1 /*	$OpenBSD: regex2.h,v 1.7 2004/11/30 17:04:23 otto 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. 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  *	@(#)regex2.h	8.4 (Berkeley) 3/20/94
36  */
37 
38 /*
39  * internals of regex_t
40  */
41 #define	MAGIC1	((('r'^0200)<<8) | 'e')
42 
43 /*
44  * The internal representation is a *strip*, a sequence of
45  * operators ending with an endmarker.  (Some terminology etc. is a
46  * historical relic of earlier versions which used multiple strips.)
47  * Certain oddities in the representation are there to permit running
48  * the machinery backwards; in particular, any deviation from sequential
49  * flow must be marked at both its source and its destination.  Some
50  * fine points:
51  *
52  * - OPLUS_ and O_PLUS are *inside* the loop they create.
53  * - OQUEST_ and O_QUEST are *outside* the bypass they create.
54  * - OCH_ and O_CH are *outside* the multi-way branch they create, while
55  *   OOR1 and OOR2 are respectively the end and the beginning of one of
56  *   the branches.  Note that there is an implicit OOR2 following OCH_
57  *   and an implicit OOR1 preceding O_CH.
58  *
59  * In state representations, an operator's bit is on to signify a state
60  * immediately *preceding* "execution" of that operator.
61  */
62 typedef unsigned long sop;	/* strip operator */
63 typedef long sopno;
64 #define	OPRMASK	0xf8000000LU
65 #define	OPDMASK	0x07ffffffLU
66 #define	OPSHIFT	((unsigned)27)
67 #define	OP(n)	((n)&OPRMASK)
68 #define	OPND(n)	((n)&OPDMASK)
69 #define	SOP(op, opnd)	((op)|(opnd))
70 /* operators			   meaning	operand			*/
71 /*						(back, fwd are offsets)	*/
72 #define	OEND	(1LU<<OPSHIFT)	/* endmarker	-			*/
73 #define	OCHAR	(2LU<<OPSHIFT)	/* character	unsigned char		*/
74 #define	OBOL	(3LU<<OPSHIFT)	/* left anchor	-			*/
75 #define	OEOL	(4LU<<OPSHIFT)	/* right anchor	-			*/
76 #define	OANY	(5LU<<OPSHIFT)	/* .		-			*/
77 #define	OANYOF	(6LU<<OPSHIFT)	/* [...]	set number		*/
78 #define	OBACK_	(7LU<<OPSHIFT)	/* begin \d	paren number		*/
79 #define	O_BACK	(8LU<<OPSHIFT)	/* end \d	paren number		*/
80 #define	OPLUS_	(9LU<<OPSHIFT)	/* + prefix	fwd to suffix		*/
81 #define	O_PLUS	(10LU<<OPSHIFT)	/* + suffix	back to prefix		*/
82 #define	OQUEST_	(11LU<<OPSHIFT)	/* ? prefix	fwd to suffix		*/
83 #define	O_QUEST	(12LU<<OPSHIFT)	/* ? suffix	back to prefix		*/
84 #define	OLPAREN	(13LU<<OPSHIFT)	/* (		fwd to )		*/
85 #define	ORPAREN	(14LU<<OPSHIFT)	/* )		back to (		*/
86 #define	OCH_	(15LU<<OPSHIFT)	/* begin choice	fwd to OOR2		*/
87 #define	OOR1	(16LU<<OPSHIFT)	/* | pt. 1	back to OOR1 or OCH_	*/
88 #define	OOR2	(17LU<<OPSHIFT)	/* | pt. 2	fwd to OOR2 or O_CH	*/
89 #define	O_CH	(18LU<<OPSHIFT)	/* end choice	back to OOR1		*/
90 #define	OBOW	(19LU<<OPSHIFT)	/* begin word	-			*/
91 #define	OEOW	(20LU<<OPSHIFT)	/* end word	-			*/
92 
93 /*
94  * Structure for [] character-set representation.  Character sets are
95  * done as bit vectors, grouped 8 to a byte vector for compactness.
96  * The individual set therefore has both a pointer to the byte vector
97  * and a mask to pick out the relevant bit of each byte.  A hash code
98  * simplifies testing whether two sets could be identical.
99  *
100  * This will get trickier for multicharacter collating elements.  As
101  * preliminary hooks for dealing with such things, we also carry along
102  * a string of multi-character elements, and decide the size of the
103  * vectors at run time.
104  */
105 typedef struct {
106 	uch *ptr;		/* -> uch [csetsize] */
107 	uch mask;		/* bit within array */
108 	uch hash;		/* hash code */
109 	size_t smultis;
110 	char *multis;		/* -> char[smulti]  ab\0cd\0ef\0\0 */
111 } cset;
112 /* note that CHadd and CHsub are unsafe, and CHIN doesn't yield 0/1 */
113 #define	CHadd(cs, c)	((cs)->ptr[(uch)(c)] |= (cs)->mask, (cs)->hash += (c))
114 #define	CHsub(cs, c)	((cs)->ptr[(uch)(c)] &= ~(cs)->mask, (cs)->hash -= (c))
115 #define	CHIN(cs, c)	((cs)->ptr[(uch)(c)] & (cs)->mask)
116 #define	MCadd(p, cs, cp)	mcadd(p, cs, cp)	/* regcomp() internal fns */
117 #define	MCsub(p, cs, cp)	mcsub(p, cs, cp)
118 #define	MCin(p, cs, cp)	mcin(p, cs, cp)
119 
120 /* stuff for character categories */
121 typedef unsigned char cat_t;
122 
123 /*
124  * main compiled-expression structure
125  */
126 struct re_guts {
127 	int magic;
128 #		define	MAGIC2	((('R'^0200)<<8)|'E')
129 	sop *strip;		/* malloced area for strip */
130 	int csetsize;		/* number of bits in a cset vector */
131 	int ncsets;		/* number of csets in use */
132 	cset *sets;		/* -> cset [ncsets] */
133 	uch *setbits;		/* -> uch[csetsize][ncsets/CHAR_BIT] */
134 	int cflags;		/* copy of regcomp() cflags argument */
135 	sopno nstates;		/* = number of sops */
136 	sopno firststate;	/* the initial OEND (normally 0) */
137 	sopno laststate;	/* the final OEND */
138 	int iflags;		/* internal flags */
139 #		define	USEBOL	01	/* used ^ */
140 #		define	USEEOL	02	/* used $ */
141 #		define	BAD	04	/* something wrong */
142 	int nbol;		/* number of ^ used */
143 	int neol;		/* number of $ used */
144 	int ncategories;	/* how many character categories */
145 	cat_t *categories;	/* ->catspace[-CHAR_MIN] */
146 	char *must;		/* match must contain this string */
147 	int mlen;		/* length of must */
148 	size_t nsub;		/* copy of re_nsub */
149 	int backrefs;		/* does it use back references? */
150 	sopno nplus;		/* how deep does it nest +s? */
151 	/* catspace must be last */
152 	cat_t catspace[1];	/* actually [NC] */
153 };
154 
155 /* misc utilities */
156 #define	OUT	(CHAR_MAX+1)	/* a non-character value */
157 #define	ISWORD(c)	(isalnum(c) || (c) == '_')
158