Cygwin: replace regex with latest verbatim FreeBSD version

Signed-off-by: Corinna Vinschen <corinna@vinschen.de>

winsup/cygwin/include/regex.h

@@ -1,4 +1,6 @@
 /*-
+ * SPDX-License-Identifier: BSD-3-Clause
+ *
  * Copyright (c) 1992 Henry Spencer.
  * Copyright (c) 1992, 1993
  *	The Regents of the University of California.  All rights reserved.
@@ -14,7 +16,7 @@
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
  *    documentation and/or other materials provided with the distribution.
- * 4. Neither the name of the University nor the names of its contributors
+ * 3. Neither the name of the University nor the names of its contributors
  *    may be used to endorse or promote products derived from this software
  *    without specific prior written permission.
  *
@@ -38,32 +40,20 @@
 #define	_REGEX_H_
 
 #include <sys/cdefs.h>
-#include <_ansi.h>
 #include <sys/_types.h>
 
 /* types */
-#ifdef __CYGWIN__
-typedef	_off_t		regoff_t;
-
-#define __need_size_t
-#include <stddef.h>
-#else /* !__CYGWIN__ */
 typedef	__off_t		regoff_t;
 
 #ifndef _SIZE_T_DECLARED
 typedef	__size_t	size_t;
 #define	_SIZE_T_DECLARED
 #endif
-#endif /* !__CYGWIN__ */
 
 typedef struct {
 	int re_magic;
 	size_t re_nsub;		/* number of parenthesized subexpressions */
-#ifdef __CYGWIN__
 	const char *re_endp;	/* end pointer for REG_PEND */
-#else
-	__const char *re_endp;	/* end pointer for REG_PEND */
-#endif
 	struct re_guts *re_g;	/* none of your business :-) */
 } regex_t;
 
@@ -81,12 +71,10 @@ typedef struct {
 #define	REG_NOSPEC	0020
 #define	REG_PEND	0040
 #define	REG_DUMP	0200
+#define	REG_POSIX	0400	/* only POSIX-compliant regex (libregex) */
 
 /* regerror() flags */
 #define	REG_ENOSYS	(-1)
-#ifdef __CYGWIN__
-#define	REG_NOERROR	 0	/* GNU extension */
-#endif
 #define	REG_NOMATCH	 1
 #define	REG_BADPAT	 2
 #define	REG_ECOLLATE	 3

winsup/cygwin/regex/COPYRIGHT

@@ -32,6 +32,10 @@ to the following restrictions:
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
  *    documentation and/or other materials provided with the distribution.
+ * 3. All advertising materials mentioning features or use of this software
+ *    must display the following acknowledgement:
+ *	This product includes software developed by the University of
+ *	California, Berkeley and 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
  *    without specific prior written permission.

winsup/cygwin/regex/cname.h

@@ -1,4 +1,6 @@
 /*-
+ * SPDX-License-Identifier: BSD-3-Clause
+ *
  * Copyright (c) 1992, 1993, 1994 Henry Spencer.
  * Copyright (c) 1992, 1993, 1994
  *	The Regents of the University of California.  All rights reserved.
@@ -14,7 +16,7 @@
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
  *    documentation and/or other materials provided with the distribution.
- * 4. Neither the name of the University nor the names of its contributors
+ * 3. Neither the name of the University nor the names of its contributors
  *    may be used to endorse or promote products derived from this software
  *    without specific prior written permission.
  *
@@ -31,7 +33,7 @@
  * SUCH DAMAGE.
  *
  *	@(#)cname.h	8.3 (Berkeley) 3/20/94
- * $FreeBSD: src/lib/libc/regex/cname.h,v 1.4 2007/01/09 00:28:04 imp Exp $
+ * $FreeBSD$
  */
 
 /* character-name table */
@@ -108,7 +110,7 @@ static struct cname {
 	{"four",		'4'},
 	{"five",		'5'},
 	{"six",			'6'},
-	{"seven",		'7'},
+	{"seven",      		'7'},
 	{"eight",		'8'},
 	{"nine",		'9'},
 	{"colon",		':'},

winsup/cygwin/regex/engine.c

@@ -1,4 +1,6 @@
 /*-
+ * SPDX-License-Identifier: BSD-3-Clause
+ *
  * Copyright (c) 1992, 1993, 1994 Henry Spencer.
  * Copyright (c) 1992, 1993, 1994
  *	The Regents of the University of California.  All rights reserved.
@@ -14,7 +16,7 @@
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
  *    documentation and/or other materials provided with the distribution.
- * 4. Neither the name of the University nor the names of its contributors
+ * 3. Neither the name of the University nor the names of its contributors
  *    may be used to endorse or promote products derived from this software
  *    without specific prior written permission.
  *
@@ -34,7 +36,9 @@
  */
 
 #include <sys/cdefs.h>
-__FBSDID("$FreeBSD: src/lib/libc/regex/engine.c,v 1.23 2009/09/16 06:32:23 dds Exp $");
+__FBSDID("$FreeBSD$");
+
+#include <stdbool.h>
 
 /*
  * The matching engine and friends.  This file is #included by regexec.c
@@ -44,9 +48,9 @@ __FBSDID("$FreeBSD: src/lib/libc/regex/engine.c,v 1.23 2009/09/16 06:32:23 dds E
  */
 
 #ifdef SNAMES
+#define	stepback sstepback
 #define	matcher	smatcher
-#define	fast	sfast
+#define	walk	swalk
-#define	slow	sslow
 #define	dissect	sdissect
 #define	backref	sbackref
 #define	step	sstep
@@ -55,9 +59,9 @@ __FBSDID("$FreeBSD: src/lib/libc/regex/engine.c,v 1.23 2009/09/16 06:32:23 dds E
 #define	match	smat
 #endif
 #ifdef LNAMES
+#define	stepback lstepback
 #define	matcher	lmatcher
-#define	fast	lfast
+#define	walk	lwalk
-#define	slow	lslow
 #define	dissect	ldissect
 #define	backref	lbackref
 #define	step	lstep
@@ -66,9 +70,9 @@ __FBSDID("$FreeBSD: src/lib/libc/regex/engine.c,v 1.23 2009/09/16 06:32:23 dds E
 #define	match	lmat
 #endif
 #ifdef MNAMES
+#define	stepback mstepback
 #define	matcher	mmatcher
-#define	fast	mfast
+#define	walk	mwalk
-#define	slow	mslow
 #define	dissect	mdissect
 #define	backref	mbackref
 #define	step	mstep
@@ -104,9 +108,8 @@ extern "C" {
 static int matcher(struct re_guts *g, const char *string, size_t nmatch, regmatch_t pmatch[], int eflags);
 static const char *dissect(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst);
 static const char *backref(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst, sopno lev, int);
-static const char *fast(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst);
+static const char *walk(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst, bool fast);
-static const char *slow(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst);
+static states step(struct re_guts *g, sopno start, sopno stop, states bef, wint_t ch, states aft, int sflags);
-static states step(struct re_guts *g, sopno start, sopno stop, states bef, wint_t ch, states aft);
 #define MAX_RECURSION	100
 #define	BOL	(OUT-1)
 #define	EOL	(BOL-1)
@@ -115,10 +118,12 @@ static states step(struct re_guts *g, sopno start, sopno stop, states bef, wint_
 #define	BOW	(BOL-4)
 #define	EOW	(BOL-5)
 #define	BADCHAR	(BOL-6)
-/* When using wint_t, which is defined as unsigned int on BSD,
+#define	NWBND	(BOL-7)
-   as well as on Cygwin or Linux, the NONCHAR test is broken without
+#define	NONCHAR(c)	((c) <= OUT)
-   the below cast.  I'm wondering how this is supposed to work at all... */
+/* sflags */
-#define	NONCHAR(c)	((int)(c) <= OUT)
+#define	SBOS	0x0001
+#define	SEOS	0x0002
+
 #ifdef REDEBUG
 static void print(struct match *m, const char *caption, states st, int ch, FILE *d);
 #endif
@@ -144,6 +149,39 @@ static const char *pchar(int ch);
 #define	NOTE(s)	/* nothing */
 #endif
 
+/*
+ * Given a multibyte string pointed to by start, step back nchar characters
+ * from current position pointed to by cur.
+ */
+static const char *
+stepback(const char *start, const char *cur, int nchar)
+{
+	const char *ret;
+	int wc, mbc;
+	mbstate_t mbs;
+	size_t clen;
+
+	if (MB_CUR_MAX == 1)
+		return ((cur - nchar) > start ? cur - nchar : NULL);
+
+	ret = cur;
+	for (wc = nchar; wc > 0; wc--) {
+		for (mbc = 1; mbc <= MB_CUR_MAX; mbc++) {
+			if ((ret - mbc) < start)
+				return (NULL);
+			memset(&mbs, 0, sizeof(mbs));
+			clen = mbrtowc(NULL, ret - mbc, mbc, &mbs);
+			if (clen != (size_t)-1 && clen != (size_t)-2)
+				break;
+		}
+		if (mbc > MB_CUR_MAX)
+			return (NULL);
+		ret -= mbc;
+	}
+
+	return (ret);
+}
+
 /*
  - matcher - the actual matching engine
  == static int matcher(struct re_guts *g, const char *string, \
@@ -157,7 +195,7 @@ matcher(struct re_guts *g,
 	int eflags)
 {
 	const char *endp;
-	int i;
+	size_t i;
 	struct match mv;
 	struct match *m = &mv;
 	const char *dp = NULL;
@@ -247,17 +285,19 @@ matcher(struct re_guts *g,
 	ZAPSTATE(&m->mbs);
 
 	/* Adjust start according to moffset, to speed things up */
-#ifndef MNAMES
+	if (dp != NULL && g->moffset > -1) {
-	/* The code evaluating moffset doesn't seem to work right
+		const char *nstart;
-	   in the multibyte case. */
+
-	if (g->moffset > -1)
+		nstart = stepback(start, dp, g->moffset);
-		start = ((dp - g->moffset) < start) ? start : dp - g->moffset;
+		if (nstart != NULL)
-#endif
+			start = nstart;
+	}
+
 	SP("mloop", m->st, *start);
 
 	/* this loop does only one repetition except for backrefs */
 	for (;;) {
-		endp = fast(m, start, stop, gf, gl);
+		endp = walk(m, start, stop, gf, gl, true);
 		if (endp == NULL) {		/* a miss */
 			if (m->pmatch != NULL)
 				free((char *)m->pmatch);
@@ -273,7 +313,7 @@ matcher(struct re_guts *g,
 		assert(m->coldp != NULL);
 		for (;;) {
 			NOTE("finding start");
-			endp = slow(m, m->coldp, stop, gf, gl);
+			endp = walk(m, m->coldp, stop, gf, gl, false);
 			if (endp != NULL)
 				break;
 			assert(m->coldp < m->endp);
@@ -318,7 +358,7 @@ matcher(struct re_guts *g,
 			if (dp != NULL || endp <= m->coldp)
 				break;		/* defeat */
 			NOTE("backoff");
-			endp = slow(m, m->coldp, endp-1, gf, gl);
+			endp = walk(m, m->coldp, endp-1, gf, gl, false);
 			if (endp == NULL)
 				break;		/* defeat */
 			/* try it on a shorter possibility */
@@ -391,7 +431,7 @@ dissect(struct match *m,
 	const char *ssp;	/* start of string matched by subsubRE */
 	const char *sep;	/* end of string matched by subsubRE */
 	const char *oldssp;	/* previous ssp */
-	const char *dp __attribute__ ((unused));
+	const char *dp __unused;
 
 	AT("diss", start, stop, startst, stopst);
 	sp = start;
@@ -404,7 +444,7 @@ dissect(struct match *m,
 			es += OPND(m->g->strip[es]);
 			break;
 		case OCH_:
-			while (OP(m->g->strip[es]) != O_CH)
+			while (OP(m->g->strip[es]) != (sop)O_CH)
 				es += OPND(m->g->strip[es]);
 			break;
 		}
@@ -422,6 +462,10 @@ dissect(struct match *m,
 		case OEOL:
 		case OBOW:
 		case OEOW:
+		case OBOS:
+		case OEOS:
+		case OWBND:
+		case ONWBND:
 			break;
 		case OANY:
 		case OANYOF:
@@ -436,10 +480,10 @@ dissect(struct match *m,
 			stp = stop;
 			for (;;) {
 				/* how long could this one be? */
-				rest = slow(m, sp, stp, ss, es);
+				rest = walk(m, sp, stp, ss, es, false);
 				assert(rest != NULL);	/* it did match */
 				/* could the rest match the rest? */
-				tail = slow(m, rest, stop, es, stopst);
+				tail = walk(m, rest, stop, es, stopst, false);
 				if (tail == stop)
 					break;		/* yes! */
 				/* no -- try a shorter match for this one */
@@ -449,7 +493,7 @@ dissect(struct match *m,
 			ssub = ss + 1;
 			esub = es - 1;
 			/* did innards match? */
-			if (slow(m, sp, rest, ssub, esub) != NULL) {
+			if (walk(m, sp, rest, ssub, esub, false) != NULL) {
 				dp = dissect(m, sp, rest, ssub, esub);
 				assert(dp == rest);
 			} else		/* no */
@@ -460,10 +504,10 @@ dissect(struct match *m,
 			stp = stop;
 			for (;;) {
 				/* how long could this one be? */
-				rest = slow(m, sp, stp, ss, es);
+				rest = walk(m, sp, stp, ss, es, false);
 				assert(rest != NULL);	/* it did match */
 				/* could the rest match the rest? */
-				tail = slow(m, rest, stop, es, stopst);
+				tail = walk(m, rest, stop, es, stopst, false);
 				if (tail == stop)
 					break;		/* yes! */
 				/* no -- try a shorter match for this one */
@@ -475,7 +519,7 @@ dissect(struct match *m,
 			ssp = sp;
 			oldssp = ssp;
 			for (;;) {	/* find last match of innards */
-				sep = slow(m, ssp, rest, ssub, esub);
+				sep = walk(m, ssp, rest, ssub, esub, false);
 				if (sep == NULL || sep == ssp)
 					break;	/* failed or matched null */
 				oldssp = ssp;	/* on to next try */
@@ -487,7 +531,7 @@ dissect(struct match *m,
 				ssp = oldssp;
 			}
 			assert(sep == rest);	/* must exhaust substring */
-			assert(slow(m, ssp, sep, ssub, esub) == rest);
+			assert(walk(m, ssp, sep, ssub, esub, false) == rest);
 			dp = dissect(m, ssp, sep, ssub, esub);
 			assert(dp == sep);
 			sp = rest;
@@ -496,10 +540,10 @@ dissect(struct match *m,
 			stp = stop;
 			for (;;) {
 				/* how long could this one be? */
-				rest = slow(m, sp, stp, ss, es);
+				rest = walk(m, sp, stp, ss, es, false);
 				assert(rest != NULL);	/* it did match */
 				/* could the rest match the rest? */
-				tail = slow(m, rest, stop, es, stopst);
+				tail = walk(m, rest, stop, es, stopst, false);
 				if (tail == stop)
 					break;		/* yes! */
 				/* no -- try a shorter match for this one */
@@ -510,7 +554,7 @@ dissect(struct match *m,
 			esub = ss + OPND(m->g->strip[ss]) - 1;
 			assert(OP(m->g->strip[esub]) == OOR1);
 			for (;;) {	/* find first matching branch */
-				if (slow(m, sp, rest, ssub, esub) == rest)
+				if (walk(m, sp, rest, ssub, esub, false) == rest)
 					break;	/* it matched all of it */
 				/* that one missed, try next one */
 				assert(OP(m->g->strip[esub]) == OOR1);
@@ -518,7 +562,7 @@ dissect(struct match *m,
 				assert(OP(m->g->strip[esub]) == OOR2);
 				ssub = esub + 1;
 				esub += OPND(m->g->strip[esub]);
-				if (OP(m->g->strip[esub]) == OOR2)
+				if (OP(m->g->strip[esub]) == (sop)OOR2)
 					esub--;
 				else
 					assert(OP(m->g->strip[esub]) == O_CH);
@@ -554,6 +598,17 @@ dissect(struct match *m,
 	return(sp);
 }
 
+#define	ISBOW(m, sp)					\
+    (sp < m->endp && ISWORD(*sp) &&			\
+    ((sp == m->beginp && !(m->eflags&REG_NOTBOL)) ||	\
+    (sp > m->offp && !ISWORD(*(sp-1)))))
+#define	ISEOW(m, sp)					\
+    (((sp == m->endp && !(m->eflags&REG_NOTEOL)) ||	\
+    (sp < m->endp && *sp == '\n' &&			\
+    (m->g->cflags&REG_NEWLINE)) ||			\
+    (sp < m->endp && !ISWORD(*sp)) ) &&			\
+    (sp > m->beginp && ISWORD(*(sp-1))))		\
+
 /*
  - backref - figure out what matched what, figuring in back references
  == static const char *backref(struct match *m, const char *start, \
@@ -611,10 +666,22 @@ backref(struct match *m,
 			if (wc == BADCHAR || !CHIN(cs, wc))
 				return(NULL);
 			break;
+		case OBOS:
+			if (sp == m->beginp && (m->eflags & REG_NOTBOL) == 0)
+				{ /* yes */ }
+			else
+				return(NULL);
+			break;
+		case OEOS:
+			if (sp == m->endp && (m->eflags & REG_NOTEOL) == 0)
+				{ /* yes */ }
+			else
+				return(NULL);
+			break;
 		case OBOL:
-			if ( (sp == m->beginp && !(m->eflags&REG_NOTBOL)) ||
+			if ((sp == m->beginp && !(m->eflags&REG_NOTBOL)) ||
-					(sp < m->endp && *(sp-1) == '\n' &&
+			    (sp > m->offp && sp < m->endp &&
-						(m->g->cflags&REG_NEWLINE)) )
+			    *(sp-1) == '\n' && (m->g->cflags&REG_NEWLINE)))
 				{ /* yes */ }
 			else
 				return(NULL);
@@ -627,23 +694,29 @@ backref(struct match *m,
 			else
 				return(NULL);
 			break;
+		case OWBND:
+			if (ISBOW(m, sp) || ISEOW(m, sp))
+				{ /* yes */ }
+			else
+				return(NULL);
+			break;
+		case ONWBND:
+			if (((sp == m->beginp) && !ISWORD(*sp)) ||
+			    (sp == m->endp && !ISWORD(*(sp - 1))))
+				{ /* yes, beginning/end of subject */ }
+			else if (ISWORD(*(sp - 1)) == ISWORD(*sp))
+				{ /* yes, beginning/end of subject */ }
+			else
+				return(NULL);
+			break;
 		case OBOW:
-			if (( (sp == m->beginp && !(m->eflags&REG_NOTBOL)) ||
+			if (ISBOW(m, sp))
-					(sp < m->endp && *(sp-1) == '\n' &&
-						(m->g->cflags&REG_NEWLINE)) ||
-					(sp > m->beginp &&
-							!ISWORD(*(sp-1))) ) &&
-					(sp < m->endp && ISWORD(*sp)) )
 				{ /* yes */ }
 			else
 				return(NULL);
 			break;
 		case OEOW:
-			if (( (sp == m->endp && !(m->eflags&REG_NOTEOL)) ||
+			if (ISEOW(m, sp))
-					(sp < m->endp && *sp == '\n' &&
-						(m->g->cflags&REG_NEWLINE)) ||
-					(sp < m->endp && !ISWORD(*sp)) ) &&
-					(sp > m->beginp && ISWORD(*(sp-1))) )
 				{ /* yes */ }
 			else
 				return(NULL);
@@ -656,7 +729,7 @@ backref(struct match *m,
 			do {
 				assert(OP(s) == OOR2);
 				ss += OPND(s);
-			} while (OP(s = m->g->strip[ss]) != O_CH);
+			} while (OP(s = m->g->strip[ss]) != (sop)O_CH);
 			/* note that the ss++ gets us past the O_CH */
 			break;
 		default:	/* have to make a choice */
@@ -689,22 +762,19 @@ backref(struct match *m,
 		ssp = m->offp + m->pmatch[i].rm_so;
 		if (memcmp(sp, ssp, len) != 0)
 			return(NULL);
-		while (m->g->strip[ss] != SOP(O_BACK, i))
+		while (m->g->strip[ss] != (sop)SOP(O_BACK, i))
 			ss++;
 		return(backref(m, sp+len, stop, ss+1, stopst, lev, rec));
-		break;
 	case OQUEST_:		/* to null or not */
 		dp = backref(m, sp, stop, ss+1, stopst, lev, rec);
 		if (dp != NULL)
 			return(dp);	/* not */
 		return(backref(m, sp, stop, ss+OPND(s)+1, stopst, lev, rec));
-		break;
 	case OPLUS_:
 		assert(m->lastpos != NULL);
 		assert(lev+1 <= m->g->nplus);
 		m->lastpos[lev+1] = sp;
 		return(backref(m, sp, stop, ss+1, stopst, lev+1, rec));
-		break;
 	case O_PLUS:
 		if (sp == m->lastpos[lev])	/* last pass matched null */
 			return(backref(m, sp, stop, ss+1, stopst, lev-1, rec));
@@ -715,7 +785,6 @@ backref(struct match *m,
 			return(backref(m, sp, stop, ss+1, stopst, lev-1, rec));
 		else
 			return(dp);
-		break;
 	case OCH_:		/* find the right one, if any */
 		ssub = ss + 1;
 		esub = ss + OPND(s) - 1;
@@ -725,17 +794,18 @@ backref(struct match *m,
 			if (dp != NULL)
 				return(dp);
 			/* that one missed, try next one */
-			if (OP(m->g->strip[esub]) == O_CH)
+			if (OP(m->g->strip[esub]) == (sop)O_CH)
 				return(NULL);	/* there is none */
 			esub++;
-			assert(OP(m->g->strip[esub]) == OOR2);
+			assert(OP(m->g->strip[esub]) == (sop)OOR2);
 			ssub = esub + 1;
 			esub += OPND(m->g->strip[esub]);
-			if (OP(m->g->strip[esub]) == OOR2)
+			if (OP(m->g->strip[esub]) == (sop)OOR2)
 				esub--;
 			else
 				assert(OP(m->g->strip[esub]) == O_CH);
 		}
+		/* NOTREACHED */
 		break;
 	case OLPAREN:		/* must undo assignment if rest fails */
 		i = OPND(s);
@@ -747,7 +817,6 @@ backref(struct match *m,
 			return(dp);
 		m->pmatch[i].rm_so = offsave;
 		return(NULL);
-		break;
 	case ORPAREN:		/* must undo assignment if rest fails */
 		i = OPND(s);
 		assert(0 < i && i <= m->g->nsub);
@@ -758,7 +827,6 @@ backref(struct match *m,
 			return(dp);
 		m->pmatch[i].rm_eo = offsave;
 		return(NULL);
-		break;
 	default:		/* uh oh */
 		assert(nope);
 		break;
@@ -771,141 +839,36 @@ backref(struct match *m,
 }
 
 /*
- - fast - step through the string at top speed
+ - walk - step through the string either quickly or slowly
- == static const char *fast(struct match *m, const char *start, \
+ == static const char *walk(struct match *m, const char *start, \
- ==	const char *stop, sopno startst, sopno stopst);
+ ==	const char *stop, sopno startst, sopno stopst, bool fast);
  */
-static const char *		/* where tentative match ended, or NULL */
+static const char * /* where it ended, or NULL */
-fast(	struct match *m,
+walk(struct match *m, const char *start, const char *stop, sopno startst,
-	const char *start,
+	sopno stopst, bool fast)
-	const char *stop,
-	sopno startst,
-	sopno stopst)
 {
 	states st = m->st;
 	states fresh = m->fresh;
-	states tmp = m->tmp;
-	const char *p = start;
-	wint_t c;
-	wint_t lastc;		/* previous c */
-	wint_t flagch;
-	int i;
-	const char *coldp;	/* last p after which no match was underway */
-	size_t clen;
-
-	CLEAR(st);
-	SET1(st, startst);
-	SP("fast", st, *p);
-	st = step(m->g, startst, stopst, st, NOTHING, st);
-	ASSIGN(fresh, st);
-	SP("start", st, *p);
-	coldp = NULL;
-	if (start == m->beginp)
-		c = OUT;
-	else {
-		/*
-		 * XXX Wrong if the previous character was multi-byte.
-		 * Newline never is (in encodings supported by FreeBSD),
-		 * so this only breaks the ISWORD tests below.
-		 */
-		c = (uch)*(start - 1);
-	}
-	for (;;) {
-		/* next character */
-		lastc = c;
-		if (p == m->endp) {
-			clen = 0;
-			c = OUT;
-		} else
-			clen = XMBRTOWC(&c, p, m->endp - p, &m->mbs, BADCHAR);
-		if (EQ(st, fresh))
-			coldp = p;
-
-		/* is there an EOL and/or BOL between lastc and c? */
-		flagch = '\0';
-		i = 0;
-		if ( (lastc == '\n' && m->g->cflags&REG_NEWLINE) ||
-				(lastc == OUT && !(m->eflags&REG_NOTBOL)) ) {
-			flagch = BOL;
-			i = m->g->nbol;
-		}
-		if ( (c == '\n' && m->g->cflags&REG_NEWLINE) ||
-				(c == OUT && !(m->eflags&REG_NOTEOL)) ) {
-			flagch = (flagch == BOL) ? BOLEOL : EOL;
-			i += m->g->neol;
-		}
-		if (i != 0) {
-			for (; i > 0; i--)
-				st = step(m->g, startst, stopst, st, flagch, st);
-			SP("boleol", st, c);
-		}
-
-		/* how about a word boundary? */
-		if ( (flagch == BOL || (lastc != OUT && !ISWORD(lastc))) &&
-					(c != OUT && ISWORD(c)) ) {
-			flagch = BOW;
-		}
-		if ( (lastc != OUT && ISWORD(lastc)) &&
-				(flagch == EOL || (c != OUT && !ISWORD(c))) ) {
-			flagch = EOW;
-		}
-		if (flagch == BOW || flagch == EOW) {
-			st = step(m->g, startst, stopst, st, flagch, st);
-			SP("boweow", st, c);
-		}
-
-		/* are we done? */
-		if (ISSET(st, stopst) || p == stop || clen > stop - p)
-			break;		/* NOTE BREAK OUT */
-
-		/* no, we must deal with this character */
-		ASSIGN(tmp, st);
-		ASSIGN(st, fresh);
-		assert(c != OUT);
-		st = step(m->g, startst, stopst, tmp, c, st);
-		SP("aft", st, c);
-		assert(EQ(step(m->g, startst, stopst, st, NOTHING, st), st));
-		p += clen;
-	}
-
-	assert(coldp != NULL);
-	m->coldp = coldp;
-	if (ISSET(st, stopst))
-		return(p+XMBRTOWC(NULL, p, stop - p, &m->mbs, 0));
-	else
-		return(NULL);
-}
-
-/*
- - slow - step through the string more deliberately
- == static const char *slow(struct match *m, const char *start, \
- ==	const char *stop, sopno startst, sopno stopst);
- */
-static const char *		/* where it ended */
-slow(	struct match *m,
-	const char *start,
-	const char *stop,
-	sopno startst,
-	sopno stopst)
-{
-	states st = m->st;
 	states empty = m->empty;
 	states tmp = m->tmp;
 	const char *p = start;
 	wint_t c;
 	wint_t lastc;		/* previous c */
 	wint_t flagch;
-	int i;
+	int i, sflags;
 	const char *matchp;	/* last p at which a match ended */
 	size_t clen;
 
+	sflags = 0;
 	AT("slow", start, stop, startst, stopst);
 	CLEAR(st);
 	SET1(st, startst);
 	SP("sstart", st, *p);
-	st = step(m->g, startst, stopst, st, NOTHING, st);
+	st = step(m->g, startst, stopst, st, NOTHING, st, sflags);
+	if (fast)
+		ASSIGN(fresh, st);
 	matchp = NULL;
-	if (start == m->beginp)
+	if (start == m->offp || (start == m->beginp && !(m->eflags&REG_NOTBOL)))
 		c = OUT;
 	else {
 		/*
@@ -918,12 +881,16 @@ slow(	struct match *m,
 	for (;;) {
 		/* next character */
 		lastc = c;
+		sflags = 0;
 		if (p == m->endp) {
 			c = OUT;
 			clen = 0;
 		} else
 			clen = XMBRTOWC(&c, p, m->endp - p, &m->mbs, BADCHAR);
 
+		if (fast && EQ(st, fresh))
+			matchp = p;
+
 		/* is there an EOL and/or BOL between lastc and c? */
 		flagch = '\0';
 		i = 0;
@@ -937,9 +904,20 @@ slow(	struct match *m,
 			flagch = (flagch == BOL) ? BOLEOL : EOL;
 			i += m->g->neol;
 		}
+		if (lastc == OUT && (m->eflags & REG_NOTBOL) == 0) {
+			sflags |= SBOS;
+			/* Step one more for BOS. */
+			i++;
+		}
+		if (c == OUT && (m->eflags & REG_NOTEOL) == 0) {
+			sflags |= SEOS;
+			/* Step one more for EOS. */
+			i++;
+		}
 		if (i != 0) {
 			for (; i > 0; i--)
-				st = step(m->g, startst, stopst, st, flagch, st);
+				st = step(m->g, startst, stopst, st, flagch, st,
+				    sflags);
 			SP("sboleol", st, c);
 		}
 
@@ -953,30 +931,56 @@ slow(	struct match *m,
 			flagch = EOW;
 		}
 		if (flagch == BOW || flagch == EOW) {
-			st = step(m->g, startst, stopst, st, flagch, st);
+			st = step(m->g, startst, stopst, st, flagch, st, sflags);
 			SP("sboweow", st, c);
 		}
+		if (lastc != OUT && c != OUT &&
+		    ISWORD(lastc) == ISWORD(c)) {
+			flagch = NWBND;
+		} else if ((lastc == OUT && !ISWORD(c)) ||
+		    (c == OUT && !ISWORD(lastc))) {
+			flagch = NWBND;
+		}
+		if (flagch == NWBND) {
+			st = step(m->g, startst, stopst, st, flagch, st, sflags);
+			SP("snwbnd", st, c);
+		}
 
 		/* are we done? */
-		if (ISSET(st, stopst))
+		if (ISSET(st, stopst)) {
-			matchp = p;
+			if (fast)
-		if (EQ(st, empty) || p == stop || clen > stop - p)
+				break;
+			else
+				matchp = p;
+		}
+		if (EQ(st, empty) || p == stop || clen > (size_t)(stop - p))
 			break;		/* NOTE BREAK OUT */
 
 		/* no, we must deal with this character */
 		ASSIGN(tmp, st);
-		ASSIGN(st, empty);
+		if (fast)
+			ASSIGN(st, fresh);
+		else
+			ASSIGN(st, empty);
 		assert(c != OUT);
-		st = step(m->g, startst, stopst, tmp, c, st);
+		st = step(m->g, startst, stopst, tmp, c, st, sflags);
 		SP("saft", st, c);
-		assert(EQ(step(m->g, startst, stopst, st, NOTHING, st), st));
+		assert(EQ(step(m->g, startst, stopst, st, NOTHING, st, sflags),
+		    st));
 		p += clen;
 	}
 
-	return(matchp);
+	if (fast) {
+		assert(matchp != NULL);
+		m->coldp = matchp;
+		if (ISSET(st, stopst))
+			return (p + XMBRTOWC(NULL, p, stop - p, &m->mbs, 0));
+		else
+			return (NULL);
+	} else
+		return (matchp);
 }
 
-
 /*
  - step - map set of states reachable before char to set reachable after
  == static states step(struct re_guts *g, sopno start, sopno stop, \
@@ -996,7 +1000,8 @@ step(struct re_guts *g,
 	sopno stop,		/* state after stop state within strip */
 	states bef,		/* states reachable before */
 	wint_t ch,		/* character or NONCHAR code */
-	states aft)		/* states already known reachable after */
+	states aft,		/* states already known reachable after */
+	int sflags)		/* state flags */
 {
 	cset *cs;
 	sop s;
@@ -1017,6 +1022,14 @@ step(struct re_guts *g,
 			if (ch == OPND(s))
 				FWD(aft, bef, 1);
 			break;
+		case OBOS:
+			if ((ch == BOL || ch == BOLEOL) && (sflags & SBOS) != 0)
+				FWD(aft, bef, 1);
+			break;
+		case OEOS:
+			if ((ch == EOL || ch == BOLEOL) && (sflags & SEOS) != 0)
+				FWD(aft, bef, 1);
+			break;
 		case OBOL:
 			if (ch == BOL || ch == BOLEOL)
 				FWD(aft, bef, 1);
@@ -1033,6 +1046,14 @@ step(struct re_guts *g,
 			if (ch == EOW)
 				FWD(aft, bef, 1);
 			break;
+		case OWBND:
+			if (ch == BOW || ch == EOW)
+				FWD(aft, bef, 1);
+			break;
+		case ONWBND:
+			if (ch == NWBND)
+				FWD(aft, aft, 1);
+			break;
 		case OANY:
 			if (!NONCHAR(ch))
 				FWD(aft, bef, 1);
@@ -1072,22 +1093,22 @@ step(struct re_guts *g,
 			break;
 		case OCH_:		/* mark the first two branches */
 			FWD(aft, aft, 1);
-			assert(OP(g->strip[pc+OPND(s)]) == OOR2);
+			assert(OP(g->strip[pc+OPND(s)]) == (sop)OOR2);
 			FWD(aft, aft, OPND(s));
 			break;
 		case OOR1:		/* done a branch, find the O_CH */
 			if (ISSTATEIN(aft, here)) {
 				for (look = 1;
-						OP(s = g->strip[pc+look]) != O_CH;
+				    OP(s = g->strip[pc+look]) != (sop)O_CH;
-						look += OPND(s))
+				    look += OPND(s))
-					assert(OP(s) == OOR2);
+					assert(OP(s) == (sop)OOR2);
 				FWD(aft, aft, look + 1);
 			}
 			break;
 		case OOR2:		/* propagate OCH_'s marking */
 			FWD(aft, aft, 1);
-			if (OP(g->strip[pc+OPND(s)]) != O_CH) {
+			if (OP(g->strip[pc+OPND(s)]) != (sop)O_CH) {
-				assert(OP(g->strip[pc+OPND(s)]) == OOR2);
+				assert(OP(g->strip[pc+OPND(s)]) == (sop)OOR2);
 				FWD(aft, aft, OPND(s));
 			}
 			break;
@@ -1119,7 +1140,7 @@ print(struct match *m,
 	FILE *d)
 {
 	struct re_guts *g = m->g;
-	int i;
+	sopno i;
 	int first = 1;
 
 	if (!(m->eflags&REG_TRACE))
@@ -1130,7 +1151,7 @@ print(struct match *m,
 		fprintf(d, " %s", pchar(ch));
 	for (i = 0; i < g->nstates; i++)
 		if (ISSET(st, i)) {
-			fprintf(d, "%s%d", (first) ? "\t" : ", ", i);
+			fprintf(d, "%s%lu", (first) ? "\t" : ", ", i);
 			first = 0;
 		}
 	fprintf(d, "\n");
@@ -1186,9 +1207,9 @@ pchar(int ch)
 #endif
 #endif
 
+#undef	stepback
 #undef	matcher
-#undef	fast
+#undef	walk
-#undef	slow
 #undef	dissect
 #undef	backref
 #undef	step

winsup/cygwin/regex/regerror.c

@@ -1,4 +1,6 @@
 /*-
+ * SPDX-License-Identifier: BSD-3-Clause
+ *
  * Copyright (c) 1992, 1993, 1994 Henry Spencer.
  * Copyright (c) 1992, 1993, 1994
  *	The Regents of the University of California.  All rights reserved.
@@ -14,7 +16,7 @@
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
  *    documentation and/or other materials provided with the distribution.
- * 4. Neither the name of the University nor the names of its contributors
+ * 3. Neither the name of the University nor the names of its contributors
  *    may be used to endorse or promote products derived from this software
  *    without specific prior written permission.
  *
@@ -37,7 +39,7 @@
 static char sccsid[] = "@(#)regerror.c	8.4 (Berkeley) 3/20/94";
 #endif /* LIBC_SCCS and not lint */
 #include <sys/cdefs.h>
-__FBSDID("$FreeBSD: src/lib/libc/regex/regerror.c,v 1.11 2007/06/11 03:05:54 delphij Exp $");
+__FBSDID("$FreeBSD$");
 
 #include <sys/types.h>
 #include <stdio.h>
@@ -54,7 +56,7 @@ extern "C" {
 #endif
 
 /* === regerror.c === */
-static char *regatoi(const regex_t *preg, char *localbuf);
+static const char *regatoi(const regex_t *preg, char *localbuf);
 
 #ifdef __cplusplus
 }
@@ -83,13 +85,8 @@ static char *regatoi(const regex_t *preg, char *localbuf);
  */
 static struct rerr {
 	int code;
-#ifdef __CYGWIN__ /* Avoid whining compiler */
 	const char *name;
 	const char *explain;
-#else
-	char *name;
-	char *explain;
-#endif
 } rerrs[] = {
 	{REG_NOMATCH,	"REG_NOMATCH",	"regexec() failed to match"},
 	{REG_BADPAT,	"REG_BADPAT",	"invalid regular expression"},
@@ -125,11 +122,7 @@ regerror(int errcode,
 	struct rerr *r;
 	size_t len;
 	int target = errcode &~ REG_ITOA;
-#ifdef __CYGWIN__ /* Avoid whining compiler */
 	const char *s;
-#else
-	char *s;
-#endif
 	char convbuf[50];
 
 	if (errcode == REG_ATOI)
@@ -167,7 +160,7 @@ regerror(int errcode,
  - regatoi - internal routine to implement REG_ATOI
  == static char *regatoi(const regex_t *preg, char *localbuf);
  */
-static char *
+static const char *
 regatoi(const regex_t *preg, char *localbuf)
 {
 	struct rerr *r;
@@ -176,14 +169,7 @@ regatoi(const regex_t *preg, char *localbuf)
 		if (strcmp(r->name, preg->re_endp) == 0)
 			break;
 	if (r->code == 0)
-#ifdef __CYGWIN__ /* Avoid whining compiler */
-	    {
-		static char null[] = "0";
-		return null;
-	    }
-#else
 		return("0");
-#endif
 
 	sprintf(localbuf, "%d", r->code);
 	return(localbuf);

winsup/cygwin/regex/regex.3

@@ -13,7 +13,7 @@
 .\" 2. Redistributions in binary form must reproduce the above copyright
 .\"    notice, this list of conditions and the following disclaimer in the
 .\"    documentation and/or other materials provided with the distribution.
-.\" 4. Neither the name of the University nor the names of its contributors
+.\" 3. Neither the name of the University nor the names of its contributors
 .\"    may be used to endorse or promote products derived from this software
 .\"    without specific prior written permission.
 .\"
@@ -30,9 +30,9 @@
 .\" SUCH DAMAGE.
 .\"
 .\"	@(#)regex.3	8.4 (Berkeley) 3/20/94
-.\" $FreeBSD: src/lib/libc/regex/regex.3,v 1.21 2007/01/09 00:28:04 imp Exp $
+.\" $FreeBSD$
 .\"
-.Dd August 17, 2005
+.Dd April 15, 2017
 .Dt REGEX 3
 .Os
 .Sh NAME
@@ -183,6 +183,17 @@ compatible with but not specified by
 .St -p1003.2 ,
 and should be used with
 caution in software intended to be portable to other systems.
+.It Dv REG_POSIX
+Compile only
+.St -p1003.2
+compliant expressions.
+This flag has no effect unless linking against
+.Nm libregex .
+This is an extension,
+compatible with but not specified by
+.St -p1003.2 ,
+and should be used with
+caution in software intended to be portable to other systems.
 .El
 .Pp
 When successful,
@@ -235,11 +246,16 @@ The
 argument is the bitwise OR of zero or more of the following flags:
 .Bl -tag -width REG_STARTEND
 .It Dv REG_NOTBOL
-The first character of
+The first character of the string is treated as the continuation
-the string
+of a line.
-is not the beginning of a line, so the
+This means that the anchors
-.Ql ^\&
+.Ql ^\& ,
-anchor should not match before it.
+.Ql [[:<:]] ,
+and
+.Ql \e<
+do not match before it; but see
+.Dv REG_STARTEND
+below.
 This does not affect the behavior of newlines under
 .Dv REG_NEWLINE .
 .It Dv REG_NOTEOL
@@ -247,19 +263,16 @@ The NUL terminating
 the string
 does not end a line, so the
 .Ql $\&
-anchor should not match before it.
+anchor does not match before it.
 This does not affect the behavior of newlines under
 .Dv REG_NEWLINE .
 .It Dv REG_STARTEND
 The string is considered to start at
-.Fa string
+.Fa string No +
-+
+.Fa pmatch Ns [0]. Ns Fa rm_so
-.Fa pmatch Ns [0]. Ns Va rm_so
+and to end before the byte located at
-and to have a terminating NUL located at
+.Fa string No +
-.Fa string
+.Fa pmatch Ns [0]. Ns Fa rm_eo ,
-+
-.Fa pmatch Ns [0]. Ns Va rm_eo
-(there need not actually be a NUL at that location),
 regardless of the value of
 .Fa nmatch .
 See below for the definition of
@@ -271,13 +284,37 @@ compatible with but not specified by
 .St -p1003.2 ,
 and should be used with
 caution in software intended to be portable to other systems.
-Note that a non-zero
+.Pp
-.Va rm_so
+Without
-does not imply
+.Dv REG_NOTBOL ,
-.Dv REG_NOTBOL ;
+the position
-.Dv REG_STARTEND
+.Fa rm_so
-affects only the location of the string,
+is considered the beginning of a line, such that
-not how it is matched.
+.Ql ^
+matches before it, and the beginning of a word if there is a word
+character at this position, such that
+.Ql [[:<:]]
+and
+.Ql \e<
+match before it.
+.Pp
+With
+.Dv REG_NOTBOL ,
+the character at position
+.Fa rm_so
+is treated as the continuation of a line, and if
+.Fa rm_so
+is greater than 0, the preceding character is taken into consideration.
+If the preceding character is a newline and the regular expression was compiled
+with
+.Dv REG_NEWLINE ,
+.Ql ^
+matches before the string; if the preceding character is not a word character
+but the string starts with a word character,
+.Ql [[:<:]]
+and
+.Ql \e<
+match before the string.
 .El
 .Pp
 See
@@ -420,10 +457,12 @@ it should have been the result from the most recent
 using that
 .Ft regex_t .
 The
-.Fn ( regerror
+.Po
+.Fn regerror
 may be able to supply a more detailed message using information
 from the
-.Ft regex_t . )
+.Ft regex_t .
+.Pc
 The
 .Fn regerror
 function

winsup/cygwin/regex/regex.7

@@ -13,6 +13,10 @@
 .\" 2. Redistributions in binary form must reproduce the above copyright
 .\"    notice, this list of conditions and the following disclaimer in the
 .\"    documentation and/or other materials provided with the distribution.
+.\" 3. All advertising materials mentioning features or use of this software
+.\"    must display the following acknowledgement:
+.\"	This product includes software developed by the University of
+.\"	California, Berkeley and 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
 .\"    without specific prior written permission.
@@ -30,9 +34,9 @@
 .\" SUCH DAMAGE.
 .\"
 .\"	@(#)re_format.7	8.3 (Berkeley) 3/20/94
-.\" $FreeBSD: src/lib/libc/regex/re_format.7,v 1.12 2008/09/05 17:41:20 keramida Exp $
+.\" $FreeBSD$
 .\"
-.Dd March 20, 1994
+.Dd June 30, 2014
 .Dt RE_FORMAT 7
 .Os
 .Sh NAME
@@ -271,7 +275,6 @@ and
 stands for the list of all characters belonging to that
 class.
 Standard character class names are:
-.Pp
 .Bl -column "alnum" "digit" "xdigit" -offset indent
 .It Em "alnum	digit	punct"
 .It Em "alpha	graph	space"
@@ -311,6 +314,13 @@ compatible with but not specified by
 .St -p1003.2 ,
 and should be used with
 caution in software intended to be portable to other systems.
+The additional word delimiters
+.Ql \e<
+and
+.Ql \e>
+are provided to ease compatibility with traditional
+SVR4
+systems but are not portable and should be avoided.
 .Pp
 In the event that an RE could match more than one substring of a given
 string,
@@ -382,10 +392,12 @@ and
 .Ql ?\&
 are ordinary characters, and their functionality
 can be expressed using bounds
-.No ( Ql {1,}
+.Po
+.Ql {1,}
 or
 .Ql {0,1}
-respectively).
+respectively
+.Pc .
 Also note that
 .Ql x+
 in modern REs is equivalent to

winsup/cygwin/regex/regex2.h

@@ -1,4 +1,6 @@
 /*-
+ * SPDX-License-Identifier: BSD-3-Clause
+ *
  * Copyright (c) 1992, 1993, 1994 Henry Spencer.
  * Copyright (c) 1992, 1993, 1994
  *	The Regents of the University of California.  All rights reserved.
@@ -14,7 +16,7 @@
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
  *    documentation and/or other materials provided with the distribution.
- * 4. Neither the name of the University nor the names of its contributors
+ * 3. Neither the name of the University nor the names of its contributors
  *    may be used to endorse or promote products derived from this software
  *    without specific prior written permission.
  *
@@ -31,21 +33,21 @@
  * SUCH DAMAGE.
  *
  *	@(#)regex2.h	8.4 (Berkeley) 3/20/94
- * $FreeBSD: src/lib/libc/regex/regex2.h,v 1.11 2007/01/09 00:28:04 imp Exp $
+ * $FreeBSD$
  */
 
 /*
  * First, the stuff that ends up in the outside-world include file
  = typedef off_t regoff_t;
  = typedef struct {
- =	int re_magic;
+ = 	int re_magic;
- =	size_t re_nsub;		// number of parenthesized subexpressions
+ = 	size_t re_nsub;		// number of parenthesized subexpressions
- =	const char *re_endp;	// end pointer for REG_PEND
+ = 	const char *re_endp;	// end pointer for REG_PEND
- =	struct re_guts *re_g;	// none of your business :-)
+ = 	struct re_guts *re_g;	// none of your business :-)
  = } regex_t;
  = typedef struct {
- =	regoff_t rm_so;		// start of match
+ = 	regoff_t rm_so;		// start of match
- =	regoff_t rm_eo;		// end of match
+ = 	regoff_t rm_eo;		// end of match
  = } regmatch_t;
  */
 /*
@@ -73,7 +75,7 @@
  * immediately *preceding* "execution" of that operator.
  */
 typedef unsigned long sop;	/* strip operator */
-typedef long sopno;
+typedef unsigned long sopno;
 #define	OPRMASK	0xf8000000L
 #define	OPDMASK	0x07ffffffL
 #define	OPSHIFT	((unsigned)27)
@@ -102,6 +104,10 @@ typedef long sopno;
 #define	O_CH	(18L<<OPSHIFT)	/* end choice	back to OOR1		*/
 #define	OBOW	(19L<<OPSHIFT)	/* begin word	-			*/
 #define	OEOW	(20L<<OPSHIFT)	/* end word	-			*/
+#define	OBOS	(21L<<OPSHIFT)	/* begin subj.  -			*/
+#define	OEOS	(22L<<OPSHIFT)	/* end subj.	-			*/
+#define	OWBND	(23L<<OPSHIFT)	/* word bound	-			*/
+#define	ONWBND	(24L<<OPSHIFT)	/* not bound	-			*/
 
 /*
  * Structures for [] character-set representation.
@@ -111,13 +117,13 @@ typedef struct {
 	wint_t		max;
 } crange;
 typedef struct {
-	unsigned char	bmp[NC / 8];
+	unsigned char	bmp[NC_MAX / 8];
 	wctype_t	*types;
-	int		ntypes;
+	unsigned int	ntypes;
 	wint_t		*wides;
-	int		nwides;
+	unsigned int	nwides;
 	crange		*ranges;
-	int		nranges;
+	unsigned int	nranges;
 	int		invert;
 	int		icase;
 } cset;
@@ -125,15 +131,20 @@ typedef struct {
 static int
 CHIN1(cset *cs, wint_t ch)
 {
-	int i;
+	unsigned int i;
 
 	assert(ch >= 0);
 	if (ch < NC)
 		return (((cs->bmp[ch >> 3] & (1 << (ch & 7))) != 0) ^
 		    cs->invert);
-	for (i = 0; i < cs->nwides; i++)
+	for (i = 0; i < cs->nwides; i++) {
-		if (ch == cs->wides[i])
+		if (cs->icase) {
+			if (ch == towlower(cs->wides[i]) ||
+			    ch == towupper(cs->wides[i]))
+				return (!cs->invert);
+		} else if (ch == cs->wides[i])
 			return (!cs->invert);
+	}
 	for (i = 0; i < cs->nranges; i++)
 		if (cs->ranges[i].min <= ch && ch <= cs->ranges[i].max)
 			return (!cs->invert);
@@ -151,14 +162,10 @@ CHIN(cset *cs, wint_t ch)
 	if (ch < NC)
 		return (((cs->bmp[ch >> 3] & (1 << (ch & 7))) != 0) ^
 		    cs->invert);
-	else if (cs->icase) {
+	else if (cs->icase)
-		if (cs->invert)
+		return (CHIN1(cs, ch) || CHIN1(cs, towlower(ch)) ||
-			return (CHIN1(cs, ch) && CHIN1(cs, towlower(ch)) &&
+		    CHIN1(cs, towupper(ch)));
-			    CHIN1(cs, towupper(ch)));
+	else
-		else
-			return (CHIN1(cs, ch) || CHIN1(cs, towlower(ch)) ||
-			    CHIN1(cs, towupper(ch)));
-	} else
 		return (CHIN1(cs, ch));
 }
 
@@ -169,7 +176,7 @@ struct re_guts {
 	int magic;
 #		define	MAGIC2	((('R'^0200)<<8)|'E')
 	sop *strip;		/* malloced area for strip */
-	int ncsets;		/* number of csets in use */
+	unsigned int ncsets;	/* number of csets in use */
 	cset *sets;		/* -> cset [ncsets] */
 	int cflags;		/* copy of regcomp() cflags argument */
 	sopno nstates;		/* = number of sops */
@@ -193,4 +200,5 @@ struct re_guts {
 
 /* misc utilities */
 #define	OUT	(CHAR_MIN - 1)	/* a non-character value */
-#define ISWORD(c)       (iswalnum((wint_t)(c)) || (c) == '_')
+#define	IGN	(CHAR_MIN - 2)
+#define ISWORD(c)       (iswalnum((uch)(c)) || (c) == '_')

winsup/cygwin/regex/regexec.c

@@ -1,4 +1,6 @@
 /*-
+ * SPDX-License-Identifier: BSD-3-Clause
+ *
  * Copyright (c) 1992, 1993, 1994 Henry Spencer.
  * Copyright (c) 1992, 1993, 1994
  *	The Regents of the University of California.  All rights reserved.
@@ -14,7 +16,7 @@
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
  *    documentation and/or other materials provided with the distribution.
- * 4. Neither the name of the University nor the names of its contributors
+ * 3. Neither the name of the University nor the names of its contributors
  *    may be used to endorse or promote products derived from this software
  *    without specific prior written permission.
  *
@@ -37,7 +39,7 @@
 static char sccsid[] = "@(#)regexec.c	8.3 (Berkeley) 3/20/94";
 #endif /* LIBC_SCCS and not lint */
 #include <sys/cdefs.h>
-__FBSDID("$FreeBSD: src/lib/libc/regex/regexec.c,v 1.8 2007/06/11 03:05:54 delphij Exp $");
+__FBSDID("$FreeBSD$");
 
 /*
  * the outer shell of regexec()
@@ -46,9 +48,6 @@ __FBSDID("$FreeBSD: src/lib/libc/regex/regexec.c,v 1.8 2007/06/11 03:05:54 delph
  * macros that code uses.  This lets the same code operate on two different
  * representations for state sets and characters.
  */
-#ifdef __CYGWIN__
-#include "winsup.h"
-#endif
 #include <sys/types.h>
 #include <stdio.h>
 #include <stdlib.h>
@@ -68,9 +67,9 @@ static __inline size_t
 xmbrtowc(wint_t *wi, const char *s, size_t n, mbstate_t *mbs, wint_t dummy)
 {
 	size_t nr;
-	wint_t wc;
+	wchar_t wc;
 
-	nr = mbrtowi(&wc, s, n, mbs);
+	nr = mbrtowc(&wc, s, n, mbs);
 	if (wi != NULL)
 		*wi = wc;
 	if (nr == 0)
@@ -98,8 +97,8 @@ xmbrtowc_dummy(wint_t *wi,
 }
 
 /* macros for manipulating states, small version */
-#define	states	long
+#define	states1	long		/* for later use in regexec() decision */
-#define	states1	states		/* for later use in regexec() decision */
+#define	states	states1
 #define	CLEAR(v)	((v) = 0)
 #define	SET0(v, n)	((v) &= ~((unsigned long)1 << (n)))
 #define	SET1(v, n)	((v) |= (unsigned long)1 << (n))
@@ -228,9 +227,9 @@ regexec(const regex_t * __restrict preg,
 	eflags = GOODFLAGS(eflags);
 
 	if (MB_CUR_MAX > 1)
-		return(mmatcher(g, (char *)string, nmatch, pmatch, eflags));
+		return(mmatcher(g, string, nmatch, pmatch, eflags));
 	else if (g->nstates <= CHAR_BIT*sizeof(states1) && !(eflags&REG_LARGE))
-		return(smatcher(g, (char *)string, nmatch, pmatch, eflags));
+		return(smatcher(g, string, nmatch, pmatch, eflags));
 	else
-		return(lmatcher(g, (char *)string, nmatch, pmatch, eflags));
+		return(lmatcher(g, string, nmatch, pmatch, eflags));
 }

winsup/cygwin/regex/regfree.c

@@ -1,4 +1,6 @@
 /*-
+ * SPDX-License-Identifier: BSD-3-Clause
+ *
  * Copyright (c) 1992, 1993, 1994 Henry Spencer.
  * Copyright (c) 1992, 1993, 1994
  *	The Regents of the University of California.  All rights reserved.
@@ -14,7 +16,7 @@
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
  *    documentation and/or other materials provided with the distribution.
- * 4. Neither the name of the University nor the names of its contributors
+ * 3. Neither the name of the University nor the names of its contributors
  *    may be used to endorse or promote products derived from this software
  *    without specific prior written permission.
  *
@@ -37,7 +39,7 @@
 static char sccsid[] = "@(#)regfree.c	8.3 (Berkeley) 3/20/94";
 #endif /* LIBC_SCCS and not lint */
 #include <sys/cdefs.h>
-__FBSDID("$FreeBSD: src/lib/libc/regex/regfree.c,v 1.8 2007/06/11 03:05:54 delphij Exp $");
+__FBSDID("$FreeBSD$");
 
 #include <sys/types.h>
 #include <stdio.h>
@@ -58,7 +60,7 @@ void
 regfree(regex_t *preg)
 {
 	struct re_guts *g;
-	int i;
+	unsigned int i;
 
 	if (preg->re_magic != MAGIC1)	/* oops */
 		return;			/* nice to complain, but hard */

winsup/cygwin/regex/utils.h

@@ -1,4 +1,6 @@
 /*-
+ * SPDX-License-Identifier: BSD-3-Clause
+ *
  * Copyright (c) 1992, 1993, 1994 Henry Spencer.
  * Copyright (c) 1992, 1993, 1994
  *	The Regents of the University of California.  All rights reserved.
@@ -14,7 +16,7 @@
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
  *    documentation and/or other materials provided with the distribution.
- * 4. Neither the name of the University nor the names of its contributors
+ * 3. Neither the name of the University nor the names of its contributors
  *    may be used to endorse or promote products derived from this software
  *    without specific prior written permission.
  *
@@ -31,13 +33,15 @@
  * SUCH DAMAGE.
  *
  *	@(#)utils.h	8.3 (Berkeley) 3/20/94
- * $FreeBSD: src/lib/libc/regex/utils.h,v 1.3 2007/01/09 00:28:04 imp Exp $
+ * $FreeBSD$
  */
 
 /* utility definitions */
 #define	DUPMAX		_POSIX2_RE_DUP_MAX	/* xxx is this right? */
 #define	INFINITY	(DUPMAX + 1)
-#define	NC		(CHAR_MAX - CHAR_MIN + 1)
+
+#define	NC_MAX		(CHAR_MAX - CHAR_MIN + 1)
+#define	NC		((MB_CUR_MAX) == 1 ? (NC_MAX) : (128))
 typedef unsigned char uch;
 
 /* switch off assertions (if not already off) if no REDEBUG */