166 lines · c
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, 19946 * The Regents of the University of California. All rights reserved.7 *8 * This code is derived from software contributed to Berkeley by9 * Henry Spencer.10 *11 * Redistribution and use in source and binary forms, with or without12 * modification, are permitted provided that the following conditions13 * are met:14 * 1. Redistributions of source code must retain the above copyright15 * notice, this list of conditions and the following disclaimer.16 * 2. Redistributions in binary form must reproduce the above copyright17 * notice, this list of conditions and the following disclaimer in the18 * documentation and/or other materials provided with the distribution.19 * 3. Neither the name of the University nor the names of its contributors20 * may be used to endorse or promote products derived from this software21 * without specific prior written permission.22 *23 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND24 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE25 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE26 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE27 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL28 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS29 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)30 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT31 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY32 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF33 * SUCH DAMAGE.34 *35 * @(#)regex2.h 8.4 (Berkeley) 3/20/9436 */37 38#ifndef LLVM_SUPPORT_REGEX2_H39#define LLVM_SUPPORT_REGEX2_H40 41#include "regutils.h"42#include <stddef.h>43 44/*45 * internals of regex_t46 */47#define MAGIC1 ((('r'^0200)<<8) | 'e')48 49/*50 * The internal representation is a *strip*, a sequence of51 * operators ending with an endmarker. (Some terminology etc. is a52 * historical relic of earlier versions which used multiple strips.)53 * Certain oddities in the representation are there to permit running54 * the machinery backwards; in particular, any deviation from sequential55 * flow must be marked at both its source and its destination. Some56 * fine points:57 *58 * - OPLUS_ and O_PLUS are *inside* the loop they create.59 * - OQUEST_ and O_QUEST are *outside* the bypass they create.60 * - OCH_ and O_CH are *outside* the multi-way branch they create, while61 * OOR1 and OOR2 are respectively the end and the beginning of one of62 * the branches. Note that there is an implicit OOR2 following OCH_63 * and an implicit OOR1 preceding O_CH.64 *65 * In state representations, an operator's bit is on to signify a state66 * immediately *preceding* "execution" of that operator.67 */68typedef unsigned long sop; /* strip operator */69typedef long sopno;70#define OPRMASK 0xf8000000LU71#define OPDMASK 0x07ffffffLU72#define OPSHIFT ((unsigned)27)73#define OP(n) ((n)&OPRMASK)74#define OPND(n) ((n)&OPDMASK)75#define SOP(op, opnd) ((op)|(opnd))76/* operators meaning operand */77/* (back, fwd are offsets) */78#define OEND (1LU<<OPSHIFT) /* endmarker - */79#define OCHAR (2LU<<OPSHIFT) /* character unsigned char */80#define OBOL (3LU<<OPSHIFT) /* left anchor - */81#define OEOL (4LU<<OPSHIFT) /* right anchor - */82#define OANY (5LU<<OPSHIFT) /* . - */83#define OANYOF (6LU<<OPSHIFT) /* [...] set number */84#define OBACK_ (7LU<<OPSHIFT) /* begin \d paren number */85#define O_BACK (8LU<<OPSHIFT) /* end \d paren number */86#define OPLUS_ (9LU<<OPSHIFT) /* + prefix fwd to suffix */87#define O_PLUS (10LU<<OPSHIFT) /* + suffix back to prefix */88#define OQUEST_ (11LU<<OPSHIFT) /* ? prefix fwd to suffix */89#define O_QUEST (12LU<<OPSHIFT) /* ? suffix back to prefix */90#define OLPAREN (13LU<<OPSHIFT) /* ( fwd to ) */91#define ORPAREN (14LU<<OPSHIFT) /* ) back to ( */92#define OCH_ (15LU<<OPSHIFT) /* begin choice fwd to OOR2 */93#define OOR1 (16LU<<OPSHIFT) /* | pt. 1 back to OOR1 or OCH_ */94#define OOR2 (17LU<<OPSHIFT) /* | pt. 2 fwd to OOR2 or O_CH */95#define O_CH (18LU<<OPSHIFT) /* end choice back to OOR1 */96#define OBOW (19LU<<OPSHIFT) /* begin word - */97#define OEOW (20LU<<OPSHIFT) /* end word - */98 99/*100 * Structure for [] character-set representation. Character sets are101 * done as bit vectors, grouped 8 to a byte vector for compactness.102 * The individual set therefore has both a pointer to the byte vector103 * and a mask to pick out the relevant bit of each byte. A hash code104 * simplifies testing whether two sets could be identical.105 *106 * This will get trickier for multicharacter collating elements. As107 * preliminary hooks for dealing with such things, we also carry along108 * a string of multi-character elements, and decide the size of the109 * vectors at run time.110 */111typedef struct {112 uch *ptr; /* -> uch [csetsize] */113 uch mask; /* bit within array */114 uch hash; /* hash code */115 size_t smultis;116 char *multis; /* -> char[smulti] ab\0cd\0ef\0\0 */117} cset;118/* note that CHadd and CHsub are unsafe, and CHIN doesn't yield 0/1 */119#define CHadd(cs, c) ((cs)->ptr[(uch)(c)] |= (cs)->mask, (cs)->hash += (c))120#define CHsub(cs, c) ((cs)->ptr[(uch)(c)] &= ~(cs)->mask, (cs)->hash -= (c))121#define CHIN(cs, c) ((cs)->ptr[(uch)(c)] & (cs)->mask)122#define MCadd(p, cs, cp) mcadd(p, cs, cp) /* llvm_regcomp() internal fns */123#define MCsub(p, cs, cp) mcsub(p, cs, cp)124#define MCin(p, cs, cp) mcin(p, cs, cp)125 126/* stuff for character categories */127typedef unsigned char cat_t;128 129/*130 * main compiled-expression structure131 */132struct re_guts {133 int magic;134# define MAGIC2 ((('R'^0200)<<8)|'E')135 sop *strip; /* malloced area for strip */136 int csetsize; /* number of bits in a cset vector */137 int ncsets; /* number of csets in use */138 cset *sets; /* -> cset [ncsets] */139 uch *setbits; /* -> uch[csetsize][ncsets/CHAR_BIT] */140 int cflags; /* copy of llvm_regcomp() cflags argument */141 sopno nstates; /* = number of sops */142 sopno firststate; /* the initial OEND (normally 0) */143 sopno laststate; /* the final OEND */144 int iflags; /* internal flags */145# define USEBOL 01 /* used ^ */146# define USEEOL 02 /* used $ */147# define REGEX_BAD 04 /* something wrong */148 int nbol; /* number of ^ used */149 int neol; /* number of $ used */150 int ncategories; /* how many character categories */151 cat_t *categories; /* ->catspace[-CHAR_MIN] */152 char *must; /* match must contain this string */153 int mlen; /* length of must */154 size_t nsub; /* copy of re_nsub */155 int backrefs; /* does it use back references? */156 sopno nplus; /* how deep does it nest +s? */157 /* catspace must be last */158 cat_t catspace[1]; /* actually [NC] */159};160 161/* misc utilities */162#define OUT (CHAR_MAX+1) /* a non-character value */163#define ISWORD(c) (isalnum(c&0xff) || (c) == '_')164 165#endif166