1 /*-
2  * This code is derived from OpenBSD's libc/regex, original license follows:
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. Neither the name of the University nor the names of its contributors
20  *    may be used to endorse or promote products derived from this software
21  *    without specific prior written permission.
22  *
23  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
24  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
26  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
27  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
29  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
30  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33  * SUCH DAMAGE.
34  *
35  *	@(#)regexec.c	8.3 (Berkeley) 3/20/94
36  */
37 
38 /*
39  * the outer shell of llvm_regexec()
40  *
41  * This file includes engine.inc *twice*, after muchos fiddling with the
42  * macros that code uses.  This lets the same code operate on two different
43  * representations for state sets.
44  */
45 #include <sys/types.h>
46 #include <stdio.h>
47 #include <stdlib.h>
48 #include <string.h>
49 #include <limits.h>
50 #include <ctype.h>
51 #include "regex_impl.h"
52 
53 #include "regutils.h"
54 #include "regex2.h"
55 
56 /* macros for manipulating states, small version */
57 /* FIXME: 'states' is assumed as 'long' on small version. */
58 #define	states1	long		/* for later use in llvm_regexec() decision */
59 #define	states	states1
60 #define	CLEAR(v)	((v) = 0)
61 #define	SET0(v, n)	((v) &= ~((unsigned long)1 << (n)))
62 #define	SET1(v, n)	((v) |= (unsigned long)1 << (n))
63 #define	ISSET(v, n)	(((v) & ((unsigned long)1 << (n))) != 0)
64 #define	ASSIGN(d, s)	((d) = (s))
65 #define	EQ(a, b)	((a) == (b))
66 #define	STATEVARS	long dummy	/* dummy version */
67 #define	STATESETUP(m, n)	/* nothing */
68 #define	STATETEARDOWN(m)	/* nothing */
69 #define	SETUP(v)	((v) = 0)
70 #define	onestate	long
71 #define	INIT(o, n)	((o) = (unsigned long)1 << (n))
72 #define	INC(o)		((o) = (unsigned long)(o) << 1)
73 #define	ISSTATEIN(v, o)	(((v) & (o)) != 0)
74 /* some abbreviations; note that some of these know variable names! */
75 /* do "if I'm here, I can also be there" etc without branches */
76 #define	FWD(dst, src, n)	((dst) |= ((unsigned long)(src)&(here)) << (n))
77 #define	BACK(dst, src, n)	((dst) |= ((unsigned long)(src)&(here)) >> (n))
78 #define	ISSETBACK(v, n)		(((v) & ((unsigned long)here >> (n))) != 0)
79 /* function names */
80 #define SNAMES			/* engine.inc looks after details */
81 
82 #include "regengine.inc"
83 
84 /* now undo things */
85 #undef	states
86 #undef	CLEAR
87 #undef	SET0
88 #undef	SET1
89 #undef	ISSET
90 #undef	ASSIGN
91 #undef	EQ
92 #undef	STATEVARS
93 #undef	STATESETUP
94 #undef	STATETEARDOWN
95 #undef	SETUP
96 #undef	onestate
97 #undef	INIT
98 #undef	INC
99 #undef	ISSTATEIN
100 #undef	FWD
101 #undef	BACK
102 #undef	ISSETBACK
103 #undef	SNAMES
104 
105 /* macros for manipulating states, large version */
106 #define	states	char *
107 #define	CLEAR(v)	memset(v, 0, m->g->nstates)
108 #define	SET0(v, n)	((v)[n] = 0)
109 #define	SET1(v, n)	((v)[n] = 1)
110 #define	ISSET(v, n)	((v)[n])
111 #define	ASSIGN(d, s)	memmove(d, s, m->g->nstates)
112 #define	EQ(a, b)	(memcmp(a, b, m->g->nstates) == 0)
113 #define	STATEVARS	long vn; char *space
114 #define	STATESETUP(m, nv)	{ (m)->space = malloc((nv)*(m)->g->nstates); \
115 				if ((m)->space == NULL) return(REG_ESPACE); \
116 				(m)->vn = 0; }
117 #define	STATETEARDOWN(m)	{ free((m)->space); }
118 #define	SETUP(v)	((v) = &m->space[m->vn++ * m->g->nstates])
119 #define	onestate	long
120 #define	INIT(o, n)	((o) = (n))
121 #define	INC(o)	((o)++)
122 #define	ISSTATEIN(v, o)	((v)[o])
123 /* some abbreviations; note that some of these know variable names! */
124 /* do "if I'm here, I can also be there" etc without branches */
125 #define	FWD(dst, src, n)	((dst)[here+(n)] |= (src)[here])
126 #define	BACK(dst, src, n)	((dst)[here-(n)] |= (src)[here])
127 #define	ISSETBACK(v, n)	((v)[here - (n)])
128 /* function names */
129 #define	LNAMES			/* flag */
130 
131 #include "regengine.inc"
132 
133 /*
134  - llvm_regexec - interface for matching
135  *
136  * We put this here so we can exploit knowledge of the state representation
137  * when choosing which matcher to call.  Also, by this point the matchers
138  * have been prototyped.
139  */
140 int				/* 0 success, REG_NOMATCH failure */
llvm_regexec(const llvm_regex_t * preg,const char * string,size_t nmatch,llvm_regmatch_t pmatch[],int eflags)141 llvm_regexec(const llvm_regex_t *preg, const char *string, size_t nmatch,
142              llvm_regmatch_t pmatch[], int eflags)
143 {
144 	struct re_guts *g = preg->re_g;
145 #ifdef REDEBUG
146 #	define	GOODFLAGS(f)	(f)
147 #else
148 #	define	GOODFLAGS(f)	((f)&(REG_NOTBOL|REG_NOTEOL|REG_STARTEND))
149 #endif
150 
151 	if (preg->re_magic != MAGIC1 || g->magic != MAGIC2)
152 		return(REG_BADPAT);
153 	assert(!(g->iflags&REGEX_BAD));
154 	if (g->iflags&REGEX_BAD)		/* backstop for no-debug case */
155 		return(REG_BADPAT);
156 	eflags = GOODFLAGS(eflags);
157 
158 	if (g->nstates <= (long)(CHAR_BIT*sizeof(states1)) && !(eflags&REG_LARGE))
159 		return(smatcher(g, string, nmatch, pmatch, eflags));
160 	else
161 		return(lmatcher(g, string, nmatch, pmatch, eflags));
162 }
163