Skip to content

P
Postgres FD Implementation
Commit f7a839bc authored by Tom Lane's avatar Tom Lane
Browse files
Options

Clean up portability problems in regexp package: change all routine 

definitions from K&R to ANSI C style, and fix broken assumption that
int and long are the same datatype.  This repairs problems observed
on Alpha with regexps having between 32 and 63 states.
parent f4e4c729
Hide whitespace changes
Inline Side-by-side
Showing with 226 additions and 603 deletions
src/backend/regex/engine.c
View file @ f7a839bc
......@@ -80,37 +80,26 @@ struct match
pg_wchar *endp; /* end of string -- virtual NUL here */
pg_wchar *coldp; /* can be no match starting before here */
pg_wchar **lastpos; /* [nplus+1] */
STATEVARS;
STATEVARS;
states st; /* current states */
states fresh; /* states for a fresh start */
states tmp; /* temporary */
states empty; /* empty set of states */
};
/* ========= begin header generated by ./mkh ========= */
#ifdef __cplusplus
extern "C"
{
#endif
static int matcher(struct re_guts *g, pg_wchar *string, size_t nmatch,
regmatch_t *pmatch, int eflags);
static pg_wchar *dissect(struct match *m, pg_wchar *start, pg_wchar *stop,
sopno startst, sopno stopst);
static pg_wchar *backref(struct match *m, pg_wchar *start, pg_wchar *stop,
sopno startst, sopno stopst, sopno lev);
static pg_wchar *fast(struct match *m, pg_wchar *start, pg_wchar *stop,
sopno startst, sopno stopst);
static pg_wchar *slow(struct match *m, pg_wchar *start, pg_wchar *stop,
sopno startst, sopno stopst);
static states step(struct re_guts *g, sopno start,
sopno stop, states bef, int ch, states aft);
/* === engine.c === */
static int
matcher(struct re_guts * g, pg_wchar * string, size_t nmatch,
regmatch_t *pmatch, int eflags);
static pg_wchar *
dissect(struct match * m, pg_wchar * start, pg_wchar * stop,
sopno startst, sopno stopst);
static pg_wchar *
backref(struct match * m, pg_wchar * start, pg_wchar * stop,
sopno startst, sopno stopst, sopno lev);
static pg_wchar *
fast(struct match * m, pg_wchar * start, pg_wchar * stop,
sopno startst, sopno stopst);
static pg_wchar *
slow(struct match * m, pg_wchar * start, pg_wchar * stop, sopno startst, sopno stopst);
static states
step(struct re_guts * g, sopno start,
sopno stop, states bef, int ch, states aft);
#define BOL (OUT+1)
#define EOL (BOL+1)
#define BOLEOL (BOL+2)
......@@ -128,24 +117,13 @@ extern "C"
#endif
#ifdef REDEBUG
static void
print(struct match * m, pg_wchar * caption, states st, int ch, FILE *d);
#endif
#ifdef REDEBUG
static void
at(struct match * m, pg_wchar * title, pg_wchar * start, pg_wchar * stop,
sopno startst, sopno stopst);
#endif
#ifdef REDEBUG
static pg_wchar *
p_char(int ch);
#endif
#ifdef __cplusplus
}
static void print(struct match *m, pg_wchar *caption, states st, int ch,
FILE *d);
static void at(struct match *m, pg_wchar *title, pg_wchar *start,
pg_wchar *stop, sopno startst, sopno stopst);
static pg_wchar *pchar(int ch);
static int pg_isprint(int c);
#endif
/* ========= end header generated by ./mkh ========= */
#ifdef REDEBUG
#define SP(t, s, c) print(m, t, s, c, stdout)
......@@ -158,17 +136,11 @@ extern "C"
#endif
/*
- matcher - the actual matching engine
== static int matcher(struct re_guts *g, pg_wchar *string, \
== size_t nmatch, regmatch_t *pmatch, int eflags);
* matcher - the actual matching engine
*/
static int /* 0 success, REG_NOMATCH failure */
matcher(g, string, nmatch, pmatch, eflags)
struct re_guts *g;
pg_wchar *string;
size_t nmatch;
regmatch_t *pmatch;
int eflags;
matcher(struct re_guts *g, pg_wchar *string, size_t nmatch,
regmatch_t *pmatch, int eflags)
{
pg_wchar *endp;
int i;
......@@ -206,10 +178,11 @@ int eflags;
for (dp = start; dp < stop; dp++)
if (*dp == g->must[0] && stop - dp >= g->mlen &&
#ifdef MULTIBYTE
memcmp(dp, g->must, (size_t) (g->mlen * sizeof(pg_wchar))) == 0)
memcmp(dp, g->must, (size_t) (g->mlen * sizeof(pg_wchar))) == 0
#else
memcmp(dp, g->must, (size_t) g->mlen) == 0)
memcmp(dp, g->must, (size_t) g->mlen) == 0
#endif
)
break;
if (dp == stop) /* we didn't find g->must */
return REG_NOMATCH;
......@@ -349,17 +322,11 @@ int eflags;
}
/*
- dissect - figure out what matched what, no back references
== static char *dissect(struct match *m, char *start, \
== char *stop, sopno startst, sopno stopst);
* dissect - figure out what matched what, no back references
*/
static pg_wchar * /* == stop (success) always */
dissect(m, start, stop, startst, stopst)
struct match *m;
pg_wchar *start;
pg_wchar *stop;
sopno startst;
sopno stopst;
dissect(struct match *m, pg_wchar *start, pg_wchar *stop,
sopno startst, sopno stopst)
{
int i;
sopno ss; /* start sop of current subRE */
......@@ -549,18 +516,13 @@ sopno stopst;
}
/*
- backref - figure out what matched what, figuring in back references
== static char *backref(struct match *m, char *start, \
== char *stop, sopno startst, sopno stopst, sopno lev);
* backref - figure out what matched what, figuring in back references
*
* lev is PLUS nesting level
*/
static pg_wchar * /* == stop (success) or NULL (failure) */
backref(m, start, stop, startst, stopst, lev)
struct match *m;
pg_wchar *start;
pg_wchar *stop;
sopno startst;
sopno stopst;
sopno lev; /* PLUS nesting level */
backref(struct match *m, pg_wchar *start, pg_wchar *stop,
sopno startst, sopno stopst, sopno lev)
{
int i;
sopno ss; /* start sop of current subRE */
......@@ -763,17 +725,11 @@ sopno lev; /* PLUS nesting level */
}
/*
- fast - step through the string at top speed
== static char *fast(struct match *m, char *start, \
== char *stop, sopno startst, sopno stopst);
* fast - step through the string at top speed
*/
static pg_wchar * /* where tentative match ended, or NULL */
fast(m, start, stop, startst, stopst)
struct match *m;
pg_wchar *start;
pg_wchar *stop;
sopno startst;
sopno stopst;
fast(struct match *m, pg_wchar *start, pg_wchar *stop,
sopno startst, sopno stopst)
{
states st = m->st;
states fresh = m->fresh;
......@@ -858,17 +814,11 @@ sopno stopst;
}
/*
- slow - step through the string more deliberately
== static char *slow(struct match *m, char *start, \
== char *stop, sopno startst, sopno stopst);
* slow - step through the string more deliberately
*/
static pg_wchar * /* where it ended */
slow(m, start, stop, startst, stopst)
struct match *m;
pg_wchar *start;
pg_wchar *stop;
sopno startst;
sopno stopst;
slow(struct match *m, pg_wchar *start, pg_wchar *stop,
sopno startst, sopno stopst)
{
states st = m->st;
states empty = m->empty;
......@@ -948,27 +898,15 @@ sopno stopst;
/*
- step - map set of states reachable before char to set reachable after
== static states step(struct re_guts *g, sopno start, sopno stop, \
== states bef, int ch, states aft);
== #define BOL (OUT+1)
== #define EOL (BOL+1)
== #define BOLEOL (BOL+2)
== #define NOTHING (BOL+3)
== #define BOW (BOL+4)
== #define EOW (BOL+5)
== #define CODEMAX (BOL+5) // highest code used
== #define NONCHAR(c) ((c) > CHAR_MAX)
== #define NNONCHAR (CODEMAX-CHAR_MAX)
* step - map set of states reachable before char to set reachable after
*/
static states
step(g, start, stop, bef, ch, aft)
struct re_guts *g;
sopno start; /* start state within strip */
sopno stop; /* state after stop state within strip */
states bef; /* states reachable before */
int ch; /* character or NONCHAR code */
states aft; /* states already known reachable after */
step(struct re_guts *g,
sopno start, /* start state within strip */
sopno stop, /* state after stop state within strip */
states bef, /* states reachable before */
int ch, /* character or NONCHAR code */
states aft) /* states already known reachable after */
{
cset *cs;
sop s;
......@@ -1082,19 +1020,11 @@ states aft; /* states already known reachable after */
#ifdef REDEBUG
/*
- print - print a set of states
== #ifdef REDEBUG
== static void print(struct match *m, char *caption, states st, \
== int ch, FILE *d);
== #endif
* print - print a set of states
*/
static void
print(m, caption, st, ch, d)
struct match *m;
pg_wchar *caption;
states st;
int ch;
FILE *d;
print(struct match *m, pg_wchar *caption, states st,
int ch, FILE *d)
{
struct re_guts *g = m->g;
int i;
......@@ -1116,20 +1046,11 @@ FILE *d;
}
/*
- at - print current situation
== #ifdef REDEBUG
== static void at(struct match *m, pg_wchar *title, pg_wchar *start, pg_wchar *stop, \
== sopno startst, sopno stopst);
== #endif
* at - print current situation
*/
static void
at(m, title, start, stop, startst, stopst)
struct match *m;
pg_wchar *title;
pg_wchar *start;
pg_wchar *stop;
sopno startst;
sopno stopst;
at(struct match *m, pg_wchar *title, pg_wchar *start, pg_wchar *stop,
sopno startst, sopno stopst)
{
if (!(m->eflags & REG_TRACE))
return;
......@@ -1140,19 +1061,26 @@ sopno stopst;
}
#ifndef PCHARDONE
#define PCHARDONE /* never again */
#define PCHARDONE /* only do this once */
/*
- pchar - make a character printable
== #ifdef REDEBUG
== static char *pchar(int ch);
== #endif
* pchar - make a character printable
*
* Is this identical to regchar() over in debug.c? Well, yes. But a
* duplicate here avoids having a debugging-capable regexec.o tied to
* a matching debug.o, and this is convenient. It all disappears in
* the non-debug compilation anyway, so it doesn't matter much.
*/
static pg_wchar * /* -> representation */
pchar(int ch)
{
static pg_wchar pbuf[10];
if (pg_isprint(ch) || ch == ' ')
sprintf(pbuf, "%c", ch);
else
sprintf(pbuf, "\\%o", ch);
return pbuf;
}
static int
pg_isprint(int c)
......@@ -1164,19 +1092,6 @@ pg_isprint(int c)
#endif
}
static pg_wchar * /* -> representation */
pchar(ch)
int ch;
{
static pg_wchar pbuf[10];
if (pg_isprint(ch) || ch == ' ')
sprintf(pbuf, "%c", ch);
else
sprintf(pbuf, "\\%o", ch);
return pbuf;
}
#endif
#endif
......
src/backend/regex/regcomp.c
View file @ f7a839bc
......@@ -37,19 +37,11 @@
* @(#)regcomp.c 8.5 (Berkeley) 3/20/94
*/
#if defined(LIBC_SCCS) && !defined(lint)
static char sccsid[] = "@(#)regcomp.c 8.5 (Berkeley) 3/20/94";
#endif /* LIBC_SCCS and not lint */
#include "postgres.h"
#include <sys/types.h>
#include <stdio.h>
#include <string.h>
#include <ctype.h>
#include <limits.h>
#include <stdlib.h>
#include <assert.h>
#include "regex/regex.h"
......@@ -78,64 +70,51 @@ struct parse
sopno pend[NPAREN]; /* -> ) ([0] unused) */
};
/* ========= begin header generated by ./mkh ========= */
#ifdef __cplusplus
extern "C"
{
#endif
/* === regcomp.c === */
static void p_ere(struct parse * p, int stop);
static void p_ere_exp(struct parse * p);
static void p_str(struct parse * p);
static void p_bre(struct parse * p, int end1, int end2);
static int p_simp_re(struct parse * p, int starordinary);
static int p_count(struct parse * p);
static void p_bracket(struct parse * p);
static void p_b_term(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 pg_wchar p_b_symbol(struct parse * p);
static char p_b_coll_elem(struct parse * p, int endc);
static void p_ere(struct parse * p, int stop);
static void p_ere_exp(struct parse * p);
static void p_str(struct parse * p);
static void p_bre(struct parse * p, int end1, int end2);
static int p_simp_re(struct parse * p, int starordinary);
static int p_count(struct parse * p);
static void p_bracket(struct parse * p);
static void p_b_term(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 pg_wchar p_b_symbol(struct parse * p);
static char p_b_coll_elem(struct parse * p, int endc);
#ifdef MULTIBYTE
static unsigned char othercase(int ch);
static unsigned char othercase(int ch);
#else
static char othercase(int ch);
#endif
static void bothcases(struct parse * p, int ch);
static void ordinary(struct parse * p, int ch);
static void nonnewline(struct parse * p);
static void repeat(struct parse * p, sopno start, int from, int to);
static int seterr(struct parse * p, int e);
static cset *allocset(struct parse * p);
static void freeset(struct parse * p, cset *cs);
static int freezeset(struct parse * p, cset *cs);
static int firstch(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 mcinvert(struct parse * p, cset *cs);
static void mccase(struct parse * p, cset *cs);
static int isinsets(struct re_guts * g, int c);
static int samesets(struct re_guts * g, int c1, int c2);
static void categorize(struct parse * p, struct re_guts * g);
static sopno dupl(struct parse * p, sopno start, sopno finish);
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 dofwd(struct parse * p, sopno pos, sop value);
static void enlarge(struct parse * p, sopno size);
static void stripsnug(struct parse * p, struct re_guts * g);
static void findmust(struct parse * p, struct re_guts * g);
static sopno pluscount(struct parse * p, struct re_guts * g);
static int pg_isdigit(int c);
static int pg_isalpha(int c);
static int pg_isupper(int c);
static int pg_islower(int c);
#ifdef __cplusplus
}
static char othercase(int ch);
#endif
/* ========= end header generated by ./mkh ========= */
static void bothcases(struct parse * p, int ch);
static void ordinary(struct parse * p, int ch);
static void nonnewline(struct parse * p);
static void repeat(struct parse * p, sopno start, int from, int to);
static int seterr(struct parse * p, int e);
static cset *allocset(struct parse * p);
static void freeset(struct parse * p, cset *cs);
static int freezeset(struct parse * p, cset *cs);
static int firstch(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 mcinvert(struct parse * p, cset *cs);
static void mccase(struct parse * p, cset *cs);
static int isinsets(struct re_guts * g, int c);
static int samesets(struct re_guts * g, int c1, int c2);
static void categorize(struct parse * p, struct re_guts * g);
static sopno dupl(struct parse * p, sopno start, sopno finish);
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 dofwd(struct parse * p, sopno pos, sop value);
static void enlarge(struct parse * p, sopno size);
static void stripsnug(struct parse * p, struct re_guts * g);
static void findmust(struct parse * p, struct re_guts * g);
static sopno pluscount(struct parse * p, struct re_guts * g);
static int pg_isdigit(int c);
static int pg_isalpha(int c);
static int pg_isupper(int c);
static int pg_islower(int c);
static pg_wchar nuls[10]; /* place to point scanner in event of
* error */
......@@ -178,22 +157,10 @@ static int never = 0; /* for use in asserts; shuts lint up */
#endif
/*
- regcomp - interface for parser and compilation
= extern int regcomp(regex_t *, const char *, int);
= #define REG_BASIC 0000
= #define REG_EXTENDED 0001
= #define REG_ICASE 0002
= #define REG_NOSUB 0004
= #define REG_NEWLINE 0010
= #define REG_NOSPEC 0020
= #define REG_PEND 0040
= #define REG_DUMP 0200
* regcomp - interface for parser and compilation
*/
int /* 0 success, otherwise REG_something */
pg95_regcomp(preg, pattern, cflags)
regex_t *preg;
const char *pattern;
int cflags;
pg95_regcomp(regex_t *preg, const char *pattern, int cflags)
{
struct parse pa;
struct re_guts *g;
......@@ -325,13 +292,11 @@ int cflags;
}
/*
- p_ere - ERE parser top level, concatenation and alternation
== static void p_ere(struct parse *p, int stop);
* p_ere - ERE parser top level, concatenation and alternation
*/
static void
p_ere(p, stop)
struct parse *p;
int stop; /* character this ERE should end at */
p_ere(struct parse *p,
int stop) /* character this ERE should end at */
{
char c;
sopno prevback = 0;
......@@ -374,12 +339,10 @@ int stop; /* character this ERE should end at */
}
/*
- p_ere_exp - parse one subERE, an atom possibly followed by a repetition op
== static void p_ere_exp(struct parse *p);
* p_ere_exp - parse one subERE, an atom possibly followed by a repetition op
*/
static void
p_ere_exp(p)
struct parse *p;
p_ere_exp(struct parse *p)
{
pg_wchar c;
sopno pos;
......@@ -535,12 +498,10 @@ struct parse *p;
}
/*
- p_str - string (no metacharacters) "parser"
== static void p_str(struct parse *p);
* p_str - string (no metacharacters) "parser"
*/
static void
p_str(p)
struct parse *p;
p_str(struct parse *p)
{
REQUIRE(MORE(), REG_EMPTY);
while (MORE())
......@@ -548,9 +509,8 @@ struct parse *p;
}
/*
- p_bre - BRE parser top level, anchoring and concatenation
== static void p_bre(struct parse *p, int end1, \
== int end2);
* p_bre - BRE parser top level, anchoring and concatenation
*
* Giving end1 as OUT essentially eliminates the end1/end2 check.
*
* This implementation is a bit of a kludge, in that a trailing $ is first
......@@ -560,10 +520,9 @@ struct parse *p;
* The amount of lookahead needed to avoid this kludge is excessive.
*/
static void
p_bre(p, end1, end2)
struct parse *p;
int end1; /* first terminating character */
int end2; /* second terminating character */
p_bre(struct parse *p,
int end1, /* first terminating character */
int end2) /* second terminating character */
{
sopno start = HERE();
int first = 1; /* first subexpression? */
......@@ -592,13 +551,11 @@ int end2; /* second terminating character */
}
/*
- p_simp_re - parse a simple RE, an atom possibly followed by a repetition
== static int p_simp_re(struct parse *p, int starordinary);
* p_simp_re - parse a simple RE, an atom possibly followed by a repetition
*/
static int /* was the simple RE an unbackslashed $? */
p_simp_re(p, starordinary)
struct parse *p;
int starordinary; /* is a leading * an ordinary character? */
p_simp_re(struct parse *p,
int starordinary) /* is a leading * an ordinary character? */
{
int c;
int count;
......@@ -731,12 +688,10 @@ int starordinary; /* is a leading * an ordinary character? */
}
/*
- p_count - parse a repetition count
== static int p_count(struct parse *p);
* p_count - parse a repetition count
*/
static int /* the value */
p_count(p)
struct parse *p;
p_count(struct parse *p)
{
int count = 0;
int ndigits = 0;
......@@ -752,15 +707,13 @@ struct parse *p;
}
/*
- p_bracket - parse a bracketed character list
== static void p_bracket(struct parse *p);
* p_bracket - parse a bracketed character list
*
* Note a significant property of this code: if the allocset() did SETERROR,
* no set operations are done.
*/
static void
p_bracket(p)
struct parse *p;
p_bracket(struct parse *p)
{
cset *cs = allocset(p);
int invert = 0;
......@@ -850,13 +803,10 @@ struct parse *p;
}
/*
- p_b_term - parse one term of a bracketed character list
== static void p_b_term(struct parse *p, cset *cs);
* p_b_term - parse one term of a bracketed character list
*/
static void
p_b_term(p, cs)
struct parse *p;
cset *cs;
p_b_term(struct parse *p, cset *cs)
{
pg_wchar c;
pg_wchar start,
......@@ -925,13 +875,10 @@ cset *cs;
}
/*
- p_b_cclass - parse a character-class name and deal with it
== static void p_b_cclass(struct parse *p, cset *cs);
* p_b_cclass - parse a character-class name and deal with it
*/
static void
p_b_cclass(p, cs)
struct parse *p;
cset *cs;
p_b_cclass(struct parse *p, cset *cs)
{
pg_wchar *sp = p->next;
struct cclass *cp;
......@@ -964,15 +911,12 @@ cset *cs;
}
/*
- p_b_eclass - parse an equivalence-class name and deal with it
== static void p_b_eclass(struct parse *p, cset *cs);
* p_b_eclass - parse an equivalence-class name and deal with it
*
* This implementation is incomplete. xxx
*/
static void
p_b_eclass(p, cs)
struct parse *p;
cset *cs;
p_b_eclass(struct parse *p, cset *cs)
{
char c;
......@@ -981,12 +925,10 @@ cset *cs;
}
/*
- p_b_symbol - parse a character or [..]ed multicharacter collating symbol
== static char p_b_symbol(struct parse *p);
* p_b_symbol - parse a character or [..]ed multicharacter collating symbol
*/
static pg_wchar /* value of symbol */
p_b_symbol(p)
struct parse *p;
p_b_symbol(struct parse *p)
{
pg_wchar value;
......@@ -1001,13 +943,10 @@ struct parse *p;
}
/*
- p_b_coll_elem - parse a collating-element name and look it up
== static char p_b_coll_elem(struct parse *p, int endc);
* p_b_coll_elem - parse a collating-element name and look it up
*/
static char /* value of collating element */
p_b_coll_elem(p, endc)
struct parse *p;
int endc; /* name ended by endc,']' */
p_b_coll_elem(struct parse *p, int endc)
{
pg_wchar *sp = p->next;
struct cname *cp;
......@@ -1035,16 +974,14 @@ int endc; /* name ended by endc,']' */
}
/*
- othercase - return the case counterpart of an alphabetic
== static char othercase(int ch);
* othercase - return the case counterpart of an alphabetic
*/
#ifdef MULTIBYTE
static unsigned char /* if no counterpart, return ch */
#else
static char /* if no counterpart, return ch */
#endif
othercase(ch)
int ch;
othercase(int ch)
{
assert(pg_isalpha(ch));
if (pg_isupper(ch))
......@@ -1069,15 +1006,12 @@ int ch;
}
/*
- bothcases - emit a dualcase version of a two-case character
== static void bothcases(struct parse *p, int ch);
* bothcases - emit a dualcase version of a two-case character
*
* Boy, is this implementation ever a kludge...
*/
static void
bothcases(p, ch)
struct parse *p;
int ch;
bothcases(struct parse *p, int ch)
{
pg_wchar *oldnext = p->next;
pg_wchar *oldend = p->end;
......@@ -1096,13 +1030,10 @@ int ch;
}
/*
- ordinary - emit an ordinary character
== static void ordinary(struct parse *p, int ch);
* ordinary - emit an ordinary character
*/
static void
ordinary(p, ch)
struct parse *p;
int ch;
ordinary(struct parse *p, int ch)
{
cat_t *cap = p->g->categories;
......@@ -1121,14 +1052,12 @@ int ch;
}
/*
- nonnewline - emit REG_NEWLINE version of OANY
== static void nonnewline(struct parse *p);
* nonnewline - emit REG_NEWLINE version of OANY
*
* Boy, is this implementation ever a kludge...
*/
static void
nonnewline(p)
struct parse *p;
nonnewline(struct parse *p)
{
pg_wchar *oldnext = p->next;
pg_wchar *oldend = p->end;
......@@ -1147,15 +1076,13 @@ struct parse *p;
}
/*
- repeat - generate code for a bounded repetition, recursively if needed
== static void repeat(struct parse *p, sopno start, int from, int to);
* repeat - generate code for a bounded repetition, recursively if needed
*/
static void
repeat(p, start, from, to)
struct parse *p;
sopno start; /* operand from here to end of strip */
int from; /* repeated from this number */
int to; /* to this number of times (maybe
repeat(struct parse *p,
sopno start, /* operand from here to end of strip */
int from, /* repeated from this number */
int to) /* to this number of times (maybe
* INFINITY) */
{
sopno finish = HERE();
......@@ -1222,13 +1149,10 @@ int to; /* to this number of times (maybe
}
/*
- seterr - set an error condition
== static int seterr(struct parse *p, int e);
* seterr - set an error condition
*/
static int /* useless but makes type checking happy */
seterr(p, e)
struct parse *p;
int e;
seterr(struct parse *p, int e)
{
if (p->error == 0) /* keep earliest error condition */
p->error = e;
......@@ -1238,12 +1162,10 @@ int e;
}
/*
- allocset - allocate a set of characters for []
== static cset *allocset(struct parse *p);
* allocset - allocate a set of characters for []
*/
static cset *
allocset(p)
struct parse *p;
allocset(struct parse *p)
{
int no = p->g->ncsets++;
size_t nc;
......@@ -1296,13 +1218,10 @@ struct parse *p;
}
/*
- freeset - free a now-unused set
== static void freeset(struct parse *p, cset *cs);
* freeset - free a now-unused set
*/
static void
freeset(p, cs)
struct parse *p;
cset *cs;
freeset(struct parse *p, cset *cs)
{
int i;
cset *top = &p->g->sets[p->g->ncsets];
......@@ -1315,8 +1234,7 @@ cset *cs;
}
/*
- freezeset - final processing on a set of characters
== static int freezeset(struct parse *p, cset *cs);
* freezeset - final processing on a set of characters
*
* 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
......@@ -1325,9 +1243,7 @@ cset *cs;
* the same value!
*/
static int /* set number */
freezeset(p, cs)
struct parse *p;
cset *cs;
freezeset(struct parse *p, cset *cs)
{
uch h = cs->hash;
int i;
......@@ -1357,13 +1273,10 @@ cset *cs;
}
/*
- firstch - return first character in a set (which must have at least one)
== static int firstch(struct parse *p, cset *cs);
* firstch - return first character in a set (which must have at least one)
*/
static int /* character; there is no "none" value */
firstch(p, cs)
struct parse *p;
cset *cs;
firstch(struct parse *p, cset *cs)
{
int i;
size_t css = (size_t) p->g->csetsize;
......@@ -1376,13 +1289,10 @@ cset *cs;
}
/*
- nch - number of characters in a set
== static int nch(struct parse *p, cset *cs);
* nch - number of characters in a set
*/
static int
nch(p, cs)
struct parse *p;
cset *cs;
nch(struct parse *p, cset *cs)
{
int i;
size_t css = (size_t) p->g->csetsize;
......@@ -1395,15 +1305,10 @@ cset *cs;
}
/*
- mcadd - add a collating element to a cset
== static void mcadd(struct parse *p, cset *cs, \
== char *cp);
* mcadd - add a collating element to a cset
*/
static void
mcadd(p, cs, cp)
struct parse *p;
cset *cs;
char *cp;
mcadd(struct parse *p, cset *cs, char *cp)
{
size_t oldend = cs->smultis;
......@@ -1423,106 +1328,34 @@ char *cp;
}
/*
- mcsub - subtract a collating element from a cset
== static void mcsub(cset *cs, char *cp);
*/
/*
static void
mcsub(cs, cp)
cset *cs;
char *cp;
{
char *fp = mcfind(cs, cp);
size_t len = strlen(fp);
assert(fp != NULL);
memmove(fp, fp + len + 1,
cs->smultis - (fp + len + 1 - cs->multis));
cs->smultis -= len;
if (cs->smultis == 0) {
free(cs->multis);
cs->multis = NULL;
return;
}
cs->multis = realloc(cs->multis, cs->smultis);
assert(cs->multis != NULL);
}
*/
/*
- mcin - is a collating element in a cset?
== static int mcin(cset *cs, char *cp);
*/
/*
static int
mcin(cs, cp)
cset *cs;
char *cp;
{
return(mcfind(cs, cp) != NULL);
}
*/
/*
- mcfind - find a collating element in a cset
== static char *mcfind(cset *cs, char *cp);
*/
/*
static char *
mcfind(cs, cp)
cset *cs;
char *cp;
{
char *p;
if (cs->multis == NULL)
return(NULL);
for (p = cs->multis; *p != '\0'; p += strlen(p) + 1)
if (strcmp(cp, p) == 0)
return(p);
return(NULL);
}
*/
/*
- mcinvert - invert the list of collating elements in a cset
== static void mcinvert(struct parse *p, cset *cs);
* mcinvert - invert the list of collating elements in a cset
*
* This would have to know the set of possibilities. Implementation
* is deferred.
*/
static void
mcinvert(p, cs)
struct parse *p;
cset *cs;
mcinvert(struct parse *p, cset *cs)
{
assert(cs->multis == NULL); /* xxx */
}
/*
- mccase - add case counterparts of the list of collating elements in a cset
== static void mccase(struct parse *p, cset *cs);
* mccase - add case counterparts of the list of collating elements in a cset
*
* This would have to know the set of possibilities. Implementation
* is deferred.
*/
static void
mccase(p, cs)
struct parse *p;
cset *cs;
mccase(struct parse *p, cset *cs)
{
assert(cs->multis == NULL); /* xxx */
}
/*
- isinsets - is this character in any sets?
== static int isinsets(struct re_guts *g, int c);
* isinsets - is this character in any sets?
*/
static int /* predicate */
isinsets(g, c)
struct re_guts *g;
int c;
isinsets(struct re_guts *g, int c)
{
uch *col;
int i;
......@@ -1536,14 +1369,10 @@ int c;
}
/*
- samesets - are these two characters in exactly the same sets?
== static int samesets(struct re_guts *g, int c1, int c2);
* samesets - are these two characters in exactly the same sets?
*/
static int /* predicate */
samesets(g, c1, c2)
struct re_guts *g;
int c1;
int c2;
samesets(struct re_guts *g, int c1, int c2)
{
uch *col;
int i;
......@@ -1558,13 +1387,10 @@ int c2;
}
/*
- categorize - sort out character categories
== static void categorize(struct parse *p, struct re_guts *g);
* categorize - sort out character categories
*/
static void
categorize(p, g)
struct parse *p;
struct re_guts *g;
categorize(struct parse *p, struct re_guts *g)
{
cat_t *cats = g->categories;
int c;
......@@ -1587,14 +1413,12 @@ struct re_guts *g;
}
/*
- dupl - emit a duplicate of a bunch of sops
== static sopno dupl(struct parse *p, sopno start, sopno finish);
* dupl - emit a duplicate of a bunch of sops
*/
static sopno /* start of duplicate */
dupl(p, start, finish)
struct parse *p;
sopno start; /* from here */
sopno finish; /* to this less one */
dupl(struct parse *p,
sopno start, /* from here */
sopno finish) /* to this less one */
{
sopno ret = HERE();
sopno len = finish - start;
......@@ -1611,18 +1435,14 @@ sopno finish; /* to this less one */
}
/*
- doemit - emit a strip operator
== static void doemit(struct parse *p, sop op, size_t opnd);
* doemit - emit a strip operator
*
* 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
* some changes to the data structures. Maybe later.
*/
static void
doemit(p, op, opnd)
struct parse *p;
sop op;
size_t opnd;
doemit(struct parse *p, sop op, size_t opnd)
{
/* avoid making error situations worse */
if (p->error != 0)
......@@ -1641,15 +1461,10 @@ size_t opnd;
}
/*
- doinsert - insert a sop into the strip
== static void doinsert(struct parse *p, sop op, size_t opnd, sopno pos);
* doinsert - insert a sop into the strip
*/
static void
doinsert(p, op, opnd, pos)
struct parse *p;
sop op;
size_t opnd;
sopno pos;
doinsert(struct parse *p, sop op, size_t opnd, sopno pos)
{
sopno sn;
sop s;
......@@ -1680,14 +1495,10 @@ sopno pos;
}
/*
- dofwd - complete a forward reference
== static void dofwd(struct parse *p, sopno pos, sop value);
* dofwd - complete a forward reference
*/
static void
dofwd(p, pos, value)
struct parse *p;
sopno pos;
sop value;
dofwd(struct parse *p, sopno pos, sop value)
{
/* avoid making error situations worse */
if (p->error != 0)
......@@ -1698,13 +1509,10 @@ sop value;
}
/*
- enlarge - enlarge the strip
== static void enlarge(struct parse *p, sopno size);
* enlarge - enlarge the strip
*/
static void
enlarge(p, size)
struct parse *p;
sopno size;
enlarge(struct parse *p, sopno size)
{
sop *sp;
......@@ -1722,13 +1530,10 @@ sopno size;
}
/*
- stripsnug - compact the strip
== static void stripsnug(struct parse *p, struct re_guts *g);
* stripsnug - compact the strip
*/
static void
stripsnug(p, g)
struct parse *p;
struct re_guts *g;
stripsnug(struct parse *p, struct re_guts *g)
{
g->nstates = p->slen;
g->strip = (sop *) realloc((char *) p->strip, p->slen * sizeof(sop));
......@@ -1740,8 +1545,7 @@ struct re_guts *g;
}
/*
- findmust - fill in must and mlen with longest mandatory literal string
== static void findmust(struct parse *p, struct re_guts *g);
* findmust - fill in must and mlen with longest mandatory literal string
*
* This algorithm could do fancy things like analyzing the operands of |
* for common subsequences. Someday. This code is simple and finds most
......@@ -1750,9 +1554,7 @@ struct re_guts *g;
* Note that must and mlen got initialized during setup.
*/
static void
findmust(p, g)
struct parse *p;
struct re_guts *g;
findmust(struct parse *p, struct re_guts *g)
{
sop *scan;
sop *start = 0;
......@@ -1838,13 +1640,10 @@ struct re_guts *g;
}
/*
- pluscount - count + nesting
== static sopno pluscount(struct parse *p, struct re_guts *g);
* pluscount - count + nesting
*/
static sopno /* nesting depth */
pluscount(p, g)
struct parse *p;
struct re_guts *g;
pluscount(struct parse *p, struct re_guts *g)
{
sop *scan;
sop s;
......@@ -1876,7 +1675,7 @@ struct re_guts *g;
}
/*
* some ctype functions with none-ascii-char guard
* some ctype functions with non-ascii-char guard
*/
static int
pg_isdigit(int c)
......
src/backend/regex/regerror.c
View file @ f7a839bc
......@@ -37,59 +37,19 @@
* @(#)regerror.c 8.4 (Berkeley) 3/20/94
*/
#if defined(LIBC_SCCS) && !defined(lint)
static char sccsid[] = "@(#)regerror.c 8.4 (Berkeley) 3/20/94";
#endif /* LIBC_SCCS and not lint */
#include "postgres.h"
#include <sys/types.h>
#include <stdio.h>
#include <string.h>
#include <ctype.h>
#include <limits.h>
#include <stdlib.h>
#include <assert.h>
#include "regex/regex.h"
#include "regex/utils.h"
#include "regex/regex2.h"
/* ========= begin header generated by ./mkh ========= */
#ifdef __cplusplus
extern "C"
{
#endif
/* === regerror.c === */
static char *regatoi(const regex_t *preg, char *localbuf);
static char *regatoi(const regex_t *preg, char *localbuf);
#ifdef __cplusplus
}
#endif
/* ========= end header generated by ./mkh ========= */
/*
= #define REG_NOMATCH 1
= #define REG_BADPAT 2
= #define REG_ECOLLATE 3
= #define REG_ECTYPE 4
= #define REG_EESCAPE 5
= #define REG_ESUBREG 6
= #define REG_EBRACK 7
= #define REG_EPAREN 8
= #define REG_EBRACE 9
= #define REG_BADBR 10
= #define REG_ERANGE 11
= #define REG_ESPACE 12
= #define REG_BADRPT 13
= #define REG_EMPTY 14
= #define REG_ASSERT 15
= #define REG_INVARG 16
= #define REG_ATOI 255 // convert name to number (!)
= #define REG_ITOA 0400 // convert number to name (!)
*/
static struct rerr
{
int code;
......@@ -152,16 +112,12 @@ static struct rerr
};
/*
- regerror - the interface to error numbers
= extern size_t regerror(int, const regex_t *, char *, size_t);
* regerror - the interface to error numbers
*/
/* ARGSUSED */
size_t
pg95_regerror(errcode, preg, errbuf, errbuf_size)
int errcode;
const regex_t *preg;
char *errbuf;
size_t errbuf_size;
pg95_regerror(int errcode, const regex_t *preg,
char *errbuf, size_t errbuf_size)
{
struct rerr *r;
size_t len;
......@@ -206,13 +162,10 @@ size_t errbuf_size;
}
/*
- regatoi - internal routine to implement REG_ATOI
== static char *regatoi(const regex_t *preg, char *localbuf);
* regatoi - internal routine to implement REG_ATOI
*/
static char *
regatoi(preg, localbuf)
const regex_t *preg;
char *localbuf;
regatoi(const regex_t *preg, char *localbuf)
{
struct rerr *r;
......
src/backend/regex/regexec.c
View file @ f7a839bc
......@@ -37,11 +37,6 @@
* @(#)regexec.c 8.3 (Berkeley) 3/20/94
*/
#if defined(LIBC_SCCS) && !defined(lint)
static char sccsid[] = "@(#)regexec.c 8.3 (Berkeley) 3/20/94";
#endif /* LIBC_SCCS and not lint */
#include "postgres.h"
/*
......@@ -52,9 +47,6 @@ static char sccsid[] = "@(#)regexec.c 8.3 (Berkeley) 3/20/94";
* representations for state sets.
*/
#include <sys/types.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <limits.h>
#include <ctype.h>
#include <assert.h>
......@@ -69,24 +61,24 @@ static int nope = 0; /* for use in asserts; shuts lint up */
#define states long
#define states1 states /* for later use in regexec() decision */
#define CLEAR(v) ((v) = 0)
#define SET0(v, n) ((v) &= ~(1 << (n)))
#define SET1(v, n) ((v) |= 1 << (n))
#define ISSET(v, n) ((v) & (1 << (n)))
#define SET0(v, n) ((v) &= ~(1L << (n)))
#define SET1(v, n) ((v) |= (1L << (n)))
#define ISSET(v, n) ((v) & (1L << (n)))
#define ASSIGN(d, s) ((d) = (s))
#define EQ(a, b) ((a) == (b))
#define STATEVARS int dummy /* dummy version */
#define STATESETUP(m, n) /* nothing */
#define STATETEARDOWN(m) /* nothing */
#define SETUP(v) ((v) = 0)
#define onestate int
#define INIT(o, n) ((o) = (unsigned)1 << (n))
#define INC(o) ((o) <<= 1)
#define ISSTATEIN(v, o) ((v) & (o))
#define onestate long
#define INIT(o, n) ((o) = (1L << (n)))
#define INC(o) ((o) <<= 1)
#define ISSTATEIN(v, o) ((v) & (o))
/* some abbreviations; note that some of these know variable names! */
/* do "if I'm here, I can also be there" etc without branches */
#define FWD(dst, src, n) ((dst) |= ((unsigned)(src)&(here)) << (n))
#define BACK(dst, src, n) ((dst) |= ((unsigned)(src)&(here)) >> (n))
#define ISSETBACK(v, n) ((v) & ((unsigned)here >> (n)))
#define FWD(dst, src, n) ((dst) |= ((src) & (here)) << (n))
#define BACK(dst, src, n) ((dst) |= ((src) & (here)) >> (n))
#define ISSETBACK(v, n) ((v) & (here >> (n)))
/* function names */
#define SNAMES /* engine.c looks after details */
......@@ -129,7 +121,7 @@ static int nope = 0; /* for use in asserts; shuts lint up */
#define SETUP(v) ((v) = &m->space[m->vn++ * m->g->nstates])
#define onestate int
#define INIT(o, n) ((o) = (n))
#define INC(o) ((o)++)
#define INC(o) ((o)++)
#define ISSTATEIN(v, o) ((v)[o])
/* some abbreviations; note that some of these know variable names! */
/* do "if I'm here, I can also be there" etc without branches */
......@@ -142,27 +134,14 @@ static int nope = 0; /* for use in asserts; shuts lint up */
#include "engine.c"
/*
- regexec - interface for matching
= extern int regexec(const regex_t *, const char *, size_t, \
= regmatch_t [], int);
= #define REG_NOTBOL 00001
= #define REG_NOTEOL 00002
= #define REG_STARTEND 00004
= #define REG_TRACE 00400 // tracing of execution
= #define REG_LARGE 01000 // force large representation
= #define REG_BACKR 02000 // force use of backref code
* regexec - interface for matching
*
* We put this here so we can exploit knowledge of the state representation
* when choosing which matcher to call. Also, by this point the matchers
* have been prototyped.
* when choosing which matcher to call.
*/
int /* 0 success, REG_NOMATCH failure */
pg95_regexec(preg, string, nmatch, pmatch, eflags)
const regex_t *preg;
const char *string;
size_t nmatch;
regmatch_t *pmatch;
int eflags;
pg95_regexec(const regex_t *preg, const char *string, size_t nmatch,
regmatch_t *pmatch, int eflags)
{
struct re_guts *g = preg->re_g;
......
src/backend/regex/regfree.c
View file @ f7a839bc
......@@ -37,28 +37,19 @@
* @(#)regfree.c 8.3 (Berkeley) 3/20/94
*/
#if defined(LIBC_SCCS) && !defined(lint)
static char sccsid[] = "@(#)regfree.c 8.3 (Berkeley) 3/20/94";
#endif /* LIBC_SCCS and not lint */
#include "postgres.h"
#include <sys/types.h>
#include <stdio.h>
#include <stdlib.h>
#include "regex/regex.h"
#include "regex/utils.h"
#include "regex/regex2.h"
/*
- regfree - free everything
= extern void regfree(regex_t *);
* regfree - free everything
*/
void
pg95_regfree(preg)
regex_t *preg;
pg95_regfree(regex_t *preg)
{
struct re_guts *g;
......
src/include/regex/cclass.h
View file @ f7a839bc
......@@ -94,6 +94,6 @@ static struct cclass
""
},
{
NULL, 0, ""
NULL, NULL, ""
}
};
src/include/regex/regex.h
View file @ f7a839bc
......@@ -102,10 +102,12 @@ typedef struct
#define REG_LARGE 01000 /* force large representation */
#define REG_BACKR 02000 /* force use of backref code */
int pg95_regcomp(regex_t *, const char *, int);
size_t pg95_regerror(int, const regex_t *, char *, size_t);
int pg95_regexec(const regex_t *,
const char *, size_t, regmatch_t[], int);
void pg95_regfree(regex_t *);
extern int pg95_regcomp(regex_t *preg, const char *pattern, int cflags);
extern size_t pg95_regerror(int errcode, const regex_t *preg,
char *errbuf, size_t errbuf_size);
extern int pg95_regexec(const regex_t *preg, const char *string,
size_t nmatch,
regmatch_t *pmatch, int eflags);
extern void pg95_regfree(regex_t *preg);
#endif /* !_REGEX_H_ */
src/include/regex/regex2.h
View file @ f7a839bc
......@@ -39,22 +39,6 @@
#include <limits.h>
/*
* First, the stuff that ends up in the outside-world include file
*/
/*
typedef off_t regoff_t;
typedef struct {
int re_magic;
size_t re_nsub; // number of parenthesized subexpressions
const char *re_endp; // end pointer for REG_PEND
struct re_guts *re_g; // none of your business :-)
} regex_t;
typedef struct {
regoff_t rm_so; // start of match
regoff_t rm_eo; // end of match
} regmatch_t;
*/
/*
* internals of regex_t
*/
......@@ -82,8 +66,8 @@
typedef unsigned long sop; /* strip operator */
typedef long sopno;
#define OPRMASK 0xf8000000
#define OPDMASK 0x07ffffff
#define OPRMASK ((sop) 0xf8000000)
#define OPDMASK ((sop) 0x07ffffff)
#define OPSHIFT ((unsigned)27)
#define OP(n) ((n)&OPRMASK)
#define OPND(n) ((n)&OPDMASK)
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment