Add regular expression matching support, based on OpenBSD regexec()/regcomp()

implementation.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80493 91177308-0d34-0410-b5e6-96231b3b80d8

14 files changed, 3647 insertions, 0 deletions

diff --git a/lib/Support/CMakeLists.txt b/lib/Support/CMakeLists.txt
index 05f1ac7d98..0144b28d2e 100644
--- a/lib/Support/CMakeLists.txt
+++ b/lib/Support/CMakeLists.txt
@@ -32,6 +32,12 @@ add_llvm_library(LLVMSupport
   Twine.cpp
   raw_os_ostream.cpp
   raw_ostream.cpp
+  Regex.cpp
+  regcomp.c
+  regerror.c
+  regexec.c
+  regfree.c
+  regstrlcpy.c
   )
 
 target_link_libraries (LLVMSupport LLVMSystem)
diff --git a/lib/Support/COPYRIGHT.regex b/lib/Support/COPYRIGHT.regex
new file mode 100644
index 0000000000..a6392fd37c
--- /dev/null
+++ b/lib/Support/COPYRIGHT.regex
@@ -0,0 +1,54 @@
+$OpenBSD: COPYRIGHT,v 1.3 2003/06/02 20:18:36 millert Exp $
+
+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.
+
+Permission is granted to anyone to use this software for any purpose on
+any computer system, and to alter it and redistribute it, subject
+to the following restrictions:
+
+1. The author is not responsible for the consequences of use of this
+   software, no matter how awful, even if they arise from flaws in it.
+
+2. The origin of this software must not be misrepresented, either by
+   explicit claim or by omission.  Since few users ever read sources,
+   credits must appear in the documentation.
+
+3. Altered versions must be plainly marked as such, and must not be
+   misrepresented as being the original software.  Since few users
+   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.
+ * 3. Neither the name of the University nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * 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
+ */
diff --git a/lib/Support/Regex.cpp b/lib/Support/Regex.cpp
new file mode 100644
index 0000000000..f8b2446e3c
--- /dev/null
+++ b/lib/Support/Regex.cpp
@@ -0,0 +1,97 @@
+//===-- Regex.cpp - Regular Expression matcher implementation -------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// This file implements a POSIX regular expression matcher.
+//
+//===----------------------------------------------------------------------===//
+#include "llvm/Support/Regex.h"
+#include "llvm/Support/ErrorHandling.h"
+#include "llvm/Support/raw_ostream.h"
+#include "regex_impl.h"
+#include <string>
+
+using namespace llvm;
+Regex::Regex(const StringRef &regex, unsigned Flags)
+{
+  unsigned flags = 0;
+  preg = new struct llvm_regex;
+  preg->re_endp = regex.end();
+  if (Flags & IgnoreCase) 
+    flags |= REG_ICASE;
+  if (Flags & NoSub) {
+    flags |= REG_NOSUB;
+    sub = false;
+  } else {
+    sub = true;
+  }
+  if (Flags & Newline)
+    flags |= REG_NEWLINE;
+  error = llvm_regcomp(preg, regex.data(), flags|REG_EXTENDED|REG_PEND);
+}
+
+bool Regex::isValid(std::string &Error)
+{
+  if (!error)
+    return true;
+
+  size_t len = llvm_regerror(error, preg, NULL, 0);
+  char *errbuff = new char[len];
+  llvm_regerror(error, preg, errbuff, len);
+  Error.assign(errbuff);
+  return false;
+}
+
+Regex::~Regex()
+{
+  llvm_regfree(preg);
+  delete preg;
+}
+
+bool Regex::match(const StringRef &String, SmallVectorImpl<StringRef> *Matches)
+{
+  unsigned nmatch = Matches ? preg->re_nsub+1 : 0;
+
+  if (Matches) {
+    assert(sub && "Substring matching requested but pattern compiled without");
+    Matches->clear();
+  }
+
+  // pmatch needs to have at least one element.
+  SmallVector<llvm_regmatch_t, 2> pm;
+  pm.resize(nmatch > 0 ? nmatch : 1);
+  pm[0].rm_so = 0;
+  pm[0].rm_eo = String.size();
+
+  int rc = llvm_regexec(preg, String.data(), nmatch, pm.data(), REG_STARTEND);
+
+  if (rc == REG_NOMATCH)
+    return false;
+  if (rc != 0) {
+    // regexec can fail due to invalid pattern or running out of memory.
+    error = rc;
+    return false;
+  }
+
+  // There was a match.
+
+  if (Matches) { // match position requested
+    for (unsigned i=0;i<nmatch; i++) {
+      if (pm[i].rm_so == -1) {
+        // this group didn't match
+        Matches->push_back(StringRef());
+        continue;
+      }
+      assert(pm[i].rm_eo > pm[i].rm_so);
+      Matches->push_back(StringRef(String.data()+pm[i].rm_so,
+                                   pm[i].rm_eo-pm[i].rm_so));
+    }
+  }
+
+  return true;
+}
diff --git a/lib/Support/regcclass.h b/lib/Support/regcclass.h
new file mode 100644
index 0000000000..2cea3e4e54
--- /dev/null
+++ b/lib/Support/regcclass.h
@@ -0,0 +1,70 @@
+/*-
+ * This code is derived from OpenBSD's libc/regex, original license follows:
+ *
+ * This code is derived from OpenBSD's libc/regex, original license follows:
+ *
+ * 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.
+ * 3. Neither the name of the University nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * 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.
+ *
+ *	@(#)cclass.h	8.3 (Berkeley) 3/20/94
+ */
+
+/* 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,		"" }
+};
diff --git a/lib/Support/regcname.h b/lib/Support/regcname.h
new file mode 100644
index 0000000000..3c0bb248ff
--- /dev/null
+++ b/lib/Support/regcname.h
@@ -0,0 +1,139 @@
+/*-
+ * This code is derived from OpenBSD's libc/regex, original license follows:
+ *
+ * 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.
+ * 3. Neither the name of the University nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * 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
+ */
+
+/* character-name table */
+static struct cname {
+	const char *name;
+	char code;
+} cnames[] = {
+	{ "NUL",			'\0' },
+	{ "SOH",			'\001' },
+	{ "STX",			'\002' },
+	{ "ETX",			'\003' },
+	{ "EOT",			'\004' },
+	{ "ENQ",			'\005' },
+	{ "ACK",			'\006' },
+	{ "BEL",			'\007' },
+	{ "alert",			'\007' },
+	{ "BS",				'\010' },
+	{ "backspace",			'\b' },
+	{ "HT",				'\011' },
+	{ "tab",			'\t' },
+	{ "LF",				'\012' },
+	{ "newline",			'\n' },
+	{ "VT",				'\013' },
+	{ "vertical-tab",		'\v' },
+	{ "FF",				'\014' },
+	{ "form-feed",			'\f' },
+	{ "CR",				'\015' },
+	{ "carriage-return",		'\r' },
+	{ "SO",				'\016' },
+	{ "SI",				'\017' },
+	{ "DLE",			'\020' },
+	{ "DC1",			'\021' },
+	{ "DC2",			'\022' },
+	{ "DC3",			'\023' },
+	{ "DC4",			'\024' },
+	{ "NAK",			'\025' },
+	{ "SYN",			'\026' },
+	{ "ETB",			'\027' },
+	{ "CAN",			'\030' },
+	{ "EM",				'\031' },
+	{ "SUB",			'\032' },
+	{ "ESC",			'\033' },
+	{ "IS4",			'\034' },
+	{ "FS",				'\034' },
+	{ "IS3",			'\035' },
+	{ "GS",				'\035' },
+	{ "IS2",			'\036' },
+	{ "RS",				'\036' },
+	{ "IS1",			'\037' },
+	{ "US",				'\037' },
+	{ "space",			' ' },
+	{ "exclamation-mark",		'!' },
+	{ "quotation-mark",		'"' },
+	{ "number-sign",		'#' },
+	{ "dollar-sign",		'$' },
+	{ "percent-sign",		'%' },
+	{ "ampersand",			'&' },
+	{ "apostrophe",			'\'' },
+	{ "left-parenthesis",		'(' },
+	{ "right-parenthesis",		')' },
+	{ "asterisk",			'*' },
+	{ "plus-sign",			'+' },
+	{ "comma",			',' },
+	{ "hyphen",			'-' },
+	{ "hyphen-minus",		'-' },
+	{ "period",			'.' },
+	{ "full-stop",			'.' },
+	{ "slash",			'/' },
+	{ "solidus",			'/' },
+	{ "zero",			'0' },
+	{ "one",			'1' },
+	{ "two",			'2' },
+	{ "three",			'3' },
+	{ "four",			'4' },
+	{ "five",			'5' },
+	{ "six",			'6' },
+	{ "seven",			'7' },
+	{ "eight",			'8' },
+	{ "nine",			'9' },
+	{ "colon",			':' },
+	{ "semicolon",			';' },
+	{ "less-than-sign",		'<' },
+	{ "equals-sign",		'=' },
+	{ "greater-than-sign",		'>' },
+	{ "question-mark",		'?' },
+	{ "commercial-at",		'@' },
+	{ "left-square-bracket",	'[' },
+	{ "backslash",			'\\' },
+	{ "reverse-solidus",		'\\' },
+	{ "right-square-bracket",	']' },
+	{ "circumflex",			'^' },
+	{ "circumflex-accent",		'^' },
+	{ "underscore",			'_' },
+	{ "low-line",			'_' },
+	{ "grave-accent",		'`' },
+	{ "left-brace",			'{' },
+	{ "left-curly-bracket",		'{' },
+	{ "vertical-line",		'|' },
+	{ "right-brace",		'}' },
+	{ "right-curly-bracket",	'}' },
+	{ "tilde",			'~' },
+	{ "DEL",			'\177' },
+	{ NULL,				0 }
+};
diff --git a/lib/Support/regcomp.c b/lib/Support/regcomp.c
new file mode 100644
index 0000000000..4fd88b1249
--- /dev/null
+++ b/lib/Support/regcomp.c
@@ -0,0 +1,1524 @@
+/*-
+ * This code is derived from OpenBSD's libc/regex, original license follows:
+ *
+ * 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.
+ * 3. Neither the name of the University nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * 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.
+ *
+ *	@(#)regcomp.c	8.5 (Berkeley) 3/20/94
+ */
+
+#include <sys/types.h>
+#include <stdio.h>
+#include <string.h>
+#include <ctype.h>
+#include <limits.h>
+#include <stdlib.h>
+#include "regex_impl.h"
+
+#include "regutils.h"
+#include "regex2.h"
+
+#include "regcclass.h"
+#include "regcname.h"
+
+/*
+ * parse structure, passed up and down to avoid global variables and
+ * other clumsinesses
+ */
+struct parse {
+	char *next;		/* next character in RE */
+	char *end;		/* end of string (-> NUL normally) */
+	int error;		/* has an error been seen? */
+	sop *strip;		/* malloced strip */
+	sopno ssize;		/* malloced strip size (allocated) */
+	sopno slen;		/* malloced strip length (used) */
+	int ncsalloc;		/* number of csets allocated */
+	struct re_guts *g;
+#	define	NPAREN	10	/* we need to remember () 1-9 for back refs */
+	sopno pbegin[NPAREN];	/* -> ( ([0] unused) */
+	sopno pend[NPAREN];	/* -> ) ([0] unused) */
+};
+
+static void p_ere(struct parse *, int);
+static void p_ere_exp(struct parse *);
+static void p_str(struct parse *);
+static void p_bre(struct parse *, int, int);
+static int p_simp_re(struct parse *, int);
+static int p_count(struct parse *);
+static void p_bracket(struct parse *);
+static void p_b_term(struct parse *, cset *);
+static void p_b_cclass(struct parse *, cset *);
+static void p_b_eclass(struct parse *, cset *);
+static char p_b_symbol(struct parse *);
+static char p_b_coll_elem(struct parse *, int);
+static char othercase(int);
+static void bothcases(struct parse *, int);
+static void ordinary(struct parse *, int);
+static void nonnewline(struct parse *);
+static void repeat(struct parse *, sopno, int, int);
+static int seterr(struct parse *, int);
+static cset *allocset(struct parse *);
+static void freeset(struct parse *, cset *);
+static int freezeset(struct parse *, cset *);
+static int firstch(struct parse *, cset *);
+static int nch(struct parse *, cset *);
+static void mcadd(struct parse *, cset *, const char *);
+static void mcinvert(struct parse *, cset *);
+static void mccase(struct parse *, cset *);
+static int isinsets(struct re_guts *, int);
+static int samesets(struct re_guts *, int, int);
+static void categorize(struct parse *, struct re_guts *);
+static sopno dupl(struct parse *, sopno, sopno);
+static void doemit(struct parse *, sop, size_t);
+static void doinsert(struct parse *, sop, size_t, sopno);
+static void dofwd(struct parse *, sopno, sop);
+static void enlarge(struct parse *, sopno);
+static void stripsnug(struct parse *, struct re_guts *);
+static void findmust(struct parse *, struct re_guts *);
+static sopno pluscount(struct parse *, struct re_guts *);
+
+static char nuls[10];		/* place to point scanner in event of error */
+
+/*
+ * macros for use with parse structure
+ * BEWARE:  these know that the parse structure is named `p' !!!
+ */
+#define	PEEK()	(*p->next)
+#define	PEEK2()	(*(p->next+1))
+#define	MORE()	(p->next < p->end)
+#define	MORE2()	(p->next+1 < p->end)
+#define	SEE(c)	(MORE() && PEEK() == (c))
+#define	SEETWO(a, b)	(MORE() && MORE2() && PEEK() == (a) && PEEK2() == (b))
+#define	EAT(c)	((SEE(c)) ? (NEXT(), 1) : 0)
+#define	EATTWO(a, b)	((SEETWO(a, b)) ? (NEXT2(), 1) : 0)
+#define	NEXT()	(p->next++)
+#define	NEXT2()	(p->next += 2)
+#define	NEXTn(n)	(p->next += (n))
+#define	GETNEXT()	(*p->next++)
+#define	SETERROR(e)	seterr(p, (e))
+#define	REQUIRE(co, e)	(void)((co) || SETERROR(e))
+#define	MUSTSEE(c, e)	(REQUIRE(MORE() && PEEK() == (c), e))
+#define	MUSTEAT(c, e)	(REQUIRE(MORE() && GETNEXT() == (c), e))
+#define	MUSTNOTSEE(c, e)	(REQUIRE(!MORE() || PEEK() != (c), e))
+#define	EMIT(op, sopnd)	doemit(p, (sop)(op), (size_t)(sopnd))
+#define	INSERT(op, pos)	doinsert(p, (sop)(op), HERE()-(pos)+1, pos)
+#define	AHEAD(pos)		dofwd(p, pos, HERE()-(pos))
+#define	ASTERN(sop, pos)	EMIT(sop, HERE()-pos)
+#define	HERE()		(p->slen)
+#define	THERE()		(p->slen - 1)
+#define	THERETHERE()	(p->slen - 2)
+#define	DROP(n)	(p->slen -= (n))
+
+#ifdef	_POSIX2_RE_DUP_MAX
+#define	DUPMAX	_POSIX2_RE_DUP_MAX
+#else
+#define	DUPMAX	255
+#endif
+
+#ifndef NDEBUG
+static int never = 0;		/* for use in asserts; shuts lint up */
+#else
+#define	never	0		/* some <assert.h>s have bugs too */
+#endif
+
+/*
+ - llvm_regcomp - interface for parser and compilation
+ */
+int				/* 0 success, otherwise REG_something */
+llvm_regcomp(llvm_regex_t *preg, const char *pattern, int cflags)
+{
+	struct parse pa;
+	struct re_guts *g;
+	struct parse *p = &pa;
+	int i;
+	size_t len;
+#ifdef REDEBUG
+#	define	GOODFLAGS(f)	(f)
+#else
+#	define	GOODFLAGS(f)	((f)&~REG_DUMP)
+#endif
+
+	cflags = GOODFLAGS(cflags);
+	if ((cflags&REG_EXTENDED) && (cflags&REG_NOSPEC))
+		return(REG_INVARG);
+
+	if (cflags&REG_PEND) {
+		if (preg->re_endp < pattern)
+			return(REG_INVARG);
+		len = preg->re_endp - pattern;
+	} else
+		len = strlen((const char *)pattern);
+
+	/* do the mallocs early so failure handling is easy */
+	g = (struct re_guts *)malloc(sizeof(struct re_guts) +
+							(NC-1)*sizeof(cat_t));
+	if (g == NULL)
+		return(REG_ESPACE);
+	p->ssize = len/(size_t)2*(size_t)3 + (size_t)1;	/* ugh */
+	p->strip = (sop *)calloc(p->ssize, sizeof(sop));
+	p->slen = 0;
+	if (p->strip == NULL) {
+		free((char *)g);
+		return(REG_ESPACE);
+	}
+
+	/* set things up */
+	p->g = g;
+	p->next = (char *)pattern;	/* convenience; we do not modify it */
+	p->end = p->next + len;
+	p->error = 0;
+	p->ncsalloc = 0;
+	for (i = 0; i < NPAREN; i++) {
+		p->pbegin[i] = 0;
+		p->pend[i] = 0;
+	}
+	g->csetsize = NC;
+	g->sets = NULL;
+	g->setbits = NULL;
+	g->ncsets = 0;
+	g->cflags = cflags;
+	g->iflags = 0;
+	g->nbol = 0;
+	g->neol = 0;
+	g->must = NULL;
+	g->mlen = 0;
+	g->nsub = 0;
+	g->ncategories = 1;	/* category 0 is "everything else" */
+	g->categories = &g->catspace[-(CHAR_MIN)];
+	(void) memset((char *)g->catspace, 0, NC*sizeof(cat_t));
+	g->backrefs = 0;
+
+	/* do it */
+	EMIT(OEND, 0);
+	g->firststate = THERE();
+	if (cflags&REG_EXTENDED)
+		p_ere(p, OUT);
+	else if (cflags&REG_NOSPEC)
+		p_str(p);
+	else
+		p_bre(p, OUT, OUT);
+	EMIT(OEND, 0);
+	g->laststate = THERE();
+
+	/* tidy up loose ends and fill things in */
+	categorize(p, g);
+	stripsnug(p, g);
+	findmust(p, g);
+	g->nplus = pluscount(p, g);
+	g->magic = MAGIC2;
+	preg->re_nsub = g->nsub;
+	preg->re_g = g;
+	preg->re_magic = MAGIC1;
+#ifndef REDEBUG
+	/* not debugging, so can't rely on the assert() in llvm_regexec() */
+	if (g->iflags&REGEX_BAD)
+		SETERROR(REG_ASSERT);
+#endif
+
+	/* win or lose, we're done */
+	if (p->error != 0)	/* lose */
+		llvm_regfree(preg);
+	return(p->error);
+}
+
+/*
+ - p_ere - ERE parser top level, concatenation and alternation
+ */
+static void
+p_ere(struct parse *p, int stop)	/* character this ERE should end at */
+{
+	char c;
+	sopno prevback = prevback;
+	sopno prevfwd = prevfwd;
+	sopno conc;
+	int first = 1;		/* is this the first alternative? */
+
+	for (;;) {
+		/* do a bunch of concatenated expressions */
+		conc = HERE();
+		while (MORE() && (c = PEEK()) != '|' && c != stop)
+			p_ere_exp(p);
+		REQUIRE(HERE() != conc, REG_EMPTY);	/* require nonempty */
+
+		if (!EAT('|'))
+			break;		/* NOTE BREAK OUT */
+
+		if (first) {
+			INSERT(OCH_, conc);	/* offset is wrong */
+			prevfwd = conc;
+			prevback = conc;
+			first = 0;
+		}
+		ASTERN(OOR1, prevback);
+		prevback = THERE();
+		AHEAD(prevfwd);			/* fix previous offset */
+		prevfwd = HERE();
+		EMIT(OOR2, 0);			/* offset is very wrong */
+	}
+
+	if (!first) {		/* tail-end fixups */
+		AHEAD(prevfwd);
+		ASTERN(O_CH, prevback);
+	}
+
+	assert(!MORE() || SEE(stop));
+}
+
+/*
+ - p_ere_exp - parse one subERE, an atom possibly followed by a repetition op
+ */
+static void
+p_ere_exp(struct parse *p)
+{
+	char c;
+	sopno pos;
+	int count;
+	int count2;
+	sopno subno;
+	int wascaret = 0;
+
+	assert(MORE());		/* caller should have ensured this */
+	c = GETNEXT();
+
+	pos = HERE();
+	switch (c) {
+	case '(':
+		REQUIRE(MORE(), REG_EPAREN);
+		p->g->nsub++;
+		subno = p->g->nsub;
+		if (subno < NPAREN)
+			p->pbegin[subno] = HERE();
+		EMIT(OLPAREN, subno);
+		if (!SEE(')'))
+			p_ere(p, ')');
+		if (subno < NPAREN) {
+			p->pend[subno] = HERE();
+			assert(p->pend[subno] != 0);
+		}
+		EMIT(ORPAREN, subno);
+		MUSTEAT(')', REG_EPAREN);
+		break;
+#ifndef POSIX_MISTAKE
+	case ')':		/* happens only if no current unmatched ( */
+		/*
+		 * You may ask, why the ifndef?  Because I didn't notice
+		 * this until slightly too late for 1003.2, and none of the
+		 * other 1003.2 regular-expression reviewers noticed it at
+		 * all.  So an unmatched ) is legal POSIX, at least until
+		 * we can get it fixed.
+		 */
+		SETERROR(REG_EPAREN);
+		break;
+#endif
+	case '^':
+		EMIT(OBOL, 0);
+		p->g->iflags |= USEBOL;
+		p->g->nbol++;
+		wascaret = 1;
+		break;
+	case '$':
+		EMIT(OEOL, 0);
+		p->g->iflags |= USEEOL;
+		p->g->neol++;
+		break;
+	case '|':
+		SETERROR(REG_EMPTY);
+		break;
+	case '*':
+	case '+':
+	case '?':
+		SETERROR(REG_BADRPT);
+		break;
+	case '.':
+		if (p->g->cflags&REG_NEWLINE)
+			nonnewline(p);
+		else
+			EMIT(OANY, 0);
+		break;
+	case '[':
+		p_bracket(p);
+		break;
+	case '\\':
+		REQUIRE(MORE(), REG_EESCAPE);
+		c = GETNEXT();
+		ordinary(p, c);
+		break;
+	case '{':		/* okay as ordinary except if digit follows */
+		REQUIRE(!MORE() || !isdigit((uch)PEEK()), REG_BADRPT);
+		/* FALLTHROUGH */
+	default:
+		ordinary(p, c);
+		break;
+	}
+
+	if (!MORE())
+		return;
+	c = PEEK();
+	/* we call { a repetition if followed by a digit */
+	if (!( c == '*' || c == '+' || c == '?' ||
+				(c == '{' && MORE2() && isdigit((uch)PEEK2())) ))
+		return;		/* no repetition, we're done */
+	NEXT();
+
+	REQUIRE(!wascaret, REG_BADRPT);
+	switch (c) {
+	case '*':	/* implemented as +? */
+		/* this case does not require the (y|) trick, noKLUDGE */
+		INSERT(OPLUS_, pos);
+		ASTERN(O_PLUS, pos);
+		INSERT(OQUEST_, pos);
+		ASTERN(O_QUEST, pos);
+		break;
+	case '+':
+		INSERT(OPLUS_, pos);
+		ASTERN(O_PLUS, pos);
+		break;
+	case '?':
+		/* KLUDGE: emit y? as (y|) until subtle bug gets fixed */
+		INSERT(OCH_, pos);		/* offset slightly wrong */
+		ASTERN(OOR1, pos);		/* this one's right */
+		AHEAD(pos);			/* fix the OCH_ */
+		EMIT(OOR2, 0);			/* offset very wrong... */
+		AHEAD(THERE());			/* ...so fix it */
+		ASTERN(O_CH, THERETHERE());
+		break;
+	case '{':
+		count = p_count(p);
+		if (EAT(',')) {
+			if (isdigit((uch)PEEK())) {
+				count2 = p_count(p);
+				REQUIRE(count <= count2, REG_BADBR);
+			} else		/* single number with comma */
+				count2 = INFINITY;
+		} else		/* just a single number */
+			count2 = count;
+		repeat(p, pos, count, count2);
+		if (!EAT('}')) {	/* error heuristics */
+			while (MORE() && PEEK() != '}')
+				NEXT();
+			REQUIRE(MORE(), REG_EBRACE);
+			SETERROR(REG_BADBR);
+		}
+		break;
+	}
+
+	if (!MORE())
+		return;
+	c = PEEK();
+	if (!( c == '*' || c == '+' || c == '?' ||
+				(c == '{' && MORE2() && isdigit((uch)PEEK2())) ) )
+		return;
+	SETERROR(REG_BADRPT);
+}
+
+/*
+ - p_str - string (no metacharacters) "parser"
+ */
+static void
+p_str(struct parse *p)
+{
+	REQUIRE(MORE(), REG_EMPTY);
+	while (MORE())
+		ordinary(p, GETNEXT());
+}
+
+/*
+ - p_bre - BRE parser top level, anchoring and concatenation
+ * Giving end1 as OUT essentially eliminates the end1/end2 check.
+ *
+ * This implementation is a bit of a kludge, in that a trailing $ is first
+ * taken as an ordinary character and then revised to be an anchor.  The
+ * only undesirable side effect is that '$' gets included as a character
+ * category in such cases.  This is fairly harmless; not worth fixing.
+ * The amount of lookahead needed to avoid this kludge is excessive.
+ */
+static void
+p_bre(struct parse *p,
+    int end1,		/* first terminating character */
+    int end2)		/* second terminating character */
+{
+	sopno start = HERE();
+	int first = 1;			/* first subexpression? */
+	int wasdollar = 0;
+
+	if (EAT('^')) {
+		EMIT(OBOL, 0);
+		p->g->iflags |= USEBOL;
+		p->g->nbol++;
+	}
+	while (MORE() && !SEETWO(end1, end2)) {
+		wasdollar = p_simp_re(p, first);
+		first = 0;
+	}
+	if (wasdollar) {	/* oops, that was a trailing anchor */
+		DROP(1);
+		EMIT(OEOL, 0);
+		p->g->iflags |= USEEOL;
+		p->g->neol++;
+	}
+
+	REQUIRE(HERE() != start, REG_EMPTY);	/* require nonempty */
+}
+
+/*
+ - p_simp_re - parse a simple RE, an atom possibly followed by a repetition
+ */
+static int			/* was the simple RE an unbackslashed $? */
+p_simp_re(struct parse *p,
+    int starordinary)		/* is a leading * an ordinary character? */
+{
+	int c;
+	int count;
+	int count2;
+	sopno pos;
+	int i;
+	sopno subno;
+#	define	BACKSL	(1<<CHAR_BIT)
+
+	pos = HERE();		/* repetion op, if any, covers from here */
+
+	assert(MORE());		/* caller should have ensured this */
+	c = GETNEXT();
+	if (c == '\\') {
+		REQUIRE(MORE(), REG_EESCAPE);
+		c = BACKSL | GETNEXT();
+	}
+	switch (c) {
+	case '.':
+		if (p->g->cflags&REG_NEWLINE)
+			nonnewline(p);
+		else
+			EMIT(OANY, 0);
+		break;
+	case '[':
+		p_bracket(p);
+		break;
+	case BACKSL|'{':
+		SETERROR(REG_BADRPT);
+		break;
+	case BACKSL|'(':
+		p->g->nsub++;
+		subno = p->g->nsub;
+		if (subno < NPAREN)
+			p->pbegin[subno] = HERE();
+		EMIT(OLPAREN, subno);
+		/* the MORE here is an error heuristic */
+		if (MORE() && !SEETWO('\\', ')'))
+			p_bre(p, '\\', ')');
+		if (subno < NPAREN) {
+			p->pend[subno] = HERE();
+			assert(p->pend[subno] != 0);
+		}
+		EMIT(ORPAREN, subno);
+		REQUIRE(EATTWO('\\', ')'), REG_EPAREN);
+		break;
+	case BACKSL|')':	/* should not get here -- must be user */
+	case BACKSL|'}':
+		SETERROR(REG_EPAREN);
+		break;
+	case BACKSL|'1':
+	case BACKSL|'2':
+	case BACKSL|'3':
+	case BACKSL|'4':
+	case BACKSL|'5':
+	case BACKSL|'6':
+	case BACKSL|'7':
+	case BACKSL|'8':
+	case BACKSL|'9':
+		i = (c&~BACKSL) - '0';
+		assert(i < NPAREN);
+		if (p->pend[i] != 0) {
+			assert(i <= p->g->nsub);
+			EMIT(OBACK_, i);
+			asse