/*-
|
/*-
|
* Copyright (c) 1992, 1993, 1994 Henry Spencer.
|
* Copyright (c) 1992, 1993, 1994 Henry Spencer.
|
* Copyright (c) 1992, 1993, 1994
|
* Copyright (c) 1992, 1993, 1994
|
* The Regents of the University of California. All rights reserved.
|
* The Regents of the University of California. All rights reserved.
|
*
|
*
|
* This code is derived from software contributed to Berkeley by
|
* This code is derived from software contributed to Berkeley by
|
* Henry Spencer.
|
* Henry Spencer.
|
*
|
*
|
* Redistribution and use in source and binary forms, with or without
|
* Redistribution and use in source and binary forms, with or without
|
* modification, are permitted provided that the following conditions
|
* modification, are permitted provided that the following conditions
|
* are met:
|
* are met:
|
* 1. Redistributions of source code must retain the above copyright
|
* 1. Redistributions of source code must retain the above copyright
|
* notice, this list of conditions and the following disclaimer.
|
* notice, this list of conditions and the following disclaimer.
|
* 2. Redistributions in binary form must reproduce the above copyright
|
* 2. Redistributions in binary form must reproduce the above copyright
|
* notice, this list of conditions and the following disclaimer in the
|
* notice, this list of conditions and the following disclaimer in the
|
* documentation and/or other materials provided with the distribution.
|
* documentation and/or other materials provided with the distribution.
|
* 4. Neither the name of the University nor the names of its contributors
|
* 4. Neither the name of the University nor the names of its contributors
|
* may be used to endorse or promote products derived from this software
|
* may be used to endorse or promote products derived from this software
|
* without specific prior written permission.
|
* without specific prior written permission.
|
*
|
*
|
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
|
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
* SUCH DAMAGE.
|
* SUCH DAMAGE.
|
*
|
*
|
* @(#)regcomp.c 8.5 (Berkeley) 3/20/94
|
* @(#)regcomp.c 8.5 (Berkeley) 3/20/94
|
*/
|
*/
|
|
|
#ifndef _NO_REGEX
|
#ifndef _NO_REGEX
|
|
|
#if defined(LIBC_SCCS) && !defined(lint)
|
#if defined(LIBC_SCCS) && !defined(lint)
|
static char sccsid[] = "@(#)regcomp.c 8.5 (Berkeley) 3/20/94";
|
static char sccsid[] = "@(#)regcomp.c 8.5 (Berkeley) 3/20/94";
|
#endif /* LIBC_SCCS and not lint */
|
#endif /* LIBC_SCCS and not lint */
|
#include <sys/cdefs.h>
|
#include <sys/cdefs.h>
|
|
|
#include <sys/types.h>
|
#include <sys/types.h>
|
#include <stdio.h>
|
#include <stdio.h>
|
#include <string.h>
|
#include <string.h>
|
#include <ctype.h>
|
#include <ctype.h>
|
#include <limits.h>
|
#include <limits.h>
|
#include <stdlib.h>
|
#include <stdlib.h>
|
#include <regex.h>
|
#include <regex.h>
|
|
|
#include "collate.h"
|
#include "collate.h"
|
|
|
#include "utils.h"
|
#include "utils.h"
|
#include "regex2.h"
|
#include "regex2.h"
|
|
|
#include "cclass.h"
|
#include "cclass.h"
|
#include "cname.h"
|
#include "cname.h"
|
|
|
/*
|
/*
|
* parse structure, passed up and down to avoid global variables and
|
* parse structure, passed up and down to avoid global variables and
|
* other clumsinesses
|
* other clumsinesses
|
*/
|
*/
|
struct parse {
|
struct parse {
|
char *next; /* next character in RE */
|
char *next; /* next character in RE */
|
char *end; /* end of string (-> NUL normally) */
|
char *end; /* end of string (-> NUL normally) */
|
int error; /* has an error been seen? */
|
int error; /* has an error been seen? */
|
sop *strip; /* malloced strip */
|
sop *strip; /* malloced strip */
|
sopno ssize; /* malloced strip size (allocated) */
|
sopno ssize; /* malloced strip size (allocated) */
|
sopno slen; /* malloced strip length (used) */
|
sopno slen; /* malloced strip length (used) */
|
int ncsalloc; /* number of csets allocated */
|
int ncsalloc; /* number of csets allocated */
|
struct re_guts *g;
|
struct re_guts *g;
|
# define NPAREN 10 /* we need to remember () 1-9 for back refs */
|
# define NPAREN 10 /* we need to remember () 1-9 for back refs */
|
sopno pbegin[NPAREN]; /* -> ( ([0] unused) */
|
sopno pbegin[NPAREN]; /* -> ( ([0] unused) */
|
sopno pend[NPAREN]; /* -> ) ([0] unused) */
|
sopno pend[NPAREN]; /* -> ) ([0] unused) */
|
};
|
};
|
|
|
/* ========= begin header generated by ./mkh ========= */
|
/* ========= begin header generated by ./mkh ========= */
|
#ifdef __cplusplus
|
#ifdef __cplusplus
|
extern "C" {
|
extern "C" {
|
#endif
|
#endif
|
|
|
/* === regcomp.c === */
|
/* === regcomp.c === */
|
static void p_ere(struct parse *p, int stop);
|
static void p_ere(struct parse *p, int stop);
|
static void p_ere_exp(struct parse *p);
|
static void p_ere_exp(struct parse *p);
|
static void p_str(struct parse *p);
|
static void p_str(struct parse *p);
|
static void p_bre(struct parse *p, int end1, int end2);
|
static void p_bre(struct parse *p, int end1, int end2);
|
static int p_simp_re(struct parse *p, int starordinary);
|
static int p_simp_re(struct parse *p, int starordinary);
|
static int p_count(struct parse *p);
|
static int p_count(struct parse *p);
|
static void p_bracket(struct parse *p);
|
static void p_bracket(struct parse *p);
|
static void p_b_term(struct parse *p, cset *cs);
|
static void p_b_term(struct parse *p, cset *cs);
|
static void p_b_cclass(struct parse *p, cset *cs);
|
static void p_b_cclass(struct parse *p, cset *cs);
|
static void p_b_eclass(struct parse *p, cset *cs);
|
static void p_b_eclass(struct parse *p, cset *cs);
|
static char p_b_symbol(struct parse *p);
|
static char p_b_symbol(struct parse *p);
|
static char p_b_coll_elem(struct parse *p, int endc);
|
static char p_b_coll_elem(struct parse *p, int endc);
|
static char othercase(int ch);
|
static char othercase(int ch);
|
static void bothcases(struct parse *p, int ch);
|
static void bothcases(struct parse *p, int ch);
|
static void ordinary(struct parse *p, int ch);
|
static void ordinary(struct parse *p, int ch);
|
static void nonnewline(struct parse *p);
|
static void nonnewline(struct parse *p);
|
static void repeat(struct parse *p, sopno start, int from, int to);
|
static void repeat(struct parse *p, sopno start, int from, int to);
|
static int seterr(struct parse *p, int e);
|
static int seterr(struct parse *p, int e);
|
static cset *allocset(struct parse *p);
|
static cset *allocset(struct parse *p);
|
static void freeset(struct parse *p, cset *cs);
|
static void freeset(struct parse *p, cset *cs);
|
static int freezeset(struct parse *p, cset *cs);
|
static int freezeset(struct parse *p, cset *cs);
|
static int firstch(struct parse *p, cset *cs);
|
static int firstch(struct parse *p, cset *cs);
|
static int nch(struct parse *p, cset *cs);
|
static int nch(struct parse *p, cset *cs);
|
static void mcadd(struct parse *p, cset *cs, char *cp);
|
static void mcadd(struct parse *p, cset *cs, char *cp);
|
#if used
|
#if used
|
static void mcsub(cset *cs, char *cp);
|
static void mcsub(cset *cs, char *cp);
|
static int mcin(cset *cs, char *cp);
|
static int mcin(cset *cs, char *cp);
|
static char *mcfind(cset *cs, char *cp);
|
static char *mcfind(cset *cs, char *cp);
|
#endif
|
#endif
|
static void mcinvert(struct parse *p, cset *cs);
|
static void mcinvert(struct parse *p, cset *cs);
|
static void mccase(struct parse *p, cset *cs);
|
static void mccase(struct parse *p, cset *cs);
|
static int isinsets(struct re_guts *g, int c);
|
static int isinsets(struct re_guts *g, int c);
|
static int samesets(struct re_guts *g, int c1, int c2);
|
static int samesets(struct re_guts *g, int c1, int c2);
|
static void categorize(struct parse *p, struct re_guts *g);
|
static void categorize(struct parse *p, struct re_guts *g);
|
static sopno dupl(struct parse *p, sopno start, sopno finish);
|
static sopno dupl(struct parse *p, sopno start, sopno finish);
|
static void doemit(struct parse *p, sop op, size_t opnd);
|
static void doemit(struct parse *p, sop op, size_t opnd);
|
static void doinsert(struct parse *p, sop op, size_t opnd, sopno pos);
|
static void doinsert(struct parse *p, sop op, size_t opnd, sopno pos);
|
static void dofwd(struct parse *p, sopno pos, sop value);
|
static void dofwd(struct parse *p, sopno pos, sop value);
|
static void enlarge(struct parse *p, sopno size);
|
static void enlarge(struct parse *p, sopno size);
|
static void stripsnug(struct parse *p, struct re_guts *g);
|
static void stripsnug(struct parse *p, struct re_guts *g);
|
static void findmust(struct parse *p, struct re_guts *g);
|
static void findmust(struct parse *p, struct re_guts *g);
|
static int altoffset(sop *scan, int offset, int mccs);
|
static int altoffset(sop *scan, int offset, int mccs);
|
static void computejumps(struct parse *p, struct re_guts *g);
|
static void computejumps(struct parse *p, struct re_guts *g);
|
static void computematchjumps(struct parse *p, struct re_guts *g);
|
static void computematchjumps(struct parse *p, struct re_guts *g);
|
static sopno pluscount(struct parse *p, struct re_guts *g);
|
static sopno pluscount(struct parse *p, struct re_guts *g);
|
|
|
#ifdef __cplusplus
|
#ifdef __cplusplus
|
}
|
}
|
#endif
|
#endif
|
/* ========= end header generated by ./mkh ========= */
|
/* ========= end header generated by ./mkh ========= */
|
|
|
static char nuls[10]; /* place to point scanner in event of error */
|
static char nuls[10]; /* place to point scanner in event of error */
|
|
|
/*
|
/*
|
* macros for use with parse structure
|
* macros for use with parse structure
|
* BEWARE: these know that the parse structure is named `p' !!!
|
* BEWARE: these know that the parse structure is named `p' !!!
|
*/
|
*/
|
#define PEEK() (*p->next)
|
#define PEEK() (*p->next)
|
#define PEEK2() (*(p->next+1))
|
#define PEEK2() (*(p->next+1))
|
#define MORE() (p->next < p->end)
|
#define MORE() (p->next < p->end)
|
#define MORE2() (p->next+1 < p->end)
|
#define MORE2() (p->next+1 < p->end)
|
#define SEE(c) (MORE() && PEEK() == (c))
|
#define SEE(c) (MORE() && PEEK() == (c))
|
#define SEETWO(a, b) (MORE() && MORE2() && PEEK() == (a) && PEEK2() == (b))
|
#define SEETWO(a, b) (MORE() && MORE2() && PEEK() == (a) && PEEK2() == (b))
|
#define EAT(c) ((SEE(c)) ? (NEXT(), 1) : 0)
|
#define EAT(c) ((SEE(c)) ? (NEXT(), 1) : 0)
|
#define EATTWO(a, b) ((SEETWO(a, b)) ? (NEXT2(), 1) : 0)
|
#define EATTWO(a, b) ((SEETWO(a, b)) ? (NEXT2(), 1) : 0)
|
#define NEXT() (p->next++)
|
#define NEXT() (p->next++)
|
#define NEXT2() (p->next += 2)
|
#define NEXT2() (p->next += 2)
|
#define NEXTn(n) (p->next += (n))
|
#define NEXTn(n) (p->next += (n))
|
#define GETNEXT() (*p->next++)
|
#define GETNEXT() (*p->next++)
|
#define SETERROR(e) seterr(p, (e))
|
#define SETERROR(e) seterr(p, (e))
|
#define REQUIRE(co, e) ((co) || SETERROR(e))
|
#define REQUIRE(co, e) ((co) || SETERROR(e))
|
#define MUSTSEE(c, e) (REQUIRE(MORE() && PEEK() == (c), e))
|
#define MUSTSEE(c, e) (REQUIRE(MORE() && PEEK() == (c), e))
|
#define MUSTEAT(c, e) (REQUIRE(MORE() && GETNEXT() == (c), e))
|
#define MUSTEAT(c, e) (REQUIRE(MORE() && GETNEXT() == (c), e))
|
#define MUSTNOTSEE(c, e) (REQUIRE(!MORE() || PEEK() != (c), e))
|
#define MUSTNOTSEE(c, e) (REQUIRE(!MORE() || PEEK() != (c), e))
|
#define EMIT(op, sopnd) doemit(p, (sop)(op), (size_t)(sopnd))
|
#define EMIT(op, sopnd) doemit(p, (sop)(op), (size_t)(sopnd))
|
#define INSERT(op, pos) doinsert(p, (sop)(op), HERE()-(pos)+1, pos)
|
#define INSERT(op, pos) doinsert(p, (sop)(op), HERE()-(pos)+1, pos)
|
#define AHEAD(pos) dofwd(p, pos, HERE()-(pos))
|
#define AHEAD(pos) dofwd(p, pos, HERE()-(pos))
|
#define ASTERN(sop, pos) EMIT(sop, HERE()-pos)
|
#define ASTERN(sop, pos) EMIT(sop, HERE()-pos)
|
#define HERE() (p->slen)
|
#define HERE() (p->slen)
|
#define THERE() (p->slen - 1)
|
#define THERE() (p->slen - 1)
|
#define THERETHERE() (p->slen - 2)
|
#define THERETHERE() (p->slen - 2)
|
#define DROP(n) (p->slen -= (n))
|
#define DROP(n) (p->slen -= (n))
|
|
|
#ifndef NDEBUG
|
#ifndef NDEBUG
|
static int never = 0; /* for use in asserts; shuts lint up */
|
static int never = 0; /* for use in asserts; shuts lint up */
|
#else
|
#else
|
#define never 0 /* some <assert.h>s have bugs too */
|
#define never 0 /* some <assert.h>s have bugs too */
|
#endif
|
#endif
|
|
|
/* Macro used by computejump()/computematchjump() */
|
/* Macro used by computejump()/computematchjump() */
|
#define MIN(a,b) ((a)<(b)?(a):(b))
|
#define MIN(a,b) ((a)<(b)?(a):(b))
|
|
|
/*
|
/*
|
- regcomp - interface for parser and compilation
|
- regcomp - interface for parser and compilation
|
= extern int regcomp(regex_t *, const char *, int);
|
= extern int regcomp(regex_t *, const char *, int);
|
= #define REG_BASIC 0000
|
= #define REG_BASIC 0000
|
= #define REG_EXTENDED 0001
|
= #define REG_EXTENDED 0001
|
= #define REG_ICASE 0002
|
= #define REG_ICASE 0002
|
= #define REG_NOSUB 0004
|
= #define REG_NOSUB 0004
|
= #define REG_NEWLINE 0010
|
= #define REG_NEWLINE 0010
|
= #define REG_NOSPEC 0020
|
= #define REG_NOSPEC 0020
|
= #define REG_PEND 0040
|
= #define REG_PEND 0040
|
= #define REG_DUMP 0200
|
= #define REG_DUMP 0200
|
*/
|
*/
|
int /* 0 success, otherwise REG_something */
|
int /* 0 success, otherwise REG_something */
|
regcomp(preg, pattern, cflags)
|
regcomp(preg, pattern, cflags)
|
regex_t *preg;
|
regex_t *preg;
|
const char *pattern;
|
const char *pattern;
|
int cflags;
|
int cflags;
|
{
|
{
|
struct parse pa;
|
struct parse pa;
|
struct re_guts *g;
|
struct re_guts *g;
|
struct parse *p = &pa;
|
struct parse *p = &pa;
|
int i;
|
int i;
|
size_t len;
|
size_t len;
|
#ifdef REDEBUG
|
#ifdef REDEBUG
|
# define GOODFLAGS(f) (f)
|
# define GOODFLAGS(f) (f)
|
#else
|
#else
|
# define GOODFLAGS(f) ((f)&~REG_DUMP)
|
# define GOODFLAGS(f) ((f)&~REG_DUMP)
|
#endif
|
#endif
|
|
|
cflags = GOODFLAGS(cflags);
|
cflags = GOODFLAGS(cflags);
|
if ((cflags®_EXTENDED) && (cflags®_NOSPEC))
|
if ((cflags®_EXTENDED) && (cflags®_NOSPEC))
|
return(REG_INVARG);
|
return(REG_INVARG);
|
|
|
if (cflags®_PEND) {
|
if (cflags®_PEND) {
|
if (preg->re_endp < pattern)
|
if (preg->re_endp < pattern)
|
return(REG_INVARG);
|
return(REG_INVARG);
|
len = preg->re_endp - pattern;
|
len = preg->re_endp - pattern;
|
} else
|
} else
|
len = strlen((char *)pattern);
|
len = strlen((char *)pattern);
|
|
|
/* do the mallocs early so failure handling is easy */
|
/* do the mallocs early so failure handling is easy */
|
g = (struct re_guts *)malloc(sizeof(struct re_guts) +
|
g = (struct re_guts *)malloc(sizeof(struct re_guts) +
|
(NC-1)*sizeof(cat_t));
|
(NC-1)*sizeof(cat_t));
|
if (g == NULL)
|
if (g == NULL)
|
return(REG_ESPACE);
|
return(REG_ESPACE);
|
p->ssize = len/(size_t)2*(size_t)3 + (size_t)1; /* ugh */
|
p->ssize = len/(size_t)2*(size_t)3 + (size_t)1; /* ugh */
|
p->strip = (sop *)malloc(p->ssize * sizeof(sop));
|
p->strip = (sop *)malloc(p->ssize * sizeof(sop));
|
p->slen = 0;
|
p->slen = 0;
|
if (p->strip == NULL) {
|
if (p->strip == NULL) {
|
free((char *)g);
|
free((char *)g);
|
return(REG_ESPACE);
|
return(REG_ESPACE);
|
}
|
}
|
|
|
/* set things up */
|
/* set things up */
|
p->g = g;
|
p->g = g;
|
p->next = (char *)pattern; /* convenience; we do not modify it */
|
p->next = (char *)pattern; /* convenience; we do not modify it */
|
p->end = p->next + len;
|
p->end = p->next + len;
|
p->error = 0;
|
p->error = 0;
|
p->ncsalloc = 0;
|
p->ncsalloc = 0;
|
for (i = 0; i < NPAREN; i++) {
|
for (i = 0; i < NPAREN; i++) {
|
p->pbegin[i] = 0;
|
p->pbegin[i] = 0;
|
p->pend[i] = 0;
|
p->pend[i] = 0;
|
}
|
}
|
g->csetsize = NC;
|
g->csetsize = NC;
|
g->sets = NULL;
|
g->sets = NULL;
|
g->setbits = NULL;
|
g->setbits = NULL;
|
g->ncsets = 0;
|
g->ncsets = 0;
|
g->cflags = cflags;
|
g->cflags = cflags;
|
g->iflags = 0;
|
g->iflags = 0;
|
g->nbol = 0;
|
g->nbol = 0;
|
g->neol = 0;
|
g->neol = 0;
|
g->must = NULL;
|
g->must = NULL;
|
g->moffset = -1;
|
g->moffset = -1;
|
g->charjump = NULL;
|
g->charjump = NULL;
|
g->matchjump = NULL;
|
g->matchjump = NULL;
|
g->mlen = 0;
|
g->mlen = 0;
|
g->nsub = 0;
|
g->nsub = 0;
|
g->ncategories = 1; /* category 0 is "everything else" */
|
g->ncategories = 1; /* category 0 is "everything else" */
|
g->categories = &g->catspace[-(CHAR_MIN)];
|
g->categories = &g->catspace[-(CHAR_MIN)];
|
(void) memset((char *)g->catspace, 0, NC*sizeof(cat_t));
|
(void) memset((char *)g->catspace, 0, NC*sizeof(cat_t));
|
g->backrefs = 0;
|
g->backrefs = 0;
|
|
|
/* do it */
|
/* do it */
|
EMIT(OEND, 0);
|
EMIT(OEND, 0);
|
g->firststate = THERE();
|
g->firststate = THERE();
|
if (cflags®_EXTENDED)
|
if (cflags®_EXTENDED)
|
p_ere(p, OUT);
|
p_ere(p, OUT);
|
else if (cflags®_NOSPEC)
|
else if (cflags®_NOSPEC)
|
p_str(p);
|
p_str(p);
|
else
|
else
|
p_bre(p, OUT, OUT);
|
p_bre(p, OUT, OUT);
|
EMIT(OEND, 0);
|
EMIT(OEND, 0);
|
g->laststate = THERE();
|
g->laststate = THERE();
|
|
|
/* tidy up loose ends and fill things in */
|
/* tidy up loose ends and fill things in */
|
categorize(p, g);
|
categorize(p, g);
|
stripsnug(p, g);
|
stripsnug(p, g);
|
findmust(p, g);
|
findmust(p, g);
|
/* only use Boyer-Moore algorithm if the pattern is bigger
|
/* only use Boyer-Moore algorithm if the pattern is bigger
|
* than three characters
|
* than three characters
|
*/
|
*/
|
if(g->mlen > 3) {
|
if(g->mlen > 3) {
|
computejumps(p, g);
|
computejumps(p, g);
|
computematchjumps(p, g);
|
computematchjumps(p, g);
|
if(g->matchjump == NULL && g->charjump != NULL) {
|
if(g->matchjump == NULL && g->charjump != NULL) {
|
free(g->charjump);
|
free(g->charjump);
|
g->charjump = NULL;
|
g->charjump = NULL;
|
}
|
}
|
}
|
}
|
g->nplus = pluscount(p, g);
|
g->nplus = pluscount(p, g);
|
g->magic = MAGIC2;
|
g->magic = MAGIC2;
|
preg->re_nsub = g->nsub;
|
preg->re_nsub = g->nsub;
|
preg->re_g = g;
|
preg->re_g = g;
|
preg->re_magic = MAGIC1;
|
preg->re_magic = MAGIC1;
|
#ifndef REDEBUG
|
#ifndef REDEBUG
|
/* not debugging, so can't rely on the assert() in regexec() */
|
/* not debugging, so can't rely on the assert() in regexec() */
|
if (g->iflags&BAD)
|
if (g->iflags&BAD)
|
SETERROR(REG_ASSERT);
|
SETERROR(REG_ASSERT);
|
#endif
|
#endif
|
|
|
/* win or lose, we're done */
|
/* win or lose, we're done */
|
if (p->error != 0) /* lose */
|
if (p->error != 0) /* lose */
|
regfree(preg);
|
regfree(preg);
|
return(p->error);
|
return(p->error);
|
}
|
}
|
|
|
/*
|
/*
|
- p_ere - ERE parser top level, concatenation and alternation
|
- p_ere - ERE parser top level, concatenation and alternation
|
== static void p_ere(struct parse *p, int stop);
|
== static void p_ere(struct parse *p, int stop);
|
*/
|
*/
|
static void
|
static void
|
p_ere(p, stop)
|
p_ere(p, stop)
|
struct parse *p;
|
struct parse *p;
|
int stop; /* character this ERE should end at */
|
int stop; /* character this ERE should end at */
|
{
|
{
|
char c;
|
char c;
|
sopno prevback;
|
sopno prevback;
|
sopno prevfwd;
|
sopno prevfwd;
|
sopno conc;
|
sopno conc;
|
int first = 1; /* is this the first alternative? */
|
int first = 1; /* is this the first alternative? */
|
|
|
for (;;) {
|
for (;;) {
|
/* do a bunch of concatenated expressions */
|
/* do a bunch of concatenated expressions */
|
conc = HERE();
|
conc = HERE();
|
while (MORE() && (c = PEEK()) != '|' && c != stop)
|
while (MORE() && (c = PEEK()) != '|' && c != stop)
|
p_ere_exp(p);
|
p_ere_exp(p);
|
(void)REQUIRE(HERE() != conc, REG_EMPTY); /* require nonempty */
|
(void)REQUIRE(HERE() != conc, REG_EMPTY); /* require nonempty */
|
|
|
if (!EAT('|'))
|
if (!EAT('|'))
|
break; /* NOTE BREAK OUT */
|
break; /* NOTE BREAK OUT */
|
|
|
if (first) {
|
if (first) {
|
INSERT(OCH_, conc); /* offset is wrong */
|
INSERT(OCH_, conc); /* offset is wrong */
|
prevfwd = conc;
|
prevfwd = conc;
|
prevback = conc;
|
prevback = conc;
|
first = 0;
|
first = 0;
|
}
|
}
|
ASTERN(OOR1, prevback);
|
ASTERN(OOR1, prevback);
|
prevback = THERE();
|
prevback = THERE();
|
AHEAD(prevfwd); /* fix previous offset */
|
AHEAD(prevfwd); /* fix previous offset */
|
prevfwd = HERE();
|
prevfwd = HERE();
|
EMIT(OOR2, 0); /* offset is very wrong */
|
EMIT(OOR2, 0); /* offset is very wrong */
|
}
|
}
|
|
|
if (!first) { /* tail-end fixups */
|
if (!first) { /* tail-end fixups */
|
AHEAD(prevfwd);
|
AHEAD(prevfwd);
|
ASTERN(O_CH, prevback);
|
ASTERN(O_CH, prevback);
|
}
|
}
|
|
|
assert(!MORE() || SEE(stop));
|
assert(!MORE() || SEE(stop));
|
}
|
}
|
|
|
/*
|
/*
|
- p_ere_exp - parse one subERE, an atom possibly followed by a repetition op
|
- p_ere_exp - parse one subERE, an atom possibly followed by a repetition op
|
== static void p_ere_exp(struct parse *p);
|
== static void p_ere_exp(struct parse *p);
|
*/
|
*/
|
static void
|
static void
|
p_ere_exp(p)
|
p_ere_exp(p)
|
struct parse *p;
|
struct parse *p;
|
{
|
{
|
char c;
|
char c;
|
sopno pos;
|
sopno pos;
|
int count;
|
int count;
|
int count2;
|
int count2;
|
sopno subno;
|
sopno subno;
|
int wascaret = 0;
|
int wascaret = 0;
|
|
|
assert(MORE()); /* caller should have ensured this */
|
assert(MORE()); /* caller should have ensured this */
|
c = GETNEXT();
|
c = GETNEXT();
|
|
|
pos = HERE();
|
pos = HERE();
|
switch (c) {
|
switch (c) {
|
case '(':
|
case '(':
|
(void)REQUIRE(MORE(), REG_EPAREN);
|
(void)REQUIRE(MORE(), REG_EPAREN);
|
p->g->nsub++;
|
p->g->nsub++;
|
subno = p->g->nsub;
|
subno = p->g->nsub;
|
if (subno < NPAREN)
|
if (subno < NPAREN)
|
p->pbegin[subno] = HERE();
|
p->pbegin[subno] = HERE();
|
EMIT(OLPAREN, subno);
|
EMIT(OLPAREN, subno);
|
if (!SEE(')'))
|
if (!SEE(')'))
|
p_ere(p, ')');
|
p_ere(p, ')');
|
if (subno < NPAREN) {
|
if (subno < NPAREN) {
|
p->pend[subno] = HERE();
|
p->pend[subno] = HERE();
|
assert(p->pend[subno] != 0);
|
assert(p->pend[subno] != 0);
|
}
|
}
|
EMIT(ORPAREN, subno);
|
EMIT(ORPAREN, subno);
|
(void)MUSTEAT(')', REG_EPAREN);
|
(void)MUSTEAT(')', REG_EPAREN);
|
break;
|
break;
|
#ifndef POSIX_MISTAKE
|
#ifndef POSIX_MISTAKE
|
case ')': /* happens only if no current unmatched ( */
|
case ')': /* happens only if no current unmatched ( */
|
/*
|
/*
|
* You may ask, why the ifndef? Because I didn't notice
|
* You may ask, why the ifndef? Because I didn't notice
|
* this until slightly too late for 1003.2, and none of the
|
* this until slightly too late for 1003.2, and none of the
|
* other 1003.2 regular-expression reviewers noticed it at
|
* other 1003.2 regular-expression reviewers noticed it at
|
* all. So an unmatched ) is legal POSIX, at least until
|
* all. So an unmatched ) is legal POSIX, at least until
|
* we can get it fixed.
|
* we can get it fixed.
|
*/
|
*/
|
SETERROR(REG_EPAREN);
|
SETERROR(REG_EPAREN);
|
break;
|
break;
|
#endif
|
#endif
|
case '^':
|
case '^':
|
EMIT(OBOL, 0);
|
EMIT(OBOL, 0);
|
p->g->iflags |= USEBOL;
|
p->g->iflags |= USEBOL;
|
p->g->nbol++;
|
p->g->nbol++;
|
wascaret = 1;
|
wascaret = 1;
|
break;
|
break;
|
case '$':
|
case '$':
|
EMIT(OEOL, 0);
|
EMIT(OEOL, 0);
|
p->g->iflags |= USEEOL;
|
p->g->iflags |= USEEOL;
|
p->g->neol++;
|
p->g->neol++;
|
break;
|
break;
|
case '|':
|
case '|':
|
SETERROR(REG_EMPTY);
|
SETERROR(REG_EMPTY);
|
break;
|
break;
|
case '*':
|
case '*':
|
case '+':
|
case '+':
|
case '?':
|
case '?':
|
SETERROR(REG_BADRPT);
|
SETERROR(REG_BADRPT);
|
break;
|
break;
|
case '.':
|
case '.':
|
if (p->g->cflags®_NEWLINE)
|
if (p->g->cflags®_NEWLINE)
|
nonnewline(p);
|
nonnewline(p);
|
else
|
else
|
EMIT(OANY, 0);
|
EMIT(OANY, 0);
|
break;
|
break;
|
case '[':
|
case '[':
|
p_bracket(p);
|
p_bracket(p);
|
break;
|
break;
|
case '\\':
|
case '\\':
|
(void)REQUIRE(MORE(), REG_EESCAPE);
|
(void)REQUIRE(MORE(), REG_EESCAPE);
|
c = GETNEXT();
|
c = GETNEXT();
|
ordinary(p, c);
|
ordinary(p, c);
|
break;
|
break;
|
case '{': /* okay as ordinary except if digit follows */
|
case '{': /* okay as ordinary except if digit follows */
|
(void)REQUIRE(!MORE() || !isdigit((uch)PEEK()), REG_BADRPT);
|
(void)REQUIRE(!MORE() || !isdigit((uch)PEEK()), REG_BADRPT);
|
/* FALLTHROUGH */
|
/* FALLTHROUGH */
|
default:
|
default:
|
ordinary(p, c);
|
ordinary(p, c);
|
break;
|
break;
|
}
|
}
|
|
|
if (!MORE())
|
if (!MORE())
|
return;
|
return;
|
c = PEEK();
|
c = PEEK();
|
/* we call { a repetition if followed by a digit */
|
/* we call { a repetition if followed by a digit */
|
if (!( c == '*' || c == '+' || c == '?' ||
|
if (!( c == '*' || c == '+' || c == '?' ||
|
(c == '{' && MORE2() && isdigit((uch)PEEK2())) ))
|
(c == '{' && MORE2() && isdigit((uch)PEEK2())) ))
|
return; /* no repetition, we're done */
|
return; /* no repetition, we're done */
|
NEXT();
|
NEXT();
|
|
|
(void)REQUIRE(!wascaret, REG_BADRPT);
|
(void)REQUIRE(!wascaret, REG_BADRPT);
|
switch (c) {
|
switch (c) {
|
case '*': /* implemented as +? */
|
case '*': /* implemented as +? */
|
/* this case does not require the (y|) trick, noKLUDGE */
|
/* this case does not require the (y|) trick, noKLUDGE */
|
INSERT(OPLUS_, pos);
|
INSERT(OPLUS_, pos);
|
ASTERN(O_PLUS, pos);
|
ASTERN(O_PLUS, pos);
|
INSERT(OQUEST_, pos);
|
INSERT(OQUEST_, pos);
|
ASTERN(O_QUEST, pos);
|
ASTERN(O_QUEST, pos);
|
break;
|
break;
|
case '+':
|
case '+':
|
INSERT(OPLUS_, pos);
|
INSERT(OPLUS_, pos);
|
ASTERN(O_PLUS, pos);
|
ASTERN(O_PLUS, pos);
|
break;
|
break;
|
case '?':
|
case '?':
|
/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
|
/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
|
INSERT(OCH_, pos); /* offset slightly wrong */
|
INSERT(OCH_, pos); /* offset slightly wrong */
|
ASTERN(OOR1, pos); /* this one's right */
|
ASTERN(OOR1, pos); /* this one's right */
|
AHEAD(pos); /* fix the OCH_ */
|
AHEAD(pos); /* fix the OCH_ */
|
EMIT(OOR2, 0); /* offset very wrong... */
|
EMIT(OOR2, 0); /* offset very wrong... */
|
AHEAD(THERE()); /* ...so fix it */
|
AHEAD(THERE()); /* ...so fix it */
|
ASTERN(O_CH, THERETHERE());
|
ASTERN(O_CH, THERETHERE());
|
break;
|
break;
|
case '{':
|
case '{':
|
count = p_count(p);
|
count = p_count(p);
|
if (EAT(',')) {
|
if (EAT(',')) {
|
if (isdigit((uch)PEEK())) {
|
if (isdigit((uch)PEEK())) {
|
count2 = p_count(p);
|
count2 = p_count(p);
|
(void)REQUIRE(count <= count2, REG_BADBR);
|
(void)REQUIRE(count <= count2, REG_BADBR);
|
} else /* single number with comma */
|
} else /* single number with comma */
|
count2 = INFINITY;
|
count2 = INFINITY;
|
} else /* just a single number */
|
} else /* just a single number */
|
count2 = count;
|
count2 = count;
|
repeat(p, pos, count, count2);
|
repeat(p, pos, count, count2);
|
if (!EAT('}')) { /* error heuristics */
|
if (!EAT('}')) { /* error heuristics */
|
while (MORE() && PEEK() != '}')
|
while (MORE() && PEEK() != '}')
|
NEXT();
|
NEXT();
|
(void)REQUIRE(MORE(), REG_EBRACE);
|
(void)REQUIRE(MORE(), REG_EBRACE);
|
SETERROR(REG_BADBR);
|
SETERROR(REG_BADBR);
|
}
|
}
|
break;
|
break;
|
}
|
}
|
|
|
if (!MORE())
|
if (!MORE())
|
return;
|
return;
|
c = PEEK();
|
c = PEEK();
|
if (!( c == '*' || c == '+' || c == '?' ||
|
if (!( c == '*' || c == '+' || c == '?' ||
|
(c == '{' && MORE2() && isdigit((uch)PEEK2())) ) )
|
(c == '{' && MORE2() && isdigit((uch)PEEK2())) ) )
|
return;
|
return;
|
SETERROR(REG_BADRPT);
|
SETERROR(REG_BADRPT);
|
}
|
}
|
|
|
/*
|
/*
|
- p_str - string (no metacharacters) "parser"
|
- p_str - string (no metacharacters) "parser"
|
== static void p_str(struct parse *p);
|
== static void p_str(struct parse *p);
|
*/
|
*/
|
static void
|
static void
|
p_str(p)
|
p_str(p)
|
struct parse *p;
|
struct parse *p;
|
{
|
{
|
(void)REQUIRE(MORE(), REG_EMPTY);
|
(void)REQUIRE(MORE(), REG_EMPTY);
|
while (MORE())
|
while (MORE())
|
ordinary(p, GETNEXT());
|
ordinary(p, GETNEXT());
|
}
|
}
|
|
|
/*
|
/*
|
- p_bre - BRE parser top level, anchoring and concatenation
|
- p_bre - BRE parser top level, anchoring and concatenation
|
== static void p_bre(struct parse *p, int end1, \
|
== static void p_bre(struct parse *p, int end1, \
|
== int end2);
|
== int end2);
|
* Giving end1 as OUT essentially eliminates the end1/end2 check.
|
* Giving end1 as OUT essentially eliminates the end1/end2 check.
|
*
|
*
|
* This implementation is a bit of a kludge, in that a trailing $ is first
|
* This implementation is a bit of a kludge, in that a trailing $ is first
|
* taken as an ordinary character and then revised to be an anchor. The
|
* taken as an ordinary character and then revised to be an anchor. The
|
* only undesirable side effect is that '$' gets included as a character
|
* only undesirable side effect is that '$' gets included as a character
|
* category in such cases. This is fairly harmless; not worth fixing.
|
* category in such cases. This is fairly harmless; not worth fixing.
|
* The amount of lookahead needed to avoid this kludge is excessive.
|
* The amount of lookahead needed to avoid this kludge is excessive.
|
*/
|
*/
|
static void
|
static void
|
p_bre(p, end1, end2)
|
p_bre(p, end1, end2)
|
struct parse *p;
|
struct parse *p;
|
int end1; /* first terminating character */
|
int end1; /* first terminating character */
|
int end2; /* second terminating character */
|
int end2; /* second terminating character */
|
{
|
{
|
sopno start = HERE();
|
sopno start = HERE();
|
int first = 1; /* first subexpression? */
|
int first = 1; /* first subexpression? */
|
int wasdollar = 0;
|
int wasdollar = 0;
|
|
|
if (EAT('^')) {
|
if (EAT('^')) {
|
EMIT(OBOL, 0);
|
EMIT(OBOL, 0);
|
p->g->iflags |= USEBOL;
|
p->g->iflags |= USEBOL;
|
p->g->nbol++;
|
p->g->nbol++;
|
}
|
}
|
while (MORE() && !SEETWO(end1, end2)) {
|
while (MORE() && !SEETWO(end1, end2)) {
|
wasdollar = p_simp_re(p, first);
|
wasdollar = p_simp_re(p, first);
|
first = 0;
|
first = 0;
|
}
|
}
|
if (wasdollar) { /* oops, that was a trailing anchor */
|
if (wasdollar) { /* oops, that was a trailing anchor */
|
DROP(1);
|
DROP(1);
|
EMIT(OEOL, 0);
|
EMIT(OEOL, 0);
|
p->g->iflags |= USEEOL;
|
p->g->iflags |= USEEOL;
|
p->g->neol++;
|
p->g->neol++;
|
}
|
}
|
|
|
(void)REQUIRE(HERE() != start, REG_EMPTY); /* require nonempty */
|
(void)REQUIRE(HERE() != start, REG_EMPTY); /* require nonempty */
|
}
|
}
|
|
|
/*
|
/*
|
- p_simp_re - parse a simple RE, an atom possibly followed by a repetition
|
- p_simp_re - parse a simple RE, an atom possibly followed by a repetition
|
== static int p_simp_re(struct parse *p, int starordinary);
|
== static int p_simp_re(struct parse *p, int starordinary);
|
*/
|
*/
|
static int /* was the simple RE an unbackslashed $? */
|
static int /* was the simple RE an unbackslashed $? */
|
p_simp_re(p, starordinary)
|
p_simp_re(p, starordinary)
|
struct parse *p;
|
struct parse *p;
|
int starordinary; /* is a leading * an ordinary character? */
|
int starordinary; /* is a leading * an ordinary character? */
|
{
|
{
|
int c;
|
int c;
|
int count;
|
int count;
|
int count2;
|
int count2;
|
sopno pos;
|
sopno pos;
|
int i;
|
int i;
|
sopno subno;
|
sopno subno;
|
# define BACKSL (1<<CHAR_BIT)
|
# define BACKSL (1<<CHAR_BIT)
|
|
|
pos = HERE(); /* repetion op, if any, covers from here */
|
pos = HERE(); /* repetion op, if any, covers from here */
|
|
|
assert(MORE()); /* caller should have ensured this */
|
assert(MORE()); /* caller should have ensured this */
|
c = GETNEXT();
|
c = GETNEXT();
|
if (c == '\\') {
|
if (c == '\\') {
|
(void)REQUIRE(MORE(), REG_EESCAPE);
|
(void)REQUIRE(MORE(), REG_EESCAPE);
|
c = BACKSL | GETNEXT();
|
c = BACKSL | GETNEXT();
|
}
|
}
|
switch (c) {
|
switch (c) {
|
case '.':
|
case '.':
|
if (p->g->cflags®_NEWLINE)
|
if (p->g->cflags®_NEWLINE)
|
nonnewline(p);
|
nonnewline(p);
|
else
|
else
|
EMIT(OANY, 0);
|
EMIT(OANY, 0);
|
break;
|
break;
|
case '[':
|
case '[':
|
p_bracket(p);
|
p_bracket(p);
|
break;
|
break;
|
case BACKSL|'{':
|
case BACKSL|'{':
|
SETERROR(REG_BADRPT);
|
SETERROR(REG_BADRPT);
|
break;
|
break;
|
case BACKSL|'(':
|
case BACKSL|'(':
|
p->g->nsub++;
|
p->g->nsub++;
|
subno = p->g->nsub;
|
subno = p->g->nsub;
|
if (subno < NPAREN)
|
if (subno < NPAREN)
|
p->pbegin[subno] = HERE();
|
p->pbegin[subno] = HERE();
|
EMIT(OLPAREN, subno);
|
EMIT(OLPAREN, subno);
|
/* the MORE here is an error heuristic */
|
/* the MORE here is an error heuristic */
|
if (MORE() && !SEETWO('\\', ')'))
|
if (MORE() && !SEETWO('\\', ')'))
|
p_bre(p, '\\', ')');
|
p_bre(p, '\\', ')');
|
if (subno < NPAREN) {
|
if (subno < NPAREN) {
|
p->pend[subno] = HERE();
|
p->pend[subno] = HERE();
|
assert(p->pend[subno] != 0);
|
assert(p->pend[subno] != 0);
|
}
|
}
|
EMIT(ORPAREN, subno);
|
EMIT(ORPAREN, subno);
|
(void)REQUIRE(EATTWO('\\', ')'), REG_EPAREN);
|
(void)REQUIRE(EATTWO('\\', ')'), REG_EPAREN);
|
break;
|
break;
|
case BACKSL|')': /* should not get here -- must be user */
|
case BACKSL|')': /* should not get here -- must be user */
|
case BACKSL|'}':
|
case BACKSL|'}':
|
SETERROR(REG_EPAREN);
|
SETERROR(REG_EPAREN);
|
break;
|
break;
|
case BACKSL|'1':
|
case BACKSL|'1':
|
case BACKSL|'2':
|
case BACKSL|'2':
|
case BACKSL|'3':
|
case BACKSL|'3':
|
case BACKSL|'4':
|
case BACKSL|'4':
|
case BACKSL|'5':
|
case BACKSL|'5':
|
case BACKSL|'6':
|
case BACKSL|'6':
|
case BACKSL|'7':
|
case BACKSL|'7':
|
case BACKSL|'8':
|
case BACKSL|'8':
|
case BACKSL|'9':
|
case BACKSL|'9':
|
i = (c&~BACKSL) - '0';
|
i = (c&~BACKSL) - '0';
|
assert(i < NPAREN);
|
assert(i < NPAREN);
|
if (p->pend[i] != 0) {
|
if (p->pend[i] != 0) {
|
assert(i <= p->g->nsub);
|
assert(i <= p->g->nsub);
|
EMIT(OBACK_, i);
|
EMIT(OBACK_, i);
|
assert(p->pbegin[i] != 0);
|
assert(p->pbegin[i] != 0);
|
assert(OP(p->strip[p->pbegin[i]]) == OLPAREN);
|
assert(OP(p->strip[p->pbegin[i]]) == OLPAREN);
|
assert(OP(p->strip[p->pend[i]]) == ORPAREN);
|
assert(OP(p->strip[p->pend[i]]) == ORPAREN);
|
(void) dupl(p, p->pbegin[i]+1, p->pend[i]);
|
(void) dupl(p, p->pbegin[i]+1, p->pend[i]);
|
EMIT(O_BACK, i);
|
EMIT(O_BACK, i);
|
} else
|
} else
|
SETERROR(REG_ESUBREG);
|
SETERROR(REG_ESUBREG);
|
p->g->backrefs = 1;
|
p->g->backrefs = 1;
|
break;
|
break;
|
case '*':
|
case '*':
|
(void)REQUIRE(starordinary, REG_BADRPT);
|
(void)REQUIRE(starordinary, REG_BADRPT);
|
/* FALLTHROUGH */
|
/* FALLTHROUGH */
|
default:
|
default:
|
ordinary(p, (char)c);
|
ordinary(p, (char)c);
|
break;
|
break;
|
}
|
}
|
|
|
if (EAT('*')) { /* implemented as +? */
|
if (EAT('*')) { /* implemented as +? */
|
/* this case does not require the (y|) trick, noKLUDGE */
|
/* this case does not require the (y|) trick, noKLUDGE */
|
INSERT(OPLUS_, pos);
|
INSERT(OPLUS_, pos);
|
ASTERN(O_PLUS, pos);
|
ASTERN(O_PLUS, pos);
|
INSERT(OQUEST_, pos);
|
INSERT(OQUEST_, pos);
|
ASTERN(O_QUEST, pos);
|
ASTERN(O_QUEST, pos);
|
} else if (EATTWO('\\', '{')) {
|
} else if (EATTWO('\\', '{')) {
|
count = p_count(p);
|
count = p_count(p);
|
if (EAT(',')) {
|
if (EAT(',')) {
|
if (MORE() && isdigit((uch)PEEK())) {
|
if (MORE() && isdigit((uch)PEEK())) {
|
count2 = p_count(p);
|
count2 = p_count(p);
|
(void)REQUIRE(count <= count2, REG_BADBR);
|
(void)REQUIRE(count <= count2, REG_BADBR);
|
} else /* single number with comma */
|
} else /* single number with comma */
|
count2 = INFINITY;
|
count2 = INFINITY;
|
} else /* just a single number */
|
} else /* just a single number */
|
count2 = count;
|
count2 = count;
|
repeat(p, pos, count, count2);
|
repeat(p, pos, count, count2);
|
if (!EATTWO('\\', '}')) { /* error heuristics */
|
if (!EATTWO('\\', '}')) { /* error heuristics */
|
while (MORE() && !SEETWO('\\', '}'))
|
while (MORE() && !SEETWO('\\', '}'))
|
NEXT();
|
NEXT();
|
(void)REQUIRE(MORE(), REG_EBRACE);
|
(void)REQUIRE(MORE(), REG_EBRACE);
|
SETERROR(REG_BADBR);
|
SETERROR(REG_BADBR);
|
}
|
}
|
} else if (c == '$') /* $ (but not \$) ends it */
|
} else if (c == '$') /* $ (but not \$) ends it */
|
return(1);
|
return(1);
|
|
|
return(0);
|
return(0);
|
}
|
}
|
|
|
/*
|
/*
|
- p_count - parse a repetition count
|
- p_count - parse a repetition count
|
== static int p_count(struct parse *p);
|
== static int p_count(struct parse *p);
|
*/
|
*/
|
static int /* the value */
|
static int /* the value */
|
p_count(p)
|
p_count(p)
|
struct parse *p;
|
struct parse *p;
|
{
|
{
|
int count = 0;
|
int count = 0;
|
int ndigits = 0;
|
int ndigits = 0;
|
|
|
while (MORE() && isdigit((uch)PEEK()) && count <= DUPMAX) {
|
while (MORE() && isdigit((uch)PEEK()) && count <= DUPMAX) {
|
count = count*10 + (GETNEXT() - '0');
|
count = count*10 + (GETNEXT() - '0');
|
ndigits++;
|
ndigits++;
|
}
|
}
|
|
|
(void)REQUIRE(ndigits > 0 && count <= DUPMAX, REG_BADBR);
|
(void)REQUIRE(ndigits > 0 && count <= DUPMAX, REG_BADBR);
|
return(count);
|
return(count);
|
}
|
}
|
|
|
/*
|
/*
|
- p_bracket - parse a bracketed character list
|
- p_bracket - parse a bracketed character list
|
== static void p_bracket(struct parse *p);
|
== static void p_bracket(struct parse *p);
|
*
|
*
|
* Note a significant property of this code: if the allocset() did SETERROR,
|
* Note a significant property of this code: if the allocset() did SETERROR,
|
* no set operations are done.
|
* no set operations are done.
|
*/
|
*/
|
static void
|
static void
|
p_bracket(p)
|
p_bracket(p)
|
struct parse *p;
|
struct parse *p;
|
{
|
{
|
cset *cs = allocset(p);
|
cset *cs = allocset(p);
|
int invert = 0;
|
int invert = 0;
|
|
|
/* Dept of Truly Sickening Special-Case Kludges */
|
/* Dept of Truly Sickening Special-Case Kludges */
|
if (p->next + 5 < p->end && strncmp(p->next, "[:<:]]", 6) == 0) {
|
if (p->next + 5 < p->end && strncmp(p->next, "[:<:]]", 6) == 0) {
|
EMIT(OBOW, 0);
|
EMIT(OBOW, 0);
|
NEXTn(6);
|
NEXTn(6);
|
return;
|
return;
|
}
|
}
|
if (p->next + 5 < p->end && strncmp(p->next, "[:>:]]", 6) == 0) {
|
if (p->next + 5 < p->end && strncmp(p->next, "[:>:]]", 6) == 0) {
|
EMIT(OEOW, 0);
|
EMIT(OEOW, 0);
|
NEXTn(6);
|
NEXTn(6);
|
return;
|
return;
|
}
|
}
|
|
|
if (EAT('^'))
|
if (EAT('^'))
|
invert++; /* make note to invert set at end */
|
invert++; /* make note to invert set at end */
|
if (EAT(']'))
|
if (EAT(']'))
|
CHadd(cs, ']');
|
CHadd(cs, ']');
|
else if (EAT('-'))
|
else if (EAT('-'))
|
CHadd(cs, '-');
|
CHadd(cs, '-');
|
while (MORE() && PEEK() != ']' && !SEETWO('-', ']'))
|
while (MORE() && PEEK() != ']' && !SEETWO('-', ']'))
|
p_b_term(p, cs);
|
p_b_term(p, cs);
|
if (EAT('-'))
|
if (EAT('-'))
|
CHadd(cs, '-');
|
CHadd(cs, '-');
|
(void)MUSTEAT(']', REG_EBRACK);
|
(void)MUSTEAT(']', REG_EBRACK);
|
|
|
if (p->error != 0) /* don't mess things up further */
|
if (p->error != 0) /* don't mess things up further */
|
return;
|
return;
|
|
|
if (p->g->cflags®_ICASE) {
|
if (p->g->cflags®_ICASE) {
|
int i;
|
int i;
|
int ci;
|
int ci;
|
|
|
for (i = p->g->csetsize - 1; i >= 0; i--)
|
for (i = p->g->csetsize - 1; i >= 0; i--)
|
if (CHIN(cs, i) && isalpha(i)) {
|
if (CHIN(cs, i) && isalpha(i)) {
|
ci = othercase(i);
|
ci = othercase(i);
|
if (ci != i)
|
if (ci != i)
|
CHadd(cs, ci);
|
CHadd(cs, ci);
|
}
|
}
|
if (cs->multis != NULL)
|
if (cs->multis != NULL)
|
mccase(p, cs);
|
mccase(p, cs);
|
}
|
}
|
if (invert) {
|
if (invert) {
|
int i;
|
int i;
|
|
|
for (i = p->g->csetsize - 1; i >= 0; i--)
|
for (i = p->g->csetsize - 1; i >= 0; i--)
|
if (CHIN(cs, i))
|
if (CHIN(cs, i))
|
CHsub(cs, i);
|
CHsub(cs, i);
|
else
|
else
|
CHadd(cs, i);
|
CHadd(cs, i);
|
if (p->g->cflags®_NEWLINE)
|
if (p->g->cflags®_NEWLINE)
|
CHsub(cs, '\n');
|
CHsub(cs, '\n');
|
if (cs->multis != NULL)
|
if (cs->multis != NULL)
|
mcinvert(p, cs);
|
mcinvert(p, cs);
|
}
|
}
|
|
|
assert(cs->multis == NULL); /* xxx */
|
assert(cs->multis == NULL); /* xxx */
|
|
|
if (nch(p, cs) == 1) { /* optimize singleton sets */
|
if (nch(p, cs) == 1) { /* optimize singleton sets */
|
ordinary(p, firstch(p, cs));
|
ordinary(p, firstch(p, cs));
|
freeset(p, cs);
|
freeset(p, cs);
|
} else
|
} else
|
EMIT(OANYOF, freezeset(p, cs));
|
EMIT(OANYOF, freezeset(p, cs));
|
}
|
}
|
|
|
/*
|
/*
|
- p_b_term - parse one term of a bracketed character list
|
- p_b_term - parse one term of a bracketed character list
|
== static void p_b_term(struct parse *p, cset *cs);
|
== static void p_b_term(struct parse *p, cset *cs);
|
*/
|
*/
|
static void
|
static void
|
p_b_term(p, cs)
|
p_b_term(p, cs)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
{
|
{
|
char c;
|
char c;
|
char start, finish;
|
char start, finish;
|
int i;
|
int i;
|
|
|
/* classify what we've got */
|
/* classify what we've got */
|
switch ((MORE()) ? PEEK() : '\0') {
|
switch ((MORE()) ? PEEK() : '\0') {
|
case '[':
|
case '[':
|
c = (MORE2()) ? PEEK2() : '\0';
|
c = (MORE2()) ? PEEK2() : '\0';
|
break;
|
break;
|
case '-':
|
case '-':
|
SETERROR(REG_ERANGE);
|
SETERROR(REG_ERANGE);
|
return; /* NOTE RETURN */
|
return; /* NOTE RETURN */
|
break;
|
break;
|
default:
|
default:
|
c = '\0';
|
c = '\0';
|
break;
|
break;
|
}
|
}
|
|
|
switch (c) {
|
switch (c) {
|
case ':': /* character class */
|
case ':': /* character class */
|
NEXT2();
|
NEXT2();
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
c = PEEK();
|
c = PEEK();
|
(void)REQUIRE(c != '-' && c != ']', REG_ECTYPE);
|
(void)REQUIRE(c != '-' && c != ']', REG_ECTYPE);
|
p_b_cclass(p, cs);
|
p_b_cclass(p, cs);
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
(void)REQUIRE(EATTWO(':', ']'), REG_ECTYPE);
|
(void)REQUIRE(EATTWO(':', ']'), REG_ECTYPE);
|
break;
|
break;
|
case '=': /* equivalence class */
|
case '=': /* equivalence class */
|
NEXT2();
|
NEXT2();
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
c = PEEK();
|
c = PEEK();
|
(void)REQUIRE(c != '-' && c != ']', REG_ECOLLATE);
|
(void)REQUIRE(c != '-' && c != ']', REG_ECOLLATE);
|
p_b_eclass(p, cs);
|
p_b_eclass(p, cs);
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
(void)REQUIRE(EATTWO('=', ']'), REG_ECOLLATE);
|
(void)REQUIRE(EATTWO('=', ']'), REG_ECOLLATE);
|
break;
|
break;
|
default: /* symbol, ordinary character, or range */
|
default: /* symbol, ordinary character, or range */
|
/* xxx revision needed for multichar stuff */
|
/* xxx revision needed for multichar stuff */
|
start = p_b_symbol(p);
|
start = p_b_symbol(p);
|
if (SEE('-') && MORE2() && PEEK2() != ']') {
|
if (SEE('-') && MORE2() && PEEK2() != ']') {
|
/* range */
|
/* range */
|
NEXT();
|
NEXT();
|
if (EAT('-'))
|
if (EAT('-'))
|
finish = '-';
|
finish = '-';
|
else
|
else
|
finish = p_b_symbol(p);
|
finish = p_b_symbol(p);
|
} else
|
} else
|
finish = start;
|
finish = start;
|
if (start == finish)
|
if (start == finish)
|
CHadd(cs, start);
|
CHadd(cs, start);
|
else {
|
else {
|
if (__collate_load_error) {
|
if (__collate_load_error) {
|
(void)REQUIRE((uch)start <= (uch)finish, REG_ERANGE);
|
(void)REQUIRE((uch)start <= (uch)finish, REG_ERANGE);
|
for (i = (uch)start; i <= (uch)finish; i++)
|
for (i = (uch)start; i <= (uch)finish; i++)
|
CHadd(cs, i);
|
CHadd(cs, i);
|
} else {
|
} else {
|
(void)REQUIRE(__collate_range_cmp(start, finish) <= 0, REG_ERANGE);
|
(void)REQUIRE(__collate_range_cmp(start, finish) <= 0, REG_ERANGE);
|
for (i = CHAR_MIN; i <= CHAR_MAX; i++) {
|
for (i = CHAR_MIN; i <= CHAR_MAX; i++) {
|
if ( __collate_range_cmp(start, i) <= 0
|
if ( __collate_range_cmp(start, i) <= 0
|
&& __collate_range_cmp(i, finish) <= 0
|
&& __collate_range_cmp(i, finish) <= 0
|
)
|
)
|
CHadd(cs, i);
|
CHadd(cs, i);
|
}
|
}
|
}
|
}
|
}
|
}
|
break;
|
break;
|
}
|
}
|
}
|
}
|
|
|
/*
|
/*
|
- p_b_cclass - parse a character-class name and deal with it
|
- p_b_cclass - parse a character-class name and deal with it
|
== static void p_b_cclass(struct parse *p, cset *cs);
|
== static void p_b_cclass(struct parse *p, cset *cs);
|
*/
|
*/
|
static void
|
static void
|
p_b_cclass(p, cs)
|
p_b_cclass(p, cs)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
{
|
{
|
int c;
|
int c;
|
char *sp = p->next;
|
char *sp = p->next;
|
struct cclass *cp;
|
struct cclass *cp;
|
size_t len;
|
size_t len;
|
|
|
while (MORE() && isalpha((uch)PEEK()))
|
while (MORE() && isalpha((uch)PEEK()))
|
NEXT();
|
NEXT();
|
len = p->next - sp;
|
len = p->next - sp;
|
for (cp = cclasses; cp->name != NULL; cp++)
|
for (cp = cclasses; cp->name != NULL; cp++)
|
if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
|
if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
|
break;
|
break;
|
if (cp->name == NULL) {
|
if (cp->name == NULL) {
|
/* oops, didn't find it */
|
/* oops, didn't find it */
|
SETERROR(REG_ECTYPE);
|
SETERROR(REG_ECTYPE);
|
return;
|
return;
|
}
|
}
|
|
|
switch (cp->fidx) {
|
switch (cp->fidx) {
|
case CALNUM:
|
case CALNUM:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (isalnum((uch)c))
|
if (isalnum((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CALPHA:
|
case CALPHA:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (isalpha((uch)c))
|
if (isalpha((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CBLANK:
|
case CBLANK:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (isblank((uch)c))
|
if (isblank((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CCNTRL:
|
case CCNTRL:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (iscntrl((uch)c))
|
if (iscntrl((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CDIGIT:
|
case CDIGIT:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (isdigit((uch)c))
|
if (isdigit((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CGRAPH:
|
case CGRAPH:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (isgraph((uch)c))
|
if (isgraph((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CLOWER:
|
case CLOWER:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (islower((uch)c))
|
if (islower((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CPRINT:
|
case CPRINT:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (isprint((uch)c))
|
if (isprint((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CPUNCT:
|
case CPUNCT:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (ispunct((uch)c))
|
if (ispunct((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CSPACE:
|
case CSPACE:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (isspace((uch)c))
|
if (isspace((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CUPPER:
|
case CUPPER:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (isupper((uch)c))
|
if (isupper((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
case CXDIGIT:
|
case CXDIGIT:
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (isxdigit((uch)c))
|
if (isxdigit((uch)c))
|
CHadd(cs, c);
|
CHadd(cs, c);
|
break;
|
break;
|
}
|
}
|
#if 0
|
#if 0
|
for (u = cp->multis; *u != '\0'; u += strlen(u) + 1)
|
for (u = cp->multis; *u != '\0'; u += strlen(u) + 1)
|
MCadd(p, cs, u);
|
MCadd(p, cs, u);
|
#endif
|
#endif
|
}
|
}
|
|
|
/*
|
/*
|
- p_b_eclass - parse an equivalence-class name and deal with it
|
- p_b_eclass - parse an equivalence-class name and deal with it
|
== static void p_b_eclass(struct parse *p, cset *cs);
|
== static void p_b_eclass(struct parse *p, cset *cs);
|
*
|
*
|
* This implementation is incomplete. xxx
|
* This implementation is incomplete. xxx
|
*/
|
*/
|
static void
|
static void
|
p_b_eclass(p, cs)
|
p_b_eclass(p, cs)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
{
|
{
|
char c;
|
char c;
|
|
|
c = p_b_coll_elem(p, '=');
|
c = p_b_coll_elem(p, '=');
|
CHadd(cs, c);
|
CHadd(cs, c);
|
}
|
}
|
|
|
/*
|
/*
|
- p_b_symbol - parse a character or [..]ed multicharacter collating symbol
|
- p_b_symbol - parse a character or [..]ed multicharacter collating symbol
|
== static char p_b_symbol(struct parse *p);
|
== static char p_b_symbol(struct parse *p);
|
*/
|
*/
|
static char /* value of symbol */
|
static char /* value of symbol */
|
p_b_symbol(p)
|
p_b_symbol(p)
|
struct parse *p;
|
struct parse *p;
|
{
|
{
|
char value;
|
char value;
|
|
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
(void)REQUIRE(MORE(), REG_EBRACK);
|
if (!EATTWO('[', '.'))
|
if (!EATTWO('[', '.'))
|
return(GETNEXT());
|
return(GETNEXT());
|
|
|
/* collating symbol */
|
/* collating symbol */
|
value = p_b_coll_elem(p, '.');
|
value = p_b_coll_elem(p, '.');
|
(void)REQUIRE(EATTWO('.', ']'), REG_ECOLLATE);
|
(void)REQUIRE(EATTWO('.', ']'), REG_ECOLLATE);
|
return(value);
|
return(value);
|
}
|
}
|
|
|
/*
|
/*
|
- p_b_coll_elem - parse a collating-element name and look it up
|
- p_b_coll_elem - parse a collating-element name and look it up
|
== static char p_b_coll_elem(struct parse *p, int endc);
|
== static char p_b_coll_elem(struct parse *p, int endc);
|
*/
|
*/
|
static char /* value of collating element */
|
static char /* value of collating element */
|
p_b_coll_elem(p, endc)
|
p_b_coll_elem(p, endc)
|
struct parse *p;
|
struct parse *p;
|
int endc; /* name ended by endc,']' */
|
int endc; /* name ended by endc,']' */
|
{
|
{
|
char *sp = p->next;
|
char *sp = p->next;
|
struct cname *cp;
|
struct cname *cp;
|
int len;
|
int len;
|
|
|
while (MORE() && !SEETWO(endc, ']'))
|
while (MORE() && !SEETWO(endc, ']'))
|
NEXT();
|
NEXT();
|
if (!MORE()) {
|
if (!MORE()) {
|
SETERROR(REG_EBRACK);
|
SETERROR(REG_EBRACK);
|
return(0);
|
return(0);
|
}
|
}
|
len = p->next - sp;
|
len = p->next - sp;
|
for (cp = cnames; cp->name != NULL; cp++)
|
for (cp = cnames; cp->name != NULL; cp++)
|
if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
|
if (strncmp(cp->name, sp, len) == 0 && cp->name[len] == '\0')
|
return(cp->code); /* known name */
|
return(cp->code); /* known name */
|
if (len == 1)
|
if (len == 1)
|
return(*sp); /* single character */
|
return(*sp); /* single character */
|
SETERROR(REG_ECOLLATE); /* neither */
|
SETERROR(REG_ECOLLATE); /* neither */
|
return(0);
|
return(0);
|
}
|
}
|
|
|
/*
|
/*
|
- othercase - return the case counterpart of an alphabetic
|
- othercase - return the case counterpart of an alphabetic
|
== static char othercase(int ch);
|
== static char othercase(int ch);
|
*/
|
*/
|
static char /* if no counterpart, return ch */
|
static char /* if no counterpart, return ch */
|
othercase(ch)
|
othercase(ch)
|
int ch;
|
int ch;
|
{
|
{
|
ch = (uch)ch;
|
ch = (uch)ch;
|
assert(isalpha(ch));
|
assert(isalpha(ch));
|
if (isupper(ch))
|
if (isupper(ch))
|
return(tolower(ch));
|
return(tolower(ch));
|
else if (islower(ch))
|
else if (islower(ch))
|
return(toupper(ch));
|
return(toupper(ch));
|
else /* peculiar, but could happen */
|
else /* peculiar, but could happen */
|
return(ch);
|
return(ch);
|
}
|
}
|
|
|
/*
|
/*
|
- bothcases - emit a dualcase version of a two-case character
|
- bothcases - emit a dualcase version of a two-case character
|
== static void bothcases(struct parse *p, int ch);
|
== static void bothcases(struct parse *p, int ch);
|
*
|
*
|
* Boy, is this implementation ever a kludge...
|
* Boy, is this implementation ever a kludge...
|
*/
|
*/
|
static void
|
static void
|
bothcases(p, ch)
|
bothcases(p, ch)
|
struct parse *p;
|
struct parse *p;
|
int ch;
|
int ch;
|
{
|
{
|
char *oldnext = p->next;
|
char *oldnext = p->next;
|
char *oldend = p->end;
|
char *oldend = p->end;
|
char bracket[3];
|
char bracket[3];
|
|
|
ch = (uch)ch;
|
ch = (uch)ch;
|
assert(othercase(ch) != ch); /* p_bracket() would recurse */
|
assert(othercase(ch) != ch); /* p_bracket() would recurse */
|
p->next = bracket;
|
p->next = bracket;
|
p->end = bracket+2;
|
p->end = bracket+2;
|
bracket[0] = ch;
|
bracket[0] = ch;
|
bracket[1] = ']';
|
bracket[1] = ']';
|
bracket[2] = '\0';
|
bracket[2] = '\0';
|
p_bracket(p);
|
p_bracket(p);
|
assert(p->next == bracket+2);
|
assert(p->next == bracket+2);
|
p->next = oldnext;
|
p->next = oldnext;
|
p->end = oldend;
|
p->end = oldend;
|
}
|
}
|
|
|
/*
|
/*
|
- ordinary - emit an ordinary character
|
- ordinary - emit an ordinary character
|
== static void ordinary(struct parse *p, int ch);
|
== static void ordinary(struct parse *p, int ch);
|
*/
|
*/
|
static void
|
static void
|
ordinary(p, ch)
|
ordinary(p, ch)
|
struct parse *p;
|
struct parse *p;
|
int ch;
|
int ch;
|
{
|
{
|
cat_t *cap = p->g->categories;
|
cat_t *cap = p->g->categories;
|
|
|
if ((p->g->cflags®_ICASE) && isalpha((uch)ch) && othercase(ch) != ch)
|
if ((p->g->cflags®_ICASE) && isalpha((uch)ch) && othercase(ch) != ch)
|
bothcases(p, ch);
|
bothcases(p, ch);
|
else {
|
else {
|
EMIT(OCHAR, (uch)ch);
|
EMIT(OCHAR, (uch)ch);
|
if (cap[ch] == 0)
|
if (cap[ch] == 0)
|
cap[ch] = p->g->ncategories++;
|
cap[ch] = p->g->ncategories++;
|
}
|
}
|
}
|
}
|
|
|
/*
|
/*
|
- nonnewline - emit REG_NEWLINE version of OANY
|
- nonnewline - emit REG_NEWLINE version of OANY
|
== static void nonnewline(struct parse *p);
|
== static void nonnewline(struct parse *p);
|
*
|
*
|
* Boy, is this implementation ever a kludge...
|
* Boy, is this implementation ever a kludge...
|
*/
|
*/
|
static void
|
static void
|
nonnewline(p)
|
nonnewline(p)
|
struct parse *p;
|
struct parse *p;
|
{
|
{
|
char *oldnext = p->next;
|
char *oldnext = p->next;
|
char *oldend = p->end;
|
char *oldend = p->end;
|
char bracket[4];
|
char bracket[4];
|
|
|
p->next = bracket;
|
p->next = bracket;
|
p->end = bracket+3;
|
p->end = bracket+3;
|
bracket[0] = '^';
|
bracket[0] = '^';
|
bracket[1] = '\n';
|
bracket[1] = '\n';
|
bracket[2] = ']';
|
bracket[2] = ']';
|
bracket[3] = '\0';
|
bracket[3] = '\0';
|
p_bracket(p);
|
p_bracket(p);
|
assert(p->next == bracket+3);
|
assert(p->next == bracket+3);
|
p->next = oldnext;
|
p->next = oldnext;
|
p->end = oldend;
|
p->end = oldend;
|
}
|
}
|
|
|
/*
|
/*
|
- repeat - generate code for a bounded repetition, recursively if needed
|
- repeat - generate code for a bounded repetition, recursively if needed
|
== static void repeat(struct parse *p, sopno start, int from, int to);
|
== static void repeat(struct parse *p, sopno start, int from, int to);
|
*/
|
*/
|
static void
|
static void
|
repeat(p, start, from, to)
|
repeat(p, start, from, to)
|
struct parse *p;
|
struct parse *p;
|
sopno start; /* operand from here to end of strip */
|
sopno start; /* operand from here to end of strip */
|
int from; /* repeated from this number */
|
int from; /* repeated from this number */
|
int to; /* to this number of times (maybe INFINITY) */
|
int to; /* to this number of times (maybe INFINITY) */
|
{
|
{
|
sopno finish = HERE();
|
sopno finish = HERE();
|
# define N 2
|
# define N 2
|
# define INF 3
|
# define INF 3
|
# define REP(f, t) ((f)*8 + (t))
|
# define REP(f, t) ((f)*8 + (t))
|
# define MAP(n) (((n) <= 1) ? (n) : ((n) == INFINITY) ? INF : N)
|
# define MAP(n) (((n) <= 1) ? (n) : ((n) == INFINITY) ? INF : N)
|
sopno copy;
|
sopno copy;
|
|
|
if (p->error != 0) /* head off possible runaway recursion */
|
if (p->error != 0) /* head off possible runaway recursion */
|
return;
|
return;
|
|
|
assert(from <= to);
|
assert(from <= to);
|
|
|
switch (REP(MAP(from), MAP(to))) {
|
switch (REP(MAP(from), MAP(to))) {
|
case REP(0, 0): /* must be user doing this */
|
case REP(0, 0): /* must be user doing this */
|
DROP(finish-start); /* drop the operand */
|
DROP(finish-start); /* drop the operand */
|
break;
|
break;
|
case REP(0, 1): /* as x{1,1}? */
|
case REP(0, 1): /* as x{1,1}? */
|
case REP(0, N): /* as x{1,n}? */
|
case REP(0, N): /* as x{1,n}? */
|
case REP(0, INF): /* as x{1,}? */
|
case REP(0, INF): /* as x{1,}? */
|
/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
|
/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
|
INSERT(OCH_, start); /* offset is wrong... */
|
INSERT(OCH_, start); /* offset is wrong... */
|
repeat(p, start+1, 1, to);
|
repeat(p, start+1, 1, to);
|
ASTERN(OOR1, start);
|
ASTERN(OOR1, start);
|
AHEAD(start); /* ... fix it */
|
AHEAD(start); /* ... fix it */
|
EMIT(OOR2, 0);
|
EMIT(OOR2, 0);
|
AHEAD(THERE());
|
AHEAD(THERE());
|
ASTERN(O_CH, THERETHERE());
|
ASTERN(O_CH, THERETHERE());
|
break;
|
break;
|
case REP(1, 1): /* trivial case */
|
case REP(1, 1): /* trivial case */
|
/* done */
|
/* done */
|
break;
|
break;
|
case REP(1, N): /* as x?x{1,n-1} */
|
case REP(1, N): /* as x?x{1,n-1} */
|
/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
|
/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
|
INSERT(OCH_, start);
|
INSERT(OCH_, start);
|
ASTERN(OOR1, start);
|
ASTERN(OOR1, start);
|
AHEAD(start);
|
AHEAD(start);
|
EMIT(OOR2, 0); /* offset very wrong... */
|
EMIT(OOR2, 0); /* offset very wrong... */
|
AHEAD(THERE()); /* ...so fix it */
|
AHEAD(THERE()); /* ...so fix it */
|
ASTERN(O_CH, THERETHERE());
|
ASTERN(O_CH, THERETHERE());
|
copy = dupl(p, start+1, finish+1);
|
copy = dupl(p, start+1, finish+1);
|
assert(copy == finish+4);
|
assert(copy == finish+4);
|
repeat(p, copy, 1, to-1);
|
repeat(p, copy, 1, to-1);
|
break;
|
break;
|
case REP(1, INF): /* as x+ */
|
case REP(1, INF): /* as x+ */
|
INSERT(OPLUS_, start);
|
INSERT(OPLUS_, start);
|
ASTERN(O_PLUS, start);
|
ASTERN(O_PLUS, start);
|
break;
|
break;
|
case REP(N, N): /* as xx{m-1,n-1} */
|
case REP(N, N): /* as xx{m-1,n-1} */
|
copy = dupl(p, start, finish);
|
copy = dupl(p, start, finish);
|
repeat(p, copy, from-1, to-1);
|
repeat(p, copy, from-1, to-1);
|
break;
|
break;
|
case REP(N, INF): /* as xx{n-1,INF} */
|
case REP(N, INF): /* as xx{n-1,INF} */
|
copy = dupl(p, start, finish);
|
copy = dupl(p, start, finish);
|
repeat(p, copy, from-1, to);
|
repeat(p, copy, from-1, to);
|
break;
|
break;
|
default: /* "can't happen" */
|
default: /* "can't happen" */
|
SETERROR(REG_ASSERT); /* just in case */
|
SETERROR(REG_ASSERT); /* just in case */
|
break;
|
break;
|
}
|
}
|
}
|
}
|
|
|
/*
|
/*
|
- seterr - set an error condition
|
- seterr - set an error condition
|
== static int seterr(struct parse *p, int e);
|
== static int seterr(struct parse *p, int e);
|
*/
|
*/
|
static int /* useless but makes type checking happy */
|
static int /* useless but makes type checking happy */
|
seterr(p, e)
|
seterr(p, e)
|
struct parse *p;
|
struct parse *p;
|
int e;
|
int e;
|
{
|
{
|
if (p->error == 0) /* keep earliest error condition */
|
if (p->error == 0) /* keep earliest error condition */
|
p->error = e;
|
p->error = e;
|
p->next = nuls; /* try to bring things to a halt */
|
p->next = nuls; /* try to bring things to a halt */
|
p->end = nuls;
|
p->end = nuls;
|
return(0); /* make the return value well-defined */
|
return(0); /* make the return value well-defined */
|
}
|
}
|
|
|
/*
|
/*
|
- allocset - allocate a set of characters for []
|
- allocset - allocate a set of characters for []
|
== static cset *allocset(struct parse *p);
|
== static cset *allocset(struct parse *p);
|
*/
|
*/
|
static cset *
|
static cset *
|
allocset(p)
|
allocset(p)
|
struct parse *p;
|
struct parse *p;
|
{
|
{
|
int no = p->g->ncsets++;
|
int no = p->g->ncsets++;
|
size_t nc;
|
size_t nc;
|
size_t nbytes;
|
size_t nbytes;
|
cset *cs;
|
cset *cs;
|
size_t css = (size_t)p->g->csetsize;
|
size_t css = (size_t)p->g->csetsize;
|
int i;
|
int i;
|
|
|
if (no >= p->ncsalloc) { /* need another column of space */
|
if (no >= p->ncsalloc) { /* need another column of space */
|
p->ncsalloc += CHAR_BIT;
|
p->ncsalloc += CHAR_BIT;
|
nc = p->ncsalloc;
|
nc = p->ncsalloc;
|
assert(nc % CHAR_BIT == 0);
|
assert(nc % CHAR_BIT == 0);
|
nbytes = nc / CHAR_BIT * css;
|
nbytes = nc / CHAR_BIT * css;
|
if (p->g->sets == NULL)
|
if (p->g->sets == NULL)
|
p->g->sets = (cset *)malloc(nc * sizeof(cset));
|
p->g->sets = (cset *)malloc(nc * sizeof(cset));
|
else
|
else
|
p->g->sets = (cset *)reallocf((char *)p->g->sets,
|
p->g->sets = (cset *)reallocf((char *)p->g->sets,
|
nc * sizeof(cset));
|
nc * sizeof(cset));
|
if (p->g->setbits == NULL)
|
if (p->g->setbits == NULL)
|
p->g->setbits = (uch *)malloc(nbytes);
|
p->g->setbits = (uch *)malloc(nbytes);
|
else {
|
else {
|
p->g->setbits = (uch *)reallocf((char *)p->g->setbits,
|
p->g->setbits = (uch *)reallocf((char *)p->g->setbits,
|
nbytes);
|
nbytes);
|
/* xxx this isn't right if setbits is now NULL */
|
/* xxx this isn't right if setbits is now NULL */
|
for (i = 0; i < no; i++)
|
for (i = 0; i < no; i++)
|
p->g->sets[i].ptr = p->g->setbits + css*(i/CHAR_BIT);
|
p->g->sets[i].ptr = p->g->setbits + css*(i/CHAR_BIT);
|
}
|
}
|
if (p->g->sets != NULL && p->g->setbits != NULL)
|
if (p->g->sets != NULL && p->g->setbits != NULL)
|
(void) memset((char *)p->g->setbits + (nbytes - css),
|
(void) memset((char *)p->g->setbits + (nbytes - css),
|
0, css);
|
0, css);
|
else {
|
else {
|
no = 0;
|
no = 0;
|
SETERROR(REG_ESPACE);
|
SETERROR(REG_ESPACE);
|
/* caller's responsibility not to do set ops */
|
/* caller's responsibility not to do set ops */
|
}
|
}
|
}
|
}
|
|
|
assert(p->g->sets != NULL); /* xxx */
|
assert(p->g->sets != NULL); /* xxx */
|
cs = &p->g->sets[no];
|
cs = &p->g->sets[no];
|
cs->ptr = p->g->setbits + css*((no)/CHAR_BIT);
|
cs->ptr = p->g->setbits + css*((no)/CHAR_BIT);
|
cs->mask = 1 << ((no) % CHAR_BIT);
|
cs->mask = 1 << ((no) % CHAR_BIT);
|
cs->hash = 0;
|
cs->hash = 0;
|
cs->smultis = 0;
|
cs->smultis = 0;
|
cs->multis = NULL;
|
cs->multis = NULL;
|
|
|
return(cs);
|
return(cs);
|
}
|
}
|
|
|
/*
|
/*
|
- freeset - free a now-unused set
|
- freeset - free a now-unused set
|
== static void freeset(struct parse *p, cset *cs);
|
== static void freeset(struct parse *p, cset *cs);
|
*/
|
*/
|
static void
|
static void
|
freeset(p, cs)
|
freeset(p, cs)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
{
|
{
|
int i;
|
int i;
|
cset *top = &p->g->sets[p->g->ncsets];
|
cset *top = &p->g->sets[p->g->ncsets];
|
size_t css = (size_t)p->g->csetsize;
|
size_t css = (size_t)p->g->csetsize;
|
|
|
for (i = 0; i < css; i++)
|
for (i = 0; i < css; i++)
|
CHsub(cs, i);
|
CHsub(cs, i);
|
if (cs == top-1) /* recover only the easy case */
|
if (cs == top-1) /* recover only the easy case */
|
p->g->ncsets--;
|
p->g->ncsets--;
|
}
|
}
|
|
|
/*
|
/*
|
- freezeset - final processing on a set of characters
|
- freezeset - final processing on a set of characters
|
== static int freezeset(struct parse *p, cset *cs);
|
== static int freezeset(struct parse *p, cset *cs);
|
*
|
*
|
* The main task here is merging identical sets. This is usually a waste
|
* The main task here is merging identical sets. This is usually a waste
|
* of time (although the hash code minimizes the overhead), but can win
|
* of time (although the hash code minimizes the overhead), but can win
|
* big if REG_ICASE is being used. REG_ICASE, by the way, is why the hash
|
* big if REG_ICASE is being used. REG_ICASE, by the way, is why the hash
|
* is done using addition rather than xor -- all ASCII [aA] sets xor to
|
* is done using addition rather than xor -- all ASCII [aA] sets xor to
|
* the same value!
|
* the same value!
|
*/
|
*/
|
static int /* set number */
|
static int /* set number */
|
freezeset(p, cs)
|
freezeset(p, cs)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
{
|
{
|
short h = cs->hash;
|
short h = cs->hash;
|
int i;
|
int i;
|
cset *top = &p->g->sets[p->g->ncsets];
|
cset *top = &p->g->sets[p->g->ncsets];
|
cset *cs2;
|
cset *cs2;
|
size_t css = (size_t)p->g->csetsize;
|
size_t css = (size_t)p->g->csetsize;
|
|
|
/* look for an earlier one which is the same */
|
/* look for an earlier one which is the same */
|
for (cs2 = &p->g->sets[0]; cs2 < top; cs2++)
|
for (cs2 = &p->g->sets[0]; cs2 < top; cs2++)
|
if (cs2->hash == h && cs2 != cs) {
|
if (cs2->hash == h && cs2 != cs) {
|
/* maybe */
|
/* maybe */
|
for (i = 0; i < css; i++)
|
for (i = 0; i < css; i++)
|
if (!!CHIN(cs2, i) != !!CHIN(cs, i))
|
if (!!CHIN(cs2, i) != !!CHIN(cs, i))
|
break; /* no */
|
break; /* no */
|
if (i == css)
|
if (i == css)
|
break; /* yes */
|
break; /* yes */
|
}
|
}
|
|
|
if (cs2 < top) { /* found one */
|
if (cs2 < top) { /* found one */
|
freeset(p, cs);
|
freeset(p, cs);
|
cs = cs2;
|
cs = cs2;
|
}
|
}
|
|
|
return((int)(cs - p->g->sets));
|
return((int)(cs - p->g->sets));
|
}
|
}
|
|
|
/*
|
/*
|
- firstch - return first character in a set (which must have at least one)
|
- firstch - return first character in a set (which must have at least one)
|
== static int firstch(struct parse *p, cset *cs);
|
== static int firstch(struct parse *p, cset *cs);
|
*/
|
*/
|
static int /* character; there is no "none" value */
|
static int /* character; there is no "none" value */
|
firstch(p, cs)
|
firstch(p, cs)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
{
|
{
|
int i;
|
int i;
|
size_t css = (size_t)p->g->csetsize;
|
size_t css = (size_t)p->g->csetsize;
|
|
|
for (i = 0; i < css; i++)
|
for (i = 0; i < css; i++)
|
if (CHIN(cs, i))
|
if (CHIN(cs, i))
|
return((char)i);
|
return((char)i);
|
assert(never);
|
assert(never);
|
return(0); /* arbitrary */
|
return(0); /* arbitrary */
|
}
|
}
|
|
|
/*
|
/*
|
- nch - number of characters in a set
|
- nch - number of characters in a set
|
== static int nch(struct parse *p, cset *cs);
|
== static int nch(struct parse *p, cset *cs);
|
*/
|
*/
|
static int
|
static int
|
nch(p, cs)
|
nch(p, cs)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
{
|
{
|
int i;
|
int i;
|
size_t css = (size_t)p->g->csetsize;
|
size_t css = (size_t)p->g->csetsize;
|
int n = 0;
|
int n = 0;
|
|
|
for (i = 0; i < css; i++)
|
for (i = 0; i < css; i++)
|
if (CHIN(cs, i))
|
if (CHIN(cs, i))
|
n++;
|
n++;
|
return(n);
|
return(n);
|
}
|
}
|
|
|
/*
|
/*
|
- mcadd - add a collating element to a cset
|
- mcadd - add a collating element to a cset
|
== static void mcadd(struct parse *p, cset *cs, \
|
== static void mcadd(struct parse *p, cset *cs, \
|
== char *cp);
|
== char *cp);
|
*/
|
*/
|
static void
|
static void
|
mcadd(p, cs, cp)
|
mcadd(p, cs, cp)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
char *cp;
|
char *cp;
|
{
|
{
|
size_t oldend = cs->smultis;
|
size_t oldend = cs->smultis;
|
|
|
cs->smultis += strlen(cp) + 1;
|
cs->smultis += strlen(cp) + 1;
|
if (cs->multis == NULL)
|
if (cs->multis == NULL)
|
cs->multis = malloc(cs->smultis);
|
cs->multis = malloc(cs->smultis);
|
else
|
else
|
cs->multis = reallocf(cs->multis, cs->smultis);
|
cs->multis = reallocf(cs->multis, cs->smultis);
|
if (cs->multis == NULL) {
|
if (cs->multis == NULL) {
|
SETERROR(REG_ESPACE);
|
SETERROR(REG_ESPACE);
|
return;
|
return;
|
}
|
}
|
|
|
(void) strcpy(cs->multis + oldend - 1, cp);
|
(void) strcpy(cs->multis + oldend - 1, cp);
|
cs->multis[cs->smultis - 1] = '\0';
|
cs->multis[cs->smultis - 1] = '\0';
|
}
|
}
|
|
|
#if used
|
#if used
|
/*
|
/*
|
- mcsub - subtract a collating element from a cset
|
- mcsub - subtract a collating element from a cset
|
== static void mcsub(cset *cs, char *cp);
|
== static void mcsub(cset *cs, char *cp);
|
*/
|
*/
|
static void
|
static void
|
mcsub(cs, cp)
|
mcsub(cs, cp)
|
cset *cs;
|
cset *cs;
|
char *cp;
|
char *cp;
|
{
|
{
|
char *fp = mcfind(cs, cp);
|
char *fp = mcfind(cs, cp);
|
size_t len = strlen(fp);
|
size_t len = strlen(fp);
|
|
|
assert(fp != NULL);
|
assert(fp != NULL);
|
(void) memmove(fp, fp + len + 1,
|
(void) memmove(fp, fp + len + 1,
|
cs->smultis - (fp + len + 1 - cs->multis));
|
cs->smultis - (fp + len + 1 - cs->multis));
|
cs->smultis -= len;
|
cs->smultis -= len;
|
|
|
if (cs->smultis == 0) {
|
if (cs->smultis == 0) {
|
free(cs->multis);
|
free(cs->multis);
|
cs->multis = NULL;
|
cs->multis = NULL;
|
return;
|
return;
|
}
|
}
|
|
|
cs->multis = reallocf(cs->multis, cs->smultis);
|
cs->multis = reallocf(cs->multis, cs->smultis);
|
assert(cs->multis != NULL);
|
assert(cs->multis != NULL);
|
}
|
}
|
|
|
/*
|
/*
|
- mcin - is a collating element in a cset?
|
- mcin - is a collating element in a cset?
|
== static int mcin(cset *cs, char *cp);
|
== static int mcin(cset *cs, char *cp);
|
*/
|
*/
|
static int
|
static int
|
mcin(cs, cp)
|
mcin(cs, cp)
|
cset *cs;
|
cset *cs;
|
char *cp;
|
char *cp;
|
{
|
{
|
return(mcfind(cs, cp) != NULL);
|
return(mcfind(cs, cp) != NULL);
|
}
|
}
|
|
|
/*
|
/*
|
- mcfind - find a collating element in a cset
|
- mcfind - find a collating element in a cset
|
== static char *mcfind(cset *cs, char *cp);
|
== static char *mcfind(cset *cs, char *cp);
|
*/
|
*/
|
static char *
|
static char *
|
mcfind(cs, cp)
|
mcfind(cs, cp)
|
cset *cs;
|
cset *cs;
|
char *cp;
|
char *cp;
|
{
|
{
|
char *p;
|
char *p;
|
|
|
if (cs->multis == NULL)
|
if (cs->multis == NULL)
|
return(NULL);
|
return(NULL);
|
for (p = cs->multis; *p != '\0'; p += strlen(p) + 1)
|
for (p = cs->multis; *p != '\0'; p += strlen(p) + 1)
|
if (strcmp(cp, p) == 0)
|
if (strcmp(cp, p) == 0)
|
return(p);
|
return(p);
|
return(NULL);
|
return(NULL);
|
}
|
}
|
#endif
|
#endif
|
|
|
/*
|
/*
|
- mcinvert - invert the list of collating elements in a cset
|
- mcinvert - invert the list of collating elements in a cset
|
== static void mcinvert(struct parse *p, cset *cs);
|
== static void mcinvert(struct parse *p, cset *cs);
|
*
|
*
|
* This would have to know the set of possibilities. Implementation
|
* This would have to know the set of possibilities. Implementation
|
* is deferred.
|
* is deferred.
|
*/
|
*/
|
static void
|
static void
|
mcinvert(p, cs)
|
mcinvert(p, cs)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
{
|
{
|
assert(cs->multis == NULL); /* xxx */
|
assert(cs->multis == NULL); /* xxx */
|
}
|
}
|
|
|
/*
|
/*
|
- mccase - add case counterparts of the list of collating elements in a cset
|
- mccase - add case counterparts of the list of collating elements in a cset
|
== static void mccase(struct parse *p, cset *cs);
|
== static void mccase(struct parse *p, cset *cs);
|
*
|
*
|
* This would have to know the set of possibilities. Implementation
|
* This would have to know the set of possibilities. Implementation
|
* is deferred.
|
* is deferred.
|
*/
|
*/
|
static void
|
static void
|
mccase(p, cs)
|
mccase(p, cs)
|
struct parse *p;
|
struct parse *p;
|
cset *cs;
|
cset *cs;
|
{
|
{
|
assert(cs->multis == NULL); /* xxx */
|
assert(cs->multis == NULL); /* xxx */
|
}
|
}
|
|
|
/*
|
/*
|
- isinsets - is this character in any sets?
|
- isinsets - is this character in any sets?
|
== static int isinsets(struct re_guts *g, int c);
|
== static int isinsets(struct re_guts *g, int c);
|
*/
|
*/
|
static int /* predicate */
|
static int /* predicate */
|
isinsets(g, c)
|
isinsets(g, c)
|
struct re_guts *g;
|
struct re_guts *g;
|
int c;
|
int c;
|
{
|
{
|
uch *col;
|
uch *col;
|
int i;
|
int i;
|
int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
|
int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
|
unsigned uc = (uch)c;
|
unsigned uc = (uch)c;
|
|
|
for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
|
for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
|
if (col[uc] != 0)
|
if (col[uc] != 0)
|
return(1);
|
return(1);
|
return(0);
|
return(0);
|
}
|
}
|
|
|
/*
|
/*
|
- samesets - are these two characters in exactly the same sets?
|
- samesets - are these two characters in exactly the same sets?
|
== static int samesets(struct re_guts *g, int c1, int c2);
|
== static int samesets(struct re_guts *g, int c1, int c2);
|
*/
|
*/
|
static int /* predicate */
|
static int /* predicate */
|
samesets(g, c1, c2)
|
samesets(g, c1, c2)
|
struct re_guts *g;
|
struct re_guts *g;
|
int c1;
|
int c1;
|
int c2;
|
int c2;
|
{
|
{
|
uch *col;
|
uch *col;
|
int i;
|
int i;
|
int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
|
int ncols = (g->ncsets+(CHAR_BIT-1)) / CHAR_BIT;
|
unsigned uc1 = (uch)c1;
|
unsigned uc1 = (uch)c1;
|
unsigned uc2 = (uch)c2;
|
unsigned uc2 = (uch)c2;
|
|
|
for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
|
for (i = 0, col = g->setbits; i < ncols; i++, col += g->csetsize)
|
if (col[uc1] != col[uc2])
|
if (col[uc1] != col[uc2])
|
return(0);
|
return(0);
|
return(1);
|
return(1);
|
}
|
}
|
|
|
/*
|
/*
|
- categorize - sort out character categories
|
- categorize - sort out character categories
|
== static void categorize(struct parse *p, struct re_guts *g);
|
== static void categorize(struct parse *p, struct re_guts *g);
|
*/
|
*/
|
static void
|
static void
|
categorize(p, g)
|
categorize(p, g)
|
struct parse *p;
|
struct parse *p;
|
struct re_guts *g;
|
struct re_guts *g;
|
{
|
{
|
cat_t *cats = g->categories;
|
cat_t *cats = g->categories;
|
int c;
|
int c;
|
int c2;
|
int c2;
|
cat_t cat;
|
cat_t cat;
|
|
|
/* avoid making error situations worse */
|
/* avoid making error situations worse */
|
if (p->error != 0)
|
if (p->error != 0)
|
return;
|
return;
|
|
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
for (c = CHAR_MIN; c <= CHAR_MAX; c++)
|
if (cats[c] == 0 && isinsets(g, c)) {
|
if (cats[c] == 0 && isinsets(g, c)) {
|
cat = g->ncategories++;
|
cat = g->ncategories++;
|
cats[c] = cat;
|
cats[c] = cat;
|
for (c2 = c+1; c2 <= CHAR_MAX; c2++)
|
for (c2 = c+1; c2 <= CHAR_MAX; c2++)
|
if (cats[c2] == 0 && samesets(g, c, c2))
|
if (cats[c2] == 0 && samesets(g, c, c2))
|
cats[c2] = cat;
|
cats[c2] = cat;
|
}
|
}
|
}
|
}
|
|
|
/*
|
/*
|
- dupl - emit a duplicate of a bunch of sops
|
- dupl - emit a duplicate of a bunch of sops
|
== static sopno dupl(struct parse *p, sopno start, sopno finish);
|
== static sopno dupl(struct parse *p, sopno start, sopno finish);
|
*/
|
*/
|
static sopno /* start of duplicate */
|
static sopno /* start of duplicate */
|
dupl(p, start, finish)
|
dupl(p, start, finish)
|
struct parse *p;
|
struct parse *p;
|
sopno start; /* from here */
|
sopno start; /* from here */
|
sopno finish; /* to this less one */
|
sopno finish; /* to this less one */
|
{
|
{
|
sopno ret = HERE();
|
sopno ret = HERE();
|
sopno len = finish - start;
|
sopno len = finish - start;
|
|
|
assert(finish >= start);
|
assert(finish >= start);
|
if (len == 0)
|
if (len == 0)
|
return(ret);
|
return(ret);
|
enlarge(p, p->ssize + len); /* this many unexpected additions */
|
enlarge(p, p->ssize + len); /* this many unexpected additions */
|
assert(p->ssize >= p->slen + len);
|
assert(p->ssize >= p->slen + len);
|
(void) memcpy((char *)(p->strip + p->slen),
|
(void) memcpy((char *)(p->strip + p->slen),
|
(char *)(p->strip + start), (size_t)len*sizeof(sop));
|
(char *)(p->strip + start), (size_t)len*sizeof(sop));
|
p->slen += len;
|
p->slen += len;
|
return(ret);
|
return(ret);
|
}
|
}
|
|
|
/*
|
/*
|
- doemit - emit a strip operator
|
- doemit - emit a strip operator
|
== static void doemit(struct parse *p, sop op, size_t opnd);
|
== static void doemit(struct parse *p, sop op, size_t opnd);
|
*
|
*
|
* It might seem better to implement this as a macro with a function as
|
* It might seem better to implement this as a macro with a function as
|
* hard-case backup, but it's just too big and messy unless there are
|
* hard-case backup, but it's just too big and messy unless there are
|
* some changes to the data structures. Maybe later.
|
* some changes to the data structures. Maybe later.
|
*/
|
*/
|
static void
|
static void
|
doemit(p, op, opnd)
|
doemit(p, op, opnd)
|
struct parse *p;
|
struct parse *p;
|
sop op;
|
sop op;
|
size_t opnd;
|
size_t opnd;
|
{
|
{
|
/* avoid making error situations worse */
|
/* avoid making error situations worse */
|
if (p->error != 0)
|
if (p->error != 0)
|
return;
|
return;
|
|
|
/* deal with oversize operands ("can't happen", more or less) */
|
/* deal with oversize operands ("can't happen", more or less) */
|
assert(opnd < 1<<OPSHIFT);
|
assert(opnd < 1<<OPSHIFT);
|
|
|
/* deal with undersized strip */
|
/* deal with undersized strip */
|
if (p->slen >= p->ssize)
|
if (p->slen >= p->ssize)
|
enlarge(p, (p->ssize+1) / 2 * 3); /* +50% */
|
enlarge(p, (p->ssize+1) / 2 * 3); /* +50% */
|
assert(p->slen < p->ssize);
|
assert(p->slen < p->ssize);
|
|
|
/* finally, it's all reduced to the easy case */
|
/* finally, it's all reduced to the easy case */
|
p->strip[p->slen++] = SOP(op, opnd);
|
p->strip[p->slen++] = SOP(op, opnd);
|
}
|
}
|
|
|
/*
|
/*
|
- doinsert - insert a sop into the strip
|
- doinsert - insert a sop into the strip
|
== static void doinsert(struct parse *p, sop op, size_t opnd, sopno pos);
|
== static void doinsert(struct parse *p, sop op, size_t opnd, sopno pos);
|
*/
|
*/
|
static void
|
static void
|
doinsert(p, op, opnd, pos)
|
doinsert(p, op, opnd, pos)
|
struct parse *p;
|
struct parse *p;
|
sop op;
|
sop op;
|
size_t opnd;
|
size_t opnd;
|
sopno pos;
|
sopno pos;
|
{
|
{
|
sopno sn;
|
sopno sn;
|
sop s;
|
sop s;
|
int i;
|
int i;
|
|
|
/* avoid making error situations worse */
|
/* avoid making error situations worse */
|
if (p->error != 0)
|
if (p->error != 0)
|
return;
|
return;
|
|
|
sn = HERE();
|
sn = HERE();
|
EMIT(op, opnd); /* do checks, ensure space */
|
EMIT(op, opnd); /* do checks, ensure space */
|
assert(HERE() == sn+1);
|
assert(HERE() == sn+1);
|
s = p->strip[sn];
|
s = p->strip[sn];
|
|
|
/* adjust paren pointers */
|
/* adjust paren pointers */
|
assert(pos > 0);
|
assert(pos > 0);
|
for (i = 1; i < NPAREN; i++) {
|
for (i = 1; i < NPAREN; i++) {
|
if (p->pbegin[i] >= pos) {
|
if (p->pbegin[i] >= pos) {
|
p->pbegin[i]++;
|
p->pbegin[i]++;
|
}
|
}
|
if (p->pend[i] >= pos) {
|
if (p->pend[i] >= pos) {
|
p->pend[i]++;
|
p->pend[i]++;
|
}
|
}
|
}
|
}
|
|
|
memmove((char *)&p->strip[pos+1], (char *)&p->strip[pos],
|
memmove((char *)&p->strip[pos+1], (char *)&p->strip[pos],
|
(HERE()-pos-1)*sizeof(sop));
|
(HERE()-pos-1)*sizeof(sop));
|
p->strip[pos] = s;
|
p->strip[pos] = s;
|
}
|
}
|
|
|
/*
|
/*
|
- dofwd - complete a forward reference
|
- dofwd - complete a forward reference
|
== static void dofwd(struct parse *p, sopno pos, sop value);
|
== static void dofwd(struct parse *p, sopno pos, sop value);
|
*/
|
*/
|
static void
|
static void
|
dofwd(p, pos, value)
|
dofwd(p, pos, value)
|
struct parse *p;
|
struct parse *p;
|
sopno pos;
|
sopno pos;
|
sop value;
|
sop value;
|
{
|
{
|
/* avoid making error situations worse */
|
/* avoid making error situations worse */
|
if (p->error != 0)
|
if (p->error != 0)
|
return;
|
return;
|
|
|
assert(value < 1<<OPSHIFT);
|
assert(value < 1<<OPSHIFT);
|
p->strip[pos] = OP(p->strip[pos]) | value;
|
p->strip[pos] = OP(p->strip[pos]) | value;
|
}
|
}
|
|
|
/*
|
/*
|
- enlarge - enlarge the strip
|
- enlarge - enlarge the strip
|
== static void enlarge(struct parse *p, sopno size);
|
== static void enlarge(struct parse *p, sopno size);
|
*/
|
*/
|
static void
|
static void
|
enlarge(p, size)
|
enlarge(p, size)
|
struct parse *p;
|
struct parse *p;
|
sopno size;
|
sopno size;
|
{
|
{
|
sop *sp;
|
sop *sp;
|
|
|
if (p->ssize >= size)
|
if (p->ssize >= size)
|
return;
|
return;
|
|
|
sp = (sop *)realloc(p->strip, size*sizeof(sop));
|
sp = (sop *)realloc(p->strip, size*sizeof(sop));
|
if (sp == NULL) {
|
if (sp == NULL) {
|
SETERROR(REG_ESPACE);
|
SETERROR(REG_ESPACE);
|
return;
|
return;
|
}
|
}
|
p->strip = sp;
|
p->strip = sp;
|
p->ssize = size;
|
p->ssize = size;
|
}
|
}
|
|
|
/*
|
/*
|
- stripsnug - compact the strip
|
- stripsnug - compact the strip
|
== static void stripsnug(struct parse *p, struct re_guts *g);
|
== static void stripsnug(struct parse *p, struct re_guts *g);
|
*/
|
*/
|
static void
|
static void
|
stripsnug(p, g)
|
stripsnug(p, g)
|
struct parse *p;
|
struct parse *p;
|
struct re_guts *g;
|
struct re_guts *g;
|
{
|
{
|
g->nstates = p->slen;
|
g->nstates = p->slen;
|
g->strip = (sop *)realloc((char *)p->strip, p->slen * sizeof(sop));
|
g->strip = (sop *)realloc((char *)p->strip, p->slen * sizeof(sop));
|
if (g->strip == NULL) {
|
if (g->strip == NULL) {
|
SETERROR(REG_ESPACE);
|
SETERROR(REG_ESPACE);
|
g->strip = p->strip;
|
g->strip = p->strip;
|
}
|
}
|
}
|
}
|
|
|
/*
|
/*
|
- findmust - fill in must and mlen with longest mandatory literal string
|
- findmust - fill in must and mlen with longest mandatory literal string
|
== static void findmust(struct parse *p, struct re_guts *g);
|
== static void findmust(struct parse *p, struct re_guts *g);
|
*
|
*
|
* This algorithm could do fancy things like analyzing the operands of |
|
* This algorithm could do fancy things like analyzing the operands of |
|
* for common subsequences. Someday. This code is simple and finds most
|
* for common subsequences. Someday. This code is simple and finds most
|
* of the interesting cases.
|
* of the interesting cases.
|
*
|
*
|
* Note that must and mlen got initialized during setup.
|
* Note that must and mlen got initialized during setup.
|
*/
|
*/
|
static void
|
static void
|
findmust(p, g)
|
findmust(p, g)
|
struct parse *p;
|
struct parse *p;
|
struct re_guts *g;
|
struct re_guts *g;
|
{
|
{
|
sop *scan;
|
sop *scan;
|
sop *start;
|
sop *start;
|
sop *newstart;
|
sop *newstart;
|
sopno newlen;
|
sopno newlen;
|
sop s;
|
sop s;
|
char *cp;
|
char *cp;
|
sopno i;
|
sopno i;
|
int offset;
|
int offset;
|
int cs, mccs;
|
int cs, mccs;
|
|
|
/* avoid making error situations worse */
|
/* avoid making error situations worse */
|
if (p->error != 0)
|
if (p->error != 0)
|
return;
|
return;
|
|
|
/* Find out if we can handle OANYOF or not */
|
/* Find out if we can handle OANYOF or not */
|
mccs = 0;
|
mccs = 0;
|
for (cs = 0; cs < g->ncsets; cs++)
|
for (cs = 0; cs < g->ncsets; cs++)
|
if (g->sets[cs].multis != NULL)
|
if (g->sets[cs].multis != NULL)
|
mccs = 1;
|
mccs = 1;
|
|
|
/* find the longest OCHAR sequence in strip */
|
/* find the longest OCHAR sequence in strip */
|
newlen = 0;
|
newlen = 0;
|
offset = 0;
|
offset = 0;
|
g->moffset = 0;
|
g->moffset = 0;
|
scan = g->strip + 1;
|
scan = g->strip + 1;
|
do {
|
do {
|
s = *scan++;
|
s = *scan++;
|
switch (OP(s)) {
|
switch (OP(s)) {
|
case OCHAR: /* sequence member */
|
case OCHAR: /* sequence member */
|
if (newlen == 0) /* new sequence */
|
if (newlen == 0) /* new sequence */
|
newstart = scan - 1;
|
newstart = scan - 1;
|
newlen++;
|
newlen++;
|
break;
|
break;
|
case OPLUS_: /* things that don't break one */
|
case OPLUS_: /* things that don't break one */
|
case OLPAREN:
|
case OLPAREN:
|
case ORPAREN:
|
case ORPAREN:
|
break;
|
break;
|
case OQUEST_: /* things that must be skipped */
|
case OQUEST_: /* things that must be skipped */
|
case OCH_:
|
case OCH_:
|
offset = altoffset(scan, offset, mccs);
|
offset = altoffset(scan, offset, mccs);
|
scan--;
|
scan--;
|
do {
|
do {
|
scan += OPND(s);
|
scan += OPND(s);
|
s = *scan;
|
s = *scan;
|
/* assert() interferes w debug printouts */
|
/* assert() interferes w debug printouts */
|
if (OP(s) != O_QUEST && OP(s) != O_CH &&
|
if (OP(s) != O_QUEST && OP(s) != O_CH &&
|
OP(s) != OOR2) {
|
OP(s) != OOR2) {
|
g->iflags |= BAD;
|
g->iflags |= BAD;
|
return;
|
return;
|
}
|
}
|
} while (OP(s) != O_QUEST && OP(s) != O_CH);
|
} while (OP(s) != O_QUEST && OP(s) != O_CH);
|
/* fallthrough */
|
/* fallthrough */
|
case OBOW: /* things that break a sequence */
|
case OBOW: /* things that break a sequence */
|
case OEOW:
|
case OEOW:
|
case OBOL:
|
case OBOL:
|
case OEOL:
|
case OEOL:
|
case O_QUEST:
|
case O_QUEST:
|
case O_CH:
|
case O_CH:
|
case OEND:
|
case OEND:
|
if (newlen > g->mlen) { /* ends one */
|
if (newlen > g->mlen) { /* ends one */
|
start = newstart;
|
start = newstart;
|
g->mlen = newlen;
|
g->mlen = newlen;
|
if (offset > -1) {
|
if (offset > -1) {
|
g->moffset += offset;
|
g->moffset += offset;
|
offset = newlen;
|
offset = newlen;
|
} else
|
} else
|
g->moffset = offset;
|
g->moffset = offset;
|
} else {
|
} else {
|
if (offset > -1)
|
if (offset > -1)
|
offset += newlen;
|
offset += newlen;
|
}
|
}
|
newlen = 0;
|
newlen = 0;
|
break;
|
break;
|
case OANY:
|
case OANY:
|
if (newlen > g->mlen) { /* ends one */
|
if (newlen > g->mlen) { /* ends one */
|
start = newstart;
|
start = newstart;
|
g->mlen = newlen;
|
g->mlen = newlen;
|
if (offset > -1) {
|
if (offset > -1) {
|
g->moffset += offset;
|
g->moffset += offset;
|
offset = newlen;
|
offset = newlen;
|
} else
|
} else
|
g->moffset = offset;
|
g->moffset = offset;
|
} else {
|
} else {
|
if (offset > -1)
|
if (offset > -1)
|
offset += newlen;
|
offset += newlen;
|
}
|
}
|
if (offset > -1)
|
if (offset > -1)
|
offset++;
|
offset++;
|
newlen = 0;
|
newlen = 0;
|
break;
|
break;
|
case OANYOF: /* may or may not invalidate offset */
|
case OANYOF: /* may or may not invalidate offset */
|
/* First, everything as OANY */
|
/* First, everything as OANY */
|
if (newlen > g->mlen) { /* ends one */
|
if (newlen > g->mlen) { /* ends one */
|
start = newstart;
|
start = newstart;
|
g->mlen = newlen;
|
g->mlen = newlen;
|
if (offset > -1) {
|
if (offset > -1) {
|
g->moffset += offset;
|
g->moffset += offset;
|
offset = newlen;
|
offset = newlen;
|
} else
|
} else
|
g->moffset = offset;
|
g->moffset = offset;
|
} else {
|
} else {
|
if (offset > -1)
|
if (offset > -1)
|
offset += newlen;
|
offset += newlen;
|
}
|
}
|
if (offset > -1)
|
if (offset > -1)
|
offset++;
|
offset++;
|
newlen = 0;
|
newlen = 0;
|
/* And, now, if we found out we can't deal with
|
/* And, now, if we found out we can't deal with
|
* it, make offset = -1.
|
* it, make offset = -1.
|
*/
|
*/
|
if (mccs)
|
if (mccs)
|
offset = -1;
|
offset = -1;
|
break;
|
break;
|
default:
|
default:
|
/* Anything here makes it impossible or too hard
|
/* Anything here makes it impossible or too hard
|
* to calculate the offset -- so we give up;
|
* to calculate the offset -- so we give up;
|
* save the last known good offset, in case the
|
* save the last known good offset, in case the
|
* must sequence doesn't occur later.
|
* must sequence doesn't occur later.
|
*/
|
*/
|
if (newlen > g->mlen) { /* ends one */
|
if (newlen > g->mlen) { /* ends one */
|
start = newstart;
|
start = newstart;
|
g->mlen = newlen;
|
g->mlen = newlen;
|
if (offset > -1)
|
if (offset > -1)
|
g->moffset += offset;
|
g->moffset += offset;
|
else
|
else
|
g->moffset = offset;
|
g->moffset = offset;
|
}
|
}
|
offset = -1;
|
offset = -1;
|
newlen = 0;
|
newlen = 0;
|
break;
|
break;
|
}
|
}
|
} while (OP(s) != OEND);
|
} while (OP(s) != OEND);
|
|
|
if (g->mlen == 0) { /* there isn't one */
|
if (g->mlen == 0) { /* there isn't one */
|
g->moffset = -1;
|
g->moffset = -1;
|
return;
|
return;
|
}
|
}
|
|
|
/* turn it into a character string */
|
/* turn it into a character string */
|
g->must = malloc((size_t)g->mlen + 1);
|
g->must = malloc((size_t)g->mlen + 1);
|
if (g->must == NULL) { /* argh; just forget it */
|
if (g->must == NULL) { /* argh; just forget it */
|
g->mlen = 0;
|
g->mlen = 0;
|
g->moffset = -1;
|
g->moffset = -1;
|
return;
|
return;
|
}
|
}
|
cp = g->must;
|
cp = g->must;
|
scan = start;
|
scan = start;
|
for (i = g->mlen; i > 0; i--) {
|
for (i = g->mlen; i > 0; i--) {
|
while (OP(s = *scan++) != OCHAR)
|
while (OP(s = *scan++) != OCHAR)
|
continue;
|
continue;
|
assert(cp < g->must + g->mlen);
|
assert(cp < g->must + g->mlen);
|
*cp++ = (char)OPND(s);
|
*cp++ = (char)OPND(s);
|
}
|
}
|
assert(cp == g->must + g->mlen);
|
assert(cp == g->must + g->mlen);
|
*cp++ = '\0'; /* just on general principles */
|
*cp++ = '\0'; /* just on general principles */
|
}
|
}
|
|
|
/*
|
/*
|
- altoffset - choose biggest offset among multiple choices
|
- altoffset - choose biggest offset among multiple choices
|
== static int altoffset(sop *scan, int offset, int mccs);
|
== static int altoffset(sop *scan, int offset, int mccs);
|
*
|
*
|
* Compute, recursively if necessary, the largest offset among multiple
|
* Compute, recursively if necessary, the largest offset among multiple
|
* re paths.
|
* re paths.
|
*/
|
*/
|
static int
|
static int
|
altoffset(scan, offset, mccs)
|
altoffset(scan, offset, mccs)
|
sop *scan;
|
sop *scan;
|
int offset;
|
int offset;
|
int mccs;
|
int mccs;
|
{
|
{
|
int largest;
|
int largest;
|
int try;
|
int try;
|
sop s;
|
sop s;
|
|
|
/* If we gave up already on offsets, return */
|
/* If we gave up already on offsets, return */
|
if (offset == -1)
|
if (offset == -1)
|
return -1;
|
return -1;
|
|
|
largest = 0;
|
largest = 0;
|
try = 0;
|
try = 0;
|
s = *scan++;
|
s = *scan++;
|
while (OP(s) != O_QUEST && OP(s) != O_CH) {
|
while (OP(s) != O_QUEST && OP(s) != O_CH) {
|
switch (OP(s)) {
|
switch (OP(s)) {
|
case OOR1:
|
case OOR1:
|
if (try > largest)
|
if (try > largest)
|
largest = try;
|
largest = try;
|
try = 0;
|
try = 0;
|
break;
|
break;
|
case OQUEST_:
|
case OQUEST_:
|
case OCH_:
|
case OCH_:
|
try = altoffset(scan, try, mccs);
|
try = altoffset(scan, try, mccs);
|
if (try == -1)
|
if (try == -1)
|
return -1;
|
return -1;
|
scan--;
|
scan--;
|
do {
|
do {
|
scan += OPND(s);
|
scan += OPND(s);
|
s = *scan;
|
s = *scan;
|
if (OP(s) != O_QUEST && OP(s) != O_CH &&
|
if (OP(s) != O_QUEST && OP(s) != O_CH &&
|
OP(s) != OOR2)
|
OP(s) != OOR2)
|
return -1;
|
return -1;
|
} while (OP(s) != O_QUEST && OP(s) != O_CH);
|
} while (OP(s) != O_QUEST && OP(s) != O_CH);
|
/* We must skip to the next position, or we'll
|
/* We must skip to the next position, or we'll
|
* leave altoffset() too early.
|
* leave altoffset() too early.
|
*/
|
*/
|
scan++;
|
scan++;
|
break;
|
break;
|
case OANYOF:
|
case OANYOF:
|
if (mccs)
|
if (mccs)
|
return -1;
|
return -1;
|
case OCHAR:
|
case OCHAR:
|
case OANY:
|
case OANY:
|
try++;
|
try++;
|
case OBOW:
|
case OBOW:
|
case OEOW:
|
case OEOW:
|
case OLPAREN:
|
case OLPAREN:
|
case ORPAREN:
|
case ORPAREN:
|
case OOR2:
|
case OOR2:
|
break;
|
break;
|
default:
|
default:
|
try = -1;
|
try = -1;
|
break;
|
break;
|
}
|
}
|
if (try == -1)
|
if (try == -1)
|
return -1;
|
return -1;
|
s = *scan++;
|
s = *scan++;
|
}
|
}
|
|
|
if (try > largest)
|
if (try > largest)
|
largest = try;
|
largest = try;
|
|
|
return largest+offset;
|
return largest+offset;
|
}
|
}
|
|
|
/*
|
/*
|
- computejumps - compute char jumps for BM scan
|
- computejumps - compute char jumps for BM scan
|
== static void computejumps(struct parse *p, struct re_guts *g);
|
== static void computejumps(struct parse *p, struct re_guts *g);
|
*
|
*
|
* This algorithm assumes g->must exists and is has size greater than
|
* This algorithm assumes g->must exists and is has size greater than
|
* zero. It's based on the algorithm found on Computer Algorithms by
|
* zero. It's based on the algorithm found on Computer Algorithms by
|
* Sara Baase.
|
* Sara Baase.
|
*
|
*
|
* A char jump is the number of characters one needs to jump based on
|
* A char jump is the number of characters one needs to jump based on
|
* the value of the character from the text that was mismatched.
|
* the value of the character from the text that was mismatched.
|
*/
|
*/
|
static void
|
static void
|
computejumps(p, g)
|
computejumps(p, g)
|
struct parse *p;
|
struct parse *p;
|
struct re_guts *g;
|
struct re_guts *g;
|
{
|
{
|
int ch;
|
int ch;
|
int mindex;
|
int mindex;
|
|
|
/* Avoid making errors worse */
|
/* Avoid making errors worse */
|
if (p->error != 0)
|
if (p->error != 0)
|
return;
|
return;
|
|
|
g->charjump = (int*) malloc((NC + 1) * sizeof(int));
|
g->charjump = (int*) malloc((NC + 1) * sizeof(int));
|
if (g->charjump == NULL) /* Not a fatal error */
|
if (g->charjump == NULL) /* Not a fatal error */
|
return;
|
return;
|
/* Adjust for signed chars, if necessary */
|
/* Adjust for signed chars, if necessary */
|
g->charjump = &g->charjump[-(CHAR_MIN)];
|
g->charjump = &g->charjump[-(CHAR_MIN)];
|
|
|
/* If the character does not exist in the pattern, the jump
|
/* If the character does not exist in the pattern, the jump
|
* is equal to the number of characters in the pattern.
|
* is equal to the number of characters in the pattern.
|
*/
|
*/
|
for (ch = CHAR_MIN; ch < (CHAR_MAX + 1); ch++)
|
for (ch = CHAR_MIN; ch < (CHAR_MAX + 1); ch++)
|
g->charjump[ch] = g->mlen;
|
g->charjump[ch] = g->mlen;
|
|
|
/* If the character does exist, compute the jump that would
|
/* If the character does exist, compute the jump that would
|
* take us to the last character in the pattern equal to it
|
* take us to the last character in the pattern equal to it
|
* (notice that we match right to left, so that last character
|
* (notice that we match right to left, so that last character
|
* is the first one that would be matched).
|
* is the first one that would be matched).
|
*/
|
*/
|
for (mindex = 0; mindex < g->mlen; mindex++)
|
for (mindex = 0; mindex < g->mlen; mindex++)
|
g->charjump[g->must[mindex]] = g->mlen - mindex - 1;
|
g->charjump[g->must[mindex]] = g->mlen - mindex - 1;
|
}
|
}
|
|
|
/*
|
/*
|
- computematchjumps - compute match jumps for BM scan
|
- computematchjumps - compute match jumps for BM scan
|
== static void computematchjumps(struct parse *p, struct re_guts *g);
|
== static void computematchjumps(struct parse *p, struct re_guts *g);
|
*
|
*
|
* This algorithm assumes g->must exists and is has size greater than
|
* This algorithm assumes g->must exists and is has size greater than
|
* zero. It's based on the algorithm found on Computer Algorithms by
|
* zero. It's based on the algorithm found on Computer Algorithms by
|
* Sara Baase.
|
* Sara Baase.
|
*
|
*
|
* A match jump is the number of characters one needs to advance based
|
* A match jump is the number of characters one needs to advance based
|
* on the already-matched suffix.
|
* on the already-matched suffix.
|
* Notice that all values here are minus (g->mlen-1), because of the way
|
* Notice that all values here are minus (g->mlen-1), because of the way
|
* the search algorithm works.
|
* the search algorithm works.
|
*/
|
*/
|
static void
|
static void
|
computematchjumps(p, g)
|
computematchjumps(p, g)
|
struct parse *p;
|
struct parse *p;
|
struct re_guts *g;
|
struct re_guts *g;
|
{
|
{
|
int mindex; /* General "must" iterator */
|
int mindex; /* General "must" iterator */
|
int suffix; /* Keeps track of matching suffix */
|
int suffix; /* Keeps track of matching suffix */
|
int ssuffix; /* Keeps track of suffixes' suffix */
|
int ssuffix; /* Keeps track of suffixes' suffix */
|
int* pmatches; /* pmatches[k] points to the next i
|
int* pmatches; /* pmatches[k] points to the next i
|
* such that i+1...mlen is a substring
|
* such that i+1...mlen is a substring
|
* of k+1...k+mlen-i-1
|
* of k+1...k+mlen-i-1
|
*/
|
*/
|
|
|
/* Avoid making errors worse */
|
/* Avoid making errors worse */
|
if (p->error != 0)
|
if (p->error != 0)
|
return;
|
return;
|
|
|
pmatches = (int*) malloc(g->mlen * sizeof(unsigned int));
|
pmatches = (int*) malloc(g->mlen * sizeof(unsigned int));
|
if (pmatches == NULL) {
|
if (pmatches == NULL) {
|
g->matchjump = NULL;
|
g->matchjump = NULL;
|
return;
|
return;
|
}
|
}
|
|
|
g->matchjump = (int*) malloc(g->mlen * sizeof(unsigned int));
|
g->matchjump = (int*) malloc(g->mlen * sizeof(unsigned int));
|
if (g->matchjump == NULL) /* Not a fatal error */
|
if (g->matchjump == NULL) /* Not a fatal error */
|
return;
|
return;
|
|
|
/* Set maximum possible jump for each character in the pattern */
|
/* Set maximum possible jump for each character in the pattern */
|
for (mindex = 0; mindex < g->mlen; mindex++)
|
for (mindex = 0; mindex < g->mlen; mindex++)
|
g->matchjump[mindex] = 2*g->mlen - mindex - 1;
|
g->matchjump[mindex] = 2*g->mlen - mindex - 1;
|
|
|
/* Compute pmatches[] */
|
/* Compute pmatches[] */
|
for (mindex = g->mlen - 1, suffix = g->mlen; mindex >= 0;
|
for (mindex = g->mlen - 1, suffix = g->mlen; mindex >= 0;
|
mindex--, suffix--) {
|
mindex--, suffix--) {
|
pmatches[mindex] = suffix;
|
pmatches[mindex] = suffix;
|
|
|
/* If a mismatch is found, interrupting the substring,
|
/* If a mismatch is found, interrupting the substring,
|
* compute the matchjump for that position. If no
|
* compute the matchjump for that position. If no
|
* mismatch is found, then a text substring mismatched
|
* mismatch is found, then a text substring mismatched
|
* against the suffix will also mismatch against the
|
* against the suffix will also mismatch against the
|
* substring.
|
* substring.
|
*/
|
*/
|
while (suffix < g->mlen
|
while (suffix < g->mlen
|
&& g->must[mindex] != g->must[suffix]) {
|
&& g->must[mindex] != g->must[suffix]) {
|
g->matchjump[suffix] = MIN(g->matchjump[suffix],
|
g->matchjump[suffix] = MIN(g->matchjump[suffix],
|
g->mlen - mindex - 1);
|
g->mlen - mindex - 1);
|
suffix = pmatches[suffix];
|
suffix = pmatches[suffix];
|
}
|
}
|
}
|
}
|
|
|
/* Compute the matchjump up to the last substring found to jump
|
/* Compute the matchjump up to the last substring found to jump
|
* to the beginning of the largest must pattern prefix matching
|
* to the beginning of the largest must pattern prefix matching
|
* it's own suffix.
|
* it's own suffix.
|
*/
|
*/
|
for (mindex = 0; mindex <= suffix; mindex++)
|
for (mindex = 0; mindex <= suffix; mindex++)
|
g->matchjump[mindex] = MIN(g->matchjump[mindex],
|
g->matchjump[mindex] = MIN(g->matchjump[mindex],
|
g->mlen + suffix - mindex);
|
g->mlen + suffix - mindex);
|
|
|
ssuffix = pmatches[suffix];
|
ssuffix = pmatches[suffix];
|
while (suffix < g->mlen) {
|
while (suffix < g->mlen) {
|
while (suffix <= ssuffix && suffix < g->mlen) {
|
while (suffix <= ssuffix && suffix < g->mlen) {
|
g->matchjump[suffix] = MIN(g->matchjump[suffix],
|
g->matchjump[suffix] = MIN(g->matchjump[suffix],
|
g->mlen + ssuffix - suffix);
|
g->mlen + ssuffix - suffix);
|
suffix++;
|
suffix++;
|
}
|
}
|
if (suffix < g->mlen)
|
if (suffix < g->mlen)
|
ssuffix = pmatches[ssuffix];
|
ssuffix = pmatches[ssuffix];
|
}
|
}
|
|
|
free(pmatches);
|
free(pmatches);
|
}
|
}
|
|
|
/*
|
/*
|
- pluscount - count + nesting
|
- pluscount - count + nesting
|
== static sopno pluscount(struct parse *p, struct re_guts *g);
|
== static sopno pluscount(struct parse *p, struct re_guts *g);
|
*/
|
*/
|
static sopno /* nesting depth */
|
static sopno /* nesting depth */
|
pluscount(p, g)
|
pluscount(p, g)
|
struct parse *p;
|
struct parse *p;
|
struct re_guts *g;
|
struct re_guts *g;
|
{
|
{
|
sop *scan;
|
sop *scan;
|
sop s;
|
sop s;
|
sopno plusnest = 0;
|
sopno plusnest = 0;
|
sopno maxnest = 0;
|
sopno maxnest = 0;
|
|
|
if (p->error != 0)
|
if (p->error != 0)
|
return(0); /* there may not be an OEND */
|
return(0); /* there may not be an OEND */
|
|
|
scan = g->strip + 1;
|
scan = g->strip + 1;
|
do {
|
do {
|
s = *scan++;
|
s = *scan++;
|
switch (OP(s)) {
|
switch (OP(s)) {
|
case OPLUS_:
|
case OPLUS_:
|
plusnest++;
|
plusnest++;
|
break;
|
break;
|
case O_PLUS:
|
case O_PLUS:
|
if (plusnest > maxnest)
|
if (plusnest > maxnest)
|
maxnest = plusnest;
|
maxnest = plusnest;
|
plusnest--;
|
plusnest--;
|
break;
|
break;
|
}
|
}
|
} while (OP(s) != OEND);
|
} while (OP(s) != OEND);
|
if (plusnest != 0)
|
if (plusnest != 0)
|
g->iflags |= BAD;
|
g->iflags |= BAD;
|
return(maxnest);
|
return(maxnest);
|
}
|
}
|
|
|
#endif /* !_NO_REGEX */
|
#endif /* !_NO_REGEX */
|
|
|
