11573Srgrimes/*- 21573Srgrimes * Copyright (c) 1992, 1993, 1994 Henry Spencer. 31573Srgrimes * Copyright (c) 1992, 1993, 1994 41573Srgrimes * The Regents of the University of California. All rights reserved. 51573Srgrimes * 61573Srgrimes * This code is derived from software contributed to Berkeley by 71573Srgrimes * Henry Spencer. 81573Srgrimes * 91573Srgrimes * Redistribution and use in source and binary forms, with or without 101573Srgrimes * modification, are permitted provided that the following conditions 111573Srgrimes * are met: 121573Srgrimes * 1. Redistributions of source code must retain the above copyright 131573Srgrimes * notice, this list of conditions and the following disclaimer. 141573Srgrimes * 2. Redistributions in binary form must reproduce the above copyright 151573Srgrimes * notice, this list of conditions and the following disclaimer in the 161573Srgrimes * documentation and/or other materials provided with the distribution. 171573Srgrimes * 4. Neither the name of the University nor the names of its contributors 181573Srgrimes * may be used to endorse or promote products derived from this software 191573Srgrimes * without specific prior written permission. 201573Srgrimes * 211573Srgrimes * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 221573Srgrimes * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 231573Srgrimes * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 241573Srgrimes * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 251573Srgrimes * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 261573Srgrimes * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 271573Srgrimes * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 281573Srgrimes * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 291573Srgrimes * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 301573Srgrimes * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 311573Srgrimes * SUCH DAMAGE. 321573Srgrimes * 331573Srgrimes * @(#)regex2.h 8.4 (Berkeley) 3/20/94 3462232Sdcs * $FreeBSD$ 351573Srgrimes */ 361573Srgrimes 371573Srgrimes/* 381573Srgrimes * First, the stuff that ends up in the outside-world include file 391573Srgrimes = typedef off_t regoff_t; 401573Srgrimes = typedef struct { 411573Srgrimes = int re_magic; 421573Srgrimes = size_t re_nsub; // number of parenthesized subexpressions 431573Srgrimes = const char *re_endp; // end pointer for REG_PEND 441573Srgrimes = struct re_guts *re_g; // none of your business :-) 451573Srgrimes = } regex_t; 461573Srgrimes = typedef struct { 471573Srgrimes = regoff_t rm_so; // start of match 481573Srgrimes = regoff_t rm_eo; // end of match 491573Srgrimes = } regmatch_t; 501573Srgrimes */ 511573Srgrimes/* 521573Srgrimes * internals of regex_t 531573Srgrimes */ 541573Srgrimes#define MAGIC1 ((('r'^0200)<<8) | 'e') 551573Srgrimes 561573Srgrimes/* 571573Srgrimes * The internal representation is a *strip*, a sequence of 581573Srgrimes * operators ending with an endmarker. (Some terminology etc. is a 591573Srgrimes * historical relic of earlier versions which used multiple strips.) 601573Srgrimes * Certain oddities in the representation are there to permit running 611573Srgrimes * the machinery backwards; in particular, any deviation from sequential 621573Srgrimes * flow must be marked at both its source and its destination. Some 631573Srgrimes * fine points: 641573Srgrimes * 651573Srgrimes * - OPLUS_ and O_PLUS are *inside* the loop they create. 661573Srgrimes * - OQUEST_ and O_QUEST are *outside* the bypass they create. 671573Srgrimes * - OCH_ and O_CH are *outside* the multi-way branch they create, while 681573Srgrimes * OOR1 and OOR2 are respectively the end and the beginning of one of 691573Srgrimes * the branches. Note that there is an implicit OOR2 following OCH_ 701573Srgrimes * and an implicit OOR1 preceding O_CH. 711573Srgrimes * 721573Srgrimes * In state representations, an operator's bit is on to signify a state 731573Srgrimes * immediately *preceding* "execution" of that operator. 741573Srgrimes */ 751573Srgrimestypedef unsigned long sop; /* strip operator */ 761573Srgrimestypedef long sopno; 7736043Sjb#define OPRMASK 0xf8000000L 7836043Sjb#define OPDMASK 0x07ffffffL 791573Srgrimes#define OPSHIFT ((unsigned)27) 801573Srgrimes#define OP(n) ((n)&OPRMASK) 811573Srgrimes#define OPND(n) ((n)&OPDMASK) 821573Srgrimes#define SOP(op, opnd) ((op)|(opnd)) 831573Srgrimes/* operators meaning operand */ 841573Srgrimes/* (back, fwd are offsets) */ 8536043Sjb#define OEND (1L<<OPSHIFT) /* endmarker - */ 86132019Stjr#define OCHAR (2L<<OPSHIFT) /* character wide character */ 8736043Sjb#define OBOL (3L<<OPSHIFT) /* left anchor - */ 8836043Sjb#define OEOL (4L<<OPSHIFT) /* right anchor - */ 8936043Sjb#define OANY (5L<<OPSHIFT) /* . - */ 9036043Sjb#define OANYOF (6L<<OPSHIFT) /* [...] set number */ 9136043Sjb#define OBACK_ (7L<<OPSHIFT) /* begin \d paren number */ 9236043Sjb#define O_BACK (8L<<OPSHIFT) /* end \d paren number */ 9336043Sjb#define OPLUS_ (9L<<OPSHIFT) /* + prefix fwd to suffix */ 9436043Sjb#define O_PLUS (10L<<OPSHIFT) /* + suffix back to prefix */ 9536043Sjb#define OQUEST_ (11L<<OPSHIFT) /* ? prefix fwd to suffix */ 9636043Sjb#define O_QUEST (12L<<OPSHIFT) /* ? suffix back to prefix */ 9736043Sjb#define OLPAREN (13L<<OPSHIFT) /* ( fwd to ) */ 9836043Sjb#define ORPAREN (14L<<OPSHIFT) /* ) back to ( */ 9936043Sjb#define OCH_ (15L<<OPSHIFT) /* begin choice fwd to OOR2 */ 10036043Sjb#define OOR1 (16L<<OPSHIFT) /* | pt. 1 back to OOR1 or OCH_ */ 10136043Sjb#define OOR2 (17L<<OPSHIFT) /* | pt. 2 fwd to OOR2 or O_CH */ 10236043Sjb#define O_CH (18L<<OPSHIFT) /* end choice back to OOR1 */ 10336043Sjb#define OBOW (19L<<OPSHIFT) /* begin word - */ 10436043Sjb#define OEOW (20L<<OPSHIFT) /* end word - */ 1051573Srgrimes 1061573Srgrimes/* 107132019Stjr * Structures for [] character-set representation. 1081573Srgrimes */ 1091573Srgrimestypedef struct { 110132019Stjr wint_t min; 111132019Stjr wint_t max; 112132019Stjr} crange; 113132019Stjrtypedef struct { 114132019Stjr unsigned char bmp[NC / 8]; 115132019Stjr wctype_t *types; 116132019Stjr int ntypes; 117132019Stjr wint_t *wides; 118132019Stjr int nwides; 119132019Stjr crange *ranges; 120132019Stjr int nranges; 121132019Stjr int invert; 122132019Stjr int icase; 1231573Srgrimes} cset; 1241573Srgrimes 125132019Stjrstatic int 126150053SstefanfCHIN1(cset *cs, wint_t ch) 127132019Stjr{ 128132019Stjr int i; 129132019Stjr 130132019Stjr assert(ch >= 0); 131132019Stjr if (ch < NC) 132132019Stjr return (((cs->bmp[ch >> 3] & (1 << (ch & 7))) != 0) ^ 133132019Stjr cs->invert); 134132019Stjr for (i = 0; i < cs->nwides; i++) 135132019Stjr if (ch == cs->wides[i]) 136132019Stjr return (!cs->invert); 137132019Stjr for (i = 0; i < cs->nranges; i++) 138132019Stjr if (cs->ranges[i].min <= ch && ch <= cs->ranges[i].max) 139132019Stjr return (!cs->invert); 140132019Stjr for (i = 0; i < cs->ntypes; i++) 141132019Stjr if (iswctype(ch, cs->types[i])) 142132019Stjr return (!cs->invert); 143132019Stjr return (cs->invert); 144132019Stjr} 145132019Stjr 146132019Stjrstatic __inline int 147150053SstefanfCHIN(cset *cs, wint_t ch) 148132019Stjr{ 149132019Stjr 150132019Stjr assert(ch >= 0); 151132019Stjr if (ch < NC) 152132019Stjr return (((cs->bmp[ch >> 3] & (1 << (ch & 7))) != 0) ^ 153132019Stjr cs->invert); 154132019Stjr else if (cs->icase) 155132019Stjr return (CHIN1(cs, ch) || CHIN1(cs, towlower(ch)) || 156132019Stjr CHIN1(cs, towupper(ch))); 157132019Stjr else 158132019Stjr return (CHIN1(cs, ch)); 159132019Stjr} 160132019Stjr 1611573Srgrimes/* 1621573Srgrimes * main compiled-expression structure 1631573Srgrimes */ 1641573Srgrimesstruct re_guts { 1651573Srgrimes int magic; 1661573Srgrimes# define MAGIC2 ((('R'^0200)<<8)|'E') 1671573Srgrimes sop *strip; /* malloced area for strip */ 1681573Srgrimes int ncsets; /* number of csets in use */ 1691573Srgrimes cset *sets; /* -> cset [ncsets] */ 1701573Srgrimes int cflags; /* copy of regcomp() cflags argument */ 1711573Srgrimes sopno nstates; /* = number of sops */ 1721573Srgrimes sopno firststate; /* the initial OEND (normally 0) */ 1731573Srgrimes sopno laststate; /* the final OEND */ 1741573Srgrimes int iflags; /* internal flags */ 1751573Srgrimes# define USEBOL 01 /* used ^ */ 1761573Srgrimes# define USEEOL 02 /* used $ */ 1771573Srgrimes# define BAD 04 /* something wrong */ 1781573Srgrimes int nbol; /* number of ^ used */ 1791573Srgrimes int neol; /* number of $ used */ 1801573Srgrimes char *must; /* match must contain this string */ 18162391Sdcs int moffset; /* latest point at which must may be located */ 18262232Sdcs int *charjump; /* Boyer-Moore char jump table */ 18362232Sdcs int *matchjump; /* Boyer-Moore match jump table */ 1841573Srgrimes int mlen; /* length of must */ 1851573Srgrimes size_t nsub; /* copy of re_nsub */ 1861573Srgrimes int backrefs; /* does it use back references? */ 1871573Srgrimes sopno nplus; /* how deep does it nest +s? */ 1881573Srgrimes}; 1891573Srgrimes 1901573Srgrimes/* misc utilities */ 191149009Stjr#define OUT (CHAR_MIN - 1) /* a non-character value */ 192132019Stjr#define ISWORD(c) (iswalnum((uch)(c)) || (c) == '_') 193