Mercurial > hg > xemacs-beta
annotate src/regex.h @ 5925:08cfc8f77fb6 cygwin
make space for long ptr, and store as such, for frame in WINDOW data,
add a bit more debugging to debug-mswindow,
Vin Shelton patch to fix M-x shell
| author | Henry Thompson <ht@markup.co.uk> |
|---|---|
| date | Fri, 27 Feb 2015 17:41:20 +0000 |
| parents | 3df910176b6a |
| children |
| 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 |
|
5653
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
33 #define RE_ISWCTYPE_ARG_DECL , struct buffer *lispbuf |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
34 #define RE_ISWCTYPE_ARG(varname) , varname |
| 446 | 35 #else |
| 36 #define RE_TRANSLATE_TYPE char * | |
| 826 | 37 #define RE_LISP_SHORT_CONTEXT_ARGS_DECL |
| 38 #define RE_LISP_SHORT_CONTEXT_ARGS | |
| 39 #define RE_LISP_CONTEXT_ARGS_DECL | |
| 2333 | 40 #define RE_LISP_CONTEXT_ARGS_MULE_DECL |
| 826 | 41 #define RE_LISP_CONTEXT_ARGS |
|
5653
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
42 #define RE_ISWCTYPE_ARG_DECL |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
43 #define RE_ISWCTYPE_ARG(varname) |
| 665 | 44 #define Elemcount ssize_t |
| 45 #define Bytecount ssize_t | |
| 446 | 46 #endif /* emacs */ |
| 47 | |
|
5183
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
48 #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
|
49 # 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
|
50 # 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
|
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 # else |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
53 # 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
|
54 # 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
|
55 # endif |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
56 #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
|
57 |
|
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
58 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
|
59 |
| 428 | 60 /* POSIX says that <sys/types.h> must be included (by the caller) before |
| 61 <regex.h>. */ | |
| 62 | |
| 63 | |
| 64 /* The following bits are used to determine the regexp syntax we | |
| 65 recognize. The not-set meaning typically corresponds to the syntax | |
| 66 used by Emacs (the exception is RE_INTERVAL, made for historical | |
| 67 reasons). The bits are given in alphabetical order, and the | |
| 68 definitions shifted by one from the previous bit; thus, when we add or | |
| 69 remove a bit, only one other definition need change. */ | |
| 70 typedef unsigned reg_syntax_t; | |
| 71 | |
| 72 /* If this bit is not set, then \ inside a bracket expression is literal. | |
| 73 If set, then such a \ quotes the following character. */ | |
| 74 #define RE_BACKSLASH_ESCAPE_IN_LISTS (1) | |
| 75 | |
| 76 /* If this bit is not set, then + and ? are operators, and \+ and \? are | |
| 77 literals. | |
| 78 If set, then \+ and \? are operators and + and ? are literals. */ | |
| 79 #define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) | |
| 80 | |
| 81 /* If this bit is set, then character classes are supported. They are: | |
| 82 [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], | |
| 83 [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. | |
| 84 If not set, then character classes are not supported. */ | |
| 85 #define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) | |
| 86 | |
| 87 /* If this bit is set, then ^ and $ are always anchors (outside bracket | |
| 88 expressions, of course). | |
| 89 If this bit is not set, then it depends: | |
| 90 ^ is an anchor if it is at the beginning of a regular | |
| 91 expression or after an open-group or an alternation operator; | |
| 92 $ is an anchor if it is at the end of a regular expression, or | |
| 93 before a close-group or an alternation operator. | |
| 94 | |
| 95 This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because | |
| 96 POSIX draft 11.2 says that * etc. in leading positions is undefined. | |
| 97 We already implemented a previous draft which made those constructs | |
| 98 invalid, though, so we haven't changed the code back. */ | |
| 99 #define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) | |
| 100 | |
| 101 /* If this bit is set, then special characters are always special | |
| 102 regardless of where they are in the pattern. | |
| 103 If this bit is not set, then special characters are special only in | |
| 104 some contexts; otherwise they are ordinary. Specifically, | |
| 105 * + ? and intervals are only special when not after the beginning, | |
| 106 open-group, or alternation operator. */ | |
| 107 #define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) | |
| 108 | |
| 109 /* If this bit is set, then *, +, ?, and { cannot be first in an re or | |
| 110 immediately after an alternation or begin-group operator. */ | |
| 111 #define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) | |
| 112 | |
| 113 /* If this bit is set, then . matches newline. | |
| 114 If not set, then it doesn't. */ | |
| 115 #define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) | |
| 116 | |
| 117 /* If this bit is set, then . doesn't match NUL. | |
| 118 If not set, then it does. */ | |
| 119 #define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) | |
| 120 | |
| 121 /* If this bit is set, nonmatching lists [^...] do not match newline. | |
| 122 If not set, they do. */ | |
| 123 #define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) | |
| 124 | |
| 125 /* If this bit is set, either \{...\} or {...} defines an | |
| 126 interval, depending on RE_NO_BK_BRACES. | |
| 127 If not set, \{, \}, {, and } are literals. */ | |
| 128 #define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) | |
| 129 | |
| 130 /* If this bit is set, +, ? and | aren't recognized as operators. | |
| 131 If not set, they are. */ | |
| 132 #define RE_LIMITED_OPS (RE_INTERVALS << 1) | |
| 133 | |
| 134 /* If this bit is set, newline is an alternation operator. | |
| 135 If not set, newline is literal. */ | |
| 136 #define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) | |
| 137 | |
| 138 /* If this bit is set, then `{...}' defines an interval, and \{ and \} | |
| 139 are literals. | |
| 140 If not set, then `\{...\}' defines an interval. */ | |
| 141 #define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) | |
| 142 | |
| 143 /* If this bit is set, (...) defines a group, and \( and \) are literals. | |
| 144 If not set, \(...\) defines a group, and ( and ) are literals. */ | |
| 145 #define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) | |
| 146 | |
| 147 /* If this bit is set, then \<digit> matches <digit>. | |
| 148 If not set, then \<digit> is a back-reference. */ | |
| 149 #define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) | |
| 150 | |
| 151 /* If this bit is set, then | is an alternation operator, and \| is literal. | |
| 152 If not set, then \| is an alternation operator, and | is literal. */ | |
| 153 #define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) | |
| 154 | |
| 155 /* If this bit is set, then an ending range point collating higher | |
| 156 than the starting range point, as in [z-a], is invalid. | |
| 157 If not set, then when ending range point collates higher than the | |
| 158 starting range point, the range is ignored. */ | |
| 159 #define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) | |
| 160 | |
| 161 /* If this bit is not set, allow minimal matching: | |
| 162 - a*? and a+? and a?? perform shortest-possible matching (compare with a* | |
| 163 and a+ and a?, respectively, which perform longest-possible matching) | |
| 164 - other juxtaposing of * + and ? is rejected. | |
| 165 If this bit is set, consecutive * + and ?'s are collapsed in a logical | |
| 166 manner: | |
| 167 - a*? and a+? are the same as a* | |
| 168 - a?? is the same as a? | |
| 169 */ | |
| 170 #define RE_NO_MINIMAL_MATCHING (RE_NO_EMPTY_RANGES << 1) | |
| 171 | |
| 172 /* If this bit is set, succeed as soon as we match the whole pattern, | |
| 173 without further backtracking. */ | |
| 174 #define RE_NO_POSIX_BACKTRACKING (RE_NO_MINIMAL_MATCHING << 1) | |
| 175 | |
| 176 /* If this bit is not set, (?:re) behaves like (re) (or \(?:re\) behaves like | |
| 177 \(re\)) except that the matched string is not registered. */ | |
| 178 #define RE_NO_SHY_GROUPS (RE_NO_POSIX_BACKTRACKING << 1) | |
| 179 | |
| 180 /* If this bit is set, then an unmatched ) is ordinary. | |
| 181 If not set, then an unmatched ) is invalid. */ | |
| 182 #define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_SHY_GROUPS << 1) | |
| 183 | |
| 502 | 184 /* If this bit is set, then \22 will read as a back reference, |
| 185 provided at least 22 non-shy groups have been seen so far. In all | |
| 186 other cases (bit not set, not 22 non-shy groups seen so far), it | |
| 187 reads as a back reference \2 followed by a digit 2. */ | |
| 188 #define RE_NO_MULTI_DIGIT_BK_REFS (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) | |
| 189 | |
| 428 | 190 /* This global variable defines the particular regexp syntax to use (for |
| 191 some interfaces). When a regexp is compiled, the syntax used is | |
| 192 stored in the pattern buffer, so changing this does not affect | |
| 193 already-compiled regexps. */ | |
| 194 extern reg_syntax_t re_syntax_options; | |
| 195 | |
| 196 /* Define combinations of the above bits for the standard possibilities. | |
| 197 (The [[[ comments delimit what gets put into the Texinfo file, so | |
| 198 don't delete them!) */ | |
| 199 /* [[[begin syntaxes]]] */ | |
|
5647
1d9f603e9125
Turn on character classes in regex.c by default; test them in regexp-tests.el
Aidan Kehoe <kehoea@parhasard.net>
parents:
5402
diff
changeset
|
200 #define RE_SYNTAX_EMACS (RE_INTERVALS | RE_CHAR_CLASSES) |
| 428 | 201 |
| 202 #define RE_SYNTAX_AWK \ | |
| 203 (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ | |
| 204 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | |
| 205 | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ | |
| 206 | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_SHY_GROUPS \ | |
| 502 | 207 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 208 |
| 209 #define RE_SYNTAX_POSIX_AWK \ | |
| 210 (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS) | |
| 211 | |
| 212 #define RE_SYNTAX_GREP \ | |
| 213 (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ | |
| 214 | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ | |
| 215 | RE_NEWLINE_ALT | RE_NO_SHY_GROUPS \ | |
| 502 | 216 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 217 |
| 218 #define RE_SYNTAX_EGREP \ | |
| 219 (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ | |
| 220 | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ | |
| 221 | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ | |
| 222 | RE_NO_BK_VBAR | RE_NO_SHY_GROUPS \ | |
| 502 | 223 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 224 |
| 225 #define RE_SYNTAX_POSIX_EGREP \ | |
| 502 | 226 (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES | \ |
| 227 RE_NO_MULTI_DIGIT_BK_REFS) | |
| 428 | 228 |
| 229 /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ | |
| 230 #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC | |
| 231 | |
| 232 #define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC | |
| 233 | |
| 234 /* Syntax bits common to both basic and extended POSIX regex syntax. */ | |
| 235 #define _RE_SYNTAX_POSIX_COMMON \ | |
| 236 (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ | |
| 237 | RE_INTERVALS | RE_NO_EMPTY_RANGES | RE_NO_SHY_GROUPS \ | |
| 502 | 238 | RE_NO_MINIMAL_MATCHING | RE_NO_MULTI_DIGIT_BK_REFS) |
| 428 | 239 |
| 240 #define RE_SYNTAX_POSIX_BASIC \ | |
| 241 (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM) | |
| 242 | |
| 243 /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes | |
| 244 RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this | |
| 245 isn't minimal, since other operators, such as \`, aren't disabled. */ | |
| 246 #define RE_SYNTAX_POSIX_MINIMAL_BASIC \ | |
| 247 (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) | |
| 248 | |
| 249 #define RE_SYNTAX_POSIX_EXTENDED \ | |
| 250 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ | |
| 251 | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ | |
| 252 | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ | |
| 253 | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
| 254 | |
| 255 /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS | |
| 256 replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added. */ | |
| 257 #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ | |
| 258 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ | |
| 259 | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ | |
| 260 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | |
| 261 | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) | |
| 262 /* [[[end syntaxes]]] */ | |
| 263 | |
| 264 /* Maximum number of duplicates an interval can allow. Some systems | |
| 265 (erroneously) define this in other header files, but we want our | |
| 266 value, so remove any previous define. */ | |
| 267 #ifdef RE_DUP_MAX | |
| 268 #undef RE_DUP_MAX | |
| 269 #endif | |
| 270 #define RE_DUP_MAX ((1 << 15) - 1) | |
| 271 | |
| 272 | |
| 273 /* POSIX `cflags' bits (i.e., information for `regcomp'). */ | |
| 274 | |
| 275 /* If this bit is set, then use extended regular expression syntax. | |
| 276 If not set, then use basic regular expression syntax. */ | |
| 277 #define REG_EXTENDED 1 | |
| 278 | |
| 279 /* If this bit is set, then ignore case when matching. | |
| 280 If not set, then case is significant. */ | |
| 281 #define REG_ICASE (REG_EXTENDED << 1) | |
| 282 | |
| 283 /* If this bit is set, then anchors do not match at newline | |
| 284 characters in the string. | |
| 285 If not set, then anchors do match at newlines. */ | |
| 286 #define REG_NEWLINE (REG_ICASE << 1) | |
| 287 | |
| 288 /* If this bit is set, then report only success or fail in regexec. | |
| 289 If not set, then returns differ between not matching and errors. */ | |
| 290 #define REG_NOSUB (REG_NEWLINE << 1) | |
| 291 | |
| 292 | |
| 293 /* POSIX `eflags' bits (i.e., information for regexec). */ | |
| 294 | |
| 295 /* If this bit is set, then the beginning-of-line operator doesn't match | |
| 296 the beginning of the string (presumably because it's not the | |
| 297 beginning of a line). | |
| 298 If not set, then the beginning-of-line operator does match the | |
| 299 beginning of the string. */ | |
| 300 #define REG_NOTBOL 1 | |
| 301 | |
| 302 /* Like REG_NOTBOL, except for the end-of-line. */ | |
| 303 #define REG_NOTEOL (1 << 1) | |
| 304 | |
| 305 | |
| 306 /* If any error codes are removed, changed, or added, update the | |
| 307 `re_error_msg' table in regex.c. */ | |
| 308 typedef enum | |
| 309 { | |
| 310 REG_NOERROR = 0, /* Success. */ | |
| 311 REG_NOMATCH, /* Didn't find a match (for regexec). */ | |
| 312 | |
| 313 /* POSIX regcomp return error codes. (In the order listed in the | |
| 314 standard.) */ | |
| 315 REG_BADPAT, /* Invalid pattern. */ | |
| 316 REG_ECOLLATE, /* Not implemented. */ | |
| 317 REG_ECTYPE, /* Invalid character class name. */ | |
| 318 REG_EESCAPE, /* Trailing backslash. */ | |
| 319 REG_ESUBREG, /* Invalid back reference. */ | |
| 320 REG_EBRACK, /* Unmatched left bracket. */ | |
| 321 REG_EPAREN, /* Parenthesis imbalance. */ | |
| 322 REG_EBRACE, /* Unmatched \{. */ | |
| 323 REG_BADBR, /* Invalid contents of \{\}. */ | |
| 324 REG_ERANGE, /* Invalid range end. */ | |
| 325 REG_ESPACE, /* Ran out of memory. */ | |
| 326 REG_BADRPT, /* No preceding re for repetition op. */ | |
| 327 | |
| 328 /* Error codes we've added. */ | |
| 329 REG_EEND, /* Premature end. */ | |
| 330 REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ | |
| 331 REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ | |
| 332 #ifdef emacs | |
| 333 ,REG_ESYNTAX /* Invalid syntax designator. */ | |
| 334 #endif | |
| 335 #ifdef MULE | |
| 336 ,REG_ERANGESPAN /* Ranges may not span charsets. */ | |
| 337 ,REG_ECATEGORY /* Invalid category designator */ | |
| 338 #endif | |
| 339 } reg_errcode_t; | |
| 340 | |
| 341 /* This data structure represents a compiled pattern. Before calling | |
| 342 the pattern compiler, the fields `buffer', `allocated', `fastmap', | |
| 343 `translate', and `no_sub' can be set. After the pattern has been | |
| 344 compiled, the `re_nsub' field is available. All other fields are | |
| 345 private to the regex routines. */ | |
| 346 | |
| 347 struct re_pattern_buffer | |
| 348 { | |
| 349 /* [[[begin pattern_buffer]]] */ | |
| 350 /* Space that holds the compiled pattern. It is declared as | |
| 351 `unsigned char *' because its elements are | |
| 352 sometimes used as array indexes. */ | |
| 353 unsigned char *buffer; | |
| 354 | |
| 355 /* Number of bytes to which `buffer' points. */ | |
| 647 | 356 long allocated; |
| 428 | 357 |
| 358 /* Number of bytes actually used in `buffer'. */ | |
| 647 | 359 long used; |
| 428 | 360 |
| 361 /* Syntax setting with which the pattern was compiled. */ | |
| 362 reg_syntax_t syntax; | |
| 363 | |
| 364 /* Pointer to a fastmap, if any, otherwise zero. re_search uses | |
| 365 the fastmap, if there is one, to skip over impossible | |
| 366 starting points for matches. */ | |
| 367 char *fastmap; | |
| 368 | |
| 369 /* Either a translate table to apply to all characters before | |
| 370 comparing them, or zero for no translation. The translation | |
| 371 is applied to a pattern when it is compiled and to a string | |
| 372 when it is matched. */ | |
| 446 | 373 RE_TRANSLATE_TYPE translate; |
| 428 | 374 |
| 2639 | 375 /* Number of subpatterns (returnable groups) found by the compiler. |
| 376 (This does not count shy groups.) */ | |
| 647 | 377 int re_nsub; |
| 428 | 378 |
| 502 | 379 /* Total number of groups found by the compiler. (Including |
| 380 shy ones.) */ | |
| 381 int re_ngroups; | |
| 382 | |
| 428 | 383 /* Zero if this pattern cannot match the empty string, one else. |
| 384 Well, in truth it's used only in `re_search_2', to see | |
| 385 whether or not we should use the fastmap, so we don't set | |
| 386 this absolutely perfectly; see `re_compile_fastmap' (the | |
| 387 `duplicate' case). */ | |
| 647 | 388 unsigned int can_be_null : 1; |
| 428 | 389 |
| 390 /* If REGS_UNALLOCATED, allocate space in the `regs' structure | |
| 391 for `max (RE_NREGS, re_nsub + 1)' groups. | |
| 392 If REGS_REALLOCATE, reallocate space if necessary. | |
| 393 If REGS_FIXED, use what's there. */ | |
| 394 #define REGS_UNALLOCATED 0 | |
| 395 #define REGS_REALLOCATE 1 | |
| 396 #define REGS_FIXED 2 | |
| 647 | 397 unsigned int regs_allocated : 2; |
| 428 | 398 |
| 399 /* Set to zero when `regex_compile' compiles a pattern; set to one | |
| 400 by `re_compile_fastmap' if it updates the fastmap. */ | |
| 647 | 401 unsigned int fastmap_accurate : 1; |
| 428 | 402 |
| 403 /* If set, `re_match_2' does not return information about | |
| 404 subexpressions. */ | |
| 647 | 405 unsigned int no_sub : 1; |
| 428 | 406 |
| 407 /* If set, a beginning-of-line anchor doesn't match at the | |
| 408 beginning of the string. */ | |
| 647 | 409 unsigned int not_bol : 1; |
| 428 | 410 |
| 411 /* Similarly for an end-of-line anchor. */ | |
| 647 | 412 unsigned int not_eol : 1; |
| 428 | 413 |
| 414 /* If true, an anchor at a newline matches. */ | |
| 647 | 415 unsigned int newline_anchor : 1; |
| 428 | 416 |
| 647 | 417 unsigned int warned_about_incompatible_back_references : 1; |
| 502 | 418 |
| 419 /* Mapping between back references and groups (may not be | |
| 420 equivalent with shy groups). */ | |
| 421 int *external_to_internal_register; | |
| 422 | |
| 423 int external_to_internal_register_size; | |
| 424 | |
| 428 | 425 /* [[[end pattern_buffer]]] */ |
| 426 }; | |
| 427 | |
| 428 typedef struct re_pattern_buffer regex_t; | |
| 429 | |
| 430 /* Type for byte offsets within the string. POSIX mandates this. */ | |
| 431 typedef int regoff_t; | |
| 432 | |
| 433 | |
| 434 /* This is the structure we store register match data in. See | |
| 435 regex.texinfo for a full description of what registers match. */ | |
| 436 struct re_registers | |
| 437 { | |
| 1468 | 438 int num_regs; /* number of registers allocated */ |
| 428 | 439 regoff_t *start; |
| 440 regoff_t *end; | |
| 441 }; | |
| 442 | |
| 443 | |
| 444 /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, | |
| 445 `re_match_2' returns information about at least this many registers | |
| 446 the first time a `regs' structure is passed. */ | |
| 447 #ifndef RE_NREGS | |
| 448 #define RE_NREGS 30 | |
| 449 #endif | |
| 450 | |
| 451 | |
| 452 /* POSIX specification for registers. Aside from the different names than | |
| 453 `re_registers', POSIX uses an array of structures, instead of a | |
| 454 structure of arrays. */ | |
| 455 typedef struct | |
| 456 { | |
| 457 regoff_t rm_so; /* Byte offset from string's start to substring's start. */ | |
| 458 regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ | |
| 459 } regmatch_t; | |
| 460 | |
| 461 /* Declarations for routines. */ | |
| 462 | |
| 463 /* Sets the current default syntax to SYNTAX, and return the old syntax. | |
| 464 You can also simply assign to the `re_syntax_options' variable. */ | |
| 465 reg_syntax_t re_set_syntax (reg_syntax_t syntax); | |
| 466 | |
| 467 /* Compile the regular expression PATTERN, with length LENGTH | |
| 468 and syntax given by the global `re_syntax_options', into the buffer | |
| 469 BUFFER. Return NULL if successful, and an error string if not. */ | |
| 442 | 470 const char *re_compile_pattern (const char *pattern, int length, |
| 428 | 471 struct re_pattern_buffer *buffer); |
| 472 | |
| 473 | |
| 474 /* Compile a fastmap for the compiled pattern in BUFFER; used to | |
| 475 accelerate searches. Return 0 if successful and -2 if was an | |
| 476 internal error. */ | |
| 826 | 477 int re_compile_fastmap (struct re_pattern_buffer *buffer |
| 478 RE_LISP_SHORT_CONTEXT_ARGS_DECL); | |
| 428 | 479 |
| 480 | |
| 481 /* Search in the string STRING (with length LENGTH) for the pattern | |
| 482 compiled into BUFFER. Start searching at position START, for RANGE | |
| 483 characters. Return the starting position of the match, -1 for no | |
| 484 match, or -2 for an internal error. Also return register | |
| 485 information in REGS (if REGS and BUFFER->no_sub are nonzero). */ | |
| 442 | 486 int re_search (struct re_pattern_buffer *buffer, const char *string, |
| 428 | 487 int length, int start, int range, |
| 826 | 488 struct re_registers *regs RE_LISP_CONTEXT_ARGS_DECL); |
| 428 | 489 |
| 490 | |
| 491 /* Like `re_search', but search in the concatenation of STRING1 and | |
| 492 STRING2. Also, stop searching at index START + STOP. */ | |
| 442 | 493 int re_search_2 (struct re_pattern_buffer *buffer, const char *string1, |
| 494 int length1, const char *string2, int length2, int start, | |
| 826 | 495 int range, struct re_registers *regs, int stop |
| 496 RE_LISP_CONTEXT_ARGS_DECL); | |
| 428 | 497 |
| 826 | 498 #ifndef emacs /* never used by XEmacs */ |
| 428 | 499 |
| 500 /* Like `re_search', but return how many characters in STRING the regexp | |
| 501 in BUFFER matched, starting at position START. */ | |
| 442 | 502 int re_match (struct re_pattern_buffer *buffer, const char *string, |
| 826 | 503 int length, int start, struct re_registers *regs |
| 504 RE_LISP_CONTEXT_ARGS_DECL); | |
| 428 | 505 |
| 826 | 506 #endif /* not emacs */ |
| 428 | 507 |
| 508 /* Relates to `re_match' as `re_search_2' relates to `re_search'. */ | |
| 442 | 509 int re_match_2 (struct re_pattern_buffer *buffer, const char *string1, |
| 510 int length1, const char *string2, int length2, | |
| 826 | 511 int start, struct re_registers *regs, int stop |
| 512 RE_LISP_CONTEXT_ARGS_DECL); | |
| 428 | 513 |
| 514 /* Set REGS to hold NUM_REGS registers, storing them in STARTS and | |
| 515 ENDS. Subsequent matches using BUFFER and REGS will use this memory | |
| 516 for recording register information. STARTS and ENDS must be | |
| 517 allocated with malloc, and must each be at least `NUM_REGS * sizeof | |
| 518 (regoff_t)' bytes long. | |
| 519 | |
| 520 If NUM_REGS == 0, then subsequent matches should allocate their own | |
| 521 register data. | |
| 522 | |
| 523 Unless this function is called, the first search or match using | |
| 524 PATTERN_BUFFER will allocate its own register data, without | |
| 525 freeing the old data. */ | |
| 526 void re_set_registers (struct re_pattern_buffer *buffer, | |
| 647 | 527 struct re_registers *regs, int num_regs, |
| 428 | 528 regoff_t *starts, regoff_t *ends); |
| 529 | |
| 530 #ifdef _REGEX_RE_COMP | |
| 531 /* 4.2 bsd compatibility. */ | |
| 442 | 532 char *re_comp (const char *); |
| 533 int re_exec (const char *); | |
| 428 | 534 #endif |
| 535 | |
| 536 /* POSIX compatibility. */ | |
| 442 | 537 int regcomp (regex_t *preg, const char *pattern, int cflags); |
| 538 int regexec (const regex_t *preg, const char *string, size_t nmatch, | |
| 428 | 539 regmatch_t pmatch[], int eflags); |
| 442 | 540 size_t regerror (int errcode, const regex_t *preg, char *errbuf, |
| 428 | 541 size_t errbuf_size); |
| 542 void regfree (regex_t *preg); | |
| 543 | |
| 5041 | 544 enum regex_debug |
| 545 { | |
| 546 RE_DEBUG_COMPILATION = 1 << 0, | |
| 547 RE_DEBUG_FAILURE_POINT = 1 << 1, | |
| 548 RE_DEBUG_MATCHING = 1 << 2, | |
| 549 }; | |
| 550 | |
| 551 extern int debug_regexps; | |
| 552 | |
|
5648
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
553 typedef enum |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
554 { |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
555 RECC_ERROR = 0, |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
556 RECC_ALNUM, RECC_ALPHA, RECC_WORD, |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
557 RECC_GRAPH, RECC_PRINT, |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
558 RECC_LOWER, RECC_UPPER, |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
559 RECC_PUNCT, RECC_CNTRL, |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
560 RECC_DIGIT, RECC_XDIGIT, |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
561 RECC_BLANK, RECC_SPACE, |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
562 RECC_MULTIBYTE, RECC_NONASCII, |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
563 RECC_ASCII, RECC_UNIBYTE |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
564 } re_wctype_t; |
|
3f4a234f4672
Support non-ASCII correctly in character classes, test this.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5647
diff
changeset
|
565 |
|
5653
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
566 #define CHAR_CLASS_MAX_LENGTH 9 /* Namely, `multibyte'. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
567 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
568 /* Map a string to the char class it names (if any). */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
569 re_wctype_t re_wctype (const char *); |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
570 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
571 /* Is character CH a member of the character class CC? */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
572 int re_iswctype (int ch, re_wctype_t cc RE_ISWCTYPE_ARG_DECL); |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
573 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
574 /* Bits used to implement the multibyte-part of the various character |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
575 classes such as [:alnum:] in a charset's range table. XEmacs; use an |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
576 enum, so they're visible in the debugger. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
577 enum |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
578 { |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
579 BIT_WORD = (1 << 0), |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
580 BIT_LOWER = (1 << 1), |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
581 BIT_PUNCT = (1 << 2), |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
582 BIT_SPACE = (1 << 3), |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
583 BIT_UPPER = (1 << 4), |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
584 /* XEmacs; we need this, because we unify treatment of ASCII and non-ASCII |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
585 (possible matches) in charset_mule. [:alpha:] matches all characters |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
586 with word syntax, with the exception of [0-9]. We don't need |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
587 BIT_MULTIBYTE. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
588 BIT_ALPHA = (1 << 5) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
589 }; |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
590 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
591 #ifdef emacs |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
592 reg_errcode_t compile_char_class (re_wctype_t cc, Lisp_Object rtab, |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
593 Bitbyte *flags_out); |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
594 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
595 #endif |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
596 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
597 /* isalpha etc. are used for the character classes. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
598 #include <ctype.h> |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
599 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
600 #ifdef emacs |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
601 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
602 /* 1 if C is an ASCII character. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
603 #define ISASCII(c) ((c) < 0x80) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
604 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
605 /* 1 if C is a unibyte character. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
606 #define ISUNIBYTE ISASCII |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
607 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
608 /* The Emacs definitions should not be directly affected by locales. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
609 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
610 /* In Emacs, these are only used for single-byte characters. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
611 #define ISDIGIT(c) ((c) >= '0' && (c) <= '9') |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
612 #define ISCNTRL(c) ((c) < ' ') |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
613 #define ISXDIGIT(c) (ISDIGIT (c) || ((c) >= 'a' && (c) <= 'f') \ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
614 || ((c) >= 'A' && (c) <= 'F')) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
615 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
616 /* This is only used for single-byte characters. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
617 #define ISBLANK(c) ((c) == ' ' || (c) == '\t') |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
618 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
619 /* The rest must handle multibyte characters. */ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
620 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
621 #define ISGRAPH(c) ((c) > ' ' && (c) != 0x7f) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
622 #define ISPRINT(c) ((c) == ' ' || ISGRAPH (c)) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
623 #define ISALPHA(c) (ISASCII (c) ? (((c) >= 'a' && (c) <= 'z') \ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
624 || ((c) >= 'A' && (c) <= 'Z')) \ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
625 : ISWORD (c)) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
626 #define ISALNUM(c) (ISALPHA (c) || ISDIGIT (c)) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
627 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
628 #define ISLOWER(c) LOWERCASEP (lispbuf, c) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
629 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
630 #define ISPUNCT(c) (ISASCII (c) \ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
631 ? ((c) > ' ' && (c) < 0x7F \ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
632 && !(((c) >= 'a' && (c) <= 'z') \ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
633 || ((c) >= 'A' && (c) <= 'Z') \ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
634 || ((c) >= '0' && (c) <= '9'))) \ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
635 : !ISWORD (c)) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
636 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
637 #define ISSPACE(c) \ |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
638 (SYNTAX (BUFFER_MIRROR_SYNTAX_TABLE (lispbuf), c) == Swhitespace) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
639 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
640 #define ISUPPER(c) UPPERCASEP (lispbuf, c) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
641 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
642 #define ISWORD(c) (SYNTAX (BUFFER_MIRROR_SYNTAX_TABLE (lispbuf), c) == Sword) |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
643 |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
644 #endif |
|
3df910176b6a
Support predefined character classes in #'skip-chars-{forward,backward}, too
Aidan Kehoe <kehoea@parhasard.net>
parents:
5648
diff
changeset
|
645 |
|
5183
f283b08ff0c9
Avoid build failure, Apple's g++-4.0.1, Mac OS 10.4.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5041
diff
changeset
|
646 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
|
647 |
| 440 | 648 #endif /* INCLUDED_regex_h_ */ |