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