File:  [ELWIX - Embedded LightWeight unIX -] / embedaddon / smartmontools / regex / regex.h
Revision 1.1.1.1 (vendor branch): download - view: text, annotated - select for diffs - revision graph
Tue Feb 21 16:32:16 2012 UTC (12 years, 4 months ago) by misho
Branches: smartmontools, elwix, MAIN
CVS tags: v5_43, v5_42, HEAD
smartmontools

    1: /* Definitions for data structures and routines for the regular
    2:    expression library.
    3:    Copyright (C) 1985,1989-93,1995-98,2000,2001,2002
    4:    Free Software Foundation, Inc.
    5:    This file is part of the GNU C Library.
    6: 
    7:    The GNU C Library is free software; you can redistribute it and/or
    8:    modify it under the terms of the GNU Lesser General Public
    9:    License as published by the Free Software Foundation; either
   10:    version 2.1 of the License, or (at your option) any later version.
   11: 
   12:    The GNU C Library is distributed in the hope that it will be useful,
   13:    but WITHOUT ANY WARRANTY; without even the implied warranty of
   14:    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
   15:    Lesser General Public License for more details.
   16: 
   17:    You should have received a copy of the GNU Lesser General Public
   18:    License along with the GNU C Library; if not, write to the Free
   19:    Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
   20:    02111-1307 USA.  */
   21: 
   22: #ifndef _REGEX_H
   23: #define _REGEX_H 1
   24: 
   25: /* Allow the use in C++ code.  */
   26: #ifdef __cplusplus
   27: extern "C" {
   28: #endif
   29: 
   30: /* POSIX says that <sys/types.h> must be included (by the caller) before
   31:    <regex.h>.  */
   32: 
   33: #if !defined _POSIX_C_SOURCE && !defined _POSIX_SOURCE && (defined VMS || defined _MSC_VER)
   34: /* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it
   35:    should be there. Same for Microsoft Visual C++ 6.0 */
   36: # include <stddef.h>
   37: #endif
   38: 
   39: /* The following two types have to be signed and unsigned integer type
   40:    wide enough to hold a value of a pointer.  For most ANSI compilers
   41:    ptrdiff_t and size_t should be likely OK.  Still size of these two
   42:    types is 2 for Microsoft C.  Ugh... */
   43: typedef long int s_reg_t;
   44: typedef unsigned long int active_reg_t;
   45: 
   46: /* The following bits are used to determine the regexp syntax we
   47:    recognize.  The set/not-set meanings are chosen so that Emacs syntax
   48:    remains the value 0.  The bits are given in alphabetical order, and
   49:    the definitions shifted by one from the previous bit; thus, when we
   50:    add or remove a bit, only one other definition need change.  */
   51: typedef unsigned long int reg_syntax_t;
   52: 
   53: /* If this bit is not set, then \ inside a bracket expression is literal.
   54:    If set, then such a \ quotes the following character.  */
   55: #define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1)
   56: 
   57: /* If this bit is not set, then + and ? are operators, and \+ and \? are
   58:      literals.
   59:    If set, then \+ and \? are operators and + and ? are literals.  */
   60: #define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)
   61: 
   62: /* If this bit is set, then character classes are supported.  They are:
   63:      [:alpha:], [:upper:], [:lower:],  [:digit:], [:alnum:], [:xdigit:],
   64:      [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
   65:    If not set, then character classes are not supported.  */
   66: #define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)
   67: 
   68: /* If this bit is set, then ^ and $ are always anchors (outside bracket
   69:      expressions, of course).
   70:    If this bit is not set, then it depends:
   71:         ^  is an anchor if it is at the beginning of a regular
   72:            expression or after an open-group or an alternation operator;
   73:         $  is an anchor if it is at the end of a regular expression, or
   74:            before a close-group or an alternation operator.
   75: 
   76:    This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because
   77:    POSIX draft 11.2 says that * etc. in leading positions is undefined.
   78:    We already implemented a previous draft which made those constructs
   79:    invalid, though, so we haven't changed the code back.  */
   80: #define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)
   81: 
   82: /* If this bit is set, then special characters are always special
   83:      regardless of where they are in the pattern.
   84:    If this bit is not set, then special characters are special only in
   85:      some contexts; otherwise they are ordinary.  Specifically,
   86:      * + ? and intervals are only special when not after the beginning,
   87:      open-group, or alternation operator.  */
   88: #define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)
   89: 
   90: /* If this bit is set, then *, +, ?, and { cannot be first in an re or
   91:      immediately after an alternation or begin-group operator.  */
   92: #define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)
   93: 
   94: /* If this bit is set, then . matches newline.
   95:    If not set, then it doesn't.  */
   96: #define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)
   97: 
   98: /* If this bit is set, then . doesn't match NUL.
   99:    If not set, then it does.  */
  100: #define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)
  101: 
  102: /* If this bit is set, nonmatching lists [^...] do not match newline.
  103:    If not set, they do.  */
  104: #define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)
  105: 
  106: /* If this bit is set, either \{...\} or {...} defines an
  107:      interval, depending on RE_NO_BK_BRACES.
  108:    If not set, \{, \}, {, and } are literals.  */
  109: #define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)
  110: 
  111: /* If this bit is set, +, ? and | aren't recognized as operators.
  112:    If not set, they are.  */
  113: #define RE_LIMITED_OPS (RE_INTERVALS << 1)
  114: 
  115: /* If this bit is set, newline is an alternation operator.
  116:    If not set, newline is literal.  */
  117: #define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)
  118: 
  119: /* If this bit is set, then `{...}' defines an interval, and \{ and \}
  120:      are literals.
  121:   If not set, then `\{...\}' defines an interval.  */
  122: #define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)
  123: 
  124: /* If this bit is set, (...) defines a group, and \( and \) are literals.
  125:    If not set, \(...\) defines a group, and ( and ) are literals.  */
  126: #define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)
  127: 
  128: /* If this bit is set, then \<digit> matches <digit>.
  129:    If not set, then \<digit> is a back-reference.  */
  130: #define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)
  131: 
  132: /* If this bit is set, then | is an alternation operator, and \| is literal.
  133:    If not set, then \| is an alternation operator, and | is literal.  */
  134: #define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)
  135: 
  136: /* If this bit is set, then an ending range point collating higher
  137:      than the starting range point, as in [z-a], is invalid.
  138:    If not set, then when ending range point collates higher than the
  139:      starting range point, the range is ignored.  */
  140: #define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1)
  141: 
  142: /* If this bit is set, then an unmatched ) is ordinary.
  143:    If not set, then an unmatched ) is invalid.  */
  144: #define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1)
  145: 
  146: /* If this bit is set, succeed as soon as we match the whole pattern,
  147:    without further backtracking.  */
  148: #define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1)
  149: 
  150: /* If this bit is set, do not process the GNU regex operators.
  151:    If not set, then the GNU regex operators are recognized. */
  152: #define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1)
  153: 
  154: /* If this bit is set, turn on internal regex debugging.
  155:    If not set, and debugging was on, turn it off.
  156:    This only works if regex.c is compiled -DDEBUG.
  157:    We define this bit always, so that all that's needed to turn on
  158:    debugging is to recompile regex.c; the calling code can always have
  159:    this bit set, and it won't affect anything in the normal case. */
  160: #define RE_DEBUG (RE_NO_GNU_OPS << 1)
  161: 
  162: /* If this bit is set, a syntactically invalid interval is treated as
  163:    a string of ordinary characters.  For example, the ERE 'a{1' is
  164:    treated as 'a\{1'.  */
  165: #define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1)
  166: 
  167: /* If this bit is set, then ignore case when matching.
  168:    If not set, then case is significant.  */
  169: #define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1)
  170: 
  171: /* This global variable defines the particular regexp syntax to use (for
  172:    some interfaces).  When a regexp is compiled, the syntax used is
  173:    stored in the pattern buffer, so changing this does not affect
  174:    already-compiled regexps.  */
  175: extern reg_syntax_t re_syntax_options;
  176: 
  177: /* Define combinations of the above bits for the standard possibilities.
  178:    (The [[[ comments delimit what gets put into the Texinfo file, so
  179:    don't delete them!)  */
  180: /* [[[begin syntaxes]]] */
  181: #define RE_SYNTAX_EMACS 0
  182: 
  183: #define RE_SYNTAX_AWK							\
  184:   (RE_BACKSLASH_ESCAPE_IN_LISTS   | RE_DOT_NOT_NULL			\
  185:    | RE_NO_BK_PARENS              | RE_NO_BK_REFS			\
  186:    | RE_NO_BK_VBAR                | RE_NO_EMPTY_RANGES			\
  187:    | RE_DOT_NEWLINE		  | RE_CONTEXT_INDEP_ANCHORS		\
  188:    | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS)
  189: 
  190: #define RE_SYNTAX_GNU_AWK						\
  191:   ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG)	\
  192:    & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS		\
  193:        | RE_CONTEXT_INVALID_OPS ))
  194: 
  195: #define RE_SYNTAX_POSIX_AWK 						\
  196:   (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS		\
  197:    | RE_INTERVALS	    | RE_NO_GNU_OPS)
  198: 
  199: #define RE_SYNTAX_GREP							\
  200:   (RE_BK_PLUS_QM              | RE_CHAR_CLASSES				\
  201:    | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS				\
  202:    | RE_NEWLINE_ALT)
  203: 
  204: #define RE_SYNTAX_EGREP							\
  205:   (RE_CHAR_CLASSES        | RE_CONTEXT_INDEP_ANCHORS			\
  206:    | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE			\
  207:    | RE_NEWLINE_ALT       | RE_NO_BK_PARENS				\
  208:    | RE_NO_BK_VBAR)
  209: 
  210: #define RE_SYNTAX_POSIX_EGREP						\
  211:   (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES			\
  212:    | RE_INVALID_INTERVAL_ORD)
  213: 
  214: /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff.  */
  215: #define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC
  216: 
  217: #define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC
  218: 
  219: /* Syntax bits common to both basic and extended POSIX regex syntax.  */
  220: #define _RE_SYNTAX_POSIX_COMMON						\
  221:   (RE_CHAR_CLASSES | RE_DOT_NEWLINE      | RE_DOT_NOT_NULL		\
  222:    | RE_INTERVALS  | RE_NO_EMPTY_RANGES)
  223: 
  224: #define RE_SYNTAX_POSIX_BASIC						\
  225:   (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM)
  226: 
  227: /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes
  228:    RE_LIMITED_OPS, i.e., \? \+ \| are not recognized.  Actually, this
  229:    isn't minimal, since other operators, such as \`, aren't disabled.  */
  230: #define RE_SYNTAX_POSIX_MINIMAL_BASIC					\
  231:   (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS)
  232: 
  233: #define RE_SYNTAX_POSIX_EXTENDED					\
  234:   (_RE_SYNTAX_POSIX_COMMON  | RE_CONTEXT_INDEP_ANCHORS			\
  235:    | RE_CONTEXT_INDEP_OPS   | RE_NO_BK_BRACES				\
  236:    | RE_NO_BK_PARENS        | RE_NO_BK_VBAR				\
  237:    | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD)
  238: 
  239: /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is
  240:    removed and RE_NO_BK_REFS is added.  */
  241: #define RE_SYNTAX_POSIX_MINIMAL_EXTENDED				\
  242:   (_RE_SYNTAX_POSIX_COMMON  | RE_CONTEXT_INDEP_ANCHORS			\
  243:    | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES				\
  244:    | RE_NO_BK_PARENS        | RE_NO_BK_REFS				\
  245:    | RE_NO_BK_VBAR	    | RE_UNMATCHED_RIGHT_PAREN_ORD)
  246: /* [[[end syntaxes]]] */
  247: 
  248: /* Maximum number of duplicates an interval can allow.  Some systems
  249:    (erroneously) define this in other header files, but we want our
  250:    value, so remove any previous define.  */
  251: #ifdef RE_DUP_MAX
  252: # undef RE_DUP_MAX
  253: #endif
  254: /* If sizeof(int) == 2, then ((1 << 15) - 1) overflows.  */
  255: #define RE_DUP_MAX (0x7fff)
  256: 
  257: 
  258: /* POSIX `cflags' bits (i.e., information for `regcomp').  */
  259: 
  260: /* If this bit is set, then use extended regular expression syntax.
  261:    If not set, then use basic regular expression syntax.  */
  262: #define REG_EXTENDED 1
  263: 
  264: /* If this bit is set, then ignore case when matching.
  265:    If not set, then case is significant.  */
  266: #define REG_ICASE (REG_EXTENDED << 1)
  267: 
  268: /* If this bit is set, then anchors do not match at newline
  269:      characters in the string.
  270:    If not set, then anchors do match at newlines.  */
  271: #define REG_NEWLINE (REG_ICASE << 1)
  272: 
  273: /* If this bit is set, then report only success or fail in regexec.
  274:    If not set, then returns differ between not matching and errors.  */
  275: #define REG_NOSUB (REG_NEWLINE << 1)
  276: 
  277: 
  278: /* POSIX `eflags' bits (i.e., information for regexec).  */
  279: 
  280: /* If this bit is set, then the beginning-of-line operator doesn't match
  281:      the beginning of the string (presumably because it's not the
  282:      beginning of a line).
  283:    If not set, then the beginning-of-line operator does match the
  284:      beginning of the string.  */
  285: #define REG_NOTBOL 1
  286: 
  287: /* Like REG_NOTBOL, except for the end-of-line.  */
  288: #define REG_NOTEOL (1 << 1)
  289: 
  290: 
  291: /* If any error codes are removed, changed, or added, update the
  292:    `re_error_msg' table in regex.c.  */
  293: typedef enum
  294: {
  295: #ifdef _XOPEN_SOURCE
  296:   REG_ENOSYS = -1,	/* This will never happen for this implementation.  */
  297: #endif
  298: 
  299:   REG_NOERROR = 0,	/* Success.  */
  300:   REG_NOMATCH,		/* Didn't find a match (for regexec).  */
  301: 
  302:   /* POSIX regcomp return error codes.  (In the order listed in the
  303:      standard.)  */
  304:   REG_BADPAT,		/* Invalid pattern.  */
  305:   REG_ECOLLATE,		/* Not implemented.  */
  306:   REG_ECTYPE,		/* Invalid character class name.  */
  307:   REG_EESCAPE,		/* Trailing backslash.  */
  308:   REG_ESUBREG,		/* Invalid back reference.  */
  309:   REG_EBRACK,		/* Unmatched left bracket.  */
  310:   REG_EPAREN,		/* Parenthesis imbalance.  */
  311:   REG_EBRACE,		/* Unmatched \{.  */
  312:   REG_BADBR,		/* Invalid contents of \{\}.  */
  313:   REG_ERANGE,		/* Invalid range end.  */
  314:   REG_ESPACE,		/* Ran out of memory.  */
  315:   REG_BADRPT,		/* No preceding re for repetition op.  */
  316: 
  317:   /* Error codes we've added.  */
  318:   REG_EEND,		/* Premature end.  */
  319:   REG_ESIZE,		/* Compiled pattern bigger than 2^16 bytes.  */
  320:   REG_ERPAREN		/* Unmatched ) or \); not returned from regcomp.  */
  321: } reg_errcode_t;
  322: 
  323: /* This data structure represents a compiled pattern.  Before calling
  324:    the pattern compiler, the fields `buffer', `allocated', `fastmap',
  325:    `translate', and `no_sub' can be set.  After the pattern has been
  326:    compiled, the `re_nsub' field is available.  All other fields are
  327:    private to the regex routines.  */
  328: 
  329: #ifndef RE_TRANSLATE_TYPE
  330: # define RE_TRANSLATE_TYPE char *
  331: #endif
  332: 
  333: struct re_pattern_buffer
  334: {
  335: /* [[[begin pattern_buffer]]] */
  336: 	/* Space that holds the compiled pattern.  It is declared as
  337:           `unsigned char *' because its elements are
  338:            sometimes used as array indexes.  */
  339:   unsigned char *buffer;
  340: 
  341: 	/* Number of bytes to which `buffer' points.  */
  342:   unsigned long int allocated;
  343: 
  344: 	/* Number of bytes actually used in `buffer'.  */
  345:   unsigned long int used;
  346: 
  347:         /* Syntax setting with which the pattern was compiled.  */
  348:   reg_syntax_t syntax;
  349: 
  350:         /* Pointer to a fastmap, if any, otherwise zero.  re_search uses
  351:            the fastmap, if there is one, to skip over impossible
  352:            starting points for matches.  */
  353:   char *fastmap;
  354: 
  355:         /* Either a translate table to apply to all characters before
  356:            comparing them, or zero for no translation.  The translation
  357:            is applied to a pattern when it is compiled and to a string
  358:            when it is matched.  */
  359:   RE_TRANSLATE_TYPE translate;
  360: 
  361: 	/* Number of subexpressions found by the compiler.  */
  362:   size_t re_nsub;
  363: 
  364:         /* Zero if this pattern cannot match the empty string, one else.
  365:            Well, in truth it's used only in `re_search_2', to see
  366:            whether or not we should use the fastmap, so we don't set
  367:            this absolutely perfectly; see `re_compile_fastmap' (the
  368:            `duplicate' case).  */
  369:   unsigned can_be_null : 1;
  370: 
  371:         /* If REGS_UNALLOCATED, allocate space in the `regs' structure
  372:              for `max (RE_NREGS, re_nsub + 1)' groups.
  373:            If REGS_REALLOCATE, reallocate space if necessary.
  374:            If REGS_FIXED, use what's there.  */
  375: #define REGS_UNALLOCATED 0
  376: #define REGS_REALLOCATE 1
  377: #define REGS_FIXED 2
  378:   unsigned regs_allocated : 2;
  379: 
  380:         /* Set to zero when `regex_compile' compiles a pattern; set to one
  381:            by `re_compile_fastmap' if it updates the fastmap.  */
  382:   unsigned fastmap_accurate : 1;
  383: 
  384:         /* If set, `re_match_2' does not return information about
  385:            subexpressions.  */
  386:   unsigned no_sub : 1;
  387: 
  388:         /* If set, a beginning-of-line anchor doesn't match at the
  389:            beginning of the string.  */
  390:   unsigned not_bol : 1;
  391: 
  392:         /* Similarly for an end-of-line anchor.  */
  393:   unsigned not_eol : 1;
  394: 
  395:         /* If true, an anchor at a newline matches.  */
  396:   unsigned newline_anchor : 1;
  397: 
  398: /* [[[end pattern_buffer]]] */
  399: };
  400: 
  401: typedef struct re_pattern_buffer regex_t;
  402: 
  403: /* Type for byte offsets within the string.  POSIX mandates this.  */
  404: typedef int regoff_t;
  405: 
  406: 
  407: /* This is the structure we store register match data in.  See
  408:    regex.texinfo for a full description of what registers match.  */
  409: struct re_registers
  410: {
  411:   unsigned num_regs;
  412:   regoff_t *start;
  413:   regoff_t *end;
  414: };
  415: 
  416: 
  417: /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer,
  418:    `re_match_2' returns information about at least this many registers
  419:    the first time a `regs' structure is passed.  */
  420: #ifndef RE_NREGS
  421: # define RE_NREGS 30
  422: #endif
  423: 
  424: 
  425: /* POSIX specification for registers.  Aside from the different names than
  426:    `re_registers', POSIX uses an array of structures, instead of a
  427:    structure of arrays.  */
  428: typedef struct
  429: {
  430:   regoff_t rm_so;  /* Byte offset from string's start to substring's start.  */
  431:   regoff_t rm_eo;  /* Byte offset from string's start to substring's end.  */
  432: } regmatch_t;
  433: 
  434: /* Declarations for routines.  */
  435: 
  436: /* To avoid duplicating every routine declaration -- once with a
  437:    prototype (if we are ANSI), and once without (if we aren't) -- we
  438:    use the following macro to declare argument types.  This
  439:    unfortunately clutters up the declarations a bit, but I think it's
  440:    worth it.  */
  441: 
  442: #if defined(__STDC__) || defined(__cplusplus)
  443: 
  444: # define _RE_ARGS(args) args
  445: 
  446: #else /* not __STDC__ */
  447: 
  448: # define _RE_ARGS(args) ()
  449: 
  450: #endif /* not __STDC__ */
  451: 
  452: /* Sets the current default syntax to SYNTAX, and return the old syntax.
  453:    You can also simply assign to the `re_syntax_options' variable.  */
  454: extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax));
  455: 
  456: /* Compile the regular expression PATTERN, with length LENGTH
  457:    and syntax given by the global `re_syntax_options', into the buffer
  458:    BUFFER.  Return NULL if successful, and an error string if not.  */
  459: extern const char *re_compile_pattern
  460:   _RE_ARGS ((const char *pattern, size_t length,
  461:              struct re_pattern_buffer *buffer));
  462: 
  463: 
  464: /* Compile a fastmap for the compiled pattern in BUFFER; used to
  465:    accelerate searches.  Return 0 if successful and -2 if was an
  466:    internal error.  */
  467: extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer));
  468: 
  469: 
  470: /* Search in the string STRING (with length LENGTH) for the pattern
  471:    compiled into BUFFER.  Start searching at position START, for RANGE
  472:    characters.  Return the starting position of the match, -1 for no
  473:    match, or -2 for an internal error.  Also return register
  474:    information in REGS (if REGS and BUFFER->no_sub are nonzero).  */
  475: extern int re_search
  476:   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
  477:             int length, int start, int range, struct re_registers *regs));
  478: 
  479: 
  480: /* Like `re_search', but search in the concatenation of STRING1 and
  481:    STRING2.  Also, stop searching at index START + STOP.  */
  482: extern int re_search_2
  483:   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
  484:              int length1, const char *string2, int length2,
  485:              int start, int range, struct re_registers *regs, int stop));
  486: 
  487: 
  488: /* Like `re_search', but return how many characters in STRING the regexp
  489:    in BUFFER matched, starting at position START.  */
  490: extern int re_match
  491:   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
  492:              int length, int start, struct re_registers *regs));
  493: 
  494: 
  495: /* Relates to `re_match' as `re_search_2' relates to `re_search'.  */
  496: extern int re_match_2
  497:   _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
  498:              int length1, const char *string2, int length2,
  499:              int start, struct re_registers *regs, int stop));
  500: 
  501: 
  502: /* Set REGS to hold NUM_REGS registers, storing them in STARTS and
  503:    ENDS.  Subsequent matches using BUFFER and REGS will use this memory
  504:    for recording register information.  STARTS and ENDS must be
  505:    allocated with malloc, and must each be at least `NUM_REGS * sizeof
  506:    (regoff_t)' bytes long.
  507: 
  508:    If NUM_REGS == 0, then subsequent matches should allocate their own
  509:    register data.
  510: 
  511:    Unless this function is called, the first search or match using
  512:    PATTERN_BUFFER will allocate its own register data, without
  513:    freeing the old data.  */
  514: extern void re_set_registers
  515:   _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs,
  516:              unsigned num_regs, regoff_t *starts, regoff_t *ends));
  517: 
  518: #if defined _REGEX_RE_COMP || defined _LIBC
  519: # ifndef _CRAY
  520: /* 4.2 bsd compatibility.  */
  521: extern char *re_comp _RE_ARGS ((const char *));
  522: extern int re_exec _RE_ARGS ((const char *));
  523: # endif
  524: #endif
  525: 
  526: /* GCC 2.95 and later have "__restrict"; C99 compilers have
  527:    "restrict", and "configure" may have defined "restrict".  */
  528: #ifndef __restrict
  529: # if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__))
  530: #  if defined restrict || 199901L <= __STDC_VERSION__
  531: #   define __restrict restrict
  532: #  else
  533: #   define __restrict
  534: #  endif
  535: # endif
  536: #endif
  537: /* gcc 3.1 and up support the [restrict] syntax.  */
  538: #ifndef __restrict_arr
  539: # if (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)) && !defined(__WIN32__)
  540: #  define __restrict_arr __restrict
  541: # else
  542: #  define __restrict_arr
  543: # endif
  544: #endif
  545: 
  546: /* POSIX compatibility.  */
  547: extern int regcomp _RE_ARGS ((regex_t *__restrict __preg,
  548: 			      const char *__restrict __pattern,
  549: 			      int __cflags));
  550: 
  551: extern int regexec _RE_ARGS ((const regex_t *__restrict __preg,
  552: 			      const char *__restrict __string, size_t __nmatch,
  553: 			      regmatch_t __pmatch[__restrict_arr],
  554: 			      int __eflags));
  555: 
  556: extern size_t regerror _RE_ARGS ((int __errcode, const regex_t *__preg,
  557: 				  char *__errbuf, size_t __errbuf_size));
  558: 
  559: extern void regfree _RE_ARGS ((regex_t *__preg));
  560: 
  561: 
  562: #ifdef __cplusplus
  563: }
  564: #endif	/* C++ */
  565: 
  566: #endif /* regex.h */
  567: 
  568: /*
  569: Local variables:
  570: make-backup-files: t
  571: version-control: t
  572: trim-versions-without-asking: nil
  573: End:
  574: */

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>