regexec.c revision 1.16
1/* $NetBSD: regexec.c,v 1.16 2001/12/17 16:31:11 christos Exp $ */ 2 3/*- 4 * Copyright (c) 1992, 1993, 1994 Henry Spencer. 5 * Copyright (c) 1992, 1993, 1994 6 * The Regents of the University of California. All rights reserved. 7 * 8 * This code is derived from software contributed to Berkeley by 9 * Henry Spencer. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 14 * 1. Redistributions of source code must retain the above copyright 15 * notice, this list of conditions and the following disclaimer. 16 * 2. Redistributions in binary form must reproduce the above copyright 17 * notice, this list of conditions and the following disclaimer in the 18 * documentation and/or other materials provided with the distribution. 19 * 3. All advertising materials mentioning features or use of this software 20 * must display the following acknowledgement: 21 * This product includes software developed by the University of 22 * California, Berkeley and its contributors. 23 * 4. Neither the name of the University nor the names of its contributors 24 * may be used to endorse or promote products derived from this software 25 * without specific prior written permission. 26 * 27 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 28 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 29 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 30 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 31 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 32 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 33 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 34 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 35 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 36 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 37 * SUCH DAMAGE. 38 * 39 * @(#)regexec.c 8.3 (Berkeley) 3/20/94 40 */ 41 42#include <sys/cdefs.h> 43#if defined(LIBC_SCCS) && !defined(lint) 44#if 0 45static char sccsid[] = "@(#)regexec.c 8.3 (Berkeley) 3/20/94"; 46#else 47__RCSID("$NetBSD: regexec.c,v 1.16 2001/12/17 16:31:11 christos Exp $"); 48#endif 49#endif /* LIBC_SCCS and not lint */ 50 51/* 52 * the outer shell of regexec() 53 * 54 * This file includes engine.c *twice*, after muchos fiddling with the 55 * macros that code uses. This lets the same code operate on two different 56 * representations for state sets. 57 */ 58#include "namespace.h" 59#include <sys/types.h> 60 61#include <assert.h> 62#include <ctype.h> 63#include <limits.h> 64#include <regex.h> 65#include <stdio.h> 66#include <stdlib.h> 67#include <string.h> 68 69#ifdef __weak_alias 70__weak_alias(regexec,_regexec) 71#endif 72 73#include "utils.h" 74#include "regex2.h" 75 76/* macros for manipulating states, small version */ 77#define states unsigned long 78#define states1 unsigned long /* for later use in regexec() decision */ 79#define CLEAR(v) ((v) = 0) 80#define SET0(v, n) ((v) &= ~((unsigned long)1 << (n))) 81#define SET1(v, n) ((v) |= (unsigned long)1 << (n)) 82#define ISSET(v, n) (((v) & ((unsigned long)1 << (n))) != 0) 83#define ASSIGN(d, s) ((d) = (s)) 84#define EQ(a, b) ((a) == (b)) 85#define STATEVARS int dummy /* dummy version */ 86#define STATESETUP(m, n) /* nothing */ 87#define STATETEARDOWN(m) /* nothing */ 88#define SETUP(v) ((v) = 0) 89#define onestate unsigned long 90#define INIT(o, n) ((o) = (unsigned long)1 << (n)) 91#define INC(o) ((o) <<= 1) 92#define ISSTATEIN(v, o) (((v) & (o)) != 0) 93/* some abbreviations; note that some of these know variable names! */ 94/* do "if I'm here, I can also be there" etc without branches */ 95#define FWD(dst, src, n) ((dst) |= ((unsigned long)(src)&(here)) << (n)) 96#define BACK(dst, src, n) ((dst) |= ((unsigned long)(src)&(here)) >> (n)) 97#define ISSETBACK(v, n) (((v) & ((unsigned long)here >> (n))) != 0) 98/* function names */ 99#define SNAMES /* engine.c looks after details */ 100 101#include "engine.c" 102 103/* now undo things */ 104#undef states 105#undef CLEAR 106#undef SET0 107#undef SET1 108#undef ISSET 109#undef ASSIGN 110#undef EQ 111#undef STATEVARS 112#undef STATESETUP 113#undef STATETEARDOWN 114#undef SETUP 115#undef onestate 116#undef INIT 117#undef INC 118#undef ISSTATEIN 119#undef FWD 120#undef BACK 121#undef ISSETBACK 122#undef SNAMES 123 124/* macros for manipulating states, large version */ 125#define states char * 126#define CLEAR(v) memset(v, 0, (size_t)m->g->nstates) 127#define SET0(v, n) ((v)[n] = 0) 128#define SET1(v, n) ((v)[n] = 1) 129#define ISSET(v, n) ((v)[n]) 130#define ASSIGN(d, s) memcpy(d, s, (size_t)m->g->nstates) 131#define EQ(a, b) (memcmp(a, b, (size_t)m->g->nstates) == 0) 132#define STATEVARS int vn; char *space 133#define STATESETUP(m, nv) \ 134 if (((m)->space = malloc((size_t)((nv)*(m)->g->nstates))) == NULL) \ 135 return(REG_ESPACE); \ 136 else \ 137 (m)->vn = 0 138 139#define STATETEARDOWN(m) { free((m)->space); m->space = NULL; } 140#define SETUP(v) ((v) = &m->space[(size_t)(m->vn++ * m->g->nstates)]) 141#define onestate int 142#define INIT(o, n) ((o) = (n)) 143#define INC(o) ((o)++) 144#define ISSTATEIN(v, o) ((v)[o]) 145/* some abbreviations; note that some of these know variable names! */ 146/* do "if I'm here, I can also be there" etc without branches */ 147#define FWD(dst, src, n) ((dst)[here+(n)] |= (src)[here]) 148#define BACK(dst, src, n) ((dst)[here-(n)] |= (src)[here]) 149#define ISSETBACK(v, n) ((v)[here - (n)]) 150/* function names */ 151#define LNAMES /* flag */ 152 153#include "engine.c" 154 155/* 156 - regexec - interface for matching 157 = extern int regexec(const regex_t *, const char *, size_t, \ 158 = regmatch_t [], int); 159 = #define REG_NOTBOL 00001 160 = #define REG_NOTEOL 00002 161 = #define REG_STARTEND 00004 162 = #define REG_TRACE 00400 // tracing of execution 163 = #define REG_LARGE 01000 // force large representation 164 = #define REG_BACKR 02000 // force use of backref code 165 * 166 * We put this here so we can exploit knowledge of the state representation 167 * when choosing which matcher to call. Also, by this point the matchers 168 * have been prototyped. 169 */ 170int /* 0 success, REG_NOMATCH failure */ 171regexec(preg, string, nmatch, pmatch, eflags) 172const regex_t *preg; 173const char *string; 174size_t nmatch; 175regmatch_t pmatch[]; 176int eflags; 177{ 178 struct re_guts *g = preg->re_g; 179 char *s; 180#ifdef REDEBUG 181# define GOODFLAGS(f) (f) 182#else 183# define GOODFLAGS(f) ((f)&(REG_NOTBOL|REG_NOTEOL|REG_STARTEND)) 184#endif 185 186 _DIAGASSERT(preg != NULL); 187 _DIAGASSERT(string != NULL); 188 189 if (preg->re_magic != MAGIC1 || g->magic != MAGIC2) 190 return(REG_BADPAT); 191 assert(!(g->iflags&BAD)); 192 if (g->iflags&BAD) /* backstop for no-debug case */ 193 return(REG_BADPAT); 194 eflags = GOODFLAGS(eflags); 195 196 /* LINTED we believe that the regex routines do not change the string */ 197 s = (char *)string; 198 199 if (g->nstates <= CHAR_BIT*sizeof(states1) && !(eflags®_LARGE)) 200 return(smatcher(g, s, nmatch, pmatch, eflags)); 201 else 202 return(lmatcher(g, s, nmatch, pmatch, eflags)); 203} 204