src/regex.h

   1 /* Definitions for data structures and routines for the regular
   2    expression library, version 0.12.
   3
   4    Copyright (C) 1985,89,90,91,92,93,95,2000 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 where historically chosen so
  36    that Emacs syntax had the value 0.
  37    The bits are given in alphabetical order, and
  38    the definitions shifted by one from the previous bit; thus, when we
  39    add or remove a bit, only one other definition need change.  */
  40 typedef unsigned reg_syntax_t;
  41
  42 /* If this bit is set, then even *?, +? and ?? match greedily. */
  43 #define RE_ALL_GREEDY (1)
  44
  45 /* If this bit is not set, then \ inside a bracket expression is literal.
  46    If set, then such a \ quotes the following character.  */
  47 #define RE_BACKSLASH_ESCAPE_IN_LISTS (RE_ALL_GREEDY << 1)
  48
  49 /* If this bit is not set, then + and ? are operators, and \+ and \? are
  50      literals.
  51    If set, then \+ and \? are operators and + and ? are literals.  */
  52 #define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)
  53
  54 /* If this bit is set, then character classes are supported.  They are:
  55      [:alpha:], [:upper:], [:lower:],  [:digit:], [:alnum:], [:xdigit:],
  56      [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
  57    If not set, then character classes are not supported.  */
  58 #define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)
  59
  60 /* If this bit is set, then ^ and $ are always anchors (outside bracket
  61      expressions, of course).
  62    If this bit is not set, then it depends:
  63         ^  is an anchor if it is at the beginning of a regular
  64            expression or after an open-group or an alternation operator;
  65         $  is an anchor if it is at the end of a regular expression, or
  66            before a close-group or an alternation operator.
  67
  68    This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because
  69    POSIX draft 11.2 says that * etc. in leading positions is undefined.
  70    We already implemented a previous draft which made those constructs
  71    invalid, though, so we haven't changed the code back.  */
  72 #define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)
  73
  74 /* If this bit is set, then special characters are always special
  75      regardless of where they are in the pattern.
  76    If this bit is not set, then special characters are special only in
  77      some contexts; otherwise they are ordinary.  Specifically,
  78      * + ? and intervals are only special when not after the beginning,
  79      open-group, or alternation operator.  */
  80 #define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)
  81
  82 /* If this bit is set, then *, +, ?, and { cannot be first in an re or
  83      immediately after an alternation or begin-group operator.  */
  84 #define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)
  85
  86 /* If this bit is set, then . matches newline.
  87    If not set, then it doesn't.  */
  88 #define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)
  89
  90 /* If this bit is set, then . doesn't match NUL.
  91    If not set, then it does.  */
  92 #define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)
  93
  94 /* If this bit is set, nonmatching lists [^...] do not match newline.
  95    If not set, they do.  */
  96 #define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)
  97
  98 /* If this bit is set, either \{...\} or {...} defines an
  99      interval, depending on RE_NO_BK_BRACES.
 100    If not set, \{, \}, {, and } are literals.  */
 101 #define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)
 102
 103 /* If this bit is set, +, ? and | aren't recognized as operators.
 104    If not set, they are.  */
 105 #define RE_LIMITED_OPS (RE_INTERVALS << 1)
 106
 107 /* If this bit is set, newline is an alternation operator.
 108    If not set, newline is literal.  */
 109 #define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)
 110
 111 /* If this bit is set, then `{...}' defines an interval, and \{ and \}
 112      are literals.
 113   If not set, then `\{...\}' defines an interval.  */
 114 #define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)
 115
 116 /* If this bit is set, (...) defines a group, and \( and \) are literals.
 117    If not set, \(...\) defines a group, and ( and ) are literals.  */
 118 #define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)
 119
 120 /* If this bit is set, then \<digit> matches <digit>.
 121    If not set, then \<digit> is a back-reference.  */
 122 #define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)
 123
 124 /* If this bit is set, then | is an alternation operator, and \| is literal.
 125    If not set, then \| is an alternation operator, and | is literal.  */
 126 #define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)
 127
 128 /* If this bit is set, then an ending range point collating higher
 129      than the starting range point, as in [z-a], is invalid.
 130    If not set, then when ending range point collates higher than the
 131      starting range point, the range is ignored.  */
 132 #define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 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_NO_EMPTY_RANGES << 1)
 137
 138 /* If this bit is set, then (?:...) is treated as a shy group.  */
 139 #define RE_SHY_GROUPS (RE_NO_POSIX_BACKTRACKING << 1)
 140
 141 /* If this bit is set, then an unmatched ) is ordinary.
 142    If not set, then an unmatched ) is invalid.  */
 143 #define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_SHY_GROUPS << 1)
 144
 145 /* This global variable defines the particular regexp syntax to use (for
 146    some interfaces).  When a regexp is compiled, the syntax used is
 147    stored in the pattern buffer, so changing this does not affect
 148    already-compiled regexps.  */
 149 extern reg_syntax_t re_syntax_options;
 150
 151 #ifdef emacs
 152 /* In Emacs, this is the string or buffer in which we
 153    are matching.  It is used for looking up syntax properties.  */
 154 extern Lisp_Object re_match_object;
 155 #endif
 156
 157 \f
 158 /* Define combinations of the above bits for the standard possibilities.
 159    (The [[[ comments delimit what gets put into the Texinfo file, so
 160    don't delete them!)  */
 161 /* [[[begin syntaxes]]] */
 162 #define RE_SYNTAX_EMACS (RE_CHAR_CLASSES | RE_INTERVALS | RE_SHY_GROUPS)
 163
 164 #define RE_SYNTAX_AWK                                                   \
 165   (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL                       \
 166    | RE_NO_BK_PARENS            | RE_NO_BK_REFS                         \
 167    | RE_NO_BK_VBAR               | RE_NO_EMPTY_RANGES                   \
 168    | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_ALL_GREEDY)
 169
 170 #define RE_SYNTAX_POSIX_AWK                                             \
 171   (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS)
 172
 173 #define RE_SYNTAX_GREP                                                  \
 174   (RE_BK_PLUS_QM              | RE_CHAR_CLASSES                         \
 175    | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS                            \
 176    | RE_NEWLINE_ALT           | RE_ALL_GREEDY)
 177
 178 #define RE_SYNTAX_EGREP                                                 \
 179   (RE_CHAR_CLASSES        | RE_CONTEXT_INDEP_ANCHORS                    \
 180    | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE                    \
 181    | RE_NEWLINE_ALT       | RE_NO_BK_PARENS                             \
 182    | RE_NO_BK_VBAR        | RE_ALL_GREEDY)
 183
 184 #define RE_SYNTAX_POSIX_EGREP                                           \
 185   (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES)
 186
 187 /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff.  */
 188 #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC
 189
 190 #define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC
 191
 192 /* Syntax bits common to both basic and extended POSIX regex syntax.  */
 193 #define _RE_SYNTAX_POSIX_COMMON                                         \
 194   (RE_CHAR_CLASSES | RE_DOT_NEWLINE      | RE_DOT_NOT_NULL              \
 195    | RE_INTERVALS  | RE_NO_EMPTY_RANGES  | RE_ALL_GREEDY)
 196
 197 #define RE_SYNTAX_POSIX_BASIC                                           \
 198   (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM)
 199
 200 /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes
 201    RE_LIMITED_OPS, i.e., \? \+ \| are not recognized.  Actually, this
 202    isn't minimal, since other operators, such as \`, aren't disabled.  */
 203 #define RE_SYNTAX_POSIX_MINIMAL_BASIC                                   \
 204   (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS)
 205
 206 #define RE_SYNTAX_POSIX_EXTENDED                                        \
 207   (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS                   \
 208    | RE_CONTEXT_INDEP_OPS  | RE_NO_BK_BRACES                            \
 209    | RE_NO_BK_PARENS       | RE_NO_BK_VBAR                              \
 210    | RE_UNMATCHED_RIGHT_PAREN_ORD)
 211
 212 /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS
 213    replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added.  */
 214 #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED                                \
 215   (_RE_SYNTAX_POSIX_COMMON  | RE_CONTEXT_INDEP_ANCHORS                  \
 216    | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES                           \
 217    | RE_NO_BK_PARENS        | RE_NO_BK_REFS                             \
 218    | RE_NO_BK_VBAR          | RE_UNMATCHED_RIGHT_PAREN_ORD)
 219 /* [[[end syntaxes]]] */
 220 \f
 221 /* Maximum number of duplicates an interval can allow.  Some systems
 222    (erroneously) define this in other header files, but we want our
 223    value, so remove any previous define.  */
 224 #ifdef RE_DUP_MAX
 225 #undef RE_DUP_MAX
 226 #endif
 227 #define RE_DUP_MAX ((1 << 15) - 1)
 228
 229
 230 /* POSIX `cflags' bits (i.e., information for `regcomp').  */
 231
 232 /* If this bit is set, then use extended regular expression syntax.
 233    If not set, then use basic regular expression syntax.  */
 234 #define REG_EXTENDED 1
 235
 236 /* If this bit is set, then ignore case when matching.
 237    If not set, then case is significant.  */
 238 #define REG_ICASE (REG_EXTENDED << 1)
 239
 240 /* If this bit is set, then anchors do not match at newline
 241      characters in the string.
 242    If not set, then anchors do match at newlines.  */
 243 #define REG_NEWLINE (REG_ICASE << 1)
 244
 245 /* If this bit is set, then report only success or fail in regexec.
 246    If not set, then returns differ between not matching and errors.  */
 247 #define REG_NOSUB (REG_NEWLINE << 1)
 248
 249
 250 /* POSIX `eflags' bits (i.e., information for regexec).  */
 251
 252 /* If this bit is set, then the beginning-of-line operator doesn't match
 253      the beginning of the string (presumably because it's not the
 254      beginning of a line).
 255    If not set, then the beginning-of-line operator does match the
 256      beginning of the string.  */
 257 #define REG_NOTBOL 1
 258
 259 /* Like REG_NOTBOL, except for the end-of-line.  */
 260 #define REG_NOTEOL (1 << 1)
 261
 262
 263 /* If any error codes are removed, changed, or added, update the
 264    `re_error_msg' table in regex.c.  */
 265 typedef enum
 266 {
 267   REG_NOERROR = 0,      /* Success.  */
 268   REG_NOMATCH,          /* Didn't find a match (for regexec).  */
 269
 270   /* POSIX regcomp return error codes.  (In the order listed in the
 271      standard.)  */
 272   REG_BADPAT,           /* Invalid pattern.  */
 273   REG_ECOLLATE,         /* Not implemented.  */
 274   REG_ECTYPE,           /* Invalid character class name.  */
 275   REG_EESCAPE,          /* Trailing backslash.  */
 276   REG_ESUBREG,          /* Invalid back reference.  */
 277   REG_EBRACK,           /* Unmatched left bracket.  */
 278   REG_EPAREN,           /* Parenthesis imbalance.  */
 279   REG_EBRACE,           /* Unmatched \{.  */
 280   REG_BADBR,            /* Invalid contents of \{\}.  */
 281   REG_ERANGE,           /* Invalid range end.  */
 282   REG_ESPACE,           /* Ran out of memory.  */
 283   REG_BADRPT,           /* No preceding re for repetition op.  */
 284
 285   /* Error codes we've added.  */
 286   REG_EEND,             /* Premature end.  */
 287   REG_ESIZE,            /* Compiled pattern bigger than 2^16 bytes.  */
 288   REG_ERPAREN           /* Unmatched ) or \); not returned from regcomp.  */
 289 } reg_errcode_t;
 290 \f
 291 /* This data structure represents a compiled pattern.  Before calling
 292    the pattern compiler, the fields `buffer', `allocated', `fastmap',
 293    `translate', and `no_sub' can be set.  After the pattern has been
 294    compiled, the `re_nsub' field is available.  All other fields are
 295    private to the regex routines.  */
 296
 297 #ifndef RE_TRANSLATE_TYPE
 298 #define RE_TRANSLATE_TYPE char *
 299 #endif
 300
 301 struct re_pattern_buffer
 302 {
 303 /* [[[begin pattern_buffer]]] */
 304         /* Space that holds the compiled pattern.  It is declared as
 305           `unsigned char *' because its elements are
 306            sometimes used as array indexes.  */
 307   unsigned char *buffer;
 308
 309         /* Number of bytes to which `buffer' points.  */
 310   unsigned long allocated;
 311
 312         /* Number of bytes actually used in `buffer'.  */
 313   unsigned long used;
 314
 315         /* Syntax setting with which the pattern was compiled.  */
 316   reg_syntax_t syntax;
 317
 318         /* Pointer to a fastmap, if any, otherwise zero.  re_search uses
 319            the fastmap, if there is one, to skip over impossible
 320            starting points for matches.  */
 321   char *fastmap;
 322
 323         /* Either a translate table to apply to all characters before
 324            comparing them, or zero for no translation.  The translation
 325            is applied to a pattern when it is compiled and to a string
 326            when it is matched.  */
 327   RE_TRANSLATE_TYPE translate;
 328
 329         /* Number of subexpressions found by the compiler.  */
 330   size_t re_nsub;
 331
 332         /* Zero if this pattern cannot match the empty string, one else.
 333            Well, in truth it's used only in `re_search_2', to see
 334            whether or not we should use the fastmap, so we don't set
 335            this absolutely perfectly; see `re_compile_fastmap'.  */
 336   unsigned can_be_null : 1;
 337
 338         /* If REGS_UNALLOCATED, allocate space in the `regs' structure
 339              for `max (RE_NREGS, re_nsub + 1)' groups.
 340            If REGS_REALLOCATE, reallocate space if necessary.
 341            If REGS_FIXED, use what's there.  */
 342 #define REGS_UNALLOCATED 0
 343 #define REGS_REALLOCATE 1
 344 #define REGS_FIXED 2
 345   unsigned regs_allocated : 2;
 346
 347         /* Set to zero when `regex_compile' compiles a pattern; set to one
 348            by `re_compile_fastmap' if it updates the fastmap.  */
 349   unsigned fastmap_accurate : 1;
 350
 351         /* If set, `re_match_2' does not return information about
 352            subexpressions.  */
 353   unsigned no_sub : 1;
 354
 355         /* If set, a beginning-of-line anchor doesn't match at the
 356            beginning of the string.  */
 357   unsigned not_bol : 1;
 358
 359         /* Similarly for an end-of-line anchor.  */
 360   unsigned not_eol : 1;
 361
 362         /* If true, an anchor at a newline matches.  */
 363   unsigned newline_anchor : 1;
 364
 365   /* If true, multi-byte form in the `buffer' should be recognized as a
 366      multibyte character. */
 367   unsigned multibyte : 1;
 368
 369 /* [[[end pattern_buffer]]] */
 370 };
 371
 372 typedef struct re_pattern_buffer regex_t;
 373 \f
 374 /* Type for byte offsets within the string.  POSIX mandates this.  */
 375 typedef int regoff_t;
 376
 377
 378 /* This is the structure we store register match data in.  See
 379    regex.texinfo for a full description of what registers match.  */
 380 struct re_registers
 381 {
 382   unsigned num_regs;
 383   regoff_t *start;
 384   regoff_t *end;
 385 };
 386
 387
 388 /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer,
 389    `re_match_2' returns information about at least this many registers
 390    the first time a `regs' structure is passed.  */
 391 #ifndef RE_NREGS
 392 #define RE_NREGS 30
 393 #endif
 394
 395
 396 /* POSIX specification for registers.  Aside from the different names than
 397    `re_registers', POSIX uses an array of structures, instead of a
 398    structure of arrays.  */
 399 typedef struct
 400 {
 401   regoff_t rm_so;  /* Byte offset from string's start to substring's start.  */
 402   regoff_t rm_eo;  /* Byte offset from string's start to substring's end.  */
 403 } regmatch_t;
 404 \f
 405 /* Declarations for routines.  */
 406
 407 /* To avoid duplicating every routine declaration -- once with a
 408    prototype (if we are ANSI), and once without (if we aren't) -- we
 409    use the following macro to declare argument types.  This
 410    unfortunately clutters up the declarations a bit, but I think it's
 411    worth it.  */
 412
 413 #if __STDC__
 414
 415 #define _RE_ARGS(args) args
 416
 417 #else /* not __STDC__ */
 418
 419 #define _RE_ARGS(args) ()
 420
 421 #endif /* not __STDC__ */
 422
 423 /* Sets the current default syntax to SYNTAX, and return the old syntax.
 424    You can also simply assign to the `re_syntax_options' variable.  */
 425 extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax));
 426
 427 /* Compile the regular expression PATTERN, with length LENGTH
 428    and syntax given by the global `re_syntax_options', into the buffer
 429    BUFFER.  Return NULL if successful, and an error string if not.  */
 430 extern const char *re_compile_pattern
 431   _RE_ARGS ((const char *pattern, int length,
 432              struct re_pattern_buffer *buffer));
 433
 434
 435 /* Compile a fastmap for the compiled pattern in BUFFER; used to
 436    accelerate searches.  Return 0 if successful and -2 if was an
 437    internal error.  */
 438 extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer));
 439
 440
 441 /* Search in the string STRING (with length LENGTH) for the pattern
 442    compiled into BUFFER.  Start searching at position START, for RANGE
 443    characters.  Return the starting position of the match, -1 for no
 444    match, or -2 for an internal error.  Also return register
 445    information in REGS (if REGS and BUFFER->no_sub are nonzero).  */
 446 extern int re_search
 447   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
 448             int length, int start, int range, struct re_registers *regs));
 449
 450
 451 /* Like `re_search', but search in the concatenation of STRING1 and
 452    STRING2.  Also, stop searching at index START + STOP.  */
 453 extern int re_search_2
 454   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
 455              int length1, const char *string2, int length2,
 456              int start, int range, struct re_registers *regs, int stop));
 457
 458
 459 /* Like `re_search', but return how many characters in STRING the regexp
 460    in BUFFER matched, starting at position START.  */
 461 extern int re_match
 462   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
 463              int length, int start, struct re_registers *regs));
 464
 465
 466 /* Relates to `re_match' as `re_search_2' relates to `re_search'.  */
 467 extern int re_match_2
 468   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
 469              int length1, const char *string2, int length2,
 470              int start, struct re_registers *regs, int stop));
 471
 472
 473 /* Set REGS to hold NUM_REGS registers, storing them in STARTS and
 474    ENDS.  Subsequent matches using BUFFER and REGS will use this memory
 475    for recording register information.  STARTS and ENDS must be
 476    allocated with malloc, and must each be at least `NUM_REGS * sizeof
 477    (regoff_t)' bytes long.
 478
 479    If NUM_REGS == 0, then subsequent matches should allocate their own
 480    register data.
 481
 482    Unless this function is called, the first search or match using
 483    PATTERN_BUFFER will allocate its own register data, without
 484    freeing the old data.  */
 485 extern void re_set_registers
 486   _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs,
 487              unsigned num_regs, regoff_t *starts, regoff_t *ends));
 488
 489 #ifdef _REGEX_RE_COMP
 490 /* 4.2 bsd compatibility.  */
 491 extern char *re_comp _RE_ARGS ((const char *));
 492 extern int re_exec _RE_ARGS ((const char *));
 493 #endif
 494
 495 /* POSIX compatibility.  */
 496 extern int regcomp _RE_ARGS ((regex_t *preg, const char *pattern, int cflags));
 497 extern int regexec
 498   _RE_ARGS ((const regex_t *preg, const char *string, size_t nmatch,
 499              regmatch_t pmatch[], int eflags));
 500 extern size_t regerror
 501   _RE_ARGS ((int errcode, const regex_t *preg, char *errbuf,
 502              size_t errbuf_size));
 503 extern void regfree _RE_ARGS ((regex_t *preg));
 504
 505 #endif /* not __REGEXP_LIBRARY_H__ */
 506 \f
 507 /*
 508 Local variables:
 509 make-backup-files: t
 510 version-control: t
 511 trim-versions-without-asking: nil
 512 End:
 513 */