xref: /openbsd/lib/libc/regex/regexec.c (revision d1a1db7f)
1*d1a1db7fSmartijn /*	$OpenBSD: regexec.c,v 1.14 2018/07/11 12:38:46 martijn Exp $ */
2df930be7Sderaadt /*-
3df930be7Sderaadt  * Copyright (c) 1992, 1993, 1994 Henry Spencer.
4df930be7Sderaadt  * Copyright (c) 1992, 1993, 1994
5df930be7Sderaadt  *	The Regents of the University of California.  All rights reserved.
6df930be7Sderaadt  *
7df930be7Sderaadt  * This code is derived from software contributed to Berkeley by
8df930be7Sderaadt  * Henry Spencer.
9df930be7Sderaadt  *
10df930be7Sderaadt  * Redistribution and use in source and binary forms, with or without
11df930be7Sderaadt  * modification, are permitted provided that the following conditions
12df930be7Sderaadt  * are met:
13df930be7Sderaadt  * 1. Redistributions of source code must retain the above copyright
14df930be7Sderaadt  *    notice, this list of conditions and the following disclaimer.
15df930be7Sderaadt  * 2. Redistributions in binary form must reproduce the above copyright
16df930be7Sderaadt  *    notice, this list of conditions and the following disclaimer in the
17df930be7Sderaadt  *    documentation and/or other materials provided with the distribution.
186580fee3Smillert  * 3. Neither the name of the University nor the names of its contributors
19df930be7Sderaadt  *    may be used to endorse or promote products derived from this software
20df930be7Sderaadt  *    without specific prior written permission.
21df930be7Sderaadt  *
22df930be7Sderaadt  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23df930be7Sderaadt  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24df930be7Sderaadt  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25df930be7Sderaadt  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26df930be7Sderaadt  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27df930be7Sderaadt  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28df930be7Sderaadt  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29df930be7Sderaadt  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30df930be7Sderaadt  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31df930be7Sderaadt  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32df930be7Sderaadt  * SUCH DAMAGE.
33442a6afdSmillert  *
34442a6afdSmillert  *	@(#)regexec.c	8.3 (Berkeley) 3/20/94
35df930be7Sderaadt  */
36df930be7Sderaadt 
37df930be7Sderaadt /*
38df930be7Sderaadt  * the outer shell of regexec()
39df930be7Sderaadt  *
40df930be7Sderaadt  * This file includes engine.c *twice*, after muchos fiddling with the
41df930be7Sderaadt  * macros that code uses.  This lets the same code operate on two different
42df930be7Sderaadt  * representations for state sets.
43df930be7Sderaadt  */
44df930be7Sderaadt #include <sys/types.h>
45df930be7Sderaadt #include <stdio.h>
46df930be7Sderaadt #include <stdlib.h>
47df930be7Sderaadt #include <string.h>
48df930be7Sderaadt #include <limits.h>
49df930be7Sderaadt #include <ctype.h>
50df930be7Sderaadt #include <regex.h>
51df930be7Sderaadt 
52df930be7Sderaadt #include "utils.h"
53df930be7Sderaadt #include "regex2.h"
54df930be7Sderaadt 
55df930be7Sderaadt /* macros for manipulating states, small version */
56442a6afdSmillert #define	states	long
57df930be7Sderaadt #define	states1	states		/* for later use in regexec() decision */
58df930be7Sderaadt #define	CLEAR(v)	((v) = 0)
59442a6afdSmillert #define	SET0(v, n)	((v) &= ~((unsigned long)1 << (n)))
60442a6afdSmillert #define	SET1(v, n)	((v) |= (unsigned long)1 << (n))
61442a6afdSmillert #define	ISSET(v, n)	(((v) & ((unsigned long)1 << (n))) != 0)
62df930be7Sderaadt #define	ASSIGN(d, s)	((d) = (s))
63df930be7Sderaadt #define	EQ(a, b)	((a) == (b))
64442a6afdSmillert #define	STATEVARS	long dummy	/* dummy version */
65df930be7Sderaadt #define	STATESETUP(m, n)	/* nothing */
66df930be7Sderaadt #define	STATETEARDOWN(m)	/* nothing */
67df930be7Sderaadt #define	SETUP(v)	((v) = 0)
68442a6afdSmillert #define	onestate	long
69442a6afdSmillert #define	INIT(o, n)	((o) = (unsigned long)1 << (n))
70442a6afdSmillert #define	INC(o)		((o) <<= 1)
71df930be7Sderaadt #define	ISSTATEIN(v, o)	(((v) & (o)) != 0)
72df930be7Sderaadt /* some abbreviations; note that some of these know variable names! */
73df930be7Sderaadt /* do "if I'm here, I can also be there" etc without branches */
74442a6afdSmillert #define	FWD(dst, src, n)	((dst) |= ((unsigned long)(src)&(here)) << (n))
75442a6afdSmillert #define	BACK(dst, src, n)	((dst) |= ((unsigned long)(src)&(here)) >> (n))
76442a6afdSmillert #define	ISSETBACK(v, n)		(((v) & ((unsigned long)here >> (n))) != 0)
77df930be7Sderaadt /* function names */
78df930be7Sderaadt #define SNAMES			/* engine.c looks after details */
79df930be7Sderaadt 
80df930be7Sderaadt #include "engine.c"
81df930be7Sderaadt 
82df930be7Sderaadt /* now undo things */
83df930be7Sderaadt #undef	states
84df930be7Sderaadt #undef	CLEAR
85df930be7Sderaadt #undef	SET0
86df930be7Sderaadt #undef	SET1
87df930be7Sderaadt #undef	ISSET
88df930be7Sderaadt #undef	ASSIGN
89df930be7Sderaadt #undef	EQ
90df930be7Sderaadt #undef	STATEVARS
91df930be7Sderaadt #undef	STATESETUP
92df930be7Sderaadt #undef	STATETEARDOWN
93df930be7Sderaadt #undef	SETUP
94df930be7Sderaadt #undef	onestate
95df930be7Sderaadt #undef	INIT
96df930be7Sderaadt #undef	INC
97df930be7Sderaadt #undef	ISSTATEIN
98df930be7Sderaadt #undef	FWD
99df930be7Sderaadt #undef	BACK
100df930be7Sderaadt #undef	ISSETBACK
101df930be7Sderaadt #undef	SNAMES
102df930be7Sderaadt 
103df930be7Sderaadt /* macros for manipulating states, large version */
104df930be7Sderaadt #define	states	char *
105df930be7Sderaadt #define	CLEAR(v)	memset(v, 0, m->g->nstates)
106df930be7Sderaadt #define	SET0(v, n)	((v)[n] = 0)
107df930be7Sderaadt #define	SET1(v, n)	((v)[n] = 1)
108df930be7Sderaadt #define	ISSET(v, n)	((v)[n])
109df930be7Sderaadt #define	ASSIGN(d, s)	memcpy(d, s, m->g->nstates)
110df930be7Sderaadt #define	EQ(a, b)	(memcmp(a, b, m->g->nstates) == 0)
111442a6afdSmillert #define	STATEVARS	long vn; char *space
112e786dc0aSdoug #define	STATESETUP(m, nv)	{ (m)->space = reallocarray(NULL, 	    \
113e786dc0aSdoug     				(m)->g->nstates, (nv));			    \
114df930be7Sderaadt 				if ((m)->space == NULL) return(REG_ESPACE); \
115df930be7Sderaadt 				(m)->vn = 0; }
116df930be7Sderaadt #define	STATETEARDOWN(m)	{ free((m)->space); }
117df930be7Sderaadt #define	SETUP(v)	((v) = &m->space[m->vn++ * m->g->nstates])
118442a6afdSmillert #define	onestate	long
119df930be7Sderaadt #define	INIT(o, n)	((o) = (n))
120df930be7Sderaadt #define	INC(o)	((o)++)
121df930be7Sderaadt #define	ISSTATEIN(v, o)	((v)[o])
122df930be7Sderaadt /* some abbreviations; note that some of these know variable names! */
123df930be7Sderaadt /* do "if I'm here, I can also be there" etc without branches */
124df930be7Sderaadt #define	FWD(dst, src, n)	((dst)[here+(n)] |= (src)[here])
125df930be7Sderaadt #define	BACK(dst, src, n)	((dst)[here-(n)] |= (src)[here])
126df930be7Sderaadt #define	ISSETBACK(v, n)	((v)[here - (n)])
127df930be7Sderaadt /* function names */
128df930be7Sderaadt #define	LNAMES			/* flag */
129df930be7Sderaadt 
130df930be7Sderaadt #include "engine.c"
131df930be7Sderaadt 
132df930be7Sderaadt /*
133df930be7Sderaadt  - regexec - interface for matching
134df930be7Sderaadt  *
135df930be7Sderaadt  * We put this here so we can exploit knowledge of the state representation
136df930be7Sderaadt  * when choosing which matcher to call.  Also, by this point the matchers
137df930be7Sderaadt  * have been prototyped.
138df930be7Sderaadt  */
139df930be7Sderaadt int				/* 0 success, REG_NOMATCH failure */
regexec(const regex_t * preg,const char * string,size_t nmatch,regmatch_t pmatch[],int eflags)1403d7c72a1Sotto regexec(const regex_t *preg, const char *string, size_t nmatch,
1413d7c72a1Sotto     regmatch_t pmatch[], int eflags)
142df930be7Sderaadt {
1433d7c72a1Sotto 	struct re_guts *g = preg->re_g;
1442be36617Stedu 
145df930be7Sderaadt #ifdef REDEBUG
146df930be7Sderaadt #	define	GOODFLAGS(f)	(f)
147df930be7Sderaadt #else
148df930be7Sderaadt #	define	GOODFLAGS(f)	((f)&(REG_NOTBOL|REG_NOTEOL|REG_STARTEND))
149df930be7Sderaadt #endif
150df930be7Sderaadt 
151df930be7Sderaadt 	if (preg->re_magic != MAGIC1 || g->magic != MAGIC2)
152df930be7Sderaadt 		return(REG_BADPAT);
153df930be7Sderaadt 	assert(!(g->iflags&BAD));
154df930be7Sderaadt 	if (g->iflags&BAD)		/* backstop for no-debug case */
155df930be7Sderaadt 		return(REG_BADPAT);
156df930be7Sderaadt 	eflags = GOODFLAGS(eflags);
157df930be7Sderaadt 
158df930be7Sderaadt 	if (g->nstates <= CHAR_BIT*sizeof(states1) && !(eflags&REG_LARGE))
159*d1a1db7fSmartijn 		return(smatcher(g, string, nmatch, pmatch, eflags));
160df930be7Sderaadt 	else
161*d1a1db7fSmartijn 		return(lmatcher(g, string, nmatch, pmatch, eflags));
162df930be7Sderaadt }
163