diff options
Diffstat (limited to 'win32/regex.h')
-rw-r--r-- | win32/regex.h | 462 |
1 files changed, 277 insertions, 185 deletions
diff --git a/win32/regex.h b/win32/regex.h index 6eb64f140..61c968387 100644 --- a/win32/regex.h +++ b/win32/regex.h | |||
@@ -1,70 +1,90 @@ | |||
1 | /* Definitions for data structures and routines for the regular | 1 | #include <stdio.h> |
2 | expression library, version 0.12. | 2 | #include <stddef.h> |
3 | 3 | ||
4 | Copyright (C) 1985, 1989, 1990, 1991, 1992, 1993 Free Software Foundation, Inc. | 4 | /* Definitions for data structures and routines for the regular |
5 | expression library. | ||
6 | Copyright (C) 1985,1989-93,1995-98,2000,2001,2002,2003,2005,2006,2008 | ||
7 | Free Software Foundation, Inc. | ||
8 | This file is part of the GNU C Library. | ||
5 | 9 | ||
6 | This program is free software; you can redistribute it and/or modify | 10 | The GNU C Library is free software; you can redistribute it and/or |
7 | it under the terms of the GNU General Public License as published by | 11 | modify it under the terms of the GNU Lesser General Public |
8 | the Free Software Foundation; either version 2, or (at your option) | 12 | License as published by the Free Software Foundation; either |
9 | any later version. | 13 | version 2.1 of the License, or (at your option) any later version. |
10 | 14 | ||
11 | This program is distributed in the hope that it will be useful, | 15 | The GNU C Library is distributed in the hope that it will be useful, |
12 | but WITHOUT ANY WARRANTY; without even the implied warranty of | 16 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
13 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | 17 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
14 | GNU General Public License for more details. | 18 | Lesser General Public License for more details. |
15 | |||
16 | You should have received a copy of the GNU General Public License | ||
17 | along with this program; if not, write to the Free Software | ||
18 | Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ | ||
19 | 19 | ||
20 | #ifndef __REGEXP_LIBRARY_H__ | 20 | You should have received a copy of the GNU Lesser General Public |
21 | #define __REGEXP_LIBRARY_H__ | 21 | License along with the GNU C Library; if not, write to the Free |
22 | Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA | ||
23 | 02110-1301 USA. */ | ||
22 | 24 | ||
23 | /* POSIX says that <sys/types.h> must be included (by the caller) before | 25 | #ifndef _REGEX_H |
24 | <regex.h>. */ | 26 | #define _REGEX_H 1 |
25 | 27 | ||
26 | #ifdef VMS | 28 | #ifdef HAVE_STDDEF_H |
27 | /* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it | ||
28 | should be there. */ | ||
29 | #include <stddef.h> | 29 | #include <stddef.h> |
30 | #endif | 30 | #endif |
31 | 31 | ||
32 | #ifdef HAVE_SYS_TYPES_H | ||
33 | #include <sys/types.h> | ||
34 | #endif | ||
35 | |||
36 | #ifndef _LIBC | ||
37 | #define __USE_GNU 1 | ||
38 | #endif | ||
39 | |||
40 | /* Allow the use in C++ code. */ | ||
41 | #ifdef __cplusplus | ||
42 | extern "C" { | ||
43 | #endif | ||
44 | |||
45 | /* The following two types have to be signed and unsigned integer type | ||
46 | wide enough to hold a value of a pointer. For most ANSI compilers | ||
47 | ptrdiff_t and size_t should be likely OK. Still size of these two | ||
48 | types is 2 for Microsoft C. Ugh... */ | ||
49 | typedef long int s_reg_t; | ||
50 | typedef unsigned long int active_reg_t; | ||
32 | 51 | ||
33 | /* The following bits are used to determine the regexp syntax we | 52 | /* The following bits are used to determine the regexp syntax we |
34 | recognize. The set/not-set meanings are chosen so that Emacs syntax | 53 | recognize. The set/not-set meanings are chosen so that Emacs syntax |
35 | remains the value 0. The bits are given in alphabetical order, and | 54 | remains the value 0. The bits are given in alphabetical order, and |
36 | the definitions shifted by one from the previous bit; thus, when we | 55 | the definitions shifted by one from the previous bit; thus, when we |
37 | add or remove a bit, only one other definition need change. */ | 56 | add or remove a bit, only one other definition need change. */ |
38 | typedef unsigned reg_syntax_t; | 57 | typedef unsigned long int reg_syntax_t; |
39 | 58 | ||
59 | #ifdef __USE_GNU | ||
40 | /* If this bit is not set, then \ inside a bracket expression is literal. | 60 | /* If this bit is not set, then \ inside a bracket expression is literal. |
41 | If set, then such a \ quotes the following character. */ | 61 | If set, then such a \ quotes the following character. */ |
42 | #define RE_BACKSLASH_ESCAPE_IN_LISTS (1) | 62 | # define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1) |
43 | 63 | ||
44 | /* If this bit is not set, then + and ? are operators, and \+ and \? are | 64 | /* If this bit is not set, then + and ? are operators, and \+ and \? are |
45 | literals. | 65 | literals. |
46 | If set, then \+ and \? are operators and + and ? are literals. */ | 66 | If set, then \+ and \? are operators and + and ? are literals. */ |
47 | #define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) | 67 | # define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) |
48 | 68 | ||
49 | /* If this bit is set, then character classes are supported. They are: | 69 | /* If this bit is set, then character classes are supported. They are: |
50 | [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], | 70 | [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], |
51 | [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. | 71 | [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. |
52 | If not set, then character classes are not supported. */ | 72 | If not set, then character classes are not supported. */ |
53 | #define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) | 73 | # define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) |
54 | 74 | ||
55 | /* If this bit is set, then ^ and $ are always anchors (outside bracket | 75 | /* If this bit is set, then ^ and $ are always anchors (outside bracket |
56 | expressions, of course). | 76 | expressions, of course). |
57 | If this bit is not set, then it depends: | 77 | If this bit is not set, then it depends: |
58 | ^ is an anchor if it is at the beginning of a regular | 78 | ^ is an anchor if it is at the beginning of a regular |
59 | expression or after an open-group or an alternation operator; | 79 | expression or after an open-group or an alternation operator; |
60 | $ is an anchor if it is at the end of a regular expression, or | 80 | $ is an anchor if it is at the end of a regular expression, or |
61 | before a close-group or an alternation operator. | 81 | before a close-group or an alternation operator. |
62 | 82 | ||
63 | This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because | 83 | This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because |
64 | POSIX draft 11.2 says that * etc. in leading positions is undefined. | 84 | POSIX draft 11.2 says that * etc. in leading positions is undefined. |
65 | We already implemented a previous draft which made those constructs | 85 | We already implemented a previous draft which made those constructs |
66 | invalid, though, so we haven't changed the code back. */ | 86 | invalid, though, so we haven't changed the code back. */ |
67 | #define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) | 87 | # define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) |
68 | 88 | ||
69 | /* If this bit is set, then special characters are always special | 89 | /* If this bit is set, then special characters are always special |
70 | regardless of where they are in the pattern. | 90 | regardless of where they are in the pattern. |
@@ -72,63 +92,94 @@ typedef unsigned reg_syntax_t; | |||
72 | some contexts; otherwise they are ordinary. Specifically, | 92 | some contexts; otherwise they are ordinary. Specifically, |
73 | * + ? and intervals are only special when not after the beginning, | 93 | * + ? and intervals are only special when not after the beginning, |
74 | open-group, or alternation operator. */ | 94 | open-group, or alternation operator. */ |
75 | #define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) | 95 | # define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) |
76 | 96 | ||
77 | /* If this bit is set, then *, +, ?, and { cannot be first in an re or | 97 | /* If this bit is set, then *, +, ?, and { cannot be first in an re or |
78 | immediately after an alternation or begin-group operator. */ | 98 | immediately after an alternation or begin-group operator. */ |
79 | #define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) | 99 | # define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) |
80 | 100 | ||
81 | /* If this bit is set, then . matches newline. | 101 | /* If this bit is set, then . matches newline. |
82 | If not set, then it doesn't. */ | 102 | If not set, then it doesn't. */ |
83 | #define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) | 103 | # define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) |
84 | 104 | ||
85 | /* If this bit is set, then . doesn't match NUL. | 105 | /* If this bit is set, then . doesn't match NUL. |
86 | If not set, then it does. */ | 106 | If not set, then it does. */ |
87 | #define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) | 107 | # define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) |
88 | 108 | ||
89 | /* If this bit is set, nonmatching lists [^...] do not match newline. | 109 | /* If this bit is set, nonmatching lists [^...] do not match newline. |
90 | If not set, they do. */ | 110 | If not set, they do. */ |
91 | #define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) | 111 | # define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) |
92 | 112 | ||
93 | /* If this bit is set, either \{...\} or {...} defines an | 113 | /* If this bit is set, either \{...\} or {...} defines an |
94 | interval, depending on RE_NO_BK_BRACES. | 114 | interval, depending on RE_NO_BK_BRACES. |
95 | If not set, \{, \}, {, and } are literals. */ | 115 | If not set, \{, \}, {, and } are literals. */ |
96 | #define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) | 116 | # define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) |
97 | 117 | ||
98 | /* If this bit is set, +, ? and | aren't recognized as operators. | 118 | /* If this bit is set, +, ? and | aren't recognized as operators. |
99 | If not set, they are. */ | 119 | If not set, they are. */ |
100 | #define RE_LIMITED_OPS (RE_INTERVALS << 1) | 120 | # define RE_LIMITED_OPS (RE_INTERVALS << 1) |
101 | 121 | ||
102 | /* If this bit is set, newline is an alternation operator. | 122 | /* If this bit is set, newline is an alternation operator. |
103 | If not set, newline is literal. */ | 123 | If not set, newline is literal. */ |
104 | #define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) | 124 | # define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) |
105 | 125 | ||
106 | /* If this bit is set, then `{...}' defines an interval, and \{ and \} | 126 | /* If this bit is set, then `{...}' defines an interval, and \{ and \} |
107 | are literals. | 127 | are literals. |
108 | If not set, then `\{...\}' defines an interval. */ | 128 | If not set, then `\{...\}' defines an interval. */ |
109 | #define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) | 129 | # define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) |
110 | 130 | ||
111 | /* If this bit is set, (...) defines a group, and \( and \) are literals. | 131 | /* If this bit is set, (...) defines a group, and \( and \) are literals. |
112 | If not set, \(...\) defines a group, and ( and ) are literals. */ | 132 | If not set, \(...\) defines a group, and ( and ) are literals. */ |
113 | #define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) | 133 | # define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) |
114 | 134 | ||
115 | /* If this bit is set, then \<digit> matches <digit>. | 135 | /* If this bit is set, then \<digit> matches <digit>. |
116 | If not set, then \<digit> is a back-reference. */ | 136 | If not set, then \<digit> is a back-reference. */ |
117 | #define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) | 137 | # define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) |
118 | 138 | ||
119 | /* If this bit is set, then | is an alternation operator, and \| is literal. | 139 | /* If this bit is set, then | is an alternation operator, and \| is literal. |
120 | If not set, then \| is an alternation operator, and | is literal. */ | 140 | If not set, then \| is an alternation operator, and | is literal. */ |
121 | #define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) | 141 | # define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) |
122 | 142 | ||
123 | /* If this bit is set, then an ending range point collating higher | 143 | /* If this bit is set, then an ending range point collating higher |
124 | than the starting range point, as in [z-a], is invalid. | 144 | than the starting range point, as in [z-a], is invalid. |
125 | If not set, then when ending range point collates higher than the | 145 | If not set, then when ending range point collates higher than the |
126 | starting range point, the range is ignored. */ | 146 | starting range point, the range is ignored. */ |
127 | #define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) | 147 | # define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) |
128 | 148 | ||
129 | /* If this bit is set, then an unmatched ) is ordinary. | 149 | /* If this bit is set, then an unmatched ) is ordinary. |
130 | If not set, then an unmatched ) is invalid. */ | 150 | If not set, then an unmatched ) is invalid. */ |
131 | #define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) | 151 | # define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) |
152 | |||
153 | /* If this bit is set, succeed as soon as we match the whole pattern, | ||
154 | without further backtracking. */ | ||
155 | # define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) | ||
156 | |||
157 | /* If this bit is set, do not process the GNU regex operators. | ||
158 | If not set, then the GNU regex operators are recognized. */ | ||
159 | # define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1) | ||
160 | |||
161 | /* If this bit is set, a syntactically invalid interval is treated as | ||
162 | a string of ordinary characters. For example, the ERE 'a{1' is | ||
163 | treated as 'a\{1'. */ | ||
164 | # define RE_INVALID_INTERVAL_ORD (RE_NO_GNU_OPS << 1) | ||
165 | |||
166 | /* If this bit is set, then ignore case when matching. | ||
167 | If not set, then case is significant. */ | ||
168 | # define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1) | ||
169 | |||
170 | /* This bit is used internally like RE_CONTEXT_INDEP_ANCHORS but only | ||
171 | for ^, because it is difficult to scan the regex backwards to find | ||
172 | whether ^ should be special. */ | ||
173 | # define RE_CARET_ANCHORS_HERE (RE_ICASE << 1) | ||
174 | |||
175 | /* If this bit is set, then \{ cannot be first in an bre or | ||
176 | immediately after an alternation or begin-group operator. */ | ||
177 | # define RE_CONTEXT_INVALID_DUP (RE_CARET_ANCHORS_HERE << 1) | ||
178 | |||
179 | /* If this bit is set, then no_sub will be set to 1 during | ||
180 | re_compile_pattern. */ | ||
181 | #define RE_NO_SUB (RE_CONTEXT_INVALID_DUP << 1) | ||
182 | #endif | ||
132 | 183 | ||
133 | /* This global variable defines the particular regexp syntax to use (for | 184 | /* This global variable defines the particular regexp syntax to use (for |
134 | some interfaces). When a regexp is compiled, the syntax used is | 185 | some interfaces). When a regexp is compiled, the syntax used is |
@@ -136,6 +187,7 @@ typedef unsigned reg_syntax_t; | |||
136 | already-compiled regexps. */ | 187 | already-compiled regexps. */ |
137 | extern reg_syntax_t re_syntax_options; | 188 | extern reg_syntax_t re_syntax_options; |
138 | 189 | ||
190 | #ifdef __USE_GNU | ||
139 | /* Define combinations of the above bits for the standard possibilities. | 191 | /* Define combinations of the above bits for the standard possibilities. |
140 | (The [[[ comments delimit what gets put into the Texinfo file, so | 192 | (The [[[ comments delimit what gets put into the Texinfo file, so |
141 | don't delete them!) */ | 193 | don't delete them!) */ |
@@ -143,13 +195,22 @@ extern reg_syntax_t re_syntax_options; | |||
143 | #define RE_SYNTAX_EMACS 0 | 195 | #define RE_SYNTAX_EMACS 0 |
144 | 196 | ||
145 | #define RE_SYNTAX_AWK \ | 197 | #define RE_SYNTAX_AWK \ |
146 | (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ | 198 | (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ |
147 | | RE_NO_BK_PARENS | RE_NO_BK_REFS \ | 199 | | RE_NO_BK_PARENS | RE_NO_BK_REFS \ |
148 | | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ | 200 | | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ |
149 | | RE_UNMATCHED_RIGHT_PAREN_ORD) | 201 | | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \ |
150 | 202 | | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS) | |
151 | #define RE_SYNTAX_POSIX_AWK \ | 203 | |
152 | (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS) | 204 | #define RE_SYNTAX_GNU_AWK \ |
205 | ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ | ||
206 | | RE_INVALID_INTERVAL_ORD) \ | ||
207 | & ~(RE_DOT_NOT_NULL | RE_CONTEXT_INDEP_OPS \ | ||
208 | | RE_CONTEXT_INVALID_OPS )) | ||
209 | |||
210 | #define RE_SYNTAX_POSIX_AWK \ | ||
211 | (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ | ||
212 | | RE_INTERVALS | RE_NO_GNU_OPS \ | ||
213 | | RE_INVALID_INTERVAL_ORD) | ||
153 | 214 | ||
154 | #define RE_SYNTAX_GREP \ | 215 | #define RE_SYNTAX_GREP \ |
155 | (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ | 216 | (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ |
@@ -163,7 +224,8 @@ extern reg_syntax_t re_syntax_options; | |||
163 | | RE_NO_BK_VBAR) | 224 | | RE_NO_BK_VBAR) |
164 | 225 | ||
165 | #define RE_SYNTAX_POSIX_EGREP \ | 226 | #define RE_SYNTAX_POSIX_EGREP \ |
166 | (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES) | 227 | (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES \ |
228 | | RE_INVALID_INTERVAL_ORD) | ||
167 | 229 | ||
168 | /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ | 230 | /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ |
169 | #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC | 231 | #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC |
@@ -176,7 +238,7 @@ extern reg_syntax_t re_syntax_options; | |||
176 | | RE_INTERVALS | RE_NO_EMPTY_RANGES) | 238 | | RE_INTERVALS | RE_NO_EMPTY_RANGES) |
177 | 239 | ||
178 | #define RE_SYNTAX_POSIX_BASIC \ | 240 | #define RE_SYNTAX_POSIX_BASIC \ |
179 | (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM) | 241 | (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM | RE_CONTEXT_INVALID_DUP) |
180 | 242 | ||
181 | /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes | 243 | /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes |
182 | RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this | 244 | RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this |
@@ -185,13 +247,13 @@ extern reg_syntax_t re_syntax_options; | |||
185 | (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) | 247 | (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) |
186 | 248 | ||
187 | #define RE_SYNTAX_POSIX_EXTENDED \ | 249 | #define RE_SYNTAX_POSIX_EXTENDED \ |
188 | (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ | 250 | (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ |
189 | | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ | 251 | | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ |
190 | | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ | 252 | | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ |
191 | | RE_UNMATCHED_RIGHT_PAREN_ORD) | 253 | | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD) |
192 | 254 | ||
193 | /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS | 255 | /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is |
194 | replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added. */ | 256 | removed and RE_NO_BK_REFS is added. */ |
195 | #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ | 257 | #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ |
196 | (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ | 258 | (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ |
197 | | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ | 259 | | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ |
@@ -202,10 +264,12 @@ extern reg_syntax_t re_syntax_options; | |||
202 | /* Maximum number of duplicates an interval can allow. Some systems | 264 | /* Maximum number of duplicates an interval can allow. Some systems |
203 | (erroneously) define this in other header files, but we want our | 265 | (erroneously) define this in other header files, but we want our |
204 | value, so remove any previous define. */ | 266 | value, so remove any previous define. */ |
205 | #ifdef RE_DUP_MAX | 267 | # ifdef RE_DUP_MAX |
206 | #undef RE_DUP_MAX | 268 | # undef RE_DUP_MAX |
269 | # endif | ||
270 | /* If sizeof(int) == 2, then ((1 << 15) - 1) overflows. */ | ||
271 | # define RE_DUP_MAX (0x7fff) | ||
207 | #endif | 272 | #endif |
208 | #define RE_DUP_MAX ((1 << 15) - 1) | ||
209 | 273 | ||
210 | 274 | ||
211 | /* POSIX `cflags' bits (i.e., information for `regcomp'). */ | 275 | /* POSIX `cflags' bits (i.e., information for `regcomp'). */ |
@@ -240,18 +304,26 @@ extern reg_syntax_t re_syntax_options; | |||
240 | /* Like REG_NOTBOL, except for the end-of-line. */ | 304 | /* Like REG_NOTBOL, except for the end-of-line. */ |
241 | #define REG_NOTEOL (1 << 1) | 305 | #define REG_NOTEOL (1 << 1) |
242 | 306 | ||
307 | /* Use PMATCH[0] to delimit the start and end of the search in the | ||
308 | buffer. */ | ||
309 | #define REG_STARTEND (1 << 2) | ||
310 | |||
243 | 311 | ||
244 | /* If any error codes are removed, changed, or added, update the | 312 | /* If any error codes are removed, changed, or added, update the |
245 | `re_error_msg' table in regex.c. */ | 313 | `re_error_msg' table in regex.c. */ |
246 | typedef enum | 314 | typedef enum |
247 | { | 315 | { |
316 | #if defined _XOPEN_SOURCE || defined __USE_XOPEN2K | ||
317 | REG_ENOSYS = -1, /* This will never happen for this implementation. */ | ||
318 | #endif | ||
319 | |||
248 | REG_NOERROR = 0, /* Success. */ | 320 | REG_NOERROR = 0, /* Success. */ |
249 | REG_NOMATCH, /* Didn't find a match (for regexec). */ | 321 | REG_NOMATCH, /* Didn't find a match (for regexec). */ |
250 | 322 | ||
251 | /* POSIX regcomp return error codes. (In the order listed in the | 323 | /* POSIX regcomp return error codes. (In the order listed in the |
252 | standard.) */ | 324 | standard.) */ |
253 | REG_BADPAT, /* Invalid pattern. */ | 325 | REG_BADPAT, /* Invalid pattern. */ |
254 | REG_ECOLLATE, /* Not implemented. */ | 326 | REG_ECOLLATE, /* Inalid collating element. */ |
255 | REG_ECTYPE, /* Invalid character class name. */ | 327 | REG_ECTYPE, /* Invalid character class name. */ |
256 | REG_EESCAPE, /* Trailing backslash. */ | 328 | REG_EESCAPE, /* Trailing backslash. */ |
257 | REG_ESUBREG, /* Invalid back reference. */ | 329 | REG_ESUBREG, /* Invalid back reference. */ |
@@ -275,85 +347,92 @@ typedef enum | |||
275 | compiled, the `re_nsub' field is available. All other fields are | 347 | compiled, the `re_nsub' field is available. All other fields are |
276 | private to the regex routines. */ | 348 | private to the regex routines. */ |
277 | 349 | ||
350 | #ifndef RE_TRANSLATE_TYPE | ||
351 | # define __RE_TRANSLATE_TYPE unsigned char * | ||
352 | # ifdef __USE_GNU | ||
353 | # define RE_TRANSLATE_TYPE __RE_TRANSLATE_TYPE | ||
354 | # endif | ||
355 | #endif | ||
356 | |||
357 | #ifdef __USE_GNU | ||
358 | # define __REPB_PREFIX(name) name | ||
359 | #else | ||
360 | # define __REPB_PREFIX(name) __##name | ||
361 | #endif | ||
362 | |||
278 | struct re_pattern_buffer | 363 | struct re_pattern_buffer |
279 | { | 364 | { |
280 | /* [[[begin pattern_buffer]]] */ | 365 | /* Space that holds the compiled pattern. It is declared as |
281 | /* Space that holds the compiled pattern. It is declared as | 366 | `unsigned char *' because its elements are sometimes used as |
282 | `unsigned char *' because its elements are | 367 | array indexes. */ |
283 | sometimes used as array indexes. */ | 368 | unsigned char *__REPB_PREFIX(buffer); |
284 | unsigned char *buffer; | ||
285 | 369 | ||
286 | /* Number of bytes to which `buffer' points. */ | 370 | /* Number of bytes to which `buffer' points. */ |
287 | unsigned long allocated; | 371 | unsigned long int __REPB_PREFIX(allocated); |
288 | 372 | ||
289 | /* Number of bytes actually used in `buffer'. */ | 373 | /* Number of bytes actually used in `buffer'. */ |
290 | unsigned long used; | 374 | unsigned long int __REPB_PREFIX(used); |
291 | 375 | ||
292 | /* Syntax setting with which the pattern was compiled. */ | 376 | /* Syntax setting with which the pattern was compiled. */ |
293 | reg_syntax_t syntax; | 377 | reg_syntax_t __REPB_PREFIX(syntax); |
294 | 378 | ||
295 | /* Pointer to a fastmap, if any, otherwise zero. re_search uses | 379 | /* Pointer to a fastmap, if any, otherwise zero. re_search uses the |
296 | the fastmap, if there is one, to skip over impossible | 380 | fastmap, if there is one, to skip over impossible starting points |
297 | starting points for matches. */ | 381 | for matches. */ |
298 | char *fastmap; | 382 | char *__REPB_PREFIX(fastmap); |
299 | 383 | ||
300 | /* Either a translate table to apply to all characters before | 384 | /* Either a translate table to apply to all characters before |
301 | comparing them, or zero for no translation. The translation | 385 | comparing them, or zero for no translation. The translation is |
302 | is applied to a pattern when it is compiled and to a string | 386 | applied to a pattern when it is compiled and to a string when it |
303 | when it is matched. */ | 387 | is matched. */ |
304 | char *translate; | 388 | __RE_TRANSLATE_TYPE __REPB_PREFIX(translate); |
305 | 389 | ||
306 | /* Number of subexpressions found by the compiler. */ | 390 | /* Number of subexpressions found by the compiler. */ |
307 | size_t re_nsub; | 391 | size_t re_nsub; |
308 | 392 | ||
309 | /* Zero if this pattern cannot match the empty string, one else. | 393 | /* Zero if this pattern cannot match the empty string, one else. |
310 | Well, in truth it's used only in `re_search_2', to see | 394 | Well, in truth it's used only in `re_search_2', to see whether or |
311 | whether or not we should use the fastmap, so we don't set | 395 | not we should use the fastmap, so we don't set this absolutely |
312 | this absolutely perfectly; see `re_compile_fastmap' (the | 396 | perfectly; see `re_compile_fastmap' (the `duplicate' case). */ |
313 | `duplicate' case). */ | 397 | unsigned __REPB_PREFIX(can_be_null) : 1; |
314 | unsigned can_be_null : 1; | 398 | |
315 | 399 | /* If REGS_UNALLOCATED, allocate space in the `regs' structure | |
316 | /* If REGS_UNALLOCATED, allocate space in the `regs' structure | 400 | for `max (RE_NREGS, re_nsub + 1)' groups. |
317 | for `max (RE_NREGS, re_nsub + 1)' groups. | 401 | If REGS_REALLOCATE, reallocate space if necessary. |
318 | If REGS_REALLOCATE, reallocate space if necessary. | 402 | If REGS_FIXED, use what's there. */ |
319 | If REGS_FIXED, use what's there. */ | 403 | #ifdef __USE_GNU |
320 | #define REGS_UNALLOCATED 0 | 404 | # define REGS_UNALLOCATED 0 |
321 | #define REGS_REALLOCATE 1 | 405 | # define REGS_REALLOCATE 1 |
322 | #define REGS_FIXED 2 | 406 | # define REGS_FIXED 2 |
323 | unsigned regs_allocated : 2; | 407 | #endif |
324 | 408 | unsigned __REPB_PREFIX(regs_allocated) : 2; | |
325 | /* Set to zero when `regex_compile' compiles a pattern; set to one | ||
326 | by `re_compile_fastmap' if it updates the fastmap. */ | ||
327 | unsigned fastmap_accurate : 1; | ||
328 | |||
329 | /* If set, `re_match_2' does not return information about | ||
330 | subexpressions. */ | ||
331 | unsigned no_sub : 1; | ||
332 | |||
333 | /* If set, a beginning-of-line anchor doesn't match at the | ||
334 | beginning of the string. */ | ||
335 | unsigned not_bol : 1; | ||
336 | |||
337 | /* Similarly for an end-of-line anchor. */ | ||
338 | unsigned not_eol : 1; | ||
339 | |||
340 | /* If true, an anchor at a newline matches. */ | ||
341 | unsigned newline_anchor : 1; | ||
342 | |||
343 | /* [[[end pattern_buffer]]] */ | ||
344 | }; | ||
345 | 409 | ||
346 | typedef struct re_pattern_buffer regex_t; | 410 | /* Set to zero when `regex_compile' compiles a pattern; set to one |
411 | by `re_compile_fastmap' if it updates the fastmap. */ | ||
412 | unsigned __REPB_PREFIX(fastmap_accurate) : 1; | ||
413 | |||
414 | /* If set, `re_match_2' does not return information about | ||
415 | subexpressions. */ | ||
416 | unsigned __REPB_PREFIX(no_sub) : 1; | ||
417 | |||
418 | /* If set, a beginning-of-line anchor doesn't match at the beginning | ||
419 | of the string. */ | ||
420 | unsigned __REPB_PREFIX(not_bol) : 1; | ||
421 | |||
422 | /* Similarly for an end-of-line anchor. */ | ||
423 | unsigned __REPB_PREFIX(not_eol) : 1; | ||
347 | 424 | ||
425 | /* If true, an anchor at a newline matches. */ | ||
426 | unsigned __REPB_PREFIX(newline_anchor) : 1; | ||
427 | }; | ||
348 | 428 | ||
349 | /* search.c (search_buffer) in Emacs needs this one opcode value. It is | 429 | typedef struct re_pattern_buffer regex_t; |
350 | defined both in `regex.c' and here. */ | ||
351 | #define RE_EXACTN_VALUE 1 | ||
352 | 430 | ||
353 | /* Type for byte offsets within the string. POSIX mandates this. */ | 431 | /* Type for byte offsets within the string. POSIX mandates this. */ |
354 | typedef int regoff_t; | 432 | typedef int regoff_t; |
355 | 433 | ||
356 | 434 | ||
435 | #ifdef __USE_GNU | ||
357 | /* This is the structure we store register match data in. See | 436 | /* This is the structure we store register match data in. See |
358 | regex.texinfo for a full description of what registers match. */ | 437 | regex.texinfo for a full description of what registers match. */ |
359 | struct re_registers | 438 | struct re_registers |
@@ -367,8 +446,9 @@ struct re_registers | |||
367 | /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, | 446 | /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, |
368 | `re_match_2' returns information about at least this many registers | 447 | `re_match_2' returns information about at least this many registers |
369 | the first time a `regs' structure is passed. */ | 448 | the first time a `regs' structure is passed. */ |
370 | #ifndef RE_NREGS | 449 | # ifndef RE_NREGS |
371 | #define RE_NREGS 30 | 450 | # define RE_NREGS 30 |
451 | # endif | ||
372 | #endif | 452 | #endif |
373 | 453 | ||
374 | 454 | ||
@@ -383,38 +463,22 @@ typedef struct | |||
383 | 463 | ||
384 | /* Declarations for routines. */ | 464 | /* Declarations for routines. */ |
385 | 465 | ||
386 | /* To avoid duplicating every routine declaration -- once with a | 466 | #ifdef __USE_GNU |
387 | prototype (if we are ANSI), and once without (if we aren't) -- we | ||
388 | use the following macro to declare argument types. This | ||
389 | unfortunately clutters up the declarations a bit, but I think it's | ||
390 | worth it. */ | ||
391 | |||
392 | #if __STDC__ | ||
393 | |||
394 | #define _RE_ARGS(args) args | ||
395 | |||
396 | #else /* not __STDC__ */ | ||
397 | |||
398 | #define _RE_ARGS(args) () | ||
399 | |||
400 | #endif /* not __STDC__ */ | ||
401 | |||
402 | /* Sets the current default syntax to SYNTAX, and return the old syntax. | 467 | /* Sets the current default syntax to SYNTAX, and return the old syntax. |
403 | You can also simply assign to the `re_syntax_options' variable. */ | 468 | You can also simply assign to the `re_syntax_options' variable. */ |
404 | extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax)); | 469 | extern reg_syntax_t re_set_syntax (reg_syntax_t __syntax); |
405 | 470 | ||
406 | /* Compile the regular expression PATTERN, with length LENGTH | 471 | /* Compile the regular expression PATTERN, with length LENGTH |
407 | and syntax given by the global `re_syntax_options', into the buffer | 472 | and syntax given by the global `re_syntax_options', into the buffer |
408 | BUFFER. Return NULL if successful, and an error string if not. */ | 473 | BUFFER. Return NULL if successful, and an error string if not. */ |
409 | extern const char *re_compile_pattern | 474 | extern const char *re_compile_pattern (const char *__pattern, size_t __length, |
410 | _RE_ARGS ((const char *pattern, int length, | 475 | struct re_pattern_buffer *__buffer); |
411 | struct re_pattern_buffer *buffer)); | ||
412 | 476 | ||
413 | 477 | ||
414 | /* Compile a fastmap for the compiled pattern in BUFFER; used to | 478 | /* Compile a fastmap for the compiled pattern in BUFFER; used to |
415 | accelerate searches. Return 0 if successful and -2 if was an | 479 | accelerate searches. Return 0 if successful and -2 if was an |
416 | internal error. */ | 480 | internal error. */ |
417 | extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer)); | 481 | extern int re_compile_fastmap (struct re_pattern_buffer *__buffer); |
418 | 482 | ||
419 | 483 | ||
420 | /* Search in the string STRING (with length LENGTH) for the pattern | 484 | /* Search in the string STRING (with length LENGTH) for the pattern |
@@ -422,31 +486,30 @@ extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer)); | |||
422 | characters. Return the starting position of the match, -1 for no | 486 | characters. Return the starting position of the match, -1 for no |
423 | match, or -2 for an internal error. Also return register | 487 | match, or -2 for an internal error. Also return register |
424 | information in REGS (if REGS and BUFFER->no_sub are nonzero). */ | 488 | information in REGS (if REGS and BUFFER->no_sub are nonzero). */ |
425 | extern int re_search | 489 | extern int re_search (struct re_pattern_buffer *__buffer, const char *__cstring, |
426 | _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, | 490 | int __length, int __start, int __range, |
427 | int length, int start, int range, struct re_registers *regs)); | 491 | struct re_registers *__regs); |
428 | 492 | ||
429 | 493 | ||
430 | /* Like `re_search', but search in the concatenation of STRING1 and | 494 | /* Like `re_search', but search in the concatenation of STRING1 and |
431 | STRING2. Also, stop searching at index START + STOP. */ | 495 | STRING2. Also, stop searching at index START + STOP. */ |
432 | extern int re_search_2 | 496 | extern int re_search_2 (struct re_pattern_buffer *__buffer, |
433 | _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, | 497 | const char *__string1, int __length1, |
434 | int length1, const char *string2, int length2, | 498 | const char *__string2, int __length2, int __start, |
435 | int start, int range, struct re_registers *regs, int stop)); | 499 | int __range, struct re_registers *__regs, int __stop); |
436 | 500 | ||
437 | 501 | ||
438 | /* Like `re_search', but return how many characters in STRING the regexp | 502 | /* Like `re_search', but return how many characters in STRING the regexp |
439 | in BUFFER matched, starting at position START. */ | 503 | in BUFFER matched, starting at position START. */ |
440 | extern int re_match | 504 | extern int re_match (struct re_pattern_buffer *__buffer, const char *__cstring, |
441 | _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, | 505 | int __length, int __start, struct re_registers *__regs); |
442 | int length, int start, struct re_registers *regs)); | ||
443 | 506 | ||
444 | 507 | ||
445 | /* Relates to `re_match' as `re_search_2' relates to `re_search'. */ | 508 | /* Relates to `re_match' as `re_search_2' relates to `re_search'. */ |
446 | extern int re_match_2 | 509 | extern int re_match_2 (struct re_pattern_buffer *__buffer, |
447 | _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, | 510 | const char *__string1, int __length1, |
448 | int length1, const char *string2, int length2, | 511 | const char *__string2, int __length2, int __start, |
449 | int start, struct re_registers *regs, int stop)); | 512 | struct re_registers *__regs, int __stop); |
450 | 513 | ||
451 | 514 | ||
452 | /* Set REGS to hold NUM_REGS registers, storing them in STARTS and | 515 | /* Set REGS to hold NUM_REGS registers, storing them in STARTS and |
@@ -461,30 +524,59 @@ extern int re_match_2 | |||
461 | Unless this function is called, the first search or match using | 524 | Unless this function is called, the first search or match using |
462 | PATTERN_BUFFER will allocate its own register data, without | 525 | PATTERN_BUFFER will allocate its own register data, without |
463 | freeing the old data. */ | 526 | freeing the old data. */ |
464 | extern void re_set_registers | 527 | extern void re_set_registers (struct re_pattern_buffer *__buffer, |
465 | _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs, | 528 | struct re_registers *__regs, |
466 | unsigned num_regs, regoff_t *starts, regoff_t *ends)); | 529 | unsigned int __num_regs, |
467 | 530 | regoff_t *__starts, regoff_t *__ends); | |
531 | #endif /* Use GNU */ | ||
532 | |||
533 | #if defined _REGEX_RE_COMP || (defined _LIBC && defined __USE_BSD) | ||
534 | # ifndef _CRAY | ||
468 | /* 4.2 bsd compatibility. */ | 535 | /* 4.2 bsd compatibility. */ |
469 | extern char *re_comp _RE_ARGS ((const char *)); | 536 | extern char *re_comp (const char *); |
470 | extern int re_exec _RE_ARGS ((const char *)); | 537 | extern int re_exec (const char *); |
538 | # endif | ||
539 | #endif | ||
540 | |||
541 | /* GCC 2.95 and later have "__restrict"; C99 compilers have | ||
542 | "restrict", and "configure" may have defined "restrict". */ | ||
543 | #ifndef __restrict | ||
544 | # if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__)) | ||
545 | # if defined restrict || 199901L <= __STDC_VERSION__ | ||
546 | # define __restrict restrict | ||
547 | # else | ||
548 | # define __restrict | ||
549 | # endif | ||
550 | # endif | ||
551 | #endif | ||
552 | /* gcc 3.1 and up support the [restrict] syntax. */ | ||
553 | #ifndef __restrict_arr | ||
554 | # if (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)) \ | ||
555 | && !defined __GNUG__ | ||
556 | # define __restrict_arr __restrict | ||
557 | # else | ||
558 | # define __restrict_arr | ||
559 | # endif | ||
560 | #endif | ||
471 | 561 | ||
472 | /* POSIX compatibility. */ | 562 | /* POSIX compatibility. */ |
473 | extern int regcomp _RE_ARGS ((regex_t *preg, const char *pattern, int cflags)); | 563 | extern int regcomp (regex_t *__restrict __preg, |
474 | extern int regexec | 564 | const char *__restrict __pattern, |
475 | _RE_ARGS ((const regex_t *preg, const char *string, size_t nmatch, | 565 | int __cflags); |
476 | regmatch_t pmatch[], int eflags)); | 566 | |
477 | extern size_t regerror | 567 | extern int regexec (const regex_t *__restrict __preg, |
478 | _RE_ARGS ((int errcode, const regex_t *preg, char *errbuf, | 568 | const char *__restrict __cstring, size_t __nmatch, |
479 | size_t errbuf_size)); | 569 | regmatch_t __pmatch[__restrict_arr], |
480 | extern void regfree _RE_ARGS ((regex_t *preg)); | 570 | int __eflags); |
481 | 571 | ||
482 | #endif /* not __REGEXP_LIBRARY_H__ */ | 572 | extern size_t regerror (int __errcode, const regex_t *__restrict __preg, |
483 | 573 | char *__restrict __errbuf, size_t __errbuf_size); | |
484 | /* | 574 | |
485 | Local variables: | 575 | extern void regfree (regex_t *__preg); |
486 | make-backup-files: t | 576 | |
487 | version-control: t | 577 | |
488 | trim-versions-without-asking: nil | 578 | #ifdef __cplusplus |
489 | End: | 579 | } |
490 | */ | 580 | #endif /* C++ */ |
581 | |||
582 | #endif /* regex.h */ | ||