
| 443 #if __STDC__ | |
| 389 | 444 |
| 1283 | 445 # define _RE_ARGS(args) args |
| 389 | 446 |
| 447 #else /* not __STDC__ */ | |
| 448 | |
| 1283 | 449 # define _RE_ARGS(args) () |
| 389 | 450 |
| 451 #endif /* not __STDC__ */ | |
| 14 | 452 |
| 453 /* Sets the current default syntax to SYNTAX, and return the old syntax. | |
| 454 You can also simply assign to the `re_syntax_options' variable. */ | |
| 455 extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax)); | |
| 456 | |
| 457 /* Compile the regular expression PATTERN, with length LENGTH | |
| 458 and syntax given by the global `re_syntax_options', into the buffer | |
| 459 BUFFER. Return NULL if successful, and an error string if not. */ | |
| 460 extern const char *re_compile_pattern | |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
461 _RE_ARGS ((const char *pattern, size_t length, |
| 14 | 462 struct re_pattern_buffer *buffer)); |
| 463 | |
| 464 | |
| 465 /* Compile a fastmap for the compiled pattern in BUFFER; used to | |
| 466 accelerate searches. Return 0 if successful and -2 if was an | |
| 467 internal error. */ | |
| 468 extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer)); | |
| 469 | |
| 470 | |
| 471 /* Search in the string STRING (with length LENGTH) for the pattern | |
| 472 compiled into BUFFER. Start searching at position START, for RANGE | |
| 473 characters. Return the starting position of the match, -1 for no | |
| 474 match, or -2 for an internal error. Also return register | |
| 475 information in REGS (if REGS and BUFFER->no_sub are nonzero). */ | |
| 476 extern int re_search | |
| 477 _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, | |
| 478 int length, int start, int range, struct re_registers *regs)); | |
| 479 | |
| 480 | |
| 481 /* Like `re_search', but search in the concatenation of STRING1 and | |
| 482 STRING2. Also, stop searching at index START + STOP. */ | |
| 483 extern int re_search_2 | |
| 484 _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, | |
| 485 int length1, const char *string2, int length2, | |
| 486 int start, int range, struct re_registers *regs, int stop)); | |
| 487 | |
| 488 | |
| 489 /* Like `re_search', but return how many characters in STRING the regexp | |
| 490 in BUFFER matched, starting at position START. */ | |
| 491 extern int re_match | |
| 492 _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, | |
| 493 int length, int start, struct re_registers *regs)); | |
| 494 | |
| 495 | |
| 496 /* Relates to `re_match' as `re_search_2' relates to `re_search'. */ | |
|
649
89f4c1937ac7
update FSF address in copyright and remove any trailing blanks
Jim Meyering <jim@meyering.net>
parents:
502
diff
changeset
|
497 extern int re_match_2 |
| 14 | 498 _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, |
| 499 int length1, const char *string2, int length2, | |
| 500 int start, struct re_registers *regs, int stop)); | |
| 501 | |
| 502 | |
| 503 /* Set REGS to hold NUM_REGS registers, storing them in STARTS and | |
| 504 ENDS. Subsequent matches using BUFFER and REGS will use this memory | |
| 505 for recording register information. STARTS and ENDS must be | |
| 506 allocated with malloc, and must each be at least `NUM_REGS * sizeof | |
| 507 (regoff_t)' bytes long. | |
| 508 | |
| 509 If NUM_REGS == 0, then subsequent matches should allocate their own | |
| 510 register data. | |
| 511 | |
| 512 Unless this function is called, the first search or match using | |
| 513 PATTERN_BUFFER will allocate its own register data, without | |
| 514 freeing the old data. */ | |
| 515 extern void re_set_registers | |
| 516 _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs, | |
| 517 unsigned num_regs, regoff_t *starts, regoff_t *ends)); | |
| 518 | |
| 1639 | 519 #if defined _REGEX_RE_COMP || defined _LIBC |
| 1283 | 520 # ifndef _CRAY |
| 14 | 521 /* 4.2 bsd compatibility. */ |
| 522 extern char *re_comp _RE_ARGS ((const char *)); | |
| 523 extern int re_exec _RE_ARGS ((const char *)); | |
| 1283 | 524 # endif |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
525 #endif |
| 14 | 526 |
| 2927 | 527 /* GCC 2.95 and later have "__restrict"; C99 compilers have |
| 528 "restrict", and "configure" may have defined "restrict". */ | |
| 529 #ifndef __restrict | |
| 530 # if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__)) | |
| 531 # if defined restrict || 199901L <= __STDC_VERSION__ | |
| 532 # define __restrict restrict | |
| 533 # else | |
| 534 # define __restrict | |
| 535 # endif | |
| 536 # endif | |
|
2940
c7143052abca
(__restrict_arr): Move definition out of #ifndef block.
Jim Meyering <jim@meyering.net>
parents:
2927
diff
changeset
|
537 #endif |
|
3684
c8ed84e015e6
(__restrict_arr): Update from libc.
Jim Meyering <jim@meyering.net>
parents:
3336
diff
changeset
|
538 /* gcc 3.1 and up support the [restrict] syntax. */ |
|
c8ed84e015e6
(__restrict_arr): Update from libc.
Jim Meyering <jim@meyering.net>
parents:
3336
diff
changeset
|
539 #ifndef __restrict_arr |
|
c8ed84e015e6
(__restrict_arr): Update from libc.
Jim Meyering <jim@meyering.net>
parents:
3336
diff
changeset
|
540 # if __GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1) |
|
c8ed84e015e6
(__restrict_arr): Update from libc.
Jim Meyering <jim@meyering.net>
parents:
3336
diff
changeset
|
541 # define __restrict_arr __restrict |
|
c8ed84e015e6
(__restrict_arr): Update from libc.
Jim Meyering <jim@meyering.net>
parents:
3336
diff
changeset
|
542 # else |
|
c8ed84e015e6
(__restrict_arr): Update from libc.
Jim Meyering <jim@meyering.net>
parents:
3336
diff
changeset
|
543 # define __restrict_arr |
|
c8ed84e015e6
(__restrict_arr): Update from libc.
Jim Meyering <jim@meyering.net>
parents:
3336
diff
changeset
|
544 # endif |
|
c8ed84e015e6
(__restrict_arr): Update from libc.
Jim Meyering <jim@meyering.net>
parents:
3336
diff
changeset
|
545 #endif |
| 2927 | 546 |
| 14 | 547 /* POSIX compatibility. */ |
| 2927 | 548 extern int regcomp _RE_ARGS ((regex_t *__restrict __preg, |
| 549 const char *__restrict __pattern, | |
| 1466 | 550 int __cflags)); |
| 551 | |
| 2927 | 552 extern int regexec _RE_ARGS ((const regex_t *__restrict __preg, |
| 553 const char *__restrict __string, size_t __nmatch, | |
| 554 regmatch_t __pmatch[__restrict_arr], | |
| 555 int __eflags)); | |
| 1466 | 556 |
| 557 extern size_t regerror _RE_ARGS ((int __errcode, const regex_t *__preg, | |
| 558 char *__errbuf, size_t __errbuf_size)); | |
| 559 | |
| 560 extern void regfree _RE_ARGS ((regex_t *__preg)); | |
| 14 | 561 |
|
988
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
562 |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
563 #ifdef __cplusplus |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
564 } |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
565 #endif /* C++ */ |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
566 |
|
ba13cef00d11
replace with new version from libc
Jim Meyering <jim@meyering.net>
parents:
649
diff
changeset
|
567 #endif /* regex.h */ |
| 14 | 568 |
| 569 /* | |
| 570 Local variables: | |
| 571 make-backup-files: t | |
| 572 version-control: t | |
| 573 trim-versions-without-asking: nil | |
| 574 End: | |
| 575 */ |