Replace regex files with multibyte-aware version from FreeBSD.

* Makefile.in (install-headers): Remove extra command to install
	regex.h.
	(uninstall-headers): Remove extra command to uninstall regex.h.
	* nlsfuncs.cc (collate_lcid): Make externally available to allow
	access to collation internals from regex functions.
	(collate_charset): Ditto.
	* wchar.h: Add __cplusplus guards to make C-clean.
	* include/regex.h: New file, replacing regex/regex.h.  Remove UCB
	advertising clause.
	* regex/COPYRIGHT: Accommodate BSD license.  Remove UCB advertising
	clause.
	* regex/cclass.h: Remove.
	* regex/cname.h: New file from FreeBSD.
	* regex/engine.c: Ditto.
	(NONCHAR): Tweak for Cygwin.
	* regex/engine.ih: Remove.
	* regex/mkh: Remove.
	* regex/regcomp.c: New file from FreeBSD.  Tweak slightly for Cygwin.
	Import required collate internals from nlsfunc.cc.
	(p_ere_exp): Add GNU-specific \< and \> handling for word boundaries.
	(p_simp_re): Ditto.
	(__collate_range_cmp): Define.
	(p_b_term): Use Cygwin-specific collate internals.
	(findmust): Ditto.
	* regex/regcomp.ih: Remove.
	* regex/regerror.c: New file from FreeBSD.  Fix a few compiler warnings.
	* regex/regerror.ih: Remove.
	* regex/regex.7: New file from FreeBSD.  Remove UCB advertising clause.
	* regex/regex.h: Remove.  Replaced by include/regex.h.
	* regex/regexec.c: New file from FreeBSD.  Fix a few compiler warnings.
	* regex/regfree.c: New file from FreeBSD.
	* regex/tests: Remove.
	* regex/utils.h: New file from FreeBSD.

winsup/cygwin/ChangeLog

@@ -1,3 +1,40 @@
+2010-02-04  Corinna Vinschen  <corinna@vinschen.de>
+
+	Replace regex files with multibyte-aware version from FreeBSD.
+	* Makefile.in (install-headers): Remove extra command to install
+	regex.h.
+	(uninstall-headers): Remove extra command to uninstall regex.h.
+	* nlsfuncs.cc (collate_lcid): Make externally available to allow
+	access to collation internals from regex functions.
+	(collate_charset): Ditto.
+	* wchar.h: Add __cplusplus guards to make C-clean.
+	* include/regex.h: New file, replacing regex/regex.h.  Remove UCB
+	advertising clause.
+	* regex/COPYRIGHT: Accommodate BSD license.  Remove UCB advertising
+	clause.
+	* regex/cclass.h: Remove.
+	* regex/cname.h: New file from FreeBSD.
+	* regex/engine.c: Ditto.
+	(NONCHAR): Tweak for Cygwin.
+	* regex/engine.ih: Remove.
+	* regex/mkh: Remove.
+	* regex/regcomp.c: New file from FreeBSD.  Tweak slightly for Cygwin.
+	Import required collate internals from nlsfunc.cc.
+	(p_ere_exp): Add GNU-specific \< and \> handling for word boundaries.
+	(p_simp_re): Ditto.
+	(__collate_range_cmp): Define.
+	(p_b_term): Use Cygwin-specific collate internals.
+	(findmust): Ditto.
+	* regex/regcomp.ih: Remove.
+	* regex/regerror.c: New file from FreeBSD.  Fix a few compiler warnings.
+	* regex/regerror.ih: Remove.
+	* regex/regex.7: New file from FreeBSD.  Remove UCB advertising clause.
+	* regex/regex.h: Remove.  Replaced by include/regex.h.
+	* regex/regexec.c: New file from FreeBSD.  Fix a few compiler warnings.
+	* regex/regfree.c: New file from FreeBSD.
+	* regex/tests: Remove.
+	* regex/utils.h: New file from FreeBSD.
+
 2010-02-03  Christopher Faylor  <me+cygwin@cgf.cx>
 
 	* sigproc.cc (get_proc_lock): Fix error message typo.

winsup/cygwin/Makefile.in

@@ -329,7 +329,6 @@ install-headers:
 	      $(INSTALL_DATA) $$i $(DESTDIR)$(tooldir)/$$sub/`basename $$i` ; \
 	    done ; \
 	done ; \
-	$(INSTALL_DATA) regex/regex.h $(DESTDIR)$(tooldir)/include/regex.h
 
 install-man:
 	@$(MKDIRP) $(DESTDIR)$(mandir)/man2 $(DESTDIR)$(mandir)/man3 $(DESTDIR)$(mandir)/man5 $(DESTDIR)$(mandir)/man7
@@ -364,7 +363,6 @@ uninstall-headers:
 	      rm -f $(tooldir)/$$sub/`basename $$i` ; \
 	    done ; \
 	done ; \
-	rm -f $(tooldir)/include/regex.h
 
 uninstall-man:
 	cd $(srcdir); \

winsup/cygwin/include/regex.h

@@ -0,0 +1,130 @@
+/*-
+ * Copyright (c) 1992 Henry Spencer.
+ * Copyright (c) 1992, 1993
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * This code is derived from software contributed to Berkeley by
+ * Henry Spencer of the University of Toronto.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 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
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * 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
+ * SUCH DAMAGE.
+ *
+ *	@(#)regex.h	8.2 (Berkeley) 1/3/94
+ * $FreeBSD$
+ */
+
+#ifndef _REGEX_H_
+#define	_REGEX_H_
+
+#include <sys/cdefs.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;
+
+typedef struct {
+	regoff_t rm_so;		/* start of match */
+	regoff_t rm_eo;		/* end of match */
+} regmatch_t;
+
+/* regcomp() flags */
+#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
+
+/* 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
+#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_ILLSEQ	17
+#define	REG_ATOI	255	/* convert name to number (!) */
+#define	REG_ITOA	0400	/* convert number to name (!) */
+
+/* regexec() flags */
+#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 */
+
+__BEGIN_DECLS
+int	regcomp(regex_t * __restrict, const char * __restrict, int);
+size_t	regerror(int, const regex_t * __restrict, char * __restrict, size_t);
+/*
+ * XXX forth parameter should be `regmatch_t [__restrict]', but isn't because
+ * of a bug in GCC 3.2 (when -std=c99 is specified) which perceives this as a
+ * syntax error.
+ */
+int	regexec(const regex_t * __restrict, const char * __restrict, size_t,
+	    regmatch_t * __restrict, int);
+void	regfree(regex_t *);
+__END_DECLS
+
+#endif /* !_REGEX_H_ */

winsup/cygwin/nlsfuncs.cc

@@ -638,9 +638,9 @@ __set_lc_monetary_from_win (const char *name,
   return 1;
 }
 
-static LCID collate_lcid = 0;
+LCID collate_lcid = 0;
 static mbtowc_p collate_mbtowc = __ascii_mbtowc;
-static char collate_charset[ENCODING_LEN + 1] = "ASCII";
+char collate_charset[ENCODING_LEN + 1] = "ASCII";
 
 /* Called from newlib's setlocale() if category is LC_COLLATE.  Stores
    LC_COLLATE locale information.  This is subsequently accessed by the

winsup/cygwin/regex/COPYRIGHT

@@ -1,4 +1,4 @@
-Copyright 1992, 1993, 1994, 1997 Henry Spencer.  All rights reserved.
+Copyright 1992, 1993, 1994 Henry Spencer.  All rights reserved.
 This software is not subject to any license of the American Telephone
 and Telegraph Company or of the Regents of the University of California.
 
@@ -18,3 +18,35 @@ to the following restrictions:
    ever read sources, credits must appear in the documentation.
 
 4. This notice may not be removed or altered.
+
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
+/*-
+ * Copyright (c) 1994
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 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
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * 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
+ * SUCH DAMAGE.
+ *
+ *	@(#)COPYRIGHT	8.1 (Berkeley) 3/16/94
+ */

winsup/cygwin/regex/cclass.h

@@ -1,31 +0,0 @@
-/* character-class table */
-static struct cclass {
-	const char *name;
-	const char *chars;
-	const char *multis;
-} cclasses[] = {
-	{"alnum",	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\
-0123456789",				""},
-	{"alpha",	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz",
-						""},
-	{"blank",	" \t",		""},
-	{"cntrl",	"\007\b\t\n\v\f\r\1\2\3\4\5\6\16\17\20\21\22\23\24\
-	\25\26\27\30\31\32\33\34\35\36\37\177",	""},
-	{"digit",	"0123456789",	""},
-	{"graph",	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\
-	0123456789!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~",
-						""},
-	{"lower",	"abcdefghijklmnopqrstuvwxyz",
-						""},
-	{"print",	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\
-	0123456789!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~ ",
-						""},
-	{"punct",	"!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~",
-						""},
-	{"space",	"\t\n\v\f\r ",	""},
-	{"upper",	"ABCDEFGHIJKLMNOPQRSTUVWXYZ",
-						""},
-	{"xdigit",	"0123456789ABCDEFabcdef",
-						""},
-	{NULL,		0,		""}
-};

winsup/cygwin/regex/cname.h

@@ -1,82 +1,118 @@
+/*-
+ * Copyright (c) 1992, 1993, 1994 Henry Spencer.
+ * Copyright (c) 1992, 1993, 1994
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * This code is derived from software contributed to Berkeley by
+ * Henry Spencer.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 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
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * 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
+ * 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 $
+ */
+
 /* character-name table */
 static struct cname {
 	const char *name;
 	char code;
 } cnames[] = {
-	{"NUL",	'\0'},
+	{"NUL",			'\0'},
-	{"SOH",	'\001'},
+	{"SOH",			'\001'},
-	{"STX",	'\002'},
+	{"STX",			'\002'},
-	{"ETX",	'\003'},
+	{"ETX",			'\003'},
-	{"EOT",	'\004'},
+	{"EOT",			'\004'},
-	{"ENQ",	'\005'},
+	{"ENQ",			'\005'},
-	{"ACK",	'\006'},
+	{"ACK",			'\006'},
-	{"BEL",	'\007'},
+	{"BEL",			'\007'},
-	{"alert",	'\007'},
+	{"alert",		'\007'},
-	{"BS",		'\010'},
+	{"BS",			'\010'},
-	{"backspace",	'\b'},
+	{"backspace",		'\b'},
-	{"HT",		'\011'},
+	{"HT",			'\011'},
-	{"tab",		'\t'},
+	{"tab",			'\t'},
-	{"LF",		'\012'},
+	{"LF",			'\012'},
-	{"newline",	'\n'},
+	{"newline",		'\n'},
-	{"VT",		'\013'},
+	{"VT",			'\013'},
 	{"vertical-tab",	'\v'},
-	{"FF",		'\014'},
+	{"FF",			'\014'},
-	{"form-feed",	'\f'},
+	{"form-feed",		'\f'},
-	{"CR",		'\015'},
+	{"CR",			'\015'},
 	{"carriage-return",	'\r'},
-	{"SO",	'\016'},
+	{"SO",			'\016'},
-	{"SI",	'\017'},
+	{"SI",			'\017'},
-	{"DLE",	'\020'},
+	{"DLE",			'\020'},
-	{"DC1",	'\021'},
+	{"DC1",			'\021'},
-	{"DC2",	'\022'},
+	{"DC2",			'\022'},
-	{"DC3",	'\023'},
+	{"DC3",			'\023'},
-	{"DC4",	'\024'},
+	{"DC4",			'\024'},
-	{"NAK",	'\025'},
+	{"NAK",			'\025'},
-	{"SYN",	'\026'},
+	{"SYN",			'\026'},
-	{"ETB",	'\027'},
+	{"ETB",			'\027'},
-	{"CAN",	'\030'},
+	{"CAN",			'\030'},
-	{"EM",	'\031'},
+	{"EM",			'\031'},
-	{"SUB",	'\032'},
+	{"SUB",			'\032'},
-	{"ESC",	'\033'},
+	{"ESC",			'\033'},
-	{"IS4",	'\034'},
+	{"IS4",			'\034'},
-	{"FS",	'\034'},
+	{"FS",			'\034'},
-	{"IS3",	'\035'},
+	{"IS3",			'\035'},
-	{"GS",	'\035'},
+	{"GS",			'\035'},
-	{"IS2",	'\036'},
+	{"IS2",			'\036'},
-	{"RS",	'\036'},
+	{"RS",			'\036'},
-	{"IS1",	'\037'},
+	{"IS1",			'\037'},
-	{"US",	'\037'},
+	{"US",			'\037'},
 	{"space",		' '},
 	{"exclamation-mark",	'!'},
 	{"quotation-mark",	'"'},
 	{"number-sign",		'#'},
 	{"dollar-sign",		'$'},
-	{"percent-sign",		'%'},
+	{"percent-sign",	'%'},
 	{"ampersand",		'&'},
 	{"apostrophe",		'\''},
 	{"left-parenthesis",	'('},
 	{"right-parenthesis",	')'},
-	{"asterisk",	'*'},
+	{"asterisk",		'*'},
-	{"plus-sign",	'+'},
+	{"plus-sign",		'+'},
-	{"comma",	','},
+	{"comma",		','},
-	{"hyphen",	'-'},
+	{"hyphen",		'-'},
 	{"hyphen-minus",	'-'},
-	{"period",	'.'},
+	{"period",		'.'},
-	{"full-stop",	'.'},
+	{"full-stop",		'.'},
-	{"slash",	'/'},
+	{"slash",		'/'},
-	{"solidus",	'/'},
+	{"solidus",		'/'},
 	{"zero",		'0'},
-	{"one",		'1'},
+	{"one",			'1'},
-	{"two",		'2'},
+	{"two",			'2'},
-	{"three",	'3'},
+	{"three",		'3'},
 	{"four",		'4'},
 	{"five",		'5'},
-	{"six",		'6'},
+	{"six",			'6'},
-	{"seven",	'7'},
+	{"seven",      		'7'},
-	{"eight",	'8'},
+	{"eight",		'8'},
 	{"nine",		'9'},
-	{"colon",	':'},
+	{"colon",		':'},
-	{"semicolon",	';'},
+	{"semicolon",		';'},
 	{"less-than-sign",	'<'},
 	{"equals-sign",		'='},
 	{"greater-than-sign",	'>'},
@@ -85,12 +121,12 @@ static struct cname {
 	{"left-square-bracket",	'['},
 	{"backslash",		'\\'},
 	{"reverse-solidus",	'\\'},
-	{"right-square-bracket",	']'},
+	{"right-square-bracket",']'},
 	{"circumflex",		'^'},
 	{"circumflex-accent",	'^'},
 	{"underscore",		'_'},
 	{"low-line",		'_'},
-	{"grave-accent",		'`'},
+	{"grave-accent",	'`'},
 	{"left-brace",		'{'},
 	{"left-curly-bracket",	'{'},
 	{"vertical-line",	'|'},

winsup/cygwin/regex/engine.c

@@ -1,3 +1,41 @@
+/*-
+ * Copyright (c) 1992, 1993, 1994 Henry Spencer.
+ * Copyright (c) 1992, 1993, 1994
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * This code is derived from software contributed to Berkeley by
+ * Henry Spencer.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 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
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * 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
+ * SUCH DAMAGE.
+ *
+ *	@(#)engine.c	8.5 (Berkeley) 3/20/94
+ */
+
+#include <sys/cdefs.h>
+__FBSDID("$FreeBSD: src/lib/libc/regex/engine.c,v 1.23 2009/09/16 06:32:23 dds Exp $");
+
 /*
  * The matching engine and friends.  This file is #included by regexec.c
  * after suitable #defines of a variety of macros used herein, so that
@@ -27,25 +65,77 @@
 #define	at	lat
 #define	match	lmat
 #endif
+#ifdef MNAMES
+#define	matcher	mmatcher
+#define	fast	mfast
+#define	slow	mslow
+#define	dissect	mdissect
+#define	backref	mbackref
+#define	step	mstep
+#define	print	mprint
+#define	at	mat
+#define	match	mmat
+#endif
 
 /* another structure passed up and down to avoid zillions of parameters */
 struct match {
 	struct re_guts *g;
 	int eflags;
 	regmatch_t *pmatch;	/* [nsub+1] (0 element unused) */
-	char *offp;		/* offsets work from here */
+	const char *offp;	/* offsets work from here */
-	char *beginp;		/* start of string -- virtual NUL precedes */
+	const char *beginp;	/* start of string -- virtual NUL precedes */
-	char *endp;		/* end of string -- virtual NUL here */
+	const char *endp;	/* end of string -- virtual NUL here */
-	char *coldp;		/* can be no match starting before here */
+	const char *coldp;	/* can be no match starting before here */
-	char **lastpos;		/* [nplus+1] */
+	const char **lastpos;	/* [nplus+1] */
 	STATEVARS;
 	states st;		/* current states */
 	states fresh;		/* states for a fresh start */
 	states tmp;		/* temporary */
 	states empty;		/* empty set of states */
+	mbstate_t mbs;		/* multibyte conversion state */
 };
 
-#include "engine.ih"
+/* ========= begin header generated by ./mkh ========= */
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* === engine.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 *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);
+#define MAX_RECURSION	100
+#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	BADCHAR	(BOL-6)
+#ifdef __CYGWIN__
+/* In contrast to BSD, wint_t on Cygwin is unsigned.  This breaks this test,
+   unless the compared values are casted to signed. */
+#define	NONCHAR(c)	((int)(c) <= (int)OUT)
+#else
+#define	NONCHAR(c)	((c) <= OUT)
+#endif
+#ifdef REDEBUG
+static void print(struct match *m, const char *caption, states st, int ch, FILE *d);
+#endif
+#ifdef REDEBUG
+static void at(struct match *m, const char *title, const char *start, const char *stop, sopno startst, sopno stopst);
+#endif
+#ifdef REDEBUG
+static const char *pchar(int ch);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+/* ========= end header generated by ./mkh ========= */
 
 #ifdef REDEBUG
 #define	SP(t, s, c)	print(m, t, s, c, stdout)
@@ -59,26 +149,32 @@ struct match {
 
 /*
  - matcher - the actual matching engine
- == static int matcher(register struct re_guts *g, char *string, \
+ == static int matcher(struct re_guts *g, const char *string, \
  ==	size_t nmatch, regmatch_t pmatch[], int eflags);
  */
 static int			/* 0 success, REG_NOMATCH failure */
-matcher(g, string, nmatch, pmatch, eflags)
+matcher(struct re_guts *g,
-register struct re_guts *g;
+	const char *string,
-char *string;
+	size_t nmatch,
-size_t nmatch;
+	regmatch_t pmatch[],
-regmatch_t pmatch[];
+	int eflags)
-int eflags;
 {
-	register char *endp;
+	const char *endp;
-	register size_t i;
+	int i;
 	struct match mv;
-	register struct match *m = &mv;
+	struct match *m = &mv;
-	register char *dp;
+	const char *dp;
-	const register sopno gf = g->firststate+1;	/* +1 for OEND */
+	const sopno gf = g->firststate+1;	/* +1 for OEND */
-	const register sopno gl = g->laststate;
+	const sopno gl = g->laststate;
-	char *start;
+	const char *start;
-	char *stop;
+	const char *stop;
+	/* Boyer-Moore algorithms variables */
+	const char *pp;
+	int cj, mj;
+	const char *mustfirst;
+	const char *mustlast;
+	int *matchjump;
+	int *charjump;
 
 	/* simplify the situation where possible */
 	if (g->cflags&REG_NOSUB)
@@ -95,12 +191,46 @@ int eflags;
 
 	/* prescreening; this does wonders for this rather slow code */
 	if (g->must != NULL) {
-		for (dp = start; dp < stop; dp++)
+		if (g->charjump != NULL && g->matchjump != NULL) {
-			if (*dp == g->must[0] && stop - dp >= g->mlen &&
+			mustfirst = g->must;
-				memcmp(dp, g->must, (size_t)g->mlen) == 0)
+			mustlast = g->must + g->mlen - 1;
-				break;
+			charjump = g->charjump;
-		if (dp == stop)		/* we didn't find g->must */
+			matchjump = g->matchjump;
-			return(REG_NOMATCH);
+			pp = mustlast;
+			for (dp = start+g->mlen-1; dp < stop;) {
+				/* Fast skip non-matches */
+				while (dp < stop && charjump[(int)*dp])
+					dp += charjump[(int)*dp];
+
+				if (dp >= stop)
+					break;
+
+				/* Greedy matcher */
+				/* We depend on not being used for
+				 * for strings of length 1
+				 */
+				while (*--dp == *--pp && pp != mustfirst);
+
+				if (*dp == *pp)
+					break;
+
+				/* Jump to next possible match */
+				mj = matchjump[pp - mustfirst];
+				cj = charjump[(int)*dp];
+				dp += (cj < mj ? mj : cj);
+				pp = mustlast;
+			}
+			if (pp != mustfirst)
+				return(REG_NOMATCH);
+		} else {
+			for (dp = start; dp < stop; dp++)
+				if (*dp == g->must[0] &&
+				    stop - dp >= g->mlen &&
+				    memcmp(dp, g->must, (size_t)g->mlen) == 0)
+					break;
+			if (dp == stop)		/* we didn't find g->must */
+				return(REG_NOMATCH);
+		}
 	}
 
 	/* match struct setup */
@@ -117,11 +247,22 @@ int eflags;
 	SETUP(m->tmp);
 	SETUP(m->empty);
 	CLEAR(m->empty);
+	ZAPSTATE(&m->mbs);
+
+	/* Adjust start according to moffset, to speed things up */
+	if (g->moffset > -1)
+		start = ((dp - g->moffset) < start) ? start : dp - g->moffset;
+
+	SP("mloop", m->st, *start);
 
 	/* this loop does only one repetition except for backrefs */
 	for (;;) {
 		endp = fast(m, start, stop, gf, gl);
 		if (endp == NULL) {		/* a miss */
+			if (m->pmatch != NULL)
+				free((char *)m->pmatch);
+			if (m->lastpos != NULL)
+				free((char *)m->lastpos);
 			STATETEARDOWN(m);
 			return(REG_NOMATCH);
 		}
@@ -136,7 +277,8 @@ int eflags;
 			if (endp != NULL)
 				break;
 			assert(m->coldp < m->endp);
-			m->coldp++;
+			m->coldp += XMBRTOWC(NULL, m->coldp,
+			    m->endp - m->coldp, &m->mbs, 0);
 		}
 		if (nmatch == 1 && !g->backrefs)
 			break;		/* no further info needed */
@@ -156,15 +298,15 @@ int eflags;
 			dp = dissect(m, m->coldp, endp, gf, gl);
 		} else {
 			if (g->nplus > 0 && m->lastpos == NULL)
-				m->lastpos = (char **)malloc((g->nplus+1) *
+				m->lastpos = malloc((g->nplus+1) *
-							sizeof(char *));
+						sizeof(const char *));
 			if (g->nplus > 0 && m->lastpos == NULL) {
 				free(m->pmatch);
 				STATETEARDOWN(m);
 				return(REG_ESPACE);
 			}
 			NOTE("backref dissect");
-			dp = backref(m, m->coldp, endp, gf, gl, (sopno)0);
+			dp = backref(m, m->coldp, endp, gf, gl, (sopno)0, 0);
 		}
 		if (dp != NULL)
 			break;
@@ -187,7 +329,7 @@ int eflags;
 			}
 #endif
 			NOTE("backoff dissect");
-			dp = backref(m, m->coldp, endp, gf, gl, (sopno)0);
+			dp = backref(m, m->coldp, endp, gf, gl, (sopno)0, 0);
 		}
 		assert(dp == NULL || dp == endp);
 		if (dp != NULL)		/* found a shorter one */
@@ -195,7 +337,9 @@ int eflags;
 
 		/* despite initial appearances, there is no match here */
 		NOTE("false alarm");
-		start = m->coldp + 1;	/* recycle starting later */
+		/* recycle starting later */
+		start = m->coldp + XMBRTOWC(NULL, m->coldp,
+		    stop - m->coldp, &m->mbs, 0);
 		assert(start <= stop);
 	}
 
@@ -225,30 +369,29 @@ int eflags;
 
 /*
  - dissect - figure out what matched what, no back references
- == static char *dissect(register struct match *m, char *start, \
+ == static const char *dissect(struct match *m, const char *start, \
- ==	char *stop, sopno startst, sopno stopst);
+ ==	const char *stop, sopno startst, sopno stopst);
  */
-static char *			/* == stop (success) always */
+static const char *		/* == stop (success) always */
-dissect(m, start, stop, startst, stopst)
+dissect(struct match *m,
-register struct match *m;
+	const char *start,
-char *start;
+	const char *stop,
-char *stop;
+	sopno startst,
-sopno startst;
+	sopno stopst)
-sopno stopst;
 {
-	register int i;
+	int i;
-	register sopno ss;	/* start sop of current subRE */
+	sopno ss;		/* start sop of current subRE */
-	register sopno es;	/* end sop of current subRE */
+	sopno es;		/* end sop of current subRE */
-	register char *sp;	/* start of string matched by it */
+	const char *sp;		/* start of string matched by it */
-	register char *stp;	/* string matched by it cannot pass here */
+	const char *stp;	/* string matched by it cannot pass here */
-	register char *rest;	/* start of rest of string */
+	const char *rest;	/* start of rest of string */
-	register char *tail;	/* string unmatched by rest of RE */
+	const char *tail;	/* string unmatched by rest of RE */
-	register sopno ssub;	/* start sop of subsubRE */
+	sopno ssub;		/* start sop of subsubRE */
-	register sopno esub;	/* end sop of subsubRE */
+	sopno esub;		/* end sop of subsubRE */
-	register char *ssp;	/* start of string matched by subsubRE */
+	const char *ssp;	/* start of string matched by subsubRE */
-	register char *sep;	/* end of string matched by subsubRE */
+	const char *sep;	/* end of string matched by subsubRE */
-	register char *oldssp;	/* previous ssp */
+	const char *oldssp;	/* previous ssp */
-	register char *dp;
+	const char *dp;
 
 	AT("diss", start, stop, startst, stopst);
 	sp = start;
@@ -273,7 +416,7 @@ sopno stopst;
 			assert(nope);
 			break;
 		case OCHAR:
-			sp++;
+			sp += XMBRTOWC(NULL, sp, stop - start, &m->mbs, 0);
 			break;
 		case OBOL:
 		case OEOL:
@@ -282,7 +425,7 @@ sopno stopst;
 			break;
 		case OANY:
 		case OANYOF:
-			sp++;
+			sp += XMBRTOWC(NULL, sp, stop - start, &m->mbs, 0);
 			break;
 		case OBACK_:
 		case O_BACK:
@@ -413,30 +556,31 @@ sopno stopst;
 
 /*
  - backref - figure out what matched what, figuring in back references
- == static char *backref(register struct match *m, char *start, \
+ == static const char *backref(struct match *m, const char *start, \
- ==	char *stop, sopno startst, sopno stopst, sopno lev);
+ ==	const char *stop, sopno startst, sopno stopst, sopno lev);
  */
-static char *			/* == stop (success) or NULL (failure) */
+static const char *		/* == stop (success) or NULL (failure) */
-backref(m, start, stop, startst, stopst, lev)
+backref(struct match *m,
-register struct match *m;
+	const char *start,
-char *start;
+	const char *stop,
-char *stop;
+	sopno startst,
-sopno startst;
+	sopno stopst,
-sopno stopst;
+	sopno lev,		/* PLUS nesting level */
-sopno lev;			/* PLUS nesting level */
+	int rec)
 {
-	register int i;
+	int i;
-	register sopno ss;	/* start sop of current subRE */
+	sopno ss;		/* start sop of current subRE */
-	register char *sp;	/* start of string matched by it */
+	const char *sp;		/* start of string matched by it */
-	register sopno ssub;	/* start sop of subsubRE */
+	sopno ssub;		/* start sop of subsubRE */
-	register sopno esub;	/* end sop of subsubRE */
+	sopno esub;		/* end sop of subsubRE */
-	register char *ssp;	/* start of string matched by subsubRE */
+	const char *ssp;	/* start of string matched by subsubRE */
-	register char *dp;
+	const char *dp;
-	register size_t len;
+	size_t len;
-	register int hard;
+	int hard;
-	register sop s;
+	sop s;
-	register regoff_t offsave;
+	regoff_t offsave;
-	register cset *cs;
+	cset *cs;
+	wint_t wc;
 
 	AT("back", start, stop, startst, stopst);
 	sp = start;
@@ -446,17 +590,25 @@ sopno lev;			/* PLUS nesting level */
 	for (ss = startst; !hard && ss < stopst; ss++)
 		switch (OP(s = m->g->strip[ss])) {
 		case OCHAR:
-			if (sp == stop || *sp++ != (char)OPND(s))
+			if (sp == stop)
+				return(NULL);
+			sp += XMBRTOWC(&wc, sp, stop - sp, &m->mbs, BADCHAR);
+			if (wc != OPND(s))
 				return(NULL);
 			break;
 		case OANY:
 			if (sp == stop)
 				return(NULL);
-			sp++;
+			sp += XMBRTOWC(&wc, sp, stop - sp, &m->mbs, BADCHAR);
+			if (wc == BADCHAR)
+				return (NULL);
 			break;
 		case OANYOF:
+			if (sp == stop)
+				return (NULL);
 			cs = &m->g->sets[OPND(s)];
-			if (sp == stop || !CHIN(cs, *sp++))
+			sp += XMBRTOWC(&wc, sp, stop - sp, &m->mbs, BADCHAR);
+			if (wc == BADCHAR || !CHIN(cs, wc))
 				return(NULL);
 			break;
 		case OBOL:
@@ -529,6 +681,8 @@ sopno lev;			/* PLUS nesting level */
 			return(NULL);
 		assert(m->pmatch[i].rm_so != -1);
 		len = m->pmatch[i].rm_eo - m->pmatch[i].rm_so;
+		if (len == 0 && rec++ > MAX_RECURSION)
+			return(NULL);
 		assert(stop - m->beginp >= len);
 		if (sp > stop - len)
 			return(NULL);	/* not enough left to match */
@@ -537,28 +691,28 @@ sopno lev;			/* PLUS nesting level */
 			return(NULL);
 		while (m->g->strip[ss] != SOP(O_BACK, i))
 			ss++;
-		return(backref(m, sp+len, stop, ss+1, stopst, lev));
+		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);
+		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));
+		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));
+		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));
+			return(backref(m, sp, stop, ss+1, stopst, lev-1, rec));
 		/* try another pass */
 		m->lastpos[lev] = sp;
-		dp = backref(m, sp, stop, ss-OPND(s)+1, stopst, lev);
+		dp = backref(m, sp, stop, ss-OPND(s)+1, stopst, lev, rec);
 		if (dp == NULL)
-			return(backref(m, sp, stop, ss+1, stopst, lev-1));
+			return(backref(m, sp, stop, ss+1, stopst, lev-1, rec));
 		else
 			return(dp);
 		break;
@@ -567,7 +721,7 @@ sopno lev;			/* PLUS nesting level */
 		esub = ss + OPND(s) - 1;
 		assert(OP(m->g->strip[esub]) == OOR1);
 		for (;;) {	/* find first matching branch */
-			dp = backref(m, sp, stop, ssub, esub, lev);
+			dp = backref(m, sp, stop, ssub, esub, lev, rec);
 			if (dp != NULL)
 				return(dp);
 			/* that one missed, try next one */
@@ -588,7 +742,7 @@ sopno lev;			/* PLUS nesting level */
 		assert(0 < i && i <= m->g->nsub);
 		offsave = m->pmatch[i].rm_so;
 		m->pmatch[i].rm_so = sp - m->offp;
-		dp = backref(m, sp, stop, ss+1, stopst, lev);
+		dp = backref(m, sp, stop, ss+1, stopst, lev, rec);
 		if (dp != NULL)
 			return(dp);
 		m->pmatch[i].rm_so = offsave;
@@ -599,7 +753,7 @@ sopno lev;			/* PLUS nesting level */
 		assert(0 < i && i <= m->g->nsub);
 		offsave = m->pmatch[i].rm_eo;
 		m->pmatch[i].rm_eo = sp - m->offp;
-		dp = backref(m, sp, stop, ss+1, stopst, lev);
+		dp = backref(m, sp, stop, ss+1, stopst, lev, rec);
 		if (dp != NULL)
 			return(dp);
 		m->pmatch[i].rm_eo = offsave;
@@ -613,42 +767,57 @@ sopno lev;			/* PLUS nesting level */
 	/* "can't happen" */
 	assert(nope);
 	/* NOTREACHED */
-	return((char *)NULL);	/* dummy */
+	return "shut up gcc";
 }
 
 /*
  - fast - step through the string at top speed
- == static char *fast(register struct match *m, char *start, \
+ == static const char *fast(struct match *m, const char *start, \
- ==	char *stop, sopno startst, sopno stopst);
+ ==	const char *stop, sopno startst, sopno stopst);
  */
-static char *			/* where tentative match ended, or NULL */
+static const char *		/* where tentative match ended, or NULL */
-fast(m, start, stop, startst, stopst)
+fast(	struct match *m,
-register struct match *m;
+	const char *start,
-char *start;
+	const char *stop,
-char *stop;
+	sopno startst,
-sopno startst;
+	sopno stopst)
-sopno stopst;
 {
-	register states st = m->st;
+	states st = m->st;
-	register states fresh = m->fresh;
+	states fresh = m->fresh;
-	register states tmp = m->tmp;
+	states tmp = m->tmp;
-	register char *p = start;
+	const char *p = start;
-	register int c = (start == m->beginp) ? OUT : *(start-1);
+	wint_t c;
-	register int lastc;	/* previous c */
+	wint_t lastc;		/* previous c */
-	register int flagch;
+	wint_t flagch;
-	register int i;
+	int i;
-	register char *coldp;	/* last p after which no match was underway */
+	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;
-		c = (p == m->endp) ? OUT : *p;
+		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;
 
@@ -686,7 +855,7 @@ sopno stopst;
 		}
 
 		/* are we done? */
-		if (ISSET(st, stopst) || p == stop)
+		if (ISSET(st, stopst) || p == stop || clen > stop - p)
 			break;		/* NOTE BREAK OUT */
 
 		/* no, we must deal with this character */
@@ -696,39 +865,39 @@ sopno stopst;
 		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++;
+		p += clen;
 	}
 
 	assert(coldp != NULL);
 	m->coldp = coldp;
 	if (ISSET(st, stopst))
-		return(p+1);
+		return(p+XMBRTOWC(NULL, p, stop - p, &m->mbs, 0));
 	else
 		return(NULL);
 }
 
 /*
  - slow - step through the string more deliberately
- == static char *slow(register struct match *m, char *start, \
+ == static const char *slow(struct match *m, const char *start, \
- ==	char *stop, sopno startst, sopno stopst);
+ ==	const char *stop, sopno startst, sopno stopst);
  */
-static char *			/* where it ended */
+static const char *		/* where it ended */
-slow(m, start, stop, startst, stopst)
+slow(	struct match *m,
-register struct match *m;
+	const char *start,
-char *start;
+	const char *stop,
-char *stop;
+	sopno startst,
-sopno startst;
+	sopno stopst)
-sopno stopst;
 {
-	register states st = m->st;
+	states st = m->st;
-	register states empty = m->empty;
+	states empty = m->empty;
-	register states tmp = m->tmp;
+	states tmp = m->tmp;
-	register char *p = start;
+	const char *p = start;
-	register int c = (start == m->beginp) ? OUT : *(start-1);
+	wint_t c;
-	register int lastc;	/* previous c */
+	wint_t lastc;		/* previous c */
-	register int flagch;
+	wint_t flagch;
-	register int i;
+	int i;
-	register char *matchp;	/* last p at which a match ended */
+	const char *matchp;	/* last p at which a match ended */
+	size_t clen;
 
 	AT("slow", start, stop, startst, stopst);
 	CLEAR(st);
@@ -736,10 +905,24 @@ sopno stopst;
 	SP("sstart", st, *p);
 	st = step(m->g, startst, stopst, st, NOTHING, st);
 	matchp = 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;
-		c = (p == m->endp) ? OUT : *p;
+		if (p == m->endp) {
+			c = OUT;
+			clen = 0;
+		} else
+			clen = XMBRTOWC(&c, p, m->endp - p, &m->mbs, BADCHAR);
 
 		/* is there an EOL and/or BOL between lastc and c? */
 		flagch = '\0';
@@ -777,7 +960,7 @@ sopno stopst;
 		/* are we done? */
 		if (ISSET(st, stopst))
 			matchp = p;
-		if (EQ(st, empty) || p == stop)
+		if (EQ(st, empty) || p == stop || clen > stop - p)
 			break;		/* NOTE BREAK OUT */
 
 		/* no, we must deal with this character */
@@ -787,7 +970,7 @@ sopno stopst;
 		st = step(m->g, startst, stopst, tmp, c, st);
 		SP("saft", st, c);
 		assert(EQ(step(m->g, startst, stopst, st, NOTHING, st), st));
-		p++;
+		p += clen;
 	}
 
 	return(matchp);
@@ -796,33 +979,31 @@ sopno stopst;
 
 /*
  - step - map set of states reachable before char to set reachable after
- == static states step(register struct re_guts *g, sopno start, sopno stop, \
+ == static states step(struct re_guts *g, sopno start, sopno stop, \
- ==	register states bef, int ch, register states aft);
+ ==	states bef, int ch, states aft);
- == #define	BOL	(OUT+1)
+ == #define	BOL	(OUT-1)
- == #define	EOL	(BOL+1)
+ == #define	EOL	(BOL-1)
- == #define	BOLEOL	(BOL+2)
+ == #define	BOLEOL	(BOL-2)
- == #define	NOTHING	(BOL+3)
+ == #define	NOTHING	(BOL-3)
- == #define	BOW	(BOL+4)
+ == #define	BOW	(BOL-4)
- == #define	EOW	(BOL+5)
+ == #define	EOW	(BOL-5)
- == #define	CODEMAX	(BOL+5)		// highest code used
+ == #define	BADCHAR	(BOL-6)
- == #define	NONCHAR(c)	((c) > CHAR_MAX)
+ == #define	NONCHAR(c)	((c) <= OUT)
- == #define	NNONCHAR	(CODEMAX-CHAR_MAX)
  */
 static states
-step(g, start, stop, bef, ch, aft)
+step(struct re_guts *g,
-register struct re_guts *g;
+	sopno start,		/* start state within strip */
-sopno start;			/* start state within strip */
+	sopno stop,		/* state after stop state within strip */
-sopno stop;			/* state after stop state within strip */
+	states bef,		/* states reachable before */
-register states bef;		/* states reachable before */
+	wint_t ch,		/* character or NONCHAR code */
-int ch;				/* character or NONCHAR code */
+	states aft)		/* states already known reachable after */
-register states aft;		/* states already known reachable after */
 {
-	register cset *cs;
+	cset *cs;
-	register sop s;
+	sop s;
-	register sopno pc;
+	sopno pc;
-	register onestate here;		/* note, macros know this name */
+	onestate here;		/* note, macros know this name */
-	register sopno look;
+	sopno look;
-	register long i;
+	int i;
 
 	for (pc = start, INIT(here, pc); pc != stop; pc++, INC(here)) {
 		s = g->strip[pc];
@@ -832,8 +1013,8 @@ register states aft;		/* states already known reachable after */
 			break;
 		case OCHAR:
 			/* only characters can match */
-			assert(!NONCHAR(ch) || ch != (char)OPND(s));
+			assert(!NONCHAR(ch) || ch != OPND(s));
-			if (ch == (char)OPND(s))
+			if (ch == OPND(s))
 				FWD(aft, bef, 1);
 			break;
 		case OBOL:
@@ -900,7 +1081,7 @@ register states aft;		/* states already known reachable after */
 						OP(s = g->strip[pc+look]) != O_CH;
 						look += OPND(s))
 					assert(OP(s) == OOR2);
-				FWD(aft, aft, look);
+				FWD(aft, aft, look + 1);
 			}
 			break;
 		case OOR2:		/* propagate OCH_'s marking */
@@ -926,21 +1107,20 @@ register states aft;		/* states already known reachable after */
 /*
  - print - print a set of states
  == #ifdef REDEBUG
- == static void print(struct match *m, char *caption, states st, \
+ == static void print(struct match *m, const char *caption, states st, \
  ==	int ch, FILE *d);
  == #endif
  */
 static void
-print(m, caption, st, ch, d)
+print(struct match *m,
-struct match *m;
+	const char *caption,
-char *caption;
+	states st,
-states st;
+	int ch,
-int ch;
+	FILE *d)
-FILE *d;
 {
-	register struct re_guts *g = m->g;
+	struct re_guts *g = m->g;
-	register int i;
+	int i;
-	register int first = 1;
+	int first = 1;
 
 	if (!(m->eflags&REG_TRACE))
 		return;
@@ -959,18 +1139,17 @@ FILE *d;
 /*
  - at - print current situation
  == #ifdef REDEBUG
- == static void at(struct match *m, char *title, char *start, char *stop, \
+ == static void at(struct match *m, const char *title, const char *start, \
- ==						sopno startst, sopno stopst);
+ ==			 const char *stop, sopno startst, sopno stopst);
  == #endif
  */
 static void
-at(m, title, start, stop, startst, stopst)
+at(	struct match *m,
-struct match *m;
+	const char *title,
-char *title;
+	const char *start,
-char *start;
+	const char *stop,
-char *stop;
+	sopno startst,
-sopno startst;
+	sopno stopst)
-sopno stopst;
 {
 	if (!(m->eflags&REG_TRACE))
 		return;
@@ -985,7 +1164,7 @@ sopno stopst;
 /*
  - pchar - make a character printable
  == #ifdef REDEBUG
- == static char *pchar(int ch);
+ == static const char *pchar(int ch);
  == #endif
  *
  * Is this identical to regchar() over in debug.c?  Well, yes.  But a
@@ -993,13 +1172,12 @@ sopno stopst;
  * a matching debug.o, and this is convenient.  It all disappears in
  * the non-debug compilation anyway, so it doesn't matter much.
  */
-static char *			/* -> representation */
+static const char *		/* -> representation */
-pchar(ch)
+pchar(int ch)
-int ch;
 {
 	static char pbuf[10];
 
-	if (isprint(ch) || ch == ' ')
+	if (isprint((uch)ch) || ch == ' ')
 		sprintf(pbuf, "%c", ch);
 	else
 		sprintf(pbuf, "\\%o", ch);

winsup/cygwin/regex/engine.ih

@@ -1,35 +0,0 @@
-/* ========= begin header generated by ./mkh ========= */
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/* === engine.c === */
-static int matcher(register struct re_guts *g, char *string, size_t nmatch, regmatch_t pmatch[], int eflags);
-static char *dissect(register struct match *m, char *start, char *stop, sopno startst, sopno stopst);
-static char *backref(register struct match *m, char *start, char *stop, sopno startst, sopno stopst, sopno lev);
-static char *fast(register struct match *m, char *start, char *stop, sopno startst, sopno stopst);
-static char *slow(register struct match *m, char *start, char *stop, sopno startst, sopno stopst);
-static states step(register struct re_guts *g, sopno start, sopno stop, register states bef, int ch, register 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)
-#ifdef REDEBUG
-static void print(struct match *m, char *caption, states st, int ch, FILE *d);
-#endif
-#ifdef REDEBUG
-static void at(struct match *m, char *title, char *start, char *stop, sopno startst, sopno stopst);
-#endif
-#ifdef REDEBUG
-static char *pchar(int ch);
-#endif
-
-#ifdef __cplusplus
-}
-#endif
-/* ========= end header generated by ./mkh ========= */

winsup/cygwin/regex/mkh

@@ -1,76 +0,0 @@
-#! /bin/sh
-# mkh - pull headers out of C source
-PATH=/bin:/usr/bin ; export PATH
-
-# egrep pattern to pick out marked lines
-egrep='^ =([ 	]|$)'
-
-# Sed program to process marked lines into lines for the header file.
-# The markers have already been removed.  Two things are done here:  removal
-# of backslashed newlines, and some fudging of comments.  The first is done
-# because -o needs to have prototypes on one line to strip them down.
-# Getting comments into the output is tricky; we turn C++-style // comments
-# into /* */ comments, after altering any existing */'s to avoid trouble.
-peel='	/\\$/N
-	/\\\n[ 	]*/s///g
-	/\/\//s;\*/;* /;g
-	/\/\//s;//\(.*\);/*\1 */;'
-
-for a
-do
-	case "$a" in
-	-o)	# old (pre-function-prototype) compiler
-		# add code to comment out argument lists
-		peel="$peel
-			"'/^\([^#\/][^\/]*[a-zA-Z0-9_)]\)(\(.*\))/s;;\1(/*\2*/);'
-		shift
-		;;
-	-b)	# funny Berkeley __P macro
-		peel="$peel
-			"'/^\([^#\/][^\/]*[a-zA-Z0-9_)]\)(\(.*\))/s;;\1 __P((\2));'
-		shift
-		;;
-	-s)	# compiler doesn't like `static foo();'
-		# add code to get rid of the `static'
-		peel="$peel
-			"'/^static[ 	][^\/]*[a-zA-Z0-9_)](.*)/s;static.;;'
-		shift
-		;;
-	-p)	# private declarations
-		egrep='^ ==([ 	]|$)'
-		shift
-		;;
-	-i)	# wrap in #ifndef, argument is name
-		ifndef="$2"
-		shift ; shift
-		;;
-	*)	break
-		;;
-	esac
-done
-
-if test " $ifndef" != " "
-then
-	echo "#ifndef $ifndef"
-	echo "#define	$ifndef	/* never again */"
-fi
-echo "/* ========= begin header generated by $0 ========= */"
-echo '#ifdef __cplusplus'
-echo 'extern "C" {'
-echo '#endif'
-for f
-do
-	echo
-	echo "/* === $f === */"
-	egrep "$egrep" $f | sed 's/^ ==*[ 	]//;s/^ ==*$//' | sed "$peel"
-	echo
-done
-echo '#ifdef __cplusplus'
-echo '}'
-echo '#endif'
-echo "/* ========= end header generated by $0 ========= */"
-if test " $ifndef" != " "
-then
-	echo "#endif"
-fi
-exit 0

winsup/cygwin/regex/regcomp.ih

@@ -1,48 +0,0 @@
-/* ========= begin header generated by ./mkh ========= */
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/* === regcomp.c === */
-static void p_ere(register struct parse *p, int stop);
-static void p_ere_exp(register struct parse *p);
-static void p_str(register struct parse *p);
-static void p_bre(register struct parse *p, register int end1, register int end2);
-static int p_simp_re(register struct parse *p, int starordinary);
-static int p_count(register struct parse *p);
-static void p_bracket(register struct parse *p);
-static void p_b_term(register struct parse *p, register cset *cs);
-static void p_b_cclass(register struct parse *p, register cset *cs);
-static void p_b_eclass(register struct parse *p, register cset *cs);
-static char p_b_symbol(register struct parse *p);
-static char p_b_coll_elem(register struct parse *p, int endc);
-static char othercase(int ch);
-static void bothcases(register struct parse *p, int ch);
-static void ordinary(register struct parse *p, register int ch);
-static void nonnewline(register struct parse *p);
-static void repeat(register struct parse *p, sopno start, int from, int to);
-static int seterr(register struct parse *p, int e);
-static cset *allocset(register struct parse *p);
-static void freeset(register struct parse *p, register cset *cs);
-static int freezeset(register struct parse *p, register cset *cs);
-static int firstch(register struct parse *p, register cset *cs);
-static int nch(register struct parse *p, register cset *cs);
-static void mcadd(register struct parse *p, register cset *cs, register const char *cp);
-static void mcinvert(register struct parse *p, register cset *cs);
-static void mccase(register struct parse *p, register cset *cs);
-static int isinsets(register struct re_guts *g, int c);
-static int samesets(register struct re_guts *g, int c1, int c2);
-static void categorize(struct parse *p, register struct re_guts *g);
-static sopno dupl(register struct parse *p, sopno start, sopno finish);
-static void doemit(register struct parse *p, sop op, size_t opnd);
-static void doinsert(register struct parse *p, sop op, size_t opnd, sopno pos);
-static void dofwd(register struct parse *p, sopno pos, sop value);
-static void enlarge(register struct parse *p, sopno size);
-static void stripsnug(register struct parse *p, register struct re_guts *g);
-static void findmust(register struct parse *p, register struct re_guts *g);
-static sopno pluscount(register struct parse *p, register struct re_guts *g);
-
-#ifdef __cplusplus
-}
-#endif
-/* ========= end header generated by ./mkh ========= */

winsup/cygwin/regex/regerror.c

@@ -1,17 +1,66 @@
-#include "winsup.h"
+/*-
+ * Copyright (c) 1992, 1993, 1994 Henry Spencer.
+ * Copyright (c) 1992, 1993, 1994
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * This code is derived from software contributed to Berkeley by
+ * Henry Spencer.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 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
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * 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
+ * SUCH DAMAGE.
+ *
+ *	@(#)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 <sys/cdefs.h>
+__FBSDID("$FreeBSD: src/lib/libc/regex/regerror.c,v 1.11 2007/06/11 03:05:54 delphij Exp $");
+
 #include <sys/types.h>
 #include <stdio.h>
 #include <string.h>
-#include <ctype.h>
 #include <limits.h>
 #include <stdlib.h>
-#include "regex.h"
+#include <regex.h>
 
 #include "utils.h"
-#include "regerror.ih"
 
+/* ========= begin header generated by ./mkh ========= */
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* === regerror.c === */
+static char *regatoi(const regex_t *preg, char *localbuf);
+
+#ifdef __cplusplus
+}
+#endif
+/* ========= end header generated by ./mkh ========= */
 /*
- = #define	REG_OKAY	 0
  = #define	REG_NOMATCH	 1
  = #define	REG_BADPAT	 2
  = #define	REG_ECOLLATE	 3
@@ -28,15 +77,20 @@
  = #define	REG_EMPTY	14
  = #define	REG_ASSERT	15
  = #define	REG_INVARG	16
+ = #define	REG_ILLSEQ	17
  = #define	REG_ATOI	255	// convert name to number (!)
  = #define	REG_ITOA	0400	// convert number to name (!)
  */
 static struct rerr {
 	int code;
+#ifdef __CYGWIN__ /* Avoid whining compiler */
 	const char *name;
 	const char *explain;
+#else
+	char *name;
+	char *explain;
+#endif
 } rerrs[] = {
-	{REG_OKAY,	"REG_OKAY",	"no errors detected"},
 	{REG_NOMATCH,	"REG_NOMATCH",	"regexec() failed to match"},
 	{REG_BADPAT,	"REG_BADPAT",	"invalid regular expression"},
 	{REG_ECOLLATE,	"REG_ECOLLATE",	"invalid collating element"},
@@ -53,7 +107,8 @@ static struct rerr {
 	{REG_EMPTY,	"REG_EMPTY",	"empty (sub)expression"},
 	{REG_ASSERT,	"REG_ASSERT",	"\"can't happen\" -- you found a bug"},
 	{REG_INVARG,	"REG_INVARG",	"invalid argument to regex routine"},
-	{-1,		"",		"*** unknown regexp error code ***"},
+	{REG_ILLSEQ,	"REG_ILLSEQ",	"illegal byte sequence"},
+	{0,		"",		"*** unknown regexp error code ***"}
 };
 
 /*
@@ -62,27 +117,30 @@ static struct rerr {
  */
 /* ARGSUSED */
 size_t
-regerror(errcode, preg, errbuf, errbuf_size)
+regerror(int errcode,
-int errcode;
+	 const regex_t * __restrict preg,
-const regex_t *preg;
+	 char * __restrict errbuf,
-char *errbuf;
+	 size_t errbuf_size)
-size_t errbuf_size;
 {
-	register struct rerr *r;
+	struct rerr *r;
-	register size_t len;
+	size_t len;
-	register int target = errcode &~ REG_ITOA;
+	int target = errcode &~ REG_ITOA;
-	register const char *s;
+#ifdef __CYGWIN__ /* Avoid whining compiler */
+	const char *s;
+#else
+	char *s;
+#endif
 	char convbuf[50];
 
 	if (errcode == REG_ATOI)
 		s = regatoi(preg, convbuf);
 	else {
-		for (r = rerrs; r->code >= 0; r++)
+		for (r = rerrs; r->code != 0; r++)
 			if (r->code == target)
 				break;
 
 		if (errcode&REG_ITOA) {
-			if (r->code >= 0)
+			if (r->code != 0)
 				(void) strcpy(convbuf, r->name);
 			else
 				sprintf(convbuf, "REG_0x%x", target);
@@ -109,18 +167,23 @@ size_t errbuf_size;
  - regatoi - internal routine to implement REG_ATOI
  == static char *regatoi(const regex_t *preg, char *localbuf);
  */
-static const char *
+static char *
-regatoi(preg, localbuf)
+regatoi(const regex_t *preg, char *localbuf)
-const regex_t *preg;
-char *localbuf;
 {
-	register struct rerr *r;
+	struct rerr *r;
 
-	for (r = rerrs; r->code >= 0; r++)
+	for (r = rerrs; r->code != 0; r++)
 		if (strcmp(r->name, preg->re_endp) == 0)
 			break;
-	if (r->code < 0)
+	if (r->code == 0)
+#ifdef __CYGWIN__ /* Avoid whining compiler */
+	    {
+		static const char null[] = "0";
+		return null;
+	    }
+#else
 		return("0");
+#endif
 
 	sprintf(localbuf, "%d", r->code);
 	return(localbuf);

winsup/cygwin/regex/regerror.ih

@@ -1,12 +0,0 @@
-/* ========= begin header generated by ./mkh ========= */
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/* === regerror.c === */
-static const char *regatoi(const regex_t *preg, char *localbuf);
-
-#ifdef __cplusplus
-}
-#endif
-/* ========= end header generated by ./mkh ========= */

winsup/cygwin/regex/regex.7

@@ -1,151 +1,317 @@
-.TH REGEX 7 "25 Oct 1995"
+.\" Copyright (c) 1992, 1993, 1994 Henry Spencer.
-.BY "Henry Spencer"
+.\" Copyright (c) 1992, 1993, 1994
-.SH NAME
+.\"	The Regents of the University of California.  All rights reserved.
-regex \- POSIX 1003.2 regular expressions
+.\"
-.SH DESCRIPTION
+.\" This code is derived from software contributed to Berkeley by
-Regular expressions (``RE''s),
+.\" Henry Spencer.
-as defined in POSIX 1003.2, come in two forms:
+.\"
+.\" Redistribution and use in source and binary forms, with or without
+.\" modification, are permitted provided that the following conditions
+.\" are met:
+.\" 1. Redistributions of source code must retain the above copyright
+.\"    notice, this list of conditions and the following disclaimer.
+.\" 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
+.\"    may be used to endorse or promote products derived from this software
+.\"    without specific prior written permission.
+.\"
+.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+.\" 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
+.\" 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 $
+.\"
+.Dd March 20, 1994
+.Dt RE_FORMAT 7
+.Os
+.Sh NAME
+.Nm re_format
+.Nd POSIX 1003.2 regular expressions
+.Sh DESCRIPTION
+Regular expressions
+.Pq Dq RE Ns s ,
+as defined in
+.St -p1003.2 ,
+come in two forms:
 modern REs (roughly those of
-.IR egrep ;
+.Xr egrep 1 ;
-1003.2 calls these ``extended'' REs)
+1003.2 calls these
+.Dq extended
+REs)
 and obsolete REs (roughly those of
-.IR ed ;
+.Xr ed 1 ;
-1003.2 ``basic'' REs).
+1003.2
+.Dq basic
+REs).
 Obsolete REs mostly exist for backward compatibility in some old programs;
 they will be discussed at the end.
-1003.2 leaves some aspects of RE syntax and semantics open;
+.St -p1003.2
-`\(dg' marks decisions on these aspects that
+leaves some aspects of RE syntax and semantics open;
-may not be fully portable to other 1003.2 implementations.
+`\(dd' marks decisions on these aspects that
-.PP
+may not be fully portable to other
-A (modern) RE is one\(dg or more non-empty\(dg \fIbranches\fR,
+.St -p1003.2
-separated by `|'.
+implementations.
+.Pp
+A (modern) RE is one\(dd or more non-empty\(dd
+.Em branches ,
+separated by
+.Ql \&| .
 It matches anything that matches one of the branches.
-.PP
+.Pp
-A branch is one\(dg or more \fIpieces\fR, concatenated.
+A branch is one\(dd or more
+.Em pieces ,
+concatenated.
 It matches a match for the first, followed by a match for the second, etc.
-.PP
+.Pp
-A piece is an \fIatom\fR possibly followed
+A piece is an
-by a single\(dg `*', `+', `?', or \fIbound\fR.
+.Em atom
-An atom followed by `*' matches a sequence of 0 or more matches of the atom.
+possibly followed
-An atom followed by `+' matches a sequence of 1 or more matches of the atom.
+by a single\(dd
-An atom followed by `?' matches a sequence of 0 or 1 matches of the atom.
+.Ql \&* ,
-.PP
+.Ql \&+ ,
-A \fIbound\fR is `{' followed by an unsigned decimal integer,
+.Ql \&? ,
-possibly followed by `,'
+or
+.Em bound .
+An atom followed by
+.Ql \&*
+matches a sequence of 0 or more matches of the atom.
+An atom followed by
+.Ql \&+
+matches a sequence of 1 or more matches of the atom.
+An atom followed by
+.Ql ?\&
+matches a sequence of 0 or 1 matches of the atom.
+.Pp
+A
+.Em bound
+is
+.Ql \&{
+followed by an unsigned decimal integer,
+possibly followed by
+.Ql \&,
 possibly followed by another unsigned decimal integer,
-always followed by `}'.
+always followed by
-The integers must lie between 0 and RE_DUP_MAX (255\(dg) inclusive,
+.Ql \&} .
+The integers must lie between 0 and
+.Dv RE_DUP_MAX
+(255\(dd) inclusive,
 and if there are two of them, the first may not exceed the second.
-An atom followed by a bound containing one integer \fIi\fR
+An atom followed by a bound containing one integer
+.Em i
 and no comma matches
-a sequence of exactly \fIi\fR matches of the atom.
+a sequence of exactly
+.Em i
+matches of the atom.
 An atom followed by a bound
-containing one integer \fIi\fR and a comma matches
+containing one integer
-a sequence of \fIi\fR or more matches of the atom.
+.Em i
+and a comma matches
+a sequence of
+.Em i
+or more matches of the atom.
 An atom followed by a bound
-containing two integers \fIi\fR and \fIj\fR matches
+containing two integers
-a sequence of \fIi\fR through \fIj\fR (inclusive) matches of the atom.
+.Em i
-.PP
+and
-An atom is a regular expression enclosed in `()' (matching a match for the
+.Em j
+matches
+a sequence of
+.Em i
+through
+.Em j
+(inclusive) matches of the atom.
+.Pp
+An atom is a regular expression enclosed in
+.Ql ()
+(matching a match for the
 regular expression),
-an empty set of `()' (matching the null string)\(dg,
+an empty set of
-a \fIbracket expression\fR (see below), `.'
+.Ql ()
-(matching any single character), `^' (matching the null string at the
+(matching the null string)\(dd,
-beginning of a line), `$' (matching the null string at the
+a
-end of a line), a `\e' followed by one of the characters
+.Em bracket expression
-`^.[$()|*+?{\e'
+(see below),
+.Ql .\&
+(matching any single character),
+.Ql \&^
+(matching the null string at the beginning of a line),
+.Ql \&$
+(matching the null string at the end of a line), a
+.Ql \e
+followed by one of the characters
+.Ql ^.[$()|*+?{\e
 (matching that character taken as an ordinary character),
-a `\e' followed by any other character\(dg
+a
+.Ql \e
+followed by any other character\(dd
 (matching that character taken as an ordinary character,
-as if the `\e' had not been present\(dg),
+as if the
+.Ql \e
+had not been present\(dd),
 or a single character with no other significance (matching that character).
-A `{' followed by a character other than a digit is an ordinary
+A
-character, not the beginning of a bound\(dg.
+.Ql \&{
-It is illegal to end an RE with `\e'.
+followed by a character other than a digit is an ordinary
-.PP
+character, not the beginning of a bound\(dd.
-A \fIbracket expression\fR is a list of characters enclosed in `[]'.
+It is illegal to end an RE with
+.Ql \e .
+.Pp
+A
+.Em bracket expression
+is a list of characters enclosed in
+.Ql [] .
 It normally matches any single character from the list (but see below).
-If the list begins with `^',
+If the list begins with
+.Ql \&^ ,
 it matches any single character
-(but see below) \fInot\fR from the rest of the list.
+(but see below)
-If two characters in the list are separated by `\-', this is shorthand
+.Em not
-for the full \fIrange\fR of characters between those two (inclusive) in the
+from the rest of the list.
+If two characters in the list are separated by
+.Ql \&- ,
+this is shorthand
+for the full
+.Em range
+of characters between those two (inclusive) in the
 collating sequence,
-e.g. `[0\-9]' in ASCII matches any decimal digit.
+.No e.g. Ql [0-9]
-It is illegal\(dg for two ranges to share an
+in ASCII matches any decimal digit.
-endpoint, e.g. `a\-c\-e'.
+It is illegal\(dd for two ranges to share an
+endpoint,
+.No e.g. Ql a-c-e .
 Ranges are very collating-sequence-dependent,
 and portable programs should avoid relying on them.
-.PP
+.Pp
-To include a literal `]' in the list, make it the first character
+To include a literal
-(following a possible `^').
+.Ql \&]
-To include a literal `\-', make it the first or last character,
+in the list, make it the first character
+(following a possible
+.Ql \&^ ) .
+To include a literal
+.Ql \&- ,
+make it the first or last character,
 or the second endpoint of a range.
-To use a literal `\-' as the first endpoint of a range,
+To use a literal
-enclose it in `[.' and `.]' to make it a collating element (see below).
+.Ql \&-
-With the exception of these and some combinations using `[' (see next
+as the first endpoint of a range,
-paragraphs), all other special characters, including `\e', lose their
+enclose it in
-special significance within a bracket expression.
+.Ql [.\&
-.PP
+and
+.Ql .]\&
+to make it a collating element (see below).
+With the exception of these and some combinations using
+.Ql \&[
+(see next paragraphs), all other special characters, including
+.Ql \e ,
+lose their special significance within a bracket expression.
+.Pp
 Within a bracket expression, a collating element (a character,
 a multi-character sequence that collates as if it were a single character,
 or a collating-sequence name for either)
-enclosed in `[.' and `.]' stands for the
+enclosed in
+.Ql [.\&
+and
+.Ql .]\&
+stands for the
 sequence of characters of that collating element.
 The sequence is a single element of the bracket expression's list.
 A bracket expression containing a multi-character collating element
 can thus match more than one character,
-e.g. if the collating sequence includes a `ch' collating element,
+e.g.\& if the collating sequence includes a
-then the RE `[[.ch.]]*c' matches the first five characters
+.Ql ch
-of `chchcc'.
+collating element,
-.PP
+then the RE
-Within a bracket expression, a collating element enclosed in `[=' and
+.Ql [[.ch.]]*c
-`=]' is an equivalence class, standing for the sequences of characters
+matches the first five characters
+of
+.Ql chchcc .
+.Pp
+Within a bracket expression, a collating element enclosed in
+.Ql [=
+and
+.Ql =]
+is an equivalence class, standing for the sequences of characters
 of all collating elements equivalent to that one, including itself.
 (If there are no other equivalent collating elements,
-the treatment is as if the enclosing delimiters were `[.' and `.]'.)
+the treatment is as if the enclosing delimiters were
-For example, if o and \o'o^' are the members of an equivalence class,
+.Ql [.\&
-then `[[=o=]]', `[[=\o'o^'=]]', and `[o\o'o^']' are all synonymous.
+and
-An equivalence class may not\(dg be an endpoint
+.Ql .] . )
+For example, if
+.Ql x
+and
+.Ql y
+are the members of an equivalence class,
+then
+.Ql [[=x=]] ,
+.Ql [[=y=]] ,
+and
+.Ql [xy]
+are all synonymous.
+An equivalence class may not\(dd be an endpoint
 of a range.
-.PP
+.Pp
-Within a bracket expression, the name of a \fIcharacter class\fR enclosed
+Within a bracket expression, the name of a
-in `[:' and `:]' stands for the list of all characters belonging to that
+.Em character class
+enclosed in
+.Ql [:
+and
+.Ql :]
+stands for the list of all characters belonging to that
 class.
 Standard character class names are:
-.PP
+.Pp
-.RS
+.Bl -column "alnum" "digit" "xdigit" -offset indent
-.nf
+.It Em "alnum	digit	punct"
-.ta 3c 6c 9c
+.It Em "alpha	graph	space"
-alnum	digit	punct
+.It Em "blank	lower	upper"
-alpha	graph	space
+.It Em "cntrl	print	xdigit"
-blank	lower	upper
+.El
-cntrl	print	xdigit
+.Pp
-.fi
-.RE
-.PP
 These stand for the character classes defined in
-.IR ctype (3).
+.Xr ctype 3 .
 A locale may provide others.
 A character class may not be used as an endpoint of a range.
-.PP
+.Pp
-There are two special cases\(dg of bracket expressions:
+A bracketed expression like
-the bracket expressions `[[:<:]]' and `[[:>:]]' match the null string at
+.Ql [[:class:]]
-the beginning and end of a word respectively.
+can be used to match a single character that belongs to a character
-A word is defined as a sequence of
+class.
-word characters
+The reverse, matching any character that does not belong to a specific
+class, the negation operator of bracket expressions may be used:
+.Ql [^[:class:]] .
+.Pp
+There are two special cases\(dd of bracket expressions:
+the bracket expressions
+.Ql [[:<:]]
+and
+.Ql [[:>:]]
+match the null string at the beginning and end of a word respectively.
+A word is defined as a sequence of word characters
 which is neither preceded nor followed by
 word characters.
 A word character is an
-.I alnum
+.Em alnum
 character (as defined by
-.IR ctype (3))
+.Xr ctype 3 )
 or an underscore.
 This is an extension,
-compatible with but not specified by POSIX 1003.2,
+compatible with but not specified by
+.St -p1003.2 ,
 and should be used with
 caution in software intended to be portable to other systems.
-.PP
+.Pp
 In the event that an RE could match more than one substring of a given
 string,
 the RE matches the one starting earliest in the string.
@@ -157,79 +323,158 @@ with subexpressions starting earlier in the RE taking priority over
 ones starting later.
 Note that higher-level subexpressions thus take priority over
 their lower-level component subexpressions.
-.PP
+.Pp
 Match lengths are measured in characters, not collating elements.
 A null string is considered longer than no match at all.
 For example,
-`bb*' matches the three middle characters of `abbbc',
+.Ql bb*
-`(wee|week)(knights|nights)' matches all ten characters of `weeknights',
+matches the three middle characters of
-when `(.*).*' is matched against `abc' the parenthesized subexpression
+.Ql abbbc ,
+.Ql (wee|week)(knights|nights)
+matches all ten characters of
+.Ql weeknights ,
+when
+.Ql (.*).*\&
+is matched against
+.Ql abc
+the parenthesized subexpression
 matches all three characters, and
-when `(a*)*' is matched against `bc' both the whole RE and the parenthesized
+when
+.Ql (a*)*
+is matched against
+.Ql bc
+both the whole RE and the parenthesized
 subexpression match the null string.
-.PP
+.Pp
 If case-independent matching is specified,
 the effect is much as if all case distinctions had vanished from the
 alphabet.
 When an alphabetic that exists in multiple cases appears as an
 ordinary character outside a bracket expression, it is effectively
 transformed into a bracket expression containing both cases,
-e.g. `x' becomes `[xX]'.
+.No e.g. Ql x
+becomes
+.Ql [xX] .
 When it appears inside a bracket expression, all case counterparts
-of it are added to the bracket expression, so that (e.g.) `[x]'
+of it are added to the bracket expression, so that (e.g.)
-becomes `[xX]' and `[^x]' becomes `[^xX]'.
+.Ql [x]
-.PP
+becomes
-No particular limit is imposed on the length of REs\(dg.
+.Ql [xX]
+and
+.Ql [^x]
+becomes
+.Ql [^xX] .
+.Pp
+No particular limit is imposed on the length of REs\(dd.
 Programs intended to be portable should not employ REs longer
 than 256 bytes,
 as an implementation can refuse to accept such REs and remain
 POSIX-compliant.
-.PP
+.Pp
-Obsolete (``basic'') regular expressions differ in several respects.
+Obsolete
-`|', `+', and `?' are ordinary characters and there is no equivalent
+.Pq Dq basic
-for their functionality.
+regular expressions differ in several respects.
-The delimiters for bounds are `\e{' and `\e}',
+.Ql \&|
-with `{' and `}' by themselves ordinary characters.
+is an ordinary character and there is no equivalent
-The parentheses for nested subexpressions are `\e(' and `\e)',
+for its functionality.
-with `(' and `)' by themselves ordinary characters.
+.Ql \&+
-`^' is an ordinary character except at the beginning of the
+and
-RE or\(dg the beginning of a parenthesized subexpression,
+.Ql ?\&
-`$' is an ordinary character except at the end of the
+are ordinary characters, and their functionality
-RE or\(dg the end of a parenthesized subexpression,
+can be expressed using bounds
-and `*' is an ordinary character if it appears at the beginning of the
+.No ( Ql {1,}
+or
+.Ql {0,1}
+respectively).
+Also note that
+.Ql x+
+in modern REs is equivalent to
+.Ql xx* .
+The delimiters for bounds are
+.Ql \e{
+and
+.Ql \e} ,
+with
+.Ql \&{
+and
+.Ql \&}
+by themselves ordinary characters.
+The parentheses for nested subexpressions are
+.Ql \e(
+and
+.Ql \e) ,
+with
+.Ql \&(
+and
+.Ql \&)
+by themselves ordinary characters.
+.Ql \&^
+is an ordinary character except at the beginning of the
+RE or\(dd the beginning of a parenthesized subexpression,
+.Ql \&$
+is an ordinary character except at the end of the
+RE or\(dd the end of a parenthesized subexpression,
+and
+.Ql \&*
+is an ordinary character if it appears at the beginning of the
 RE or the beginning of a parenthesized subexpression
-(after a possible leading `^').
+(after a possible leading
-Finally, there is one new type of atom, a \fIback reference\fR:
+.Ql \&^ ) .
-`\e' followed by a non-zero decimal digit \fId\fR
+Finally, there is one new type of atom, a
+.Em back reference :
+.Ql \e
+followed by a non-zero decimal digit
+.Em d
 matches the same sequence of characters
-matched by the \fId\fRth parenthesized subexpression
+matched by the
+.Em d Ns th
+parenthesized subexpression
 (numbering subexpressions by the positions of their opening parentheses,
 left to right),
-so that (e.g.) `\e([bc]\e)\e1' matches `bb' or `cc' but not `bc'.
+so that (e.g.)
-.SH SEE ALSO
+.Ql \e([bc]\e)\e1
-regex(3)
+matches
-.PP
+.Ql bb
-POSIX 1003.2, section 2.8 (Regular Expression Notation).
+or
-.SH HISTORY
+.Ql cc
-Written by Henry Spencer, based on the 1003.2 spec.
+but not
-.SH BUGS
+.Ql bc .
+.Sh SEE ALSO
+.Xr regex 3
+.Rs
+.%T Regular Expression Notation
+.%R IEEE Std
+.%N 1003.2
+.%P section 2.8
+.Re
+.Sh BUGS
 Having two kinds of REs is a botch.
-.PP
+.Pp
-The current 1003.2 spec says that `)' is an ordinary character in
+The current
-the absence of an unmatched `(';
+.St -p1003.2
+spec says that
+.Ql \&)
+is an ordinary character in
+the absence of an unmatched
+.Ql \&( ;
 this was an unintentional result of a wording error,
 and change is likely.
 Avoid relying on it.
-.PP
+.Pp
 Back references are a dreadful botch,
 posing major problems for efficient implementations.
 They are also somewhat vaguely defined
 (does
-`a\e(\e(b\e)*\e2\e)*d' match `abbbd'?).
+.Ql a\e(\e(b\e)*\e2\e)*d
+match
+.Ql abbbd ? ) .
 Avoid using them.
-.PP
+.Pp
-1003.2's specification of case-independent matching is vague.
+.St -p1003.2
-The ``one case implies all cases'' definition given above
+specification of case-independent matching is vague.
+The
+.Dq one case implies all cases
+definition given above
 is current consensus among implementors as to the right interpretation.
-.PP
+.Pp
 The syntax for word boundaries is incredibly ugly.

winsup/cygwin/regex/regex.h

@@ -1,76 +0,0 @@
-#ifndef _REGEX_H_
-#define	_REGEX_H_	/* never again */
-#include <sys/types.h>
-/* ========= begin header generated by ./mkh ========= */
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/* === regex2.h === */
-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;
-
-
-/* === regcomp.c === */
-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
-
-
-/* === regerror.c === */
-#define	REG_OKAY	 0
-#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 (!) */
-extern size_t regerror(int, const regex_t *, char *, size_t);
-
-
-/* === regexec.c === */
-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 */
-
-
-/* === regfree.c === */
-extern void regfree(regex_t *);
-
-
-#ifdef __cplusplus
-}
-#endif
-/* ========= end header generated by ./mkh ========= */
-#endif

winsup/cygwin/regex/regex2.h

@@ -1,6 +1,42 @@
+/*-
+ * Copyright (c) 1992, 1993, 1994 Henry Spencer.
+ * Copyright (c) 1992, 1993, 1994
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * This code is derived from software contributed to Berkeley by
+ * Henry Spencer.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 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
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * 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
+ * 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 $
+ */
+
 /*
  * First, the stuff that ends up in the outside-world include file
- = typedef _off_t regoff_t;
+ = typedef off_t regoff_t;
  = typedef struct {
  = 	int re_magic;
  = 	size_t re_nsub;		// number of parenthesized subexpressions
@@ -36,66 +72,91 @@
  * In state representations, an operator's bit is on to signify a state
  * immediately *preceding* "execution" of that operator.
  */
-typedef long sop;		/* strip operator */
+typedef unsigned long sop;	/* strip operator */
 typedef long sopno;
-#define	OPRMASK	0x7c000000
+#define	OPRMASK	0xf8000000L
-#define	OPDMASK	0x03ffffff
+#define	OPDMASK	0x07ffffffL
-#define	OPSHIFT	(26)
+#define	OPSHIFT	((unsigned)27)
 #define	OP(n)	((n)&OPRMASK)
 #define	OPND(n)	((n)&OPDMASK)
 #define	SOP(op, opnd)	((op)|(opnd))
 /* operators			   meaning	operand			*/
 /*						(back, fwd are offsets)	*/
-#define	OEND	(1<<OPSHIFT)	/* endmarker	-			*/
+#define	OEND	(1L<<OPSHIFT)	/* endmarker	-			*/
-#define	OCHAR	(2<<OPSHIFT)	/* character	unsigned char		*/
+#define	OCHAR	(2L<<OPSHIFT)	/* character	wide character		*/
-#define	OBOL	(3<<OPSHIFT)	/* left anchor	-			*/
+#define	OBOL	(3L<<OPSHIFT)	/* left anchor	-			*/
-#define	OEOL	(4<<OPSHIFT)	/* right anchor	-			*/
+#define	OEOL	(4L<<OPSHIFT)	/* right anchor	-			*/
-#define	OANY	(5<<OPSHIFT)	/* .		-			*/
+#define	OANY	(5L<<OPSHIFT)	/* .		-			*/
-#define	OANYOF	(6<<OPSHIFT)	/* [...]	set number		*/
+#define	OANYOF	(6L<<OPSHIFT)	/* [...]	set number		*/
-#define	OBACK_	(7<<OPSHIFT)	/* begin \d	paren number		*/
+#define	OBACK_	(7L<<OPSHIFT)	/* begin \d	paren number		*/
-#define	O_BACK	(8<<OPSHIFT)	/* end \d	paren number		*/
+#define	O_BACK	(8L<<OPSHIFT)	/* end \d	paren number		*/
-#define	OPLUS_	(9<<OPSHIFT)	/* + prefix	fwd to suffix		*/
+#define	OPLUS_	(9L<<OPSHIFT)	/* + prefix	fwd to suffix		*/
-#define	O_PLUS	(10<<OPSHIFT)	/* + suffix	back to prefix		*/
+#define	O_PLUS	(10L<<OPSHIFT)	/* + suffix	back to prefix		*/
-#define	OQUEST_	(11<<OPSHIFT)	/* ? prefix	fwd to suffix		*/
+#define	OQUEST_	(11L<<OPSHIFT)	/* ? prefix	fwd to suffix		*/
-#define	O_QUEST	(12<<OPSHIFT)	/* ? suffix	back to prefix		*/
+#define	O_QUEST	(12L<<OPSHIFT)	/* ? suffix	back to prefix		*/
-#define	OLPAREN	(13<<OPSHIFT)	/* (		fwd to )		*/
+#define	OLPAREN	(13L<<OPSHIFT)	/* (		fwd to )		*/
-#define	ORPAREN	(14<<OPSHIFT)	/* )		back to (		*/
+#define	ORPAREN	(14L<<OPSHIFT)	/* )		back to (		*/
-#define	OCH_	(15<<OPSHIFT)	/* begin choice	fwd to OOR2		*/
+#define	OCH_	(15L<<OPSHIFT)	/* begin choice	fwd to OOR2		*/
-#define	OOR1	(16<<OPSHIFT)	/* | pt. 1	back to OOR1 or OCH_	*/
+#define	OOR1	(16L<<OPSHIFT)	/* | pt. 1	back to OOR1 or OCH_	*/
-#define	OOR2	(17<<OPSHIFT)	/* | pt. 2	fwd to OOR2 or O_CH	*/
+#define	OOR2	(17L<<OPSHIFT)	/* | pt. 2	fwd to OOR2 or O_CH	*/
-#define	O_CH	(18<<OPSHIFT)	/* end choice	back to OOR1		*/
+#define	O_CH	(18L<<OPSHIFT)	/* end choice	back to OOR1		*/
-#define	OBOW	(19<<OPSHIFT)	/* begin word	-			*/
+#define	OBOW	(19L<<OPSHIFT)	/* begin word	-			*/
-#define	OEOW	(20<<OPSHIFT)	/* end word	-			*/
+#define	OEOW	(20L<<OPSHIFT)	/* end word	-			*/
 
 /*
- * Structure for [] character-set representation.  Character sets are
+ * Structures for [] character-set representation.
- * done as bit vectors, grouped 8 to a byte vector for compactness.
- * The individual set therefore has both a pointer to the byte vector
- * and a mask to pick out the relevant bit of each byte.  A hash code
- * simplifies testing whether two sets could be identical.
- *
- * This will get trickier for multicharacter collating elements.  As
- * preliminary hooks for dealing with such things, we also carry along
- * a string of multi-character elements, and decide the size of the
- * vectors at run time.
  */
 typedef struct {
-	uch *ptr;		/* -> uch [csetsize] */
+	wint_t		min;
-	uch mask;		/* bit within array */
+	wint_t		max;
-	uch hash;		/* hash code */
+} crange;
-	size_t smultis;
+typedef struct {
-	char *multis;		/* -> char[smulti]  ab\0cd\0ef\0\0 */
+	unsigned char	bmp[NC / 8];
+	wctype_t	*types;
+	int		ntypes;
+	wint_t		*wides;
+	int		nwides;
+	crange		*ranges;
+	int		nranges;
+	int		invert;
+	int		icase;
 } cset;
-/* note that CHadd and CHsub are unsafe, and CHIN doesn't yield 0/1 */
-#define	CHadd(cs, c)	((cs)->ptr[(uch)(c)] |= (cs)->mask, (cs)->hash += (c))
-#define	CHsub(cs, c)	((cs)->ptr[(uch)(c)] &= ~(cs)->mask, (cs)->hash -= (c))
-#define	CHIN(cs, c)	((cs)->ptr[(uch)(c)] & (cs)->mask)
-#define	MCadd(p, cs, cp)	mcadd(p, cs, cp)	/* regcomp() internal fns */
-#define	MCsub(p, cs, cp)	mcsub(p, cs, cp)
-#define	MCin(p, cs, cp)	mcin(p, cs, cp)
 
-/* stuff for character categories */
+static int
-typedef unsigned char cat_t;
+CHIN1(cset *cs, wint_t ch)
+{
+	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++)
+		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);
+	for (i = 0; i < cs->ntypes; i++)
+		if (iswctype(ch, cs->types[i]))
+			return (!cs->invert);
+	return (cs->invert);
+}
+
+static __inline int
+CHIN(cset *cs, wint_t ch)
+{
+
+	assert(ch >= 0);
+	if (ch < NC)
+		return (((cs->bmp[ch >> 3] & (1 << (ch & 7))) != 0) ^
+		    cs->invert);
+	else if (cs->icase)
+		return (CHIN1(cs, ch) || CHIN1(cs, towlower(ch)) ||
+		    CHIN1(cs, towupper(ch)));
+	else
+		return (CHIN1(cs, ch));
+}
 
 /*
  * main compiled-expression structure
@@ -104,10 +165,8 @@ struct re_guts {
 	int magic;
 #		define	MAGIC2	((('R'^0200)<<8)|'E')
 	sop *strip;		/* malloced area for strip */
-	int csetsize;		/* number of bits in a cset vector */
 	int ncsets;		/* number of csets in use */
 	cset *sets;		/* -> cset [ncsets] */
-	uch *setbits;		/* -> uch[csetsize][ncsets/CHAR_BIT] */
 	int cflags;		/* copy of regcomp() cflags argument */
 	sopno nstates;		/* = number of sops */
 	sopno firststate;	/* the initial OEND (normally 0) */
@@ -118,17 +177,16 @@ struct re_guts {
 #		define	BAD	04	/* something wrong */
 	int nbol;		/* number of ^ used */
 	int neol;		/* number of $ used */
-	int ncategories;	/* how many character categories */
-	cat_t *categories;	/* ->catspace[-CHAR_MIN] */
 	char *must;		/* match must contain this string */
+	int moffset;		/* latest point at which must may be located */
+	int *charjump;		/* Boyer-Moore char jump table */
+	int *matchjump;		/* Boyer-Moore match jump table */
 	int mlen;		/* length of must */
 	size_t nsub;		/* copy of re_nsub */
 	int backrefs;		/* does it use back references? */
 	sopno nplus;		/* how deep does it nest +s? */
-	/* catspace must be last */
-	cat_t catspace[1];	/* actually [NC] */
 };
 
 /* misc utilities */
-#define	OUT	(CHAR_MAX+1)	/* a non-character value */
+#define	OUT	(CHAR_MIN - 1)	/* a non-character value */
-#define	ISWORD(c)	(isalnum((unsigned char)c) || (c) == '_')
+#define ISWORD(c)       (iswalnum((uch)(c)) || (c) == '_')

winsup/cygwin/regex/regexec.c

@@ -1,48 +1,131 @@
+/*-
+ * Copyright (c) 1992, 1993, 1994 Henry Spencer.
+ * Copyright (c) 1992, 1993, 1994
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * This code is derived from software contributed to Berkeley by
+ * Henry Spencer.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 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
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * 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
+ * SUCH DAMAGE.
+ *
+ *	@(#)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 <sys/cdefs.h>
+__FBSDID("$FreeBSD: src/lib/libc/regex/regexec.c,v 1.8 2007/06/11 03:05:54 delphij Exp $");
+
 /*
  * the outer shell of regexec()
  *
- * This file includes engine.c *twice*, after muchos fiddling with the
+ * This file includes engine.c three times, after muchos fiddling with the
  * macros that code uses.  This lets the same code operate on two different
- * representations for state sets.
+ * representations for state sets and characters.
  */
+#ifdef __CYGWIN__
 #include "winsup.h"
+#endif
 #include <sys/types.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
 #include <limits.h>
 #include <ctype.h>
-#include "regex.h"
+#include <regex.h>
+#include <wchar.h>
+#include <wctype.h>
 
 #include "utils.h"
 #include "regex2.h"
 
-#ifdef lint
+#ifdef __CYGWIN__
-static int nope = 0;		/* for use in asserts; shuts lint up */
+#define __unused	__attribute__ ((unused))
 #endif
 
+static int nope __unused = 0;	/* for use in asserts; shuts lint up */
+
+static __inline size_t
+xmbrtowc(wint_t *wi, const char *s, size_t n, mbstate_t *mbs, wint_t dummy)
+{
+	size_t nr;
+	wchar_t wc;
+
+	nr = mbrtowc(&wc, s, n, mbs);
+	if (wi != NULL)
+		*wi = wc;
+	if (nr == 0)
+		return (1);
+	else if (nr == (size_t)-1 || nr == (size_t)-2) {
+		memset(mbs, 0, sizeof(*mbs));
+		if (wi != NULL)
+			*wi = dummy;
+		return (1);
+	} else
+                return (nr);
+}
+
+static __inline size_t
+xmbrtowc_dummy(wint_t *wi,
+		const char *s,
+		size_t n __unused,
+		mbstate_t *mbs __unused,
+		wint_t dummy __unused)
+{
+
+	if (wi != NULL)
+		*wi = (unsigned char)*s;
+	return (1);
+}
+
 /* macros for manipulating states, small version */
-#define	states	unsigned
+#define	states	long
-#define	states1	unsigned	/* for later use in regexec() decision */
+#define	states1	states		/* for later use in regexec() decision */
 #define	CLEAR(v)	((v) = 0)
-#define	SET0(v, n)	((v) &= ~((unsigned)1 << (n)))
+#define	SET0(v, n)	((v) &= ~((unsigned long)1 << (n)))
-#define	SET1(v, n)	((v) |= (unsigned)1 << (n))
+#define	SET1(v, n)	((v) |= (unsigned long)1 << (n))
-#define	ISSET(v, n)	((v) & ((unsigned)1 << (n)))
+#define	ISSET(v, n)	(((v) & ((unsigned long)1 << (n))) != 0)
 #define	ASSIGN(d, s)	((d) = (s))
 #define	EQ(a, b)	((a) == (b))
-#define	STATEVARS	int dummy	/* dummy version */
+#define	STATEVARS	long dummy	/* dummy version */
 #define	STATESETUP(m, n)	/* nothing */
 #define	STATETEARDOWN(m)	/* nothing */
 #define	SETUP(v)	((v) = 0)
-#define	onestate	unsigned
+#define	onestate	long
-#define	INIT(o, n)	((o) = (unsigned)1 << (n))
+#define	INIT(o, n)	((o) = (unsigned long)1 << (n))
 #define	INC(o)	((o) <<= 1)
-#define	ISSTATEIN(v, o)	((v) & (o))
+#define	ISSTATEIN(v, o)	(((v) & (o)) != 0)
 /* 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	FWD(dst, src, n)	((dst) |= ((unsigned long)(src)&(here)) << (n))
-#define	BACK(dst, src, n)	((dst) |= ((unsigned)(src)&(here)) >> (n))
+#define	BACK(dst, src, n)	((dst) |= ((unsigned long)(src)&(here)) >> (n))
-#define	ISSETBACK(v, n)	((v) & ((unsigned)here >> (n)))
+#define	ISSETBACK(v, n)	(((v) & ((unsigned long)here >> (n))) != 0)
+/* no multibyte support */
+#define	XMBRTOWC	xmbrtowc_dummy
+#define	ZAPSTATE(mbs)	((void)(mbs))
 /* function names */
 #define SNAMES			/* engine.c looks after details */
 
@@ -68,6 +151,8 @@ static int nope = 0;		/* for use in asserts; shuts lint up */
 #undef	BACK
 #undef	ISSETBACK
 #undef	SNAMES
+#undef	XMBRTOWC
+#undef	ZAPSTATE
 
 /* macros for manipulating states, large version */
 #define	states	char *
@@ -77,13 +162,13 @@ static int nope = 0;		/* for use in asserts; shuts lint up */
 #define	ISSET(v, n)	((v)[n])
 #define	ASSIGN(d, s)	memcpy(d, s, m->g->nstates)
 #define	EQ(a, b)	(memcmp(a, b, m->g->nstates) == 0)
-#define	STATEVARS	int vn; char *space
+#define	STATEVARS	long vn; char *space
 #define	STATESETUP(m, nv)	{ (m)->space = malloc((nv)*(m)->g->nstates); \
 				if ((m)->space == NULL) return(REG_ESPACE); \
 				(m)->vn = 0; }
 #define	STATETEARDOWN(m)	{ free((m)->space); }
 #define	SETUP(v)	((v) = &m->space[m->vn++ * m->g->nstates])
-#define	onestate	int
+#define	onestate	long
 #define	INIT(o, n)	((o) = (n))
 #define	INC(o)	((o)++)
 #define	ISSTATEIN(v, o)	((v)[o])
@@ -92,11 +177,24 @@ static int nope = 0;		/* for use in asserts; shuts lint up */
 #define	FWD(dst, src, n)	((dst)[here+(n)] |= (src)[here])
 #define	BACK(dst, src, n)	((dst)[here-(n)] |= (src)[here])
 #define	ISSETBACK(v, n)	((v)[here - (n)])
+/* no multibyte support */
+#define	XMBRTOWC	xmbrtowc_dummy
+#define	ZAPSTATE(mbs)	((void)(mbs))
 /* function names */
 #define	LNAMES			/* flag */
 
 #include "engine.c"
 
+/* multibyte character & large states version */
+#undef	LNAMES
+#undef	XMBRTOWC
+#undef	ZAPSTATE
+#define	XMBRTOWC	xmbrtowc
+#define	ZAPSTATE(mbs)	memset((mbs), 0, sizeof(*(mbs)))
+#define	MNAMES
+
+#include "engine.c"
+
 /*
  - regexec - interface for matching
  = extern int regexec(const regex_t *, const char *, size_t, \
@@ -113,14 +211,13 @@ static int nope = 0;		/* for use in asserts; shuts lint up */
  * have been prototyped.
  */
 int				/* 0 success, REG_NOMATCH failure */
-regexec(preg, string, nmatch, pmatch, eflags)
+regexec(const regex_t * __restrict preg,
-const regex_t *preg;
+	const char * __restrict string,
-const char *string;
+	size_t nmatch,
-size_t nmatch;
+	regmatch_t pmatch[__restrict],
-regmatch_t pmatch[];
+	int eflags)
-int eflags;
 {
-	register struct re_guts *g = preg->re_g;
+	struct re_guts *g = preg->re_g;
 #ifdef REDEBUG
 #	define	GOODFLAGS(f)	(f)
 #else
@@ -134,7 +231,9 @@ int eflags;
 		return(REG_BADPAT);
 	eflags = GOODFLAGS(eflags);
 
-	if ((unsigned) g->nstates <= CHAR_BIT*sizeof(states1) && !(eflags&REG_LARGE))
+	if (MB_CUR_MAX > 1)
+		return(mmatcher(g, (char *)string, nmatch, pmatch, eflags));
+	else if (g->nstates <= CHAR_BIT*sizeof(states1) && !(eflags&REG_LARGE))
 		return(smatcher(g, (char *)string, nmatch, pmatch, eflags));
 	else
 		return(lmatcher(g, (char *)string, nmatch, pmatch, eflags));

winsup/cygwin/regex/regfree.c

@@ -1,8 +1,51 @@
-#include "winsup.h"
+/*-
+ * Copyright (c) 1992, 1993, 1994 Henry Spencer.
+ * Copyright (c) 1992, 1993, 1994
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * This code is derived from software contributed to Berkeley by
+ * Henry Spencer.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 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
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * 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
+ * SUCH DAMAGE.
+ *
+ *	@(#)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 <sys/cdefs.h>
+__FBSDID("$FreeBSD: src/lib/libc/regex/regfree.c,v 1.8 2007/06/11 03:05:54 delphij Exp $");
+
 #include <sys/types.h>
 #include <stdio.h>
 #include <stdlib.h>
-#include "regex.h"
+#include <limits.h>
+#include <regex.h>
+#include <wchar.h>
+#include <wctype.h>
 
 #include "utils.h"
 #include "regex2.h"
@@ -12,10 +55,10 @@
  = extern void regfree(regex_t *);
  */
 void
-regfree(preg)
+regfree(regex_t *preg)
-regex_t *preg;
 {
-	register struct re_guts *g;
+	struct re_guts *g;
+	int i;
 
 	if (preg->re_magic != MAGIC1)	/* oops */
 		return;			/* nice to complain, but hard */
@@ -28,11 +71,19 @@ regex_t *preg;
 
 	if (g->strip != NULL)
 		free((char *)g->strip);
-	if (g->sets != NULL)
+	if (g->sets != NULL) {
+		for (i = 0; i < g->ncsets; i++) {
+			free(g->sets[i].ranges);
+			free(g->sets[i].wides);
+			free(g->sets[i].types);
+		}
 		free((char *)g->sets);
-	if (g->setbits != NULL)
+	}
-		free((char *)g->setbits);
 	if (g->must != NULL)
 		free(g->must);
+	if (g->charjump != NULL)
+		free(&g->charjump[CHAR_MIN]);
+	if (g->matchjump != NULL)
+		free(g->matchjump);
 	free((char *)g);
 }

winsup/cygwin/regex/tests

@@ -1,477 +0,0 @@
-# regular expression test set
-# Lines are at least three fields, separated by one or more tabs.  "" stands
-# for an empty field.  First field is an RE.  Second field is flags.  If
-# C flag given, regcomp() is expected to fail, and the third field is the
-# error name (minus the leading REG_).
-#
-# Otherwise it is expected to succeed, and the third field is the string to
-# try matching it against.  If there is no fourth field, the match is
-# expected to fail.  If there is a fourth field, it is the substring that
-# the RE is expected to match.  If there is a fifth field, it is a comma-
-# separated list of what the subexpressions should match, with - indicating
-# no match for that one.  In both the fourth and fifth fields, a (sub)field
-# starting with @ indicates that the (sub)expression is expected to match
-# a null string followed by the stuff after the @; this provides a way to
-# test where null strings match.  The character `N' in REs and strings
-# is newline, `S' is space, `T' is tab, `Z' is NUL.
-#
-# The full list of flags:
-#	-	placeholder, does nothing
-#	b	RE is a BRE, not an ERE
-#	&	try it as both an ERE and a BRE
-#	C	regcomp() error expected, third field is error name
-#	i	REG_ICASE
-#	m	("mundane") REG_NOSPEC
-#	s	REG_NOSUB (not really testable)
-#	n	REG_NEWLINE
-#	^	REG_NOTBOL
-#	$	REG_NOTEOL
-#	#	REG_STARTEND (see below)
-#	p	REG_PEND
-#
-# For REG_STARTEND, the start/end offsets are those of the substring
-# enclosed in ().
-
-# basics
-a		&	a	a
-abc		&	abc	abc
-abc|de		-	abc	abc
-a|b|c		-	abc	a
-
-# parentheses and perversions thereof
-a(b)c		-	abc	abc
-a\(b\)c		b	abc	abc
-a(		C	EPAREN
-a(		b	a(	a(
-a\(		-	a(	a(
-a\(		bC	EPAREN
-a\(b		bC	EPAREN
-a(b		C	EPAREN
-a(b		b	a(b	a(b
-# gag me with a right parenthesis -- 1003.2 goofed here (my fault, partly)
-a)		-	a)	a)
-)		-	)	)
-# end gagging (in a just world, those *should* give EPAREN)
-a)		b	a)	a)
-a\)		bC	EPAREN
-\)		bC	EPAREN
-a()b		-	ab	ab
-a\(\)b		b	ab	ab
-
-# anchoring and REG_NEWLINE
-^abc$		&	abc	abc
-a^b		-	a^b
-a^b		b	a^b	a^b
-a$b		-	a$b
-a$b		b	a$b	a$b
-^		&	abc	@abc
-$		&	abc	@
-^$		&	""	@
-$^		-	""	@
-\($\)\(^\)	b	""	@
-# stop retching, those are legitimate (although disgusting)
-^^		-	""	@
-$$		-	""	@
-b$		&	abNc
-b$		&n	abNc	b
-^b$		&	aNbNc
-^b$		&n	aNbNc	b
-^$		&n	aNNb	@Nb
-^$		n	abc
-^$		n	abcN	@
-$^		n	aNNb	@Nb
-\($\)\(^\)	bn	aNNb	@Nb
-^^		n^	aNNb	@Nb
-$$		n	aNNb	@NN
-^a		^	a
-a$		$	a
-^a		^n	aNb
-^b		^n	aNb	b
-a$		$n	bNa
-b$		$n	bNa	b
-a*(^b$)c*	-	b	b
-a*\(^b$\)c*	b	b	b
-
-# certain syntax errors and non-errors
-|		C	EMPTY
-|		b	|	|
-*		C	BADRPT
-*		b	*	*
-+		C	BADRPT
-?		C	BADRPT
-""		&C	EMPTY
-()		-	abc	@abc
-\(\)		b	abc	@abc
-a||b		C	EMPTY
-|ab		C	EMPTY
-ab|		C	EMPTY
-(|a)b		C	EMPTY
-(a|)b		C	EMPTY
-(*a)		C	BADRPT
-(+a)		C	BADRPT
-(?a)		C	BADRPT
-({1}a)		C	BADRPT
-\(\{1\}a\)	bC	BADRPT
-(a|*b)		C	BADRPT
-(a|+b)		C	BADRPT
-(a|?b)		C	BADRPT
-(a|{1}b)	C	BADRPT
-^*		C	BADRPT
-^*		b	*	*
-^+		C	BADRPT
-^?		C	BADRPT
-^{1}		C	BADRPT
-^\{1\}		bC	BADRPT
-
-# metacharacters, backslashes
-a.c		&	abc	abc
-a[bc]d		&	abd	abd
-a\*c		&	a*c	a*c
-a\\b		&	a\b	a\b
-a\\\*b		&	a\*b	a\*b
-a\bc		&	abc	abc
-a\		&C	EESCAPE
-a\\bc		&	a\bc	a\bc
-\{		bC	BADRPT
-a\[b		&	a[b	a[b
-a[b		&C	EBRACK
-# trailing $ is a peculiar special case for the BRE code
-a$		&	a	a
-a$		&	a$
-a\$		&	a
-a\$		&	a$	a$
-a\\$		&	a
-a\\$		&	a$
-a\\$		&	a\$
-a\\$		&	a\	a\
-
-# back references, ugh
-a\(b\)\2c	bC	ESUBREG
-a\(b\1\)c	bC	ESUBREG
-a\(b*\)c\1d	b	abbcbbd	abbcbbd	bb
-a\(b*\)c\1d	b	abbcbd
-a\(b*\)c\1d	b	abbcbbbd
-^\(.\)\1	b	abc
-a\([bc]\)\1d	b	abcdabbd	abbd	b
-a\(\([bc]\)\2\)*d	b	abbccd	abbccd
-a\(\([bc]\)\2\)*d	b	abbcbd
-# actually, this next one probably ought to fail, but the spec is unclear
-a\(\(b\)*\2\)*d		b	abbbd	abbbd
-# here is a case that no NFA implementation does right
-\(ab*\)[ab]*\1	b	ababaaa	ababaaa	a
-# check out normal matching in the presence of back refs
-\(a\)\1bcd	b	aabcd	aabcd
-\(a\)\1bc*d	b	aabcd	aabcd
-\(a\)\1bc*d	b	aabd	aabd
-\(a\)\1bc*d	b	aabcccd	aabcccd
-\(a\)\1bc*[ce]d	b	aabcccd	aabcccd
-^\(a\)\1b\(c\)*cd$	b	aabcccd	aabcccd
-
-# ordinary repetitions
-ab*c		&	abc	abc
-ab+c		-	abc	abc
-ab?c		-	abc	abc
-a\(*\)b		b	a*b	a*b
-a\(**\)b	b	ab	ab
-a\(***\)b	bC	BADRPT
-*a		b	*a	*a
-**a		b	a	a
-***a		bC	BADRPT
-
-# the dreaded bounded repetitions
-{		&	{	{
-{abc		&	{abc	{abc
-{1		C	BADRPT
-{1}		C	BADRPT
-a{b		&	a{b	a{b
-a{1}b		-	ab	ab
-a\{1\}b		b	ab	ab
-a{1,}b		-	ab	ab
-a\{1,\}b	b	ab	ab
-a{1,2}b		-	aab	aab
-a\{1,2\}b	b	aab	aab
-a{1		C	EBRACE
-a\{1		bC	EBRACE
-a{1a		C	EBRACE
-a\{1a		bC	EBRACE
-a{1a}		C	BADBR
-a\{1a\}		bC	BADBR
-a{,2}		-	a{,2}	a{,2}
-a\{,2\}		bC	BADBR
-a{,}		-	a{,}	a{,}
-a\{,\}		bC	BADBR
-a{1,x}		C	BADBR
-a\{1,x\}	bC	BADBR
-a{1,x		C	EBRACE
-a\{1,x		bC	EBRACE
-a{300}		C	BADBR
-a\{300\}	bC	BADBR
-a{1,0}		C	BADBR
-a\{1,0\}	bC	BADBR
-ab{0,0}c	-	abcac	ac
-ab\{0,0\}c	b	abcac	ac
-ab{0,1}c	-	abcac	abc
-ab\{0,1\}c	b	abcac	abc
-ab{0,3}c	-	abbcac	abbc
-ab\{0,3\}c	b	abbcac	abbc
-ab{1,1}c	-	acabc	abc
-ab\{1,1\}c	b	acabc	abc
-ab{1,3}c	-	acabc	abc
-ab\{1,3\}c	b	acabc	abc
-ab{2,2}c	-	abcabbc	abbc
-ab\{2,2\}c	b	abcabbc	abbc
-ab{2,4}c	-	abcabbc	abbc
-ab\{2,4\}c	b	abcabbc	abbc
-((a{1,10}){1,10}){1,10}	-	a	a	a,a
-
-# multiple repetitions
-a**		&C	BADRPT
-a++		C	BADRPT
-a??		C	BADRPT
-a*+		C	BADRPT
-a*?		C	BADRPT
-a+*		C	BADRPT
-a+?		C	BADRPT
-a?*		C	BADRPT
-a?+		C	BADRPT
-a{1}{1}		C	BADRPT
-a*{1}		C	BADRPT
-a+{1}		C	BADRPT
-a?{1}		C	BADRPT
-a{1}*		C	BADRPT
-a{1}+		C	BADRPT
-a{1}?		C	BADRPT
-a*{b}		-	a{b}	a{b}
-a\{1\}\{1\}	bC	BADRPT
-a*\{1\}		bC	BADRPT
-a\{1\}*		bC	BADRPT
-
-# brackets, and numerous perversions thereof
-a[b]c		&	abc	abc
-a[ab]c		&	abc	abc
-a[^ab]c		&	adc	adc
-a[]b]c		&	a]c	a]c
-a[[b]c		&	a[c	a[c
-a[-b]c		&	a-c	a-c
-a[^]b]c		&	adc	adc
-a[^-b]c		&	adc	adc
-a[b-]c		&	a-c	a-c
-a[b		&C	EBRACK
-a[]		&C	EBRACK
-a[1-3]c		&	a2c	a2c
-a[3-1]c		&C	ERANGE
-a[1-3-5]c	&C	ERANGE
-a[[.-.]--]c	&	a-c	a-c
-a[1-		&C	ERANGE
-a[[.		&C	EBRACK
-a[[.x		&C	EBRACK
-a[[.x.		&C	EBRACK
-a[[.x.]		&C	EBRACK
-a[[.x.]]	&	ax	ax
-a[[.x,.]]	&C	ECOLLATE
-a[[.one.]]b	&	a1b	a1b
-a[[.notdef.]]b	&C	ECOLLATE
-a[[.].]]b	&	a]b	a]b
-a[[:alpha:]]c	&	abc	abc
-a[[:notdef:]]c	&C	ECTYPE
-a[[:		&C	EBRACK
-a[[:alpha	&C	EBRACK
-a[[:alpha:]	&C	EBRACK
-a[[:alpha,:]	&C	ECTYPE
-a[[:]:]]b	&C	ECTYPE
-a[[:-:]]b	&C	ECTYPE
-a[[:alph:]]	&C	ECTYPE
-a[[:alphabet:]]	&C	ECTYPE
-[[:alnum:]]+	-	-%@a0X-	a0X
-[[:alpha:]]+	-	-%@aX0-	aX
-[[:blank:]]+	-	aSSTb	SST
-[[:cntrl:]]+	-	aNTb	NT
-[[:digit:]]+	-	a019b	019
-[[:graph:]]+	-	Sa%bS	a%b
-[[:lower:]]+	-	AabC	ab
-[[:print:]]+	-	NaSbN	aSb
-[[:punct:]]+	-	S%-&T	%-&
-[[:space:]]+	-	aSNTb	SNT
-[[:upper:]]+	-	aBCd	BC
-[[:xdigit:]]+	-	p0f3Cq	0f3C
-a[[=b=]]c	&	abc	abc
-a[[=		&C	EBRACK
-a[[=b		&C	EBRACK
-a[[=b=		&C	EBRACK
-a[[=b=]		&C	EBRACK
-a[[=b,=]]	&C	ECOLLATE
-a[[=one=]]b	&	a1b	a1b
-
-# complexities
-a(((b)))c	-	abc	abc
-a(b|(c))d	-	abd	abd
-a(b*|c)d	-	abbd	abbd
-# just gotta have one DFA-buster, of course
-a[ab]{20}	-	aaaaabaaaabaaaabaaaab	aaaaabaaaabaaaabaaaab
-# and an inline expansion in case somebody gets tricky
-a[ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab]	-	aaaaabaaaabaaaabaaaab	aaaaabaaaabaaaabaaaab
-# and in case somebody just slips in an NFA...
-a[ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab](wee|week)(knights|night)	-	aaaaabaaaabaaaabaaaabweeknights	aaaaabaaaabaaaabaaaabweeknights
-# fish for anomalies as the number of states passes 32
-12345678901234567890123456789	-	a12345678901234567890123456789b	12345678901234567890123456789
-123456789012345678901234567890	-	a123456789012345678901234567890b	123456789012345678901234567890
-1234567890123456789012345678901	-	a1234567890123456789012345678901b	1234567890123456789012345678901
-12345678901234567890123456789012	-	a12345678901234567890123456789012b	12345678901234567890123456789012
-123456789012345678901234567890123	-	a123456789012345678901234567890123b	123456789012345678901234567890123
-# and one really big one, beyond any plausible word width
-1234567890123456789012345678901234567890123456789012345678901234567890	-	a1234567890123456789012345678901234567890123456789012345678901234567890b	1234567890123456789012345678901234567890123456789012345678901234567890
-# fish for problems as brackets go past 8
-[ab][cd][ef][gh][ij][kl][mn]	-	xacegikmoq	acegikm
-[ab][cd][ef][gh][ij][kl][mn][op]	-	xacegikmoq	acegikmo
-[ab][cd][ef][gh][ij][kl][mn][op][qr]	-	xacegikmoqy	acegikmoq
-[ab][cd][ef][gh][ij][kl][mn][op][q]	-	xacegikmoqy	acegikmoq
-
-# subtleties of matching
-abc		&	xabcy	abc
-a\(b\)?c\1d	b	acd
-aBc		i	Abc	Abc
-a[Bc]*d		i	abBCcd	abBCcd
-0[[:upper:]]1	&i	0a1	0a1
-0[[:lower:]]1	&i	0A1	0A1
-a[^b]c		&i	abc
-a[^b]c		&i	aBc
-a[^b]c		&i	adc	adc
-[a]b[c]		-	abc	abc
-[a]b[a]		-	aba	aba
-[abc]b[abc]	-	abc	abc
-[abc]b[abd]	-	abd	abd
-a(b?c)+d	-	accd	accd
-(wee|week)(knights|night)	-	weeknights	weeknights
-(we|wee|week|frob)(knights|night|day)	-	weeknights	weeknights
-a[bc]d		-	xyzaaabcaababdacd	abd
-a[ab]c		-	aaabc	abc
-abc		s	abc	abc
-a*		&	b	@b
-
-# Let's have some fun -- try to match a C comment.
-# first the obvious, which looks okay at first glance...
-/\*.*\*/	-	/*x*/	/*x*/
-# but...
-/\*.*\*/	-	/*x*/y/*z*/	/*x*/y/*z*/
-# okay, we must not match */ inside; try to do that...
-/\*([^*]|\*[^/])*\*/	-	/*x*/	/*x*/
-/\*([^*]|\*[^/])*\*/	-	/*x*/y/*z*/	/*x*/
-# but...
-/\*([^*]|\*[^/])*\*/	-	/*x**/y/*z*/	/*x**/y/*z*/
-# and a still fancier version, which does it right (I think)...
-/\*([^*]|\*+[^*/])*\*+/	-	/*x*/	/*x*/
-/\*([^*]|\*+[^*/])*\*+/	-	/*x*/y/*z*/	/*x*/
-/\*([^*]|\*+[^*/])*\*+/	-	/*x**/y/*z*/	/*x**/
-/\*([^*]|\*+[^*/])*\*+/	-	/*x****/y/*z*/	/*x****/
-/\*([^*]|\*+[^*/])*\*+/	-	/*x**x*/y/*z*/	/*x**x*/
-/\*([^*]|\*+[^*/])*\*+/	-	/*x***x/y/*z*/	/*x***x/y/*z*/
-
-# subexpressions
-.*		-	abc	abc	-
-a(b)(c)d	-	abcd	abcd	b,c
-a(((b)))c	-	abc	abc	b,b,b
-a(b|(c))d	-	abd	abd	b,-
-a(b*|c|e)d	-	abbd	abbd	bb
-a(b*|c|e)d	-	acd	acd	c
-a(b*|c|e)d	-	ad	ad	@d
-a(b?)c		-	abc	abc	b
-a(b?)c		-	ac	ac	@c
-a(b+)c		-	abc	abc	b
-a(b+)c		-	abbbc	abbbc	bbb
-a(b*)c		-	ac	ac	@c
-(a|ab)(bc([de]+)f|cde)	-	abcdef	abcdef	a,bcdef,de
-# the regression tester only asks for 9 subexpressions
-a(b)(c)(d)(e)(f)(g)(h)(i)(j)k	-	abcdefghijk	abcdefghijk	b,c,d,e,f,g,h,i,j
-a(b)(c)(d)(e)(f)(g)(h)(i)(j)(k)l	-	abcdefghijkl	abcdefghijkl	b,c,d,e,f,g,h,i,j,k
-a([bc]?)c	-	abc	abc	b
-a([bc]?)c	-	ac	ac	@c
-a([bc]+)c	-	abc	abc	b
-a([bc]+)c	-	abcc	abcc	bc
-a([bc]+)bc	-	abcbc	abcbc	bc
-a(bb+|b)b	-	abb	abb	b
-a(bbb+|bb+|b)b	-	abb	abb	b
-a(bbb+|bb+|b)b	-	abbb	abbb	bb
-a(bbb+|bb+|b)bb	-	abbb	abbb	b
-(.*).*		-	abcdef	abcdef	abcdef
-(a*)*		-	bc	@b	@b
-
-# do we get the right subexpression when it is used more than once?
-a(b|c)*d	-	ad	ad	-
-a(b|c)*d	-	abcd	abcd	c
-a(b|c)+d	-	abd	abd	b
-a(b|c)+d	-	abcd	abcd	c
-a(b|c?)+d	-	ad	ad	@d
-a(b|c?)+d	-	abcd	abcd	@d
-a(b|c){0,0}d	-	ad	ad	-
-a(b|c){0,1}d	-	ad	ad	-
-a(b|c){0,1}d	-	abd	abd	b
-a(b|c){0,2}d	-	ad	ad	-
-a(b|c){0,2}d	-	abcd	abcd	c
-a(b|c){0,}d	-	ad	ad	-
-a(b|c){0,}d	-	abcd	abcd	c
-a(b|c){1,1}d	-	abd	abd	b
-a(b|c){1,1}d	-	acd	acd	c
-a(b|c){1,2}d	-	abd	abd	b
-a(b|c){1,2}d	-	abcd	abcd	c
-a(b|c){1,}d	-	abd	abd	b
-a(b|c){1,}d	-	abcd	abcd	c
-a(b|c){2,2}d	-	acbd	acbd	b
-a(b|c){2,2}d	-	abcd	abcd	c
-a(b|c){2,4}d	-	abcd	abcd	c
-a(b|c){2,4}d	-	abcbd	abcbd	b
-a(b|c){2,4}d	-	abcbcd	abcbcd	c
-a(b|c){2,}d	-	abcd	abcd	c
-a(b|c){2,}d	-	abcbd	abcbd	b
-a(b+|((c)*))+d	-	abd	abd	@d,@d,-
-a(b+|((c)*))+d	-	abcd	abcd	@d,@d,-
-
-# check out the STARTEND option
-[abc]		&#	a(b)c	b
-[abc]		&#	a(d)c
-[abc]		&#	a(bc)d	b
-[abc]		&#	a(dc)d	c
-.		&#	a()c
-b.*c		&#	b(bc)c	bc
-b.*		&#	b(bc)c	bc
-.*c		&#	b(bc)c	bc
-
-# plain strings, with the NOSPEC flag
-abc		m	abc	abc
-abc		m	xabcy	abc
-abc		m	xyz
-a*b		m	aba*b	a*b
-a*b		m	ab
-""		mC	EMPTY
-
-# cases involving NULs
-aZb		&	a	a
-aZb		&p	a
-aZb		&p#	(aZb)	aZb
-aZ*b		&p#	(ab)	ab
-a.b		&#	(aZb)	aZb
-a.*		&#	(aZb)c	aZb
-
-# word boundaries (ick)
-[[:<:]]a	&	a	a
-[[:<:]]a	&	ba
-[[:<:]]a	&	-a	a
-a[[:>:]]	&	a	a
-a[[:>:]]	&	ab
-a[[:>:]]	&	a-	a
-[[:<:]]a.c[[:>:]]	&	axcd-dayc-dazce-abc	abc
-[[:<:]]a.c[[:>:]]	&	axcd-dayc-dazce-abc-q	abc
-[[:<:]]a.c[[:>:]]	&	axc-dayc-dazce-abc	axc
-[[:<:]]b.c[[:>:]]	&	a_bxc-byc_d-bzc-q	bzc
-[[:<:]].x..[[:>:]]	&	y_xa_-_xb_y-_xc_-axdc	_xc_
-[[:<:]]a_b[[:>:]]	&	x_a_b
-
-# past problems, and suspected problems
-(A[1])|(A[2])|(A[3])|(A[4])|(A[5])|(A[6])|(A[7])|(A[8])|(A[9])|(A[A])	-	A1	A1
-abcdefghijklmnop	i	abcdefghijklmnop	abcdefghijklmnop
-abcdefghijklmnopqrstuv	i	abcdefghijklmnopqrstuv	abcdefghijklmnopqrstuv
-(ALAK)|(ALT[AB])|(CC[123]1)|(CM[123]1)|(GAMC)|(LC[23][EO ])|(SEM[1234])|(SL[ES][12])|(SLWW)|(SLF )|(SLDT)|(VWH[12])|(WH[34][EW])|(WP1[ESN])	-	CC11	CC11
-CC[13]1|a{21}[23][EO][123][Es][12]a{15}aa[34][EW]aaaaaaa[X]a	-	CC11	CC11
-Char \([a-z0-9_]*\)\[.*	b	Char xyz[k	Char xyz[k	xyz
-a?b	-	ab	ab
--\{0,1\}[0-9]*$	b	-5	-5
-a*a*a*a*a*a*a*	&	aaaaaa	aaaaaa

winsup/cygwin/regex/utils.h

@@ -1,9 +1,41 @@
+/*-
+ * Copyright (c) 1992, 1993, 1994 Henry Spencer.
+ * Copyright (c) 1992, 1993, 1994
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * This code is derived from software contributed to Berkeley by
+ * Henry Spencer.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 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
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * 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
+ * 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 $
+ */
+
 /* utility definitions */
-#ifdef _POSIX2_RE_DUP_MAX
+#define	DUPMAX		_POSIX2_RE_DUP_MAX	/* xxx is this right? */
-#define	DUPMAX	_POSIX2_RE_DUP_MAX
-#else
-#define	DUPMAX	255
-#endif
 #define	INFINITY	(DUPMAX + 1)
 #define	NC		(CHAR_MAX - CHAR_MIN + 1)
 typedef unsigned char uch;

winsup/cygwin/wchar.h

@@ -49,6 +49,7 @@ extern char *__locale_charset ();
 #endif
 
 #ifdef __INSIDE_CYGWIN__
+#ifdef __cplusplus
 size_t __stdcall sys_cp_wcstombs (wctomb_p, const char *, char *, size_t,
 				  const wchar_t *, size_t = (size_t) -1)
        __attribute__ ((regparm(3)));
@@ -68,6 +69,7 @@ size_t __stdcall sys_mbstowcs (wchar_t * dst, size_t dlen, const char *src,
 size_t __stdcall sys_mbstowcs_alloc (wchar_t **, int, const char *,
 				     size_t = (size_t) -1)
        __attribute__ ((regparm(3)));
+#endif /* __cplusplus */
 #endif /* __INSIDE_CYGWIN__ */
 
 #endif /* _CYGWIN_WCHAR_H */