src/regex.h

   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
  18    Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
  19    USA.  */
  20
  21 #ifndef __REGEXP_LIBRARY_H__
  22 #define __REGEXP_LIBRARY_H__
  23
  24 /* POSIX says that <sys/types.h> must be included (by the caller) before
  25    <regex.h>.  */
  26
  27 #if !defined (_POSIX_C_SOURCE) && !defined (_POSIX_SOURCE) && defined (VMS)
  28 /* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it
  29    should be there.  */
  30 #include <stddef.h>
  31 #endif
  32
  33
  34 /* The following bits are used to determine the regexp syntax we
  35    recognize.  The set/not-set meanings are chosen so that Emacs syntax
  36    remains the value 0.  The bits are given in alphabetical order, and
  37    the definitions shifted by one from the previous bit; thus, when we
  38    add or remove a bit, only one other definition need change.  */
  39 typedef unsigned reg_syntax_t;
  40
  41 /* If this bit is not set, then \ inside a bracket expression is literal.
  42    If set, then such a \ quotes the following character.  */
  43 #define RE_BACKSLASH_ESCAPE_IN_LISTS (1)
  44
  45 /* If this bit is not set, then + and ? are operators, and \+ and \? are
  46      literals.
  47    If set, then \+ and \? are operators and + and ? are literals.  */
  48 #define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)
  49
  50 /* If this bit is set, then character classes are supported.  They are:
  51      [:alpha:], [:upper:], [:lower:],  [:digit:], [:alnum:], [:xdigit:],
  52      [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
  53    If not set, then character classes are not supported.  */
  54 #define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)
  55
  56 /* If this bit is set, then ^ and $ are always anchors (outside bracket
  57      expressions, of course).
  58    If this bit is not set, then it depends:
  59         ^  is an anchor if it is at the beginning of a regular
  60            expression or after an open-group or an alternation operator;
  61         $  is an anchor if it is at the end of a regular expression, or
  62            before a close-group or an alternation operator.
  63
  64    This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because
  65    POSIX draft 11.2 says that * etc. in leading positions is undefined.
  66    We already implemented a previous draft which made those constructs
  67    invalid, though, so we haven't changed the code back.  */
  68 #define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)
  69
  70 /* If this bit is set, then special characters are always special
  71      regardless of where they are in the pattern.
  72    If this bit is not set, then special characters are special only in
  73      some contexts; otherwise they are ordinary.  Specifically,
  74      * + ? and intervals are only special when not after the beginning,
  75      open-group, or alternation operator.  */
  76 #define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)
  77
  78 /* If this bit is set, then *, +, ?, and { cannot be first in an re or
  79      immediately after an alternation or begin-group operator.  */
  80 #define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)
  81
  82 /* If this bit is set, then . matches newline.
  83    If not set, then it doesn't.  */
  84 #define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)
  85
  86 /* If this bit is set, then . doesn't match NUL.
  87    If not set, then it does.  */
  88 #define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)
  89
  90 /* If this bit is set, nonmatching lists [^...] do not match newline.
  91    If not set, they do.  */
  92 #define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)
  93
  94 /* If this bit is set, either \{...\} or {...} defines an
  95      interval, depending on RE_NO_BK_BRACES.
  96    If not set, \{, \}, {, and } are literals.  */
  97 #define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)
  98
  99 /* If this bit is set, +, ? and | aren't recognized as operators.
 100    If not set, they are.  */
 101 #define RE_LIMITED_OPS (RE_INTERVALS << 1)
 102
 103 /* If this bit is set, newline is an alternation operator.
 104    If not set, newline is literal.  */
 105 #define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)
 106
 107 /* If this bit is set, then `{...}' defines an interval, and \{ and \}
 108      are literals.
 109   If not set, then `\{...\}' defines an interval.  */
 110 #define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)
 111
 112 /* If this bit is set, (...) defines a group, and \( and \) are literals.
 113    If not set, \(...\) defines a group, and ( and ) are literals.  */
 114 #define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)
 115
 116 /* If this bit is set, then \<digit> matches <digit>.
 117    If not set, then \<digit> is a back-reference.  */
 118 #define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)
 119
 120 /* If this bit is set, then | is an alternation operator, and \| is literal.
 121    If not set, then \| is an alternation operator, and | is literal.  */
 122 #define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)
 123
 124 /* If this bit is set, then an ending range point collating higher
 125      than the starting range point, as in [z-a], is invalid.
 126    If not set, then when ending range point collates higher than the
 127      starting range point, the range is ignored.  */
 128 #define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1)
 129
 130 /* If this bit is set, then an unmatched ) is ordinary.
 131    If not set, then an unmatched ) is invalid.  */
 132 #define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1)
 133
 134 /* If this bit is set, succeed as soon as we match the whole pattern,
 135    without further backtracking.  */
 136 #define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1)
 137
 138 /* This global variable defines the particular regexp syntax to use (for
 139    some interfaces).  When a regexp is compiled, the syntax used is
 140    stored in the pattern buffer, so changing this does not affect
 141    already-compiled regexps.  */
 142 extern reg_syntax_t re_syntax_options;
 143
 144 #ifdef emacs
 145 /* In Emacs, this is the string or buffer in which we
 146    are matching.  It is used for looking up syntax properties.  */
 147 extern Lisp_Object re_match_object;
 148 #endif
 149
 150 \f
 151 /* Define combinations of the above bits for the standard possibilities.
 152    (The [[[ comments delimit what gets put into the Texinfo file, so
 153    don't delete them!)  */
 154 /* [[[begin syntaxes]]] */
 155 #define RE_SYNTAX_EMACS 0
 156
 157 #define RE_SYNTAX_AWK                                                   \
 158   (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL                       \
 159    | RE_NO_BK_PARENS            | RE_NO_BK_REFS                         \
 160    | RE_NO_BK_VBAR               | RE_NO_EMPTY_RANGES                   \
 161    | RE_UNMATCHED_RIGHT_PAREN_ORD)
 162
 163 #define RE_SYNTAX_POSIX_AWK                                             \
 164   (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS)
 165
 166 #define RE_SYNTAX_GREP                                                  \
 167   (RE_BK_PLUS_QM              | RE_CHAR_CLASSES                         \
 168    | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS                            \
 169    | RE_NEWLINE_ALT)
 170
 171 #define RE_SYNTAX_EGREP                                                 \
 172   (RE_CHAR_CLASSES        | RE_CONTEXT_INDEP_ANCHORS                    \
 173    | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE                    \
 174    | RE_NEWLINE_ALT       | RE_NO_BK_PARENS                             \
 175    | RE_NO_BK_VBAR)
 176
 177 #define RE_SYNTAX_POSIX_EGREP                                           \
 178   (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES)
 179
 180 /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff.  */
 181 #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC
 182
 183 #define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC
 184
 185 /* Syntax bits common to both basic and extended POSIX regex syntax.  */
 186 #define _RE_SYNTAX_POSIX_COMMON                                         \
 187   (RE_CHAR_CLASSES | RE_DOT_NEWLINE      | RE_DOT_NOT_NULL              \
 188    | RE_INTERVALS  | RE_NO_EMPTY_RANGES)
 189
 190 #define RE_SYNTAX_POSIX_BASIC                                           \
 191   (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM)
 192
 193 /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes
 194    RE_LIMITED_OPS, i.e., \? \+ \| are not recognized.  Actually, this
 195    isn't minimal, since other operators, such as \`, aren't disabled.  */
 196 #define RE_SYNTAX_POSIX_MINIMAL_BASIC                                   \
 197   (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS)
 198
 199 #define RE_SYNTAX_POSIX_EXTENDED                                        \
 200   (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS                   \
 201    | RE_CONTEXT_INDEP_OPS  | RE_NO_BK_BRACES                            \
 202    | RE_NO_BK_PARENS       | RE_NO_BK_VBAR                              \
 203    | RE_UNMATCHED_RIGHT_PAREN_ORD)
 204
 205 /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS
 206    replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added.  */
 207 #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED                                \
 208   (_RE_SYNTAX_POSIX_COMMON  | RE_CONTEXT_INDEP_ANCHORS                  \
 209    | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES                           \
 210    | RE_NO_BK_PARENS        | RE_NO_BK_REFS                             \
 211    | RE_NO_BK_VBAR          | RE_UNMATCHED_RIGHT_PAREN_ORD)
 212 /* [[[end syntaxes]]] */
 213 \f
 214 /* Maximum number of duplicates an interval can allow.  Some systems
 215    (erroneously) define this in other header files, but we want our
 216    value, so remove any previous define.  */
 217 #ifdef RE_DUP_MAX
 218 #undef RE_DUP_MAX
 219 #endif
 220 #define RE_DUP_MAX ((1 << 15) - 1)
 221
 222
 223 /* POSIX `cflags' bits (i.e., information for `regcomp').  */
 224
 225 /* If this bit is set, then use extended regular expression syntax.
 226    If not set, then use basic regular expression syntax.  */
 227 #define REG_EXTENDED 1
 228
 229 /* If this bit is set, then ignore case when matching.
 230    If not set, then case is significant.  */
 231 #define REG_ICASE (REG_EXTENDED << 1)
 232
 233 /* If this bit is set, then anchors do not match at newline
 234      characters in the string.
 235    If not set, then anchors do match at newlines.  */
 236 #define REG_NEWLINE (REG_ICASE << 1)
 237
 238 /* If this bit is set, then report only success or fail in regexec.
 239    If not set, then returns differ between not matching and errors.  */
 240 #define REG_NOSUB (REG_NEWLINE << 1)
 241
 242
 243 /* POSIX `eflags' bits (i.e., information for regexec).  */
 244
 245 /* If this bit is set, then the beginning-of-line operator doesn't match
 246      the beginning of the string (presumably because it's not the
 247      beginning of a line).
 248    If not set, then the beginning-of-line operator does match the
 249      beginning of the string.  */
 250 #define REG_NOTBOL 1
 251
 252 /* Like REG_NOTBOL, except for the end-of-line.  */
 253 #define REG_NOTEOL (1 << 1)
 254
 255
 256 /* If any error codes are removed, changed, or added, update the
 257    `re_error_msg' table in regex.c.  */
 258 typedef enum
 259 {
 260   REG_NOERROR = 0,      /* Success.  */
 261   REG_NOMATCH,          /* Didn't find a match (for regexec).  */
 262
 263   /* POSIX regcomp return error codes.  (In the order listed in the
 264      standard.)  */
 265   REG_BADPAT,           /* Invalid pattern.  */
 266   REG_ECOLLATE,         /* Not implemented.  */
 267   REG_ECTYPE,           /* Invalid character class name.  */
 268   REG_EESCAPE,          /* Trailing backslash.  */
 269   REG_ESUBREG,          /* Invalid back reference.  */
 270   REG_EBRACK,           /* Unmatched left bracket.  */
 271   REG_EPAREN,           /* Parenthesis imbalance.  */
 272   REG_EBRACE,           /* Unmatched \{.  */
 273   REG_BADBR,            /* Invalid contents of \{\}.  */
 274   REG_ERANGE,           /* Invalid range end.  */
 275   REG_ESPACE,           /* Ran out of memory.  */
 276   REG_BADRPT,           /* No preceding re for repetition op.  */
 277
 278   /* Error codes we've added.  */
 279   REG_EEND,             /* Premature end.  */
 280   REG_ESIZE,            /* Compiled pattern bigger than 2^16 bytes.  */
 281   REG_ERPAREN           /* Unmatched ) or \); not returned from regcomp.  */
 282 } reg_errcode_t;
 283 \f
 284 /* This data structure represents a compiled pattern.  Before calling
 285    the pattern compiler, the fields `buffer', `allocated', `fastmap',
 286    `translate', and `no_sub' can be set.  After the pattern has been
 287    compiled, the `re_nsub' field is available.  All other fields are
 288    private to the regex routines.  */
 289
 290 #ifndef RE_TRANSLATE_TYPE
 291 #define RE_TRANSLATE_TYPE char *
 292 #define RE_TRANSLATE(TBL, C) ((TBL)[C])
 293 #define RE_TRANSLATE_P(TBL) (TBL)
 294 #endif
 295
 296 struct re_pattern_buffer
 297 {
 298 /* [[[begin pattern_buffer]]] */
 299         /* Space that holds the compiled pattern.  It is declared as
 300           `unsigned char *' because its elements are
 301            sometimes used as array indexes.  */
 302   unsigned char *buffer;
 303
 304         /* Number of bytes to which `buffer' points.  */
 305   unsigned long allocated;
 306
 307         /* Number of bytes actually used in `buffer'.  */
 308   unsigned long used;
 309
 310         /* Syntax setting with which the pattern was compiled.  */
 311   reg_syntax_t syntax;
 312
 313         /* Pointer to a fastmap, if any, otherwise zero.  re_search uses
 314            the fastmap, if there is one, to skip over impossible
 315            starting points for matches.  */
 316   char *fastmap;
 317
 318         /* Either a translate table to apply to all characters before
 319            comparing them, or zero for no translation.  The translation
 320            is applied to a pattern when it is compiled and to a string
 321            when it is matched.  */
 322   RE_TRANSLATE_TYPE translate;
 323
 324         /* Number of subexpressions found by the compiler.  */
 325   size_t re_nsub;
 326
 327         /* Zero if this pattern cannot match the empty string, one else.
 328            Well, in truth it's used only in `re_search_2', to see
 329            whether or not we should use the fastmap, so we don't set
 330            this absolutely perfectly; see `re_compile_fastmap' (the
 331            `duplicate' case).  */
 332   unsigned can_be_null : 1;
 333
 334         /* If REGS_UNALLOCATED, allocate space in the `regs' structure
 335              for `max (RE_NREGS, re_nsub + 1)' groups.
 336            If REGS_REALLOCATE, reallocate space if necessary.
 337            If REGS_FIXED, use what's there.  */
 338 #define REGS_UNALLOCATED 0
 339 #define REGS_REALLOCATE 1
 340 #define REGS_FIXED 2
 341   unsigned regs_allocated : 2;
 342
 343         /* Set to zero when `regex_compile' compiles a pattern; set to one
 344            by `re_compile_fastmap' if it updates the fastmap.  */
 345   unsigned fastmap_accurate : 1;
 346
 347         /* If set, `re_match_2' does not return information about
 348            subexpressions.  */
 349   unsigned no_sub : 1;
 350
 351         /* If set, a beginning-of-line anchor doesn't match at the
 352            beginning of the string.  */
 353   unsigned not_bol : 1;
 354
 355         /* Similarly for an end-of-line anchor.  */
 356   unsigned not_eol : 1;
 357
 358         /* If true, an anchor at a newline matches.  */
 359   unsigned newline_anchor : 1;
 360
 361   /* If true, multi-byte form in the `buffer' should be recognized as a
 362      multibyte character. */
 363   unsigned multibyte : 1;
 364
 365 /* [[[end pattern_buffer]]] */
 366 };
 367
 368 typedef struct re_pattern_buffer regex_t;
 369 \f
 370 /* Type for byte offsets within the string.  POSIX mandates this.  */
 371 typedef int regoff_t;
 372
 373
 374 /* This is the structure we store register match data in.  See
 375    regex.texinfo for a full description of what registers match.  */
 376 struct re_registers
 377 {
 378   unsigned num_regs;
 379   regoff_t *start;
 380   regoff_t *end;
 381 };
 382
 383
 384 /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer,
 385    `re_match_2' returns information about at least this many registers
 386    the first time a `regs' structure is passed.  */
 387 #ifndef RE_NREGS
 388 #define RE_NREGS 30
 389 #endif
 390
 391
 392 /* POSIX specification for registers.  Aside from the different names than
 393    `re_registers', POSIX uses an array of structures, instead of a
 394    structure of arrays.  */
 395 typedef struct
 396 {
 397   regoff_t rm_so;  /* Byte offset from string's start to substring's start.  */
 398   regoff_t rm_eo;  /* Byte offset from string's start to substring's end.  */
 399 } regmatch_t;
 400 \f
 401 /* Declarations for routines.  */
 402
 403 /* To avoid duplicating every routine declaration -- once with a
 404    prototype (if we are ANSI), and once without (if we aren't) -- we
 405    use the following macro to declare argument types.  This
 406    unfortunately clutters up the declarations a bit, but I think it's
 407    worth it.  */
 408
 409 #if __STDC__
 410
 411 #define _RE_ARGS(args) args
 412
 413 #else /* not __STDC__ */
 414
 415 #define _RE_ARGS(args) ()
 416
 417 #endif /* not __STDC__ */
 418
 419 /* Sets the current default syntax to SYNTAX, and return the old syntax.
 420    You can also simply assign to the `re_syntax_options' variable.  */
 421 extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax));
 422
 423 /* Compile the regular expression PATTERN, with length LENGTH
 424    and syntax given by the global `re_syntax_options', into the buffer
 425    BUFFER.  Return NULL if successful, and an error string if not.  */
 426 extern const char *re_compile_pattern
 427   _RE_ARGS ((const char *pattern, int length,
 428              struct re_pattern_buffer *buffer));
 429
 430
 431 /* Compile a fastmap for the compiled pattern in BUFFER; used to
 432    accelerate searches.  Return 0 if successful and -2 if was an
 433    internal error.  */
 434 extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer));
 435
 436
 437 /* Search in the string STRING (with length LENGTH) for the pattern
 438    compiled into BUFFER.  Start searching at position START, for RANGE
 439    characters.  Return the starting position of the match, -1 for no
 440    match, or -2 for an internal error.  Also return register
 441    information in REGS (if REGS and BUFFER->no_sub are nonzero).  */
 442 extern int re_search
 443   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
 444             int length, int start, int range, struct re_registers *regs));
 445
 446
 447 /* Like `re_search', but search in the concatenation of STRING1 and
 448    STRING2.  Also, stop searching at index START + STOP.  */
 449 extern int re_search_2
 450   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
 451              int length1, const char *string2, int length2,
 452              int start, int range, struct re_registers *regs, int stop));
 453
 454
 455 /* Like `re_search', but return how many characters in STRING the regexp
 456    in BUFFER matched, starting at position START.  */
 457 extern int re_match
 458   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
 459              int length, int start, struct re_registers *regs));
 460
 461
 462 /* Relates to `re_match' as `re_search_2' relates to `re_search'.  */
 463 extern int re_match_2
 464   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
 465              int length1, const char *string2, int length2,
 466              int start, struct re_registers *regs, int stop));
 467
 468
 469 /* Set REGS to hold NUM_REGS registers, storing them in STARTS and
 470    ENDS.  Subsequent matches using BUFFER and REGS will use this memory
 471    for recording register information.  STARTS and ENDS must be
 472    allocated with malloc, and must each be at least `NUM_REGS * sizeof
 473    (regoff_t)' bytes long.
 474
 475    If NUM_REGS == 0, then subsequent matches should allocate their own
 476    register data.
 477
 478    Unless this function is called, the first search or match using
 479    PATTERN_BUFFER will allocate its own register data, without
 480    freeing the old data.  */
 481 extern void re_set_registers
 482   _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs,
 483              unsigned num_regs, regoff_t *starts, regoff_t *ends));
 484
 485 #ifdef _REGEX_RE_COMP
 486 /* 4.2 bsd compatibility.  */
 487 extern char *re_comp _RE_ARGS ((const char *));
 488 extern int re_exec _RE_ARGS ((const char *));
 489 #endif
 490
 491 /* POSIX compatibility.  */
 492 extern int regcomp _RE_ARGS ((regex_t *preg, const char *pattern, int cflags));
 493 extern int regexec
 494   _RE_ARGS ((const regex_t *preg, const char *string, size_t nmatch,
 495              regmatch_t pmatch[], int eflags));
 496 extern size_t regerror
 497   _RE_ARGS ((int errcode, const regex_t *preg, char *errbuf,
 498              size_t errbuf_size));
 499 extern void regfree _RE_ARGS ((regex_t *preg));
 500
 501 #endif /* not __REGEXP_LIBRARY_H__ */
 502 \f
 503 /*
 504 Local variables:
 505 make-backup-files: t
 506 version-control: t
 507 trim-versions-without-asking: nil
 508 End:
 509 */