1 /* Definitions for data structures and routines for the regular 2 expression library, version 0.12. 3 4 Copyright (C) 1985, 1989, 1990, 1991, 1992, 1993 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 #ifndef __REGEXP_LIBRARY_H__ 17 #define __REGEXP_LIBRARY_H__ 18 19 /* POSIX says that <sys/types.h> must be included (by the caller) before 20 <regex.h>. */ 21 22 #ifdef VMS 23 /* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it 24 should be there. */ 25 #include <stddef.h> 26 #endif 27 28 29 /* The following bits are used to determine the regexp syntax we 30 recognize. The set/not-set meanings are chosen so that Emacs syntax 31 remains the value 0. The bits are given in alphabetical order, and 32 the definitions shifted by one from the previous bit; thus, when we 33 add or remove a bit, only one other definition need change. */ 34 typedef unsigned reg_syntax_t; 35 36 /* If this bit is not set, then \ inside a bracket expression is literal. 37 If set, then such a \ quotes the following character. */ 38 #define RE_BACKSLASH_ESCAPE_IN_LISTS (1) 39 40 /* If this bit is not set, then + and ? are operators, and \+ and \? are 41 literals. 42 If set, then \+ and \? are operators and + and ? are literals. */ 43 #define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) 44 45 /* If this bit is set, then character classes are supported. They are: 46 [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], 47 [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. 48 If not set, then character classes are not supported. */ 49 #define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) 50 51 /* If this bit is set, then ^ and $ are always anchors (outside bracket 52 expressions, of course). 53 If this bit is not set, then it depends: 54 ^ is an anchor if it is at the beginning of a regular 55 expression or after an open-group or an alternation operator; 56 $ is an anchor if it is at the end of a regular expression, or 57 before a close-group or an alternation operator. 58 59 This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because 60 POSIX draft 11.2 says that * etc. in leading positions is undefined. 61 We already implemented a previous draft which made those constructs 62 invalid, though, so we haven't changed the code back. */ 63 #define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) 64 65 /* If this bit is set, then special characters are always special 66 regardless of where they are in the pattern. 67 If this bit is not set, then special characters are special only in 68 some contexts; otherwise they are ordinary. Specifically, 69 * + ? and intervals are only special when not after the beginning, 70 open-group, or alternation operator. */ 71 #define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) 72 73 /* If this bit is set, then *, +, ?, and { cannot be first in an re or 74 immediately after an alternation or begin-group operator. */ 75 #define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) 76 77 /* If this bit is set, then . matches newline. 78 If not set, then it doesn't. */ 79 #define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) 80 81 /* If this bit is set, then . doesn't match NUL. 82 If not set, then it does. */ 83 #define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) 84 85 /* If this bit is set, nonmatching lists [^...] do not match newline. 86 If not set, they do. */ 87 #define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) 88 89 /* If this bit is set, either \{...\} or {...} defines an 90 interval, depending on RE_NO_BK_BRACES. 91 If not set, \{, \}, {, and } are literals. */ 92 #define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) 93 94 /* If this bit is set, +, ? and | aren't recognized as operators. 95 If not set, they are. */ 96 #define RE_LIMITED_OPS (RE_INTERVALS << 1) 97 98 /* If this bit is set, newline is an alternation operator. 99 If not set, newline is literal. */ 100 #define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) 101 102 /* If this bit is set, then `{...}' defines an interval, and \{ and \} 103 are literals. 104 If not set, then `\{...\}' defines an interval. */ 105 #define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) 106 107 /* If this bit is set, (...) defines a group, and \( and \) are literals. 108 If not set, \(...\) defines a group, and ( and ) are literals. */ 109 #define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) 110 111 /* If this bit is set, then \<digit> matches <digit>. 112 If not set, then \<digit> is a back-reference. */ 113 #define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) 114 115 /* If this bit is set, then | is an alternation operator, and \| is literal. 116 If not set, then \| is an alternation operator, and | is literal. */ 117 #define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) 118 119 /* If this bit is set, then an ending range point collating higher 120 than the starting range point, as in [z-a], is invalid. 121 If not set, then when ending range point collates higher than the 122 starting range point, the range is ignored. */ 123 #define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) 124 125 /* If this bit is set, then an unmatched ) is ordinary. 126 If not set, then an unmatched ) is invalid. */ 127 #define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) 128 129 /* This global variable defines the particular regexp syntax to use (for 130 some interfaces). When a regexp is compiled, the syntax used is 131 stored in the pattern buffer, so changing this does not affect 132 already-compiled regexps. */ 133 extern reg_syntax_t re_syntax_options; 134 135 /* Define combinations of the above bits for the standard possibilities. 136 (The [[[ comments delimit what gets put into the Texinfo file, so 137 don't delete them!) */ 138 /* [[[begin syntaxes]]] */ 139 #define RE_SYNTAX_EMACS 0 140 141 #define RE_SYNTAX_AWK \ 142 (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ 143 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ 144 | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ 145 | RE_UNMATCHED_RIGHT_PAREN_ORD) 146 147 #define RE_SYNTAX_POSIX_AWK \ 148 (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS) 149 150 #define RE_SYNTAX_GREP \ 151 (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ 152 | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ 153 | RE_NEWLINE_ALT) 154 155 #define RE_SYNTAX_EGREP \ 156 (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ 157 | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ 158 | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ 159 | RE_NO_BK_VBAR) 160 161 #define RE_SYNTAX_POSIX_EGREP \ 162 (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES) 163 164 /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ 165 #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC 166 167 #define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC 168 169 /* Syntax bits common to both basic and extended POSIX regex syntax. */ 170 #define _RE_SYNTAX_POSIX_COMMON \ 171 (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ 172 | RE_INTERVALS | RE_NO_EMPTY_RANGES) 173 174 #define RE_SYNTAX_POSIX_BASIC \ 175 (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM) 176 177 /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes 178 RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this 179 isn't minimal, since other operators, such as \`, aren't disabled. */ 180 #define RE_SYNTAX_POSIX_MINIMAL_BASIC \ 181 (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) 182 183 #define RE_SYNTAX_POSIX_EXTENDED \ 184 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ 185 | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ 186 | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ 187 | RE_UNMATCHED_RIGHT_PAREN_ORD) 188 189 /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS 190 replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added. */ 191 #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ 192 (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ 193 | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ 194 | RE_NO_BK_PARENS | RE_NO_BK_REFS \ 195 | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) 196 /* [[[end syntaxes]]] */ 197 198 /* Maximum number of duplicates an interval can allow. Some systems 199 (erroneously) define this in other header files, but we want our 200 value, so remove any previous define. */ 201 #ifdef RE_DUP_MAX 202 #undef RE_DUP_MAX 203 #endif 204 #define RE_DUP_MAX ((1 << 15) - 1) 205 206 207 /* POSIX `cflags' bits (i.e., information for `regcomp'). */ 208 209 /* If this bit is set, then use extended regular expression syntax. 210 If not set, then use basic regular expression syntax. */ 211 #define REG_EXTENDED 1 212 213 /* If this bit is set, then ignore case when matching. 214 If not set, then case is significant. */ 215 #define REG_ICASE (REG_EXTENDED << 1) 216 217 /* If this bit is set, then anchors do not match at newline 218 characters in the string. 219 If not set, then anchors do match at newlines. */ 220 #define REG_NEWLINE (REG_ICASE << 1) 221 222 /* If this bit is set, then report only success or fail in regexec. 223 If not set, then returns differ between not matching and errors. */ 224 #define REG_NOSUB (REG_NEWLINE << 1) 225 226 227 /* POSIX `eflags' bits (i.e., information for regexec). */ 228 229 /* If this bit is set, then the beginning-of-line operator doesn't match 230 the beginning of the string (presumably because it's not the 231 beginning of a line). 232 If not set, then the beginning-of-line operator does match the 233 beginning of the string. */ 234 #define REG_NOTBOL 1 235 236 /* Like REG_NOTBOL, except for the end-of-line. */ 237 #define REG_NOTEOL (1 << 1) 238 239 240 /* If any error codes are removed, changed, or added, update the 241 `re_error_msg' table in regex.c. */ 242 typedef enum 243 { 244 REG_NOERROR = 0, /* Success. */ 245 REG_NOMATCH, /* Didn't find a match (for regexec). */ 246 247 /* POSIX regcomp return error codes. (In the order listed in the 248 standard.) */ 249 REG_BADPAT, /* Invalid pattern. */ 250 REG_ECOLLATE, /* Not implemented. */ 251 REG_ECTYPE, /* Invalid character class name. */ 252 REG_EESCAPE, /* Trailing backslash. */ 253 REG_ESUBREG, /* Invalid back reference. */ 254 REG_EBRACK, /* Unmatched left bracket. */ 255 REG_EPAREN, /* Parenthesis imbalance. */ 256 REG_EBRACE, /* Unmatched \{. */ 257 REG_BADBR, /* Invalid contents of \{\}. */ 258 REG_ERANGE, /* Invalid range end. */ 259 REG_ESPACE, /* Ran out of memory. */ 260 REG_BADRPT, /* No preceding re for repetition op. */ 261 262 /* Error codes we've added. */ 263 REG_EEND, /* Premature end. */ 264 REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ 265 REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ 266 } reg_errcode_t; 267 268 /* This data structure represents a compiled pattern. Before calling 269 the pattern compiler, the fields `buffer', `allocated', `fastmap', 270 `translate', and `no_sub' can be set. After the pattern has been 271 compiled, the `re_nsub' field is available. All other fields are 272 private to the regex routines. */ 273 274 struct re_pattern_buffer 275 { 276 /* [[[begin pattern_buffer]]] */ 277 /* Space that holds the compiled pattern. It is declared as 278 `unsigned char *' because its elements are 279 sometimes used as array indexes. */ 280 unsigned char *buffer; 281 282 /* Number of bytes to which `buffer' points. */ 283 unsigned long allocated; 284 285 /* Number of bytes actually used in `buffer'. */ 286 unsigned long used; 287 288 /* Syntax setting with which the pattern was compiled. */ 289 reg_syntax_t syntax; 290 291 /* Pointer to a fastmap, if any, otherwise zero. re_search uses 292 the fastmap, if there is one, to skip over impossible 293 starting points for matches. */ 294 char *fastmap; 295 296 /* Either a translate table to apply to all characters before 297 comparing them, or zero for no translation. The translation 298 is applied to a pattern when it is compiled and to a string 299 when it is matched. */ 300 char *translate; 301 302 /* Number of subexpressions found by the compiler. */ 303 size_t re_nsub; 304 305 /* Zero if this pattern cannot match the empty string, one else. 306 Well, in truth it's used only in `re_search_2', to see 307 whether or not we should use the fastmap, so we don't set 308 this absolutely perfectly; see `re_compile_fastmap' (the 309 `duplicate' case). */ 310 unsigned can_be_null : 1; 311 312 /* If REGS_UNALLOCATED, allocate space in the `regs' structure 313 for `max (RE_NREGS, re_nsub + 1)' groups. 314 If REGS_REALLOCATE, reallocate space if necessary. 315 If REGS_FIXED, use what's there. */ 316 #define REGS_UNALLOCATED 0 317 #define REGS_REALLOCATE 1 318 #define REGS_FIXED 2 319 unsigned regs_allocated : 2; 320 321 /* Set to zero when `regex_compile' compiles a pattern; set to one 322 by `re_compile_fastmap' if it updates the fastmap. */ 323 unsigned fastmap_accurate : 1; 324 325 /* If set, `re_match_2' does not return information about 326 subexpressions. */ 327 unsigned no_sub : 1; 328 329 /* If set, a beginning-of-line anchor doesn't match at the 330 beginning of the string. */ 331 unsigned not_bol : 1; 332 333 /* Similarly for an end-of-line anchor. */ 334 unsigned not_eol : 1; 335 336 /* If true, an anchor at a newline matches. */ 337 unsigned newline_anchor : 1; 338 339 /* [[[end pattern_buffer]]] */ 340 }; 341 342 typedef struct re_pattern_buffer regex_t; 343 344 345 /* search.c (search_buffer) in Emacs needs this one opcode value. It is 346 defined both in `regex.c' and here. */ 347 #define RE_EXACTN_VALUE 1 348 349 /* Type for byte offsets within the string. POSIX mandates this. */ 350 typedef int regoff_t; 351 352 353 /* This is the structure we store register match data in. See 354 regex.texinfo for a full description of what registers match. */ 355 struct re_registers 356 { 357 unsigned num_regs; 358 regoff_t *start; 359 regoff_t *end; 360 }; 361 362 363 /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, 364 `re_match_2' returns information about at least this many registers 365 the first time a `regs' structure is passed. */ 366 #ifndef RE_NREGS 367 #define RE_NREGS 30 368 #endif 369 370 371 /* POSIX specification for registers. Aside from the different names than 372 `re_registers', POSIX uses an array of structures, instead of a 373 structure of arrays. */ 374 typedef struct 375 { 376 regoff_t rm_so; /* Byte offset from string's start to substring's start. */ 377 regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ 378 } regmatch_t; 379 380 /* Declarations for routines. */ 381 382 /* To avoid duplicating every routine declaration -- once with a 383 prototype (if we are ANSI), and once without (if we aren't) -- we 384 use the following macro to declare argument types. This 385 unfortunately clutters up the declarations a bit, but I think it's 386 worth it. */ 387 388 #if __STDC__ 389 390 #define _RE_ARGS(args) args 391 392 #else /* not __STDC__ */ 393 394 #define _RE_ARGS(args) () 395 396 #endif /* not __STDC__ */ 397 398 /* Sets the current default syntax to SYNTAX, and return the old syntax. 399 You can also simply assign to the `re_syntax_options' variable. */ 400 extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax)); 401 402 /* Compile the regular expression PATTERN, with length LENGTH 403 and syntax given by the global `re_syntax_options', into the buffer 404 BUFFER. Return NULL if successful, and an error string if not. */ 405 extern const char *re_compile_pattern 406 _RE_ARGS ((const char *pattern, int length, 407 struct re_pattern_buffer *buffer)); 408 409 410 /* Compile a fastmap for the compiled pattern in BUFFER; used to 411 accelerate searches. Return 0 if successful and -2 if was an 412 internal error. */ 413 extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer)); 414 415 416 /* Search in the string STRING (with length LENGTH) for the pattern 417 compiled into BUFFER. Start searching at position START, for RANGE 418 characters. Return the starting position of the match, -1 for no 419 match, or -2 for an internal error. Also return register 420 information in REGS (if REGS and BUFFER->no_sub are nonzero). */ 421 extern int re_search 422 _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, 423 int length, int start, int range, struct re_registers *regs)); 424 425 426 /* Like `re_search', but search in the concatenation of STRING1 and 427 STRING2. Also, stop searching at index START + STOP. */ 428 extern int re_search_2 429 _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, 430 int length1, const char *string2, int length2, 431 int start, int range, struct re_registers *regs, int stop)); 432 433 434 /* Like `re_search', but return how many characters in STRING the regexp 435 in BUFFER matched, starting at position START. */ 436 extern int re_match 437 _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, 438 int length, int start, struct re_registers *regs)); 439 440 441 /* Relates to `re_match' as `re_search_2' relates to `re_search'. */ 442 extern int re_match_2 443 _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, 444 int length1, const char *string2, int length2, 445 int start, struct re_registers *regs, int stop)); 446 447 448 /* Set REGS to hold NUM_REGS registers, storing them in STARTS and 449 ENDS. Subsequent matches using BUFFER and REGS will use this memory 450 for recording register information. STARTS and ENDS must be 451 allocated with malloc, and must each be at least `NUM_REGS * sizeof 452 (regoff_t)' bytes long. 453 454 If NUM_REGS == 0, then subsequent matches should allocate their own 455 register data. 456 457 Unless this function is called, the first search or match using 458 PATTERN_BUFFER will allocate its own register data, without 459 freeing the old data. */ 460 extern void re_set_registers 461 _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs, 462 unsigned num_regs, regoff_t *starts, regoff_t *ends)); 463 464 /* 4.2 bsd compatibility. System headers may declare the argument as 465 either "char *" (e.g. Cray unistd.h) or "const char *" (e.g. linux 466 regex.h), so don't prototype them here. */ 467 extern char *re_comp (); 468 extern int re_exec (); 469 470 /* POSIX compatibility. */ 471 extern int regcomp _RE_ARGS ((regex_t *preg, const char *pattern, int cflags)); 472 extern int regexec 473 _RE_ARGS ((const regex_t *preg, const char *string, size_t nmatch, 474 regmatch_t pmatch[], int eflags)); 475 extern size_t regerror 476 _RE_ARGS ((int errcode, const regex_t *preg, char *errbuf, 477 size_t errbuf_size)); 478 extern void regfree _RE_ARGS ((regex_t *preg)); 479 480 #endif /* not __REGEXP_LIBRARY_H__ */ 481 482 /* 483 Local variables: 484 make-backup-files: t 485 version-control: t 486 trim-versions-without-asking: nil 487 End: 488 */ 489