Replace regular expression package with Henry Spencer's latest version

(extracted from Tcl 8.4.1 release, as Henry still hasn't got round to
making it a separate library).  This solves a performance problem for
multibyte, as well as upgrading our regexp support to match recent Tcl
and nearly match recent Perl.

doc/src/sgml/release.sgml

@@ -1,5 +1,5 @@
 <!--
-$Header: /cvsroot/pgsql/doc/src/sgml/release.sgml,v 1.184 2003/02/02 23:46:38 tgl Exp $
+$Header: /cvsroot/pgsql/doc/src/sgml/release.sgml,v 1.185 2003/02/05 17:41:32 tgl Exp $
 -->
 
 <appendix id="release">
@@ -24,6 +24,7 @@ CDATA means the content is "SGML-free", so you can write without
 worries about funny characters.
 -->
 <literallayout><![CDATA[
+New regular expression package, many more regexp features (most of Perl5)
 Can now do EXPLAIN ... EXECUTE to see plan used for a prepared query
 Explicit JOINs no longer constrain query plan, unless JOIN_COLLAPSE_LIMIT = 1
 Performance of "foo IN (SELECT ...)" queries has been considerably improved

src/backend/regex/COPYRIGHT

@@ -1,56 +1,84 @@
-Copyright 1992, 1993, 1994 Henry Spencer.  All rights reserved.
+This regular expression package was originally developed by Henry Spencer.
-This software is not subject to any license of the American Telephone
+It bears the following copyright notice:
-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
+Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
-to the following restrictions:
+
-
+Development of this software was funded, in part, by Cray Research Inc.,
-1. The author is not responsible for the consequences of use of this
+UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
-   software, no matter how awful, even if they arise from flaws in it.
+Corporation, none of whom are responsible for the results.  The author
-
+thanks all of them. 
-2. The origin of this software must not be misrepresented, either by
+
-   explicit claim or by omission.  Since few users ever read sources,
+Redistribution and use in source and binary forms -- with or without
-   credits must appear in the documentation.
+modification -- are permitted for any purpose, provided that
-
+redistributions in source form retain this entire copyright notice and
-3. Altered versions must be plainly marked as such, and must not be
+indicate the origin and nature of any modifications.
-   misrepresented as being the original software.  Since few users
+
-   ever read sources, credits must appear in the documentation.
+I'd appreciate being given credit for this package in the documentation
-
+of software which uses it, but that is not a requirement.
-4. This notice may not be removed or altered.
+
-
+THIS SOFTWARE IS PROVIDED ``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
- * Copyright (c) 1994
+HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
- *	The Regents of the University of California.  All rights reserved.
+EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
- *
+PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
- * Redistribution and use in source and binary forms, with or without
+OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
- * modification, are permitted provided that the following conditions
+WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
- * are met:
+OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
- * 1. Redistributions of source code must retain the above copyright
+ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
- *    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.
+PostgreSQL adopted the code out of Tcl 8.4.1.  Portions of regc_locale.c
- * 3. All advertising materials mentioning features or use of this software
+and re_syntax.n were developed by Tcl developers other than Henry; these
- *    must display the following acknowledgement:
+files bear the Tcl copyright and license notice:
- *	This product includes software developed by the University of
+
- *	California, Berkeley and its contributors.
+**********************************************************************
- * 4. Neither the name of the University nor the names of its contributors
+
- *    may be used to endorse or promote products derived from this software
+This software is copyrighted by the Regents of the University of
- *    without specific prior written permission.
+California, Sun Microsystems, Inc., Scriptics Corporation, ActiveState
- *
+Corporation and other parties.  The following terms apply to all files
- * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+associated with the software unless explicitly disclaimed in
- * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+individual files.
- * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+
- * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+The authors hereby grant permission to use, copy, modify, distribute,
- * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+and license this software and its documentation for any purpose, provided
- * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+that existing copyright notices are retained in all copies and that this
- * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+notice is included verbatim in any distributions. No written agreement,
- * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+license, or royalty fee is required for any of the authorized uses.
- * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+Modifications to this software may be copyrighted by their authors
- * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+and need not follow the licensing terms described here, provided that
- * SUCH DAMAGE.
+the new terms are clearly indicated on the first page of each file where
- *
+they apply.
- *	@(#)COPYRIGHT	8.1 (Berkeley) 3/16/94
+
- */
+IN NO EVENT SHALL THE AUTHORS OR DISTRIBUTORS BE LIABLE TO ANY PARTY
+FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ARISING OUT OF THE USE OF THIS SOFTWARE, ITS DOCUMENTATION, OR ANY
+DERIVATIVES THEREOF, EVEN IF THE AUTHORS HAVE BEEN ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGE.
+
+THE AUTHORS AND DISTRIBUTORS SPECIFICALLY DISCLAIM ANY WARRANTIES,
+INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE, AND NON-INFRINGEMENT.  THIS SOFTWARE
+IS PROVIDED ON AN "AS IS" BASIS, AND THE AUTHORS AND DISTRIBUTORS HAVE
+NO OBLIGATION TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR
+MODIFICATIONS.
+
+GOVERNMENT USE: If you are acquiring this software on behalf of the
+U.S. government, the Government shall have only "Restricted Rights"
+in the software and related documentation as defined in the Federal 
+Acquisition Regulations (FARs) in Clause 52.227.19 (c) (2).  If you
+are acquiring the software on behalf of the Department of Defense, the
+software shall be classified as "Commercial Computer Software" and the
+Government shall have only "Restricted Rights" as defined in Clause
+252.227-7013 (c) (1) of DFARs.  Notwithstanding the foregoing, the
+authors grant the U.S. Government and others acting in its behalf
+permission to use and distribute the software in accordance with the
+terms specified in this license. 
+
+**********************************************************************
+
+Subsequent modifications to the code by the PostgreSQL project follow
+the same license terms as the rest of PostgreSQL.

src/backend/regex/Makefile

@@ -1,10 +1,10 @@
 #-------------------------------------------------------------------------
 #
 # Makefile--
-#    Makefile for regex
+#    Makefile for backend/regex
 #
 # IDENTIFICATION
-#    $Header: /cvsroot/pgsql/src/backend/regex/Makefile,v 1.19 2002/09/16 16:02:43 momjian Exp $
+#    $Header: /cvsroot/pgsql/src/backend/regex/Makefile,v 1.20 2003/02/05 17:41:32 tgl Exp $
 #
 #-------------------------------------------------------------------------
 
@@ -12,23 +12,17 @@ subdir = src/backend/regex
 top_builddir = ../../..
 include $(top_builddir)/src/Makefile.global
 
-override CPPFLAGS += -DPOSIX_MISTAKE 
-
 OBJS = regcomp.o regerror.o regexec.o regfree.o
-DEBUGOBJ += ../utils/mb/SUBSYS.o
 
 all: SUBSYS.o
 
 SUBSYS.o: $(OBJS)
 	$(LD) $(LDREL) $(LDOUT) SUBSYS.o $(OBJS)
 
-regexec.o: regexec.c engine.c
+# mark inclusion dependencies between .c files explicitly
+regcomp.o: regcomp.c regc_lex.c regc_color.c regc_nfa.c regc_cvec.c regc_locale.c
 
-# retest will not compile because multibyte is now enabled by default
+regexec.o: regexec.c rege_dfa.c
-# and the multibyte calls require /mmgr, /adt, and other calls that
-# are complex for linkage,  bjm 2002-09-16
-#retest: retest.o SUBSYS.o $(DEBUGOBJ)
-#	$(CC) $(CFLAGS) $(LDFLAGS) $^ $(LIBS) -o $@
 
 clean: 
-	rm -f SUBSYS.o $(OBJS) retest retest.o
+	rm -f SUBSYS.o $(OBJS)

src/backend/regex/WHATSNEW

@@ -1,94 +0,0 @@
-# @(#)WHATSNEW	8.3 (Berkeley) 3/18/94
-
-New in alpha3.4:  The complex bug alluded to below has been fixed (in a
-slightly kludgey temporary way that may hurt efficiency a bit; this is
-another "get it out the door for 4.4" release).  The tests at the end of
-the tests file have accordingly been uncommented.  The primary sign of
-the bug was that something like a?b matching ab matched b rather than ab.
-(The bug was essentially specific to this exact situation, else it would
-have shown up earlier.)
-
-New in alpha3.3:  The definition of word boundaries has been altered
-slightly, to more closely match the usual programming notion that "_"
-is an alphabetic.  Stuff used for pre-ANSI systems is now in a subdir,
-and the makefile no longer alludes to it in mysterious ways.  The
-makefile has generally been cleaned up some.  Fixes have been made
-(again!) so that the regression test will run without -DREDEBUG, at
-the cost of weaker checking.  A workaround for a bug in some folks'
-<assert.h> has been added.  And some more things have been added to
-tests, including a couple right at the end which are commented out
-because the code currently flunks them (complex bug; fix coming).
-Plus the usual minor cleanup.
-
-New in alpha3.2:  Assorted bits of cleanup and portability improvement
-(the development base is now a BSDI system using GCC instead of an ancient
-Sun system, and the newer compiler exposed some glitches).  Fix for a
-serious bug that affected REs using many [] (including REG_ICASE REs
-because of the way they are implemented), *sometimes*, depending on
-memory-allocation patterns.  The header-file prototypes no longer name
-the parameters, avoiding possible name conflicts.  The possibility that
-some clot has defined CHAR_MIN as (say) `-128' instead of `(-128)' is
-now handled gracefully.  "uchar" is no longer used as an internal type
-name (too many people have the same idea).  Still the same old lousy
-performance, alas.
-
-New in alpha3.1:  Basically nothing, this release is just a bookkeeping
-convenience.  Stay tuned.
-
-New in alpha3.0:  Performance is no better, alas, but some fixes have been
-made and some functionality has been added.  (This is basically the "get
-it out the door in time for 4.4" release.)  One bug fix:  regfree() didn't
-free the main internal structure (how embarrassing).  It is now possible
-to put NULs in either the RE or the target string, using (resp.) a new
-REG_PEND flag and the old REG_STARTEND flag.  The REG_NOSPEC flag to
-regcomp() makes all characters ordinary, so you can match a literal
-string easily (this will become more useful when performance improves!).
-There are now primitives to match beginnings and ends of words, although
-the syntax is disgusting and so is the implementation.  The REG_ATOI
-debugging interface has changed a bit.  And there has been considerable
-internal cleanup of various kinds.
-
-New in alpha2.3:  Split change list out of README, and moved flags notes
-into Makefile.  Macro-ized the name of regex(7) in regex(3), since it has
-to change for 4.4BSD.  Cleanup work in engine.c, and some new regression
-tests to catch tricky cases thereof.
-
-New in alpha2.2:  Out-of-date manpages updated.  Regerror() acquires two
-small extensions -- REG_ITOA and REG_ATOI -- which avoid debugging kludges
-in my own test program and might be useful to others for similar purposes.
-The regression test will now compile (and run) without REDEBUG.  The
-BRE \$ bug is fixed.  Most uses of "uchar" are gone; it's all chars now.
-Char/uchar parameters are now written int/unsigned, to avoid possible
-portability problems with unpromoted parameters.  Some unsigned casts have
-been introduced to minimize portability problems with shifting into sign
-bits.
-
-New in alpha2.1:  Lots of little stuff, cleanup and fixes.  The one big
-thing is that regex.h is now generated, using mkh, rather than being
-supplied in the distribution; due to circularities in dependencies,
-you have to build regex.h explicitly by "make h".  The two known bugs
-have been fixed (and the regression test now checks for them), as has a
-problem with assertions not being suppressed in the absence of REDEBUG.
-No performance work yet.
-
-New in alpha2:  Backslash-anything is an ordinary character, not an
-error (except, of course, for the handful of backslashed metacharacters
-in BREs), which should reduce script breakage.  The regression test
-checks *where* null strings are supposed to match, and has generally
-been tightened up somewhat.  Small bug fixes in parameter passing (not
-harmful, but technically errors) and some other areas.  Debugging
-invoked by defining REDEBUG rather than not defining NDEBUG.
-
-New in alpha+3:  full prototyping for internal routines, using a little
-helper program, mkh, which extracts prototypes given in stylized comments.
-More minor cleanup.  Buglet fix:  it's CHAR_BIT, not CHAR_BITS.  Simple
-pre-screening of input when a literal string is known to be part of the
-RE; this does wonders for performance.
-
-New in alpha+2:  minor bits of cleanup.  Notably, the number "32" for the
-word width isn't hardwired into regexec.c any more, the public header
-file prototypes the functions if __STDC__ is defined, and some small typos
-in the manpages have been fixed.
-
-New in alpha+1:  improvements to the manual pages, and an important
-extension, the REG_STARTEND option to regexec().

src/backend/regex/re_format.7

@@ -1,269 +0,0 @@
-.\" 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. All advertising materials mentioning features or use of this software
-.\"    must display the following acknowledgement:
-.\"	This product includes software developed by the University of
-.\"	California, Berkeley and its contributors.
-.\" 4. Neither the name of the University nor the names of its contributors
-.\"    may be used to endorse or promote products derived from this software
-.\"    without specific prior written permission.
-.\"
-.\" 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
-.\"
-.TH RE_FORMAT 7 "March 20, 1994"
-.SH NAME
-re_format \- POSIX 1003.2 regular expressions
-.SH DESCRIPTION
-Regular expressions (``RE''s),
-as defined in POSIX 1003.2, come in two forms:
-modern REs (roughly those of
-.IR egrep ;
-1003.2 calls these ``extended'' REs)
-and obsolete REs (roughly those of
-.IR ed ;
-1003.2 ``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;
-`\(dg' marks decisions on these aspects that
-may not be fully portable to other 1003.2 implementations.
-.PP
-A (modern) RE is one\(dg or more non-empty\(dg \fIbranches\fR,
-separated by `|'.
-It matches anything that matches one of the branches.
-.PP
-A branch is one\(dg or more \fIpieces\fR, concatenated.
-It matches a match for the first, followed by a match for the second, etc.
-.PP
-A piece is an \fIatom\fR possibly followed
-by a single\(dg `*', `+', `?', or \fIbound\fR.
-An atom followed by `*' matches a sequence of 0 or more matches of the atom.
-An atom followed by `+' matches a sequence of 1 or more matches of the atom.
-An atom followed by `?' matches a sequence of 0 or 1 matches of the atom.
-.PP
-A \fIbound\fR is `{' followed by an unsigned decimal integer,
-possibly followed by `,'
-possibly followed by another unsigned decimal integer,
-always followed by `}'.
-The integers must lie between 0 and RE_DUP_MAX (255\(dg) 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
-and no comma matches
-a sequence of exactly \fIi\fR matches of the atom.
-An atom followed by a bound
-containing one integer \fIi\fR and a comma matches
-a sequence of \fIi\fR or more matches of the atom.
-An atom followed by a bound
-containing two integers \fIi\fR and \fIj\fR matches
-a sequence of \fIi\fR through \fIj\fR (inclusive) matches of the atom.
-.PP
-An atom is a regular expression enclosed in `()' (matching a match for the
-regular expression),
-an empty set of `()' (matching the null string)\(dg,
-a \fIbracket expression\fR (see below), `.'
-(matching any single character), `^' (matching the null string at the
-beginning of a line), `$' (matching the null string at the
-end of a line), a `\e' followed by one of the characters
-`^.[$()|*+?{\e'
-(matching that character taken as an ordinary character),
-a `\e' followed by any other character\(dg
-(matching that character taken as an ordinary character,
-as if the `\e' had not been present\(dg),
-or a single character with no other significance (matching that character).
-A `{' followed by a character other than a digit is an ordinary
-character, not the beginning of a bound\(dg.
-It is illegal to end an RE with `\e'.
-.PP
-A \fIbracket expression\fR is a list of characters enclosed in `[]'.
-It normally matches any single character from the list (but see below).
-If the list begins with `^',
-it matches any single character
-(but see below) \fInot\fR from the rest of the list.
-If two characters in the list are separated by `\-', this is shorthand
-for the full \fIrange\fR of characters between those two (inclusive) in the
-collating sequence,
-e.g. `[0-9]' in ASCII matches any decimal digit.
-It is illegal\(dg for two ranges to share an
-endpoint, e.g. `a-c-e'.
-Ranges are very collating-sequence-dependent,
-and portable programs should avoid relying on them.
-.PP
-To include a literal `]' in the list, make it the first character
-(following a possible `^').
-To include a literal `\-', 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,
-enclose it in `[.' and `.]' to make it a collating element (see below).
-With the exception of these and some combinations using `[' (see next
-paragraphs), all other special characters, including `\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
-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,
-then the RE `[[.ch.]]*c' matches the first five characters
-of `chchcc'.
-.PP
-Within a bracket expression, a collating element enclosed in `[=' and
-`=]' 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 `.]'.)
-For example, if o and \o'o^' are the members of an equivalence class,
-then `[[=o=]]', `[[=\o'o^'=]]', and `[o\o'o^']' are all synonymous.
-An equivalence class may not\(dg be an endpoint
-of a range.
-.PP
-Within a bracket expression, the name of a \fIcharacter class\fR enclosed
-in `[:' and `:]' stands for the list of all characters belonging to that
-class.
-Standard character class names are:
-.PP
-.RS
-.nf
-.ta 3c 6c 9c
-alnum	digit	punct
-alpha	graph	space
-blank	lower	upper
-cntrl	print	xdigit
-.fi
-.RE
-.PP
-These stand for the character classes defined in
-.IR ctype (3).
-A locale may provide others.
-A character class may not be used as an endpoint of a range.
-.PP
-There are two special cases\(dg of bracket expressions:
-the bracket expressions `[[:<:]]' and `[[:>:]]' 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
-character (as defined by
-.IR ctype (3))
-or an underscore.
-This is an extension,
-compatible with but not specified by POSIX 1003.2,
-and should be used with
-caution in software intended to be portable to other systems.
-.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.
-If the RE could match more than one substring starting at that point,
-it matches the longest.
-Subexpressions also match the longest possible substrings, subject to
-the constraint that the whole match be as long as possible,
-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
-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',
-`(wee|week)(knights|nights)' matches all ten characters of `weeknights',
-when `(.*).*' is matched against `abc' the parenthesized subexpression
-matches all three characters, and
-when `(a*)*' is matched against `bc' both the whole RE and the parenthesized
-subexpression match the null string.
-.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]'.
-When it appears inside a bracket expression, all case counterparts
-of it are added to the bracket expression, so that (e.g.) `[x]'
-becomes `[xX]' and `[^x]' becomes `[^xX]'.
-.PP
-No particular limit is imposed on the length of REs\(dg.
-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
-Obsolete (``basic'') regular expressions differ in several respects.
-`|', `+', and `?' are ordinary characters and there is no equivalent
-for their functionality.
-The delimiters for bounds are `\e{' and `\e}',
-with `{' and `}' by themselves ordinary characters.
-The parentheses for nested subexpressions are `\e(' and `\e)',
-with `(' and `)' by themselves ordinary characters.
-`^' is an ordinary character except at the beginning of the
-RE or\(dg the beginning of a parenthesized subexpression,
-`$' is an ordinary character except at the end of the
-RE or\(dg the end of a parenthesized subexpression,
-and `*' is an ordinary character if it appears at the beginning of the
-RE or the beginning of a parenthesized subexpression
-(after a possible leading `^').
-Finally, there is one new type of atom, a \fIback reference\fR:
-`\e' followed by a non-zero decimal digit \fId\fR
-matches the same sequence of characters
-matched by the \fId\fRth 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'.
-.SH SEE ALSO
-regex(3)
-.PP
-POSIX 1003.2, section 2.8 (Regular Expression Notation).
-.SH BUGS
-Having two kinds of REs is a botch.
-.PP
-The current 1003.2 spec says that `)' is an ordinary character in
-the absence of an unmatched `(';
-this was an unintentional result of a wording error,
-and change is likely.
-Avoid relying on it.
-.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'?).
-Avoid using them.
-.PP
-1003.2's specification of case-independent matching is vague.
-The ``one case implies all cases'' definition given above
-is current consensus among implementors as to the right interpretation.
-.PP
-The syntax for word boundaries is incredibly ugly.

src/backend/regex/re_syntax.n

@@ -0,0 +1,970 @@
+'\"
+'\" Copyright (c) 1998 Sun Microsystems, Inc.
+'\" Copyright (c) 1999 Scriptics Corporation
+'\"
+'\" This software is copyrighted by the Regents of the University of
+'\" California, Sun Microsystems, Inc., Scriptics Corporation, ActiveState
+'\" Corporation and other parties.  The following terms apply to all files
+'\" associated with the software unless explicitly disclaimed in
+'\" individual files.
+'\" 
+'\" The authors hereby grant permission to use, copy, modify, distribute,
+'\" and license this software and its documentation for any purpose, provided
+'\" that existing copyright notices are retained in all copies and that this
+'\" notice is included verbatim in any distributions. No written agreement,
+'\" license, or royalty fee is required for any of the authorized uses.
+'\" Modifications to this software may be copyrighted by their authors
+'\" and need not follow the licensing terms described here, provided that
+'\" the new terms are clearly indicated on the first page of each file where
+'\" they apply.
+'\" 
+'\" IN NO EVENT SHALL THE AUTHORS OR DISTRIBUTORS BE LIABLE TO ANY PARTY
+'\" FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+'\" ARISING OUT OF THE USE OF THIS SOFTWARE, ITS DOCUMENTATION, OR ANY
+'\" DERIVATIVES THEREOF, EVEN IF THE AUTHORS HAVE BEEN ADVISED OF THE
+'\" POSSIBILITY OF SUCH DAMAGE.
+'\" 
+'\" THE AUTHORS AND DISTRIBUTORS SPECIFICALLY DISCLAIM ANY WARRANTIES,
+'\" INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY,
+'\" FITNESS FOR A PARTICULAR PURPOSE, AND NON-INFRINGEMENT.  THIS SOFTWARE
+'\" IS PROVIDED ON AN "AS IS" BASIS, AND THE AUTHORS AND DISTRIBUTORS HAVE
+'\" NO OBLIGATION TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR
+'\" MODIFICATIONS.
+'\" 
+'\" GOVERNMENT USE: If you are acquiring this software on behalf of the
+'\" U.S. government, the Government shall have only "Restricted Rights"
+'\" in the software and related documentation as defined in the Federal 
+'\" Acquisition Regulations (FARs) in Clause 52.227.19 (c) (2).  If you
+'\" are acquiring the software on behalf of the Department of Defense, the
+'\" software shall be classified as "Commercial Computer Software" and the
+'\" Government shall have only "Restricted Rights" as defined in Clause
+'\" 252.227-7013 (c) (1) of DFARs.  Notwithstanding the foregoing, the
+'\" authors grant the U.S. Government and others acting in its behalf
+'\" permission to use and distribute the software in accordance with the
+'\" terms specified in this license. 
+'\" 
+'\" RCS: @(#) Id: re_syntax.n,v 1.3 1999/07/14 19:09:36 jpeek Exp 
+'\"
+.so man.macros
+.TH re_syntax n "8.1" Tcl "Tcl Built-In Commands"
+.BS
+.SH NAME
+re_syntax \- Syntax of Tcl regular expressions.
+.BE
+
+.SH DESCRIPTION
+.PP
+A \fIregular expression\fR describes strings of characters.
+It's a pattern that matches certain strings and doesn't match others.
+
+.SH "DIFFERENT FLAVORS OF REs"
+Regular expressions (``RE''s), as defined by POSIX, come in two
+flavors: \fIextended\fR REs (``EREs'') and \fIbasic\fR REs (``BREs'').
+EREs are roughly those of the traditional \fIegrep\fR, while BREs are
+roughly those of the traditional \fIed\fR.  This implementation adds
+a third flavor, \fIadvanced\fR REs (``AREs''), basically EREs with
+some significant extensions.
+.PP
+This manual page primarily describes AREs.  BREs mostly exist for
+backward compatibility in some old programs; they will be discussed at
+the end.  POSIX EREs are almost an exact subset of AREs.  Features of
+AREs that are not present in EREs will be indicated.
+
+.SH "REGULAR EXPRESSION SYNTAX"
+.PP
+Tcl regular expressions are implemented using the package written by
+Henry Spencer, based on the 1003.2 spec and some (not quite all) of
+the Perl5 extensions (thanks, Henry!).  Much of the description of
+regular expressions below is copied verbatim from his manual entry.
+.PP
+An ARE is one or more \fIbranches\fR,
+separated by `\fB|\fR',
+matching anything that matches any of the branches.
+.PP
+A branch is zero or more \fIconstraints\fR or \fIquantified atoms\fR,
+concatenated.
+It matches a match for the first, followed by a match for the second, etc;
+an empty branch matches the empty string.
+.PP
+A quantified atom is an \fIatom\fR possibly followed
+by a single \fIquantifier\fR.
+Without a quantifier, it matches a match for the atom.
+The quantifiers,
+and what a so-quantified atom matches, are:
+.RS 2
+.TP 6
+\fB*\fR
+a sequence of 0 or more matches of the atom
+.TP
+\fB+\fR
+a sequence of 1 or more matches of the atom
+.TP
+\fB?\fR
+a sequence of 0 or 1 matches of the atom
+.TP
+\fB{\fIm\fB}\fR
+a sequence of exactly \fIm\fR matches of the atom
+.TP
+\fB{\fIm\fB,}\fR
+a sequence of \fIm\fR or more matches of the atom
+.TP
+\fB{\fIm\fB,\fIn\fB}\fR
+a sequence of \fIm\fR through \fIn\fR (inclusive) matches of the atom;
+\fIm\fR may not exceed \fIn\fR
+.TP
+\fB*?  +?  ??  {\fIm\fB}?  {\fIm\fB,}?  {\fIm\fB,\fIn\fB}?\fR
+\fInon-greedy\fR quantifiers,
+which match the same possibilities,
+but prefer the smallest number rather than the largest number
+of matches (see MATCHING)
+.RE
+.PP
+The forms using
+\fB{\fR and \fB}\fR
+are known as \fIbound\fRs.
+The numbers
+\fIm\fR and \fIn\fR are unsigned decimal integers
+with permissible values from 0 to 255 inclusive.
+.PP
+An atom is one of:
+.RS 2
+.TP 6
+\fB(\fIre\fB)\fR
+(where \fIre\fR is any regular expression)
+matches a match for
+\fIre\fR, with the match noted for possible reporting
+.TP
+\fB(?:\fIre\fB)\fR
+as previous,
+but does no reporting
+(a ``non-capturing'' set of parentheses)
+.TP
+\fB()\fR
+matches an empty string,
+noted for possible reporting
+.TP
+\fB(?:)\fR
+matches an empty string,
+without reporting
+.TP
+\fB[\fIchars\fB]\fR
+a \fIbracket expression\fR,
+matching any one of the \fIchars\fR (see BRACKET EXPRESSIONS for more detail)
+.TP
+ \fB.\fR
+matches any single character
+.TP
+\fB\e\fIk\fR
+(where \fIk\fR is a non-alphanumeric character)
+matches that character taken as an ordinary character,
+e.g. \e\e matches a backslash character
+.TP
+\fB\e\fIc\fR
+where \fIc\fR is alphanumeric
+(possibly followed by other characters),
+an \fIescape\fR (AREs only),
+see ESCAPES below
+.TP
+\fB{\fR
+when followed by a character other than a digit,
+matches the left-brace character `\fB{\fR';
+when followed by a digit, it is the beginning of a
+\fIbound\fR (see above)
+.TP
+\fIx\fR
+where \fIx\fR is
+a single character with no other significance, matches that character.
+.RE
+.PP
+A \fIconstraint\fR matches an empty string when specific conditions
+are met.
+A constraint may not be followed by a quantifier.
+The simple constraints are as follows; some more constraints are
+described later, under ESCAPES.
+.RS 2
+.TP 8
+\fB^\fR
+matches at the beginning of a line
+.TP
+\fB$\fR
+matches at the end of a line
+.TP
+\fB(?=\fIre\fB)\fR
+\fIpositive lookahead\fR (AREs only), matches at any point
+where a substring matching \fIre\fR begins
+.TP
+\fB(?!\fIre\fB)\fR
+\fInegative lookahead\fR (AREs only), matches at any point
+where no substring matching \fIre\fR begins
+.RE
+.PP
+The lookahead constraints may not contain back references (see later),
+and all parentheses within them are considered non-capturing.
+.PP
+An RE may not end with `\fB\e\fR'.
+
+.SH "BRACKET EXPRESSIONS"
+A \fIbracket expression\fR is a list of characters enclosed in `\fB[\|]\fR'.
+It normally matches any single character from the list (but see below).
+If the list begins with `\fB^\fR',
+it matches any single character
+(but see below) \fInot\fR from the rest of the list.
+.PP
+If two characters in the list are separated by `\fB\-\fR',
+this is shorthand
+for the full \fIrange\fR of characters between those two (inclusive) in the
+collating sequence,
+e.g.
+\fB[0\-9]\fR
+in ASCII matches any decimal digit.
+Two ranges may not share an
+endpoint, so e.g.
+\fBa\-c\-e\fR
+is illegal.
+Ranges are very collating-sequence-dependent,
+and portable programs should avoid relying on them.
+.PP
+To include a literal
+\fB]\fR
+or
+\fB\-\fR
+in the list,
+the simplest method is to
+enclose it in
+\fB[.\fR and \fB.]\fR
+to make it a collating element (see below).
+Alternatively,
+make it the first character
+(following a possible `\fB^\fR'),
+or (AREs only) precede it with `\fB\e\fR'.
+Alternatively, for `\fB\-\fR',
+make it the last character,
+or the second endpoint of a range.
+To use a literal
+\fB\-\fR
+as the first endpoint of a range,
+make it a collating element
+or (AREs only) precede it with `\fB\e\fR'.
+With the exception of these, some combinations using
+\fB[\fR
+(see next
+paragraphs), and escapes,
+all other special characters 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
+\fB[.\fR and \fB.]\fR
+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 in a locale that has
+multi-character collating elements
+can thus match more than one character.
+.VS 8.2
+So (insidiously), a bracket expression that starts with \fB^\fR
+can match multi-character collating elements even if none of them
+appear in the bracket expression!
+(\fINote:\fR Tcl currently has no multi-character collating elements.
+This information is only for illustration.)
+.PP
+For example, assume the collating sequence includes a \fBch\fR
+multi-character collating element.
+Then the RE \fB[[.ch.]]*c\fR (zero or more \fBch\fP's followed by \fBc\fP)
+matches the first five characters of `\fBchchcc\fR'.
+Also, the RE \fB[^c]b\fR matches all of `\fBchb\fR'
+(because \fB[^c]\fR matches the multi-character \fBch\fR).
+.VE 8.2
+.PP
+Within a bracket expression, a collating element enclosed in
+\fB[=\fR
+and
+\fB=]\fR
+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 `\fB[.\fR'\&
+and `\fB.]\fR'.)
+For example, if
+\fBo\fR
+and
+\fB\o'o^'\fR
+are the members of an equivalence class,
+then `\fB[[=o=]]\fR', `\fB[[=\o'o^'=]]\fR',
+and `\fB[o\o'o^']\fR'\&
+are all synonymous.
+An equivalence class may not be an endpoint
+of a range.
+.VS 8.2
+(\fINote:\fR 
+Tcl currently implements only the Unicode locale.
+It doesn't define any equivalence classes.
+The examples above are just illustrations.)
+.VE 8.2
+.PP
+Within a bracket expression, the name of a \fIcharacter class\fR enclosed
+in
+\fB[:\fR
+and
+\fB:]\fR
+stands for the list of all characters
+(not all collating elements!)
+belonging to that
+class.
+Standard character classes are:
+.PP
+.RS
+.ne 5
+.nf
+.ta 3c
+\fBalpha\fR	A letter. 
+\fBupper\fR	An upper-case letter. 
+\fBlower\fR	A lower-case letter. 
+\fBdigit\fR	A decimal digit. 
+\fBxdigit\fR	A hexadecimal digit. 
+\fBalnum\fR	An alphanumeric (letter or digit). 
+\fBprint\fR	An alphanumeric (same as alnum).
+\fBblank\fR	A space or tab character.
+\fBspace\fR	A character producing white space in displayed text. 
+\fBpunct\fR	A punctuation character. 
+\fBgraph\fR	A character with a visible representation. 
+\fBcntrl\fR	A control character. 
+.fi
+.RE
+.PP
+A locale may provide others.
+.VS 8.2
+(Note that the current Tcl implementation has only one locale:
+the Unicode locale.)
+.VE 8.2
+A character class may not be used as an endpoint of a range.
+.PP
+There are two special cases of bracket expressions:
+the bracket expressions
+\fB[[:<:]]\fR
+and
+\fB[[:>:]]\fR
+are constraints, matching empty strings at
+the beginning and end of a word respectively.
+'\" note, discussion of escapes below references this definition of word
+A word is defined as a sequence of
+word characters
+that is neither preceded nor followed by
+word characters.
+A word character is an
+\fIalnum\fR
+character
+or an underscore
+(\fB_\fR).
+These special bracket expressions are deprecated;
+users of AREs should use constraint escapes instead (see below).
+.SH ESCAPES
+Escapes (AREs only), which begin with a
+\fB\e\fR
+followed by an alphanumeric character,
+come in several varieties:
+character entry, class shorthands, constraint escapes, and back references.
+A
+\fB\e\fR
+followed by an alphanumeric character but not constituting
+a valid escape is illegal in AREs.
+In EREs, there are no escapes:
+outside a bracket expression,
+a
+\fB\e\fR
+followed by an alphanumeric character merely stands for that
+character as an ordinary character,
+and inside a bracket expression,
+\fB\e\fR
+is an ordinary character.
+(The latter is the one actual incompatibility between EREs and AREs.)
+.PP
+Character-entry escapes (AREs only) exist to make it easier to specify
+non-printing and otherwise inconvenient characters in REs:
+.RS 2
+.TP 5
+\fB\ea\fR
+alert (bell) character, as in C
+.TP
+\fB\eb\fR
+backspace, as in C
+.TP
+\fB\eB\fR
+synonym for
+\fB\e\fR
+to help reduce backslash doubling in some
+applications where there are multiple levels of backslash processing
+.TP
+\fB\ec\fIX\fR
+(where X is any character) the character whose
+low-order 5 bits are the same as those of
+\fIX\fR,
+and whose other bits are all zero
+.TP
+\fB\ee\fR
+the character whose collating-sequence name
+is `\fBESC\fR',
+or failing that, the character with octal value 033
+.TP
+\fB\ef\fR
+formfeed, as in C
+.TP
+\fB\en\fR
+newline, as in C
+.TP
+\fB\er\fR
+carriage return, as in C
+.TP
+\fB\et\fR
+horizontal tab, as in C
+.TP
+\fB\eu\fIwxyz\fR
+(where
+\fIwxyz\fR
+is exactly four hexadecimal digits)
+the Unicode character
+\fBU+\fIwxyz\fR
+in the local byte ordering
+.TP
+\fB\eU\fIstuvwxyz\fR
+(where
+\fIstuvwxyz\fR
+is exactly eight hexadecimal digits)
+reserved for a somewhat-hypothetical Unicode extension to 32 bits
+.TP
+\fB\ev\fR
+vertical tab, as in C
+are all available.
+.TP
+\fB\ex\fIhhh\fR
+(where
+\fIhhh\fR
+is any sequence of hexadecimal digits)
+the character whose hexadecimal value is
+\fB0x\fIhhh\fR
+(a single character no matter how many hexadecimal digits are used).
+.TP
+\fB\e0\fR
+the character whose value is
+\fB0\fR
+.TP
+\fB\e\fIxy\fR
+(where
+\fIxy\fR
+is exactly two octal digits,
+and is not a
+\fIback reference\fR (see below))
+the character whose octal value is
+\fB0\fIxy\fR
+.TP
+\fB\e\fIxyz\fR
+(where
+\fIxyz\fR
+is exactly three octal digits,
+and is not a
+back reference (see below))
+the character whose octal value is
+\fB0\fIxyz\fR
+.RE
+.PP
+Hexadecimal digits are `\fB0\fR'-`\fB9\fR', `\fBa\fR'-`\fBf\fR',
+and `\fBA\fR'-`\fBF\fR'.
+Octal digits are `\fB0\fR'-`\fB7\fR'.
+.PP
+The character-entry escapes are always taken as ordinary characters.
+For example,
+\fB\e135\fR
+is
+\fB]\fR
+in ASCII,
+but
+\fB\e135\fR
+does not terminate a bracket expression.
+Beware, however, that some applications (e.g., C compilers) interpret 
+such sequences themselves before the regular-expression package
+gets to see them, which may require doubling (quadrupling, etc.) the `\fB\e\fR'.
+.PP
+Class-shorthand escapes (AREs only) provide shorthands for certain commonly-used
+character classes:
+.RS 2
+.TP 10
+\fB\ed\fR
+\fB[[:digit:]]\fR
+.TP
+\fB\es\fR
+\fB[[:space:]]\fR
+.TP
+\fB\ew\fR
+\fB[[:alnum:]_]\fR
+(note underscore)
+.TP
+\fB\eD\fR
+\fB[^[:digit:]]\fR
+.TP
+\fB\eS\fR
+\fB[^[:space:]]\fR
+.TP
+\fB\eW\fR
+\fB[^[:alnum:]_]\fR
+(note underscore)
+.RE
+.PP
+Within bracket expressions, `\fB\ed\fR', `\fB\es\fR',
+and `\fB\ew\fR'\&
+lose their outer brackets,
+and `\fB\eD\fR', `\fB\eS\fR',
+and `\fB\eW\fR'\&
+are illegal.
+.VS 8.2
+(So, for example, \fB[a-c\ed]\fR is equivalent to \fB[a-c[:digit:]]\fR.
+Also, \fB[a-c\eD]\fR, which is equivalent to \fB[a-c^[:digit:]]\fR, is illegal.)
+.VE 8.2
+.PP
+A constraint escape (AREs only) is a constraint,
+matching the empty string if specific conditions are met,
+written as an escape:
+.RS 2
+.TP 6
+\fB\eA\fR
+matches only at the beginning of the string
+(see MATCHING, below, for how this differs from `\fB^\fR')
+.TP
+\fB\em\fR
+matches only at the beginning of a word
+.TP
+\fB\eM\fR
+matches only at the end of a word
+.TP
+\fB\ey\fR
+matches only at the beginning or end of a word
+.TP
+\fB\eY\fR
+matches only at a point that is not the beginning or end of a word
+.TP
+\fB\eZ\fR
+matches only at the end of the string
+(see MATCHING, below, for how this differs from `\fB$\fR')
+.TP
+\fB\e\fIm\fR
+(where
+\fIm\fR
+is a nonzero digit) a \fIback reference\fR, see below
+.TP
+\fB\e\fImnn\fR
+(where
+\fIm\fR
+is a nonzero digit, and
+\fInn\fR
+is some more digits,
+and the decimal value
+\fImnn\fR
+is not greater than the number of closing capturing parentheses seen so far)
+a \fIback reference\fR, see below
+.RE
+.PP
+A word is defined as in the specification of
+\fB[[:<:]]\fR
+and
+\fB[[:>:]]\fR
+above.
+Constraint escapes are illegal within bracket expressions.
+.PP
+A back reference (AREs only) matches the same string matched by the parenthesized
+subexpression specified by the number,
+so that (e.g.)
+\fB([bc])\e1\fR
+matches
+\fBbb\fR
+or
+\fBcc\fR
+but not `\fBbc\fR'.
+The subexpression must entirely precede the back reference in the RE.
+Subexpressions are numbered in the order of their leading parentheses.
+Non-capturing parentheses do not define subexpressions.
+.PP
+There is an inherent historical ambiguity between octal character-entry 
+escapes and back references, which is resolved by heuristics,
+as hinted at above.
+A leading zero always indicates an octal escape.
+A single non-zero digit, not followed by another digit,
+is always taken as a back reference.
+A multi-digit sequence not starting with a zero is taken as a back 
+reference if it comes after a suitable subexpression
+(i.e. the number is in the legal range for a back reference),
+and otherwise is taken as octal.
+.SH "METASYNTAX"
+In addition to the main syntax described above, there are some special
+forms and miscellaneous syntactic facilities available.
+.PP
+Normally the flavor of RE being used is specified by
+application-dependent means.
+However, this can be overridden by a \fIdirector\fR.
+If an RE of any flavor begins with `\fB***:\fR',
+the rest of the RE is an ARE.
+If an RE of any flavor begins with `\fB***=\fR',
+the rest of the RE is taken to be a literal string,
+with all characters considered ordinary characters.
+.PP
+An ARE may begin with \fIembedded options\fR:
+a sequence
+\fB(?\fIxyz\fB)\fR
+(where
+\fIxyz\fR
+is one or more alphabetic characters)
+specifies options affecting the rest of the RE.
+These supplement, and can override,
+any options specified by the application.
+The available option letters are:
+.RS 2
+.TP 3
+\fBb\fR
+rest of RE is a BRE
+.TP 3
+\fBc\fR
+case-sensitive matching (usual default)
+.TP 3
+\fBe\fR
+rest of RE is an ERE
+.TP 3
+\fBi\fR
+case-insensitive matching (see MATCHING, below)
+.TP 3
+\fBm\fR
+historical synonym for
+\fBn\fR
+.TP 3
+\fBn\fR
+newline-sensitive matching (see MATCHING, below)
+.TP 3
+\fBp\fR
+partial newline-sensitive matching (see MATCHING, below)
+.TP 3
+\fBq\fR
+rest of RE is a literal (``quoted'') string, all ordinary characters
+.TP 3
+\fBs\fR
+non-newline-sensitive matching (usual default)
+.TP 3
+\fBt\fR
+tight syntax (usual default; see below)
+.TP 3
+\fBw\fR
+inverse partial newline-sensitive (``weird'') matching (see MATCHING, below)
+.TP 3
+\fBx\fR
+expanded syntax (see below)
+.RE
+.PP
+Embedded options take effect at the
+\fB)\fR
+terminating the sequence.
+They are available only at the start of an ARE,
+and may not be used later within it.
+.PP
+In addition to the usual (\fItight\fR) RE syntax, in which all characters are
+significant, there is an \fIexpanded\fR syntax,
+available in all flavors of RE
+with the \fB-expanded\fR switch, or in AREs with the embedded x option.
+In the expanded syntax,
+white-space characters are ignored
+and all characters between a
+\fB#\fR
+and the following newline (or the end of the RE) are ignored,
+permitting paragraphing and commenting a complex RE.
+There are three exceptions to that basic rule:
+.RS 2
+.PP
+a white-space character or `\fB#\fR' preceded by `\fB\e\fR' is retained
+.PP
+white space or `\fB#\fR' within a bracket expression is retained
+.PP
+white space and comments are illegal within multi-character symbols
+like the ARE `\fB(?:\fR' or the BRE `\fB\e(\fR'
+.RE
+.PP
+Expanded-syntax white-space characters are blank, tab, newline, and
+.VS 8.2
+any character that belongs to the \fIspace\fR character class.
+.VE 8.2
+.PP
+Finally, in an ARE,
+outside bracket expressions, the sequence `\fB(?#\fIttt\fB)\fR'
+(where
+\fIttt\fR
+is any text not containing a `\fB)\fR')
+is a comment,
+completely ignored.
+Again, this is not allowed between the characters of
+multi-character symbols like `\fB(?:\fR'.
+Such comments are more a historical artifact than a useful facility,
+and their use is deprecated;
+use the expanded syntax instead.
+.PP
+\fINone\fR of these metasyntax extensions is available if the application
+(or an initial
+\fB***=\fR
+director)
+has specified that the user's input be treated as a literal string
+rather than as an RE.
+.SH MATCHING
+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.
+If the RE could match more than one substring starting at that point,
+its choice is determined by its \fIpreference\fR:
+either the longest substring, or the shortest.
+.PP
+Most atoms, and all constraints, have no preference.
+A parenthesized RE has the same preference (possibly none) as the RE.
+A quantified atom with quantifier
+\fB{\fIm\fB}\fR
+or
+\fB{\fIm\fB}?\fR
+has the same preference (possibly none) as the atom itself.
+A quantified atom with other normal quantifiers (including
+\fB{\fIm\fB,\fIn\fB}\fR
+with
+\fIm\fR
+equal to
+\fIn\fR)
+prefers longest match.
+A quantified atom with other non-greedy quantifiers (including
+\fB{\fIm\fB,\fIn\fB}?\fR
+with
+\fIm\fR
+equal to
+\fIn\fR)
+prefers shortest match.
+A branch has the same preference as the first quantified atom in it
+which has a preference.
+An RE consisting of two or more branches connected by the
+\fB|\fR
+operator prefers longest match.
+.PP
+Subject to the constraints imposed by the rules for matching the whole RE,
+subexpressions also match the longest or shortest possible substrings,
+based on their preferences,
+with subexpressions starting earlier in the RE taking priority over
+ones starting later.
+Note that outer subexpressions thus take priority over
+their component subexpressions.
+.PP
+Note that the quantifiers
+\fB{1,1}\fR
+and
+\fB{1,1}?\fR
+can be used to force longest and shortest preference, respectively,
+on a subexpression or a whole RE.
+.PP
+Match lengths are measured in characters, not collating elements.
+An empty string is considered longer than no match at all.
+For example,
+\fBbb*\fR
+matches the three middle characters of `\fBabbbc\fR',
+\fB(week|wee)(night|knights)\fR
+matches all ten characters of `\fBweeknights\fR',
+when
+\fB(.*).*\fR
+is matched against
+\fBabc\fR
+the parenthesized subexpression
+matches all three characters, and
+when
+\fB(a*)*\fR
+is matched against
+\fBbc\fR
+both the whole RE and the parenthesized
+subexpression match an empty string.
+.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,
+so that
+\fBx\fR
+becomes `\fB[xX]\fR'.
+When it appears inside a bracket expression, all case counterparts
+of it are added to the bracket expression, so that
+\fB[x]\fR
+becomes
+\fB[xX]\fR
+and
+\fB[^x]\fR
+becomes `\fB[^xX]\fR'.
+.PP
+If newline-sensitive matching is specified, \fB.\fR
+and bracket expressions using
+\fB^\fR
+will never match the newline character
+(so that matches will never cross newlines unless the RE
+explicitly arranges it)
+and
+\fB^\fR
+and
+\fB$\fR
+will match the empty string after and before a newline
+respectively, in addition to matching at beginning and end of string
+respectively.
+ARE
+\fB\eA\fR
+and
+\fB\eZ\fR
+continue to match beginning or end of string \fIonly\fR.
+.PP
+If partial newline-sensitive matching is specified,
+this affects \fB.\fR
+and bracket expressions
+as with newline-sensitive matching, but not
+\fB^\fR
+and `\fB$\fR'.
+.PP
+If inverse partial newline-sensitive matching is specified,
+this affects
+\fB^\fR
+and
+\fB$\fR
+as with
+newline-sensitive matching,
+but not \fB.\fR
+and bracket expressions.
+This isn't very useful but is provided for symmetry.
+.SH "LIMITS AND COMPATIBILITY"
+No particular limit is imposed on the length of REs.
+Programs intended to be highly portable should not employ REs longer
+than 256 bytes,
+as a POSIX-compliant implementation can refuse to accept such REs.
+.PP
+The only feature of AREs that is actually incompatible with
+POSIX EREs is that
+\fB\e\fR
+does not lose its special
+significance inside bracket expressions.
+All other ARE features use syntax which is illegal or has
+undefined or unspecified effects in POSIX EREs;
+the
+\fB***\fR
+syntax of directors likewise is outside the POSIX
+syntax for both BREs and EREs.
+.PP
+Many of the ARE extensions are borrowed from Perl, but some have
+been changed to clean them up, and a few Perl extensions are not present.
+Incompatibilities of note include `\fB\eb\fR', `\fB\eB\fR',
+the lack of special treatment for a trailing newline,
+the addition of complemented bracket expressions to the things
+affected by newline-sensitive matching,
+the restrictions on parentheses and back references in lookahead constraints,
+and the longest/shortest-match (rather than first-match) matching semantics.
+.PP
+The matching rules for REs containing both normal and non-greedy quantifiers
+have changed since early beta-test versions of this package.
+(The new rules are much simpler and cleaner,
+but don't work as hard at guessing the user's real intentions.)
+.PP
+Henry Spencer's original 1986 \fIregexp\fR package,
+still in widespread use (e.g., in pre-8.1 releases of Tcl),
+implemented an early version of today's EREs.
+There are four incompatibilities between \fIregexp\fR's near-EREs
+(`RREs' for short) and AREs.
+In roughly increasing order of significance:
+.PP
+.RS
+In AREs,
+\fB\e\fR
+followed by an alphanumeric character is either an
+escape or an error,
+while in RREs, it was just another way of writing the 
+alphanumeric.
+This should not be a problem because there was no reason to write
+such a sequence in RREs.
+.PP
+\fB{\fR
+followed by a digit in an ARE is the beginning of a bound,
+while in RREs,
+\fB{\fR
+was always an ordinary character.
+Such sequences should be rare,
+and will often result in an error because following characters
+will not look like a valid bound.
+.PP
+In AREs,
+\fB\e\fR
+remains a special character within `\fB[\|]\fR',
+so a literal
+\fB\e\fR
+within
+\fB[\|]\fR
+must be written `\fB\e\e\fR'.
+\fB\e\e\fR
+also gives a literal
+\fB\e\fR
+within
+\fB[\|]\fR
+in RREs,
+but only truly paranoid programmers routinely doubled the backslash.
+.PP
+AREs report the longest/shortest match for the RE,
+rather than the first found in a specified search order.
+This may affect some RREs which were written in the expectation that
+the first match would be reported.
+(The careful crafting of RREs to optimize the search order for fast
+matching is obsolete (AREs examine all possible matches
+in parallel, and their performance is largely insensitive to their
+complexity) but cases where the search order was exploited to deliberately 
+find a match which was \fInot\fR the longest/shortest will need rewriting.)
+.RE
+
+.SH "BASIC REGULAR EXPRESSIONS"
+BREs differ from EREs in several respects.  `\fB|\fR', `\fB+\fR',
+and
+\fB?\fR
+are ordinary characters and there is no equivalent
+for their functionality.
+The delimiters for bounds are
+\fB\e{\fR
+and `\fB\e}\fR',
+with
+\fB{\fR
+and
+\fB}\fR
+by themselves ordinary characters.
+The parentheses for nested subexpressions are
+\fB\e(\fR
+and `\fB\e)\fR',
+with
+\fB(\fR
+and
+\fB)\fR
+by themselves ordinary characters.
+\fB^\fR
+is an ordinary character except at the beginning of the
+RE or the beginning of a parenthesized subexpression,
+\fB$\fR
+is an ordinary character except at the end of the
+RE or the end of a parenthesized subexpression,
+and
+\fB*\fR
+is an ordinary character if it appears at the beginning of the
+RE or the beginning of a parenthesized subexpression
+(after a possible leading `\fB^\fR').
+Finally,
+single-digit back references are available,
+and
+\fB\e<\fR
+and
+\fB\e>\fR
+are synonyms for
+\fB[[:<:]]\fR
+and
+\fB[[:>:]]\fR
+respectively;
+no other escapes are available.
+
+.SH "SEE ALSO"
+RegExp(3), regexp(n), regsub(n), lsearch(n), switch(n), text(n)
+
+.SH KEYWORDS
+match, regular expression, string

src/backend/regex/regc_color.c

@@ -0,0 +1,728 @@
+/*
+ * colorings of characters
+ * This file is #included by regcomp.c.
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ * 
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them. 
+ * 
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ * 
+ * I'd appreciate being given credit for this package in the documentation
+ * of software which uses it, but that is not a requirement.
+ * 
+ * THIS SOFTWARE IS PROVIDED ``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
+ * HENRY SPENCER 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.
+ *
+ * $Header: /cvsroot/pgsql/src/backend/regex/regc_color.c,v 1.1 2003/02/05 17:41:32 tgl Exp $
+ *
+ *
+ * Note that there are some incestuous relationships between this code and
+ * NFA arc maintenance, which perhaps ought to be cleaned up sometime.
+ */
+
+
+
+#define	CISERR()	VISERR(cm->v)
+#define	CERR(e)		VERR(cm->v, (e))
+
+
+
+/*
+ * initcm - set up new colormap
+ */
+static void
+initcm(struct vars *v,
+	   struct colormap *cm)
+{
+	int i;
+	int j;
+	union tree *t;
+	union tree *nextt;
+	struct colordesc *cd;
+
+	cm->magic = CMMAGIC;
+	cm->v = v;
+
+	cm->ncds = NINLINECDS;
+	cm->cd = cm->cdspace;
+	cm->max = 0;
+	cm->free = 0;
+
+	cd = cm->cd;			/* cm->cd[WHITE] */
+	cd->sub = NOSUB;
+	cd->arcs = NULL;
+	cd->flags = 0;
+	cd->nchrs = CHR_MAX - CHR_MIN + 1;
+
+	/* upper levels of tree */
+	for (t = &cm->tree[0], j = NBYTS-1; j > 0; t = nextt, j--) {
+		nextt = t + 1;
+		for (i = BYTTAB-1; i >= 0; i--)
+			t->tptr[i] = nextt;
+	}
+	/* bottom level is solid white */
+	t = &cm->tree[NBYTS-1];
+	for (i = BYTTAB-1; i >= 0; i--)
+		t->tcolor[i] = WHITE;
+	cd->block = t;
+}
+
+/*
+ * freecm - free dynamically-allocated things in a colormap
+ */
+static void
+freecm(struct colormap *cm)
+{
+	size_t i;
+	union tree *cb;
+
+	cm->magic = 0;
+	if (NBYTS > 1)
+		cmtreefree(cm, cm->tree, 0);
+	for (i = 1; i <= cm->max; i++)		/* skip WHITE */
+		if (!UNUSEDCOLOR(&cm->cd[i])) {
+			cb = cm->cd[i].block;
+			if (cb != NULL)
+				FREE(cb);
+		}
+	if (cm->cd != cm->cdspace)
+		FREE(cm->cd);
+}
+
+/*
+ * cmtreefree - free a non-terminal part of a colormap tree
+ */
+static void
+cmtreefree(struct colormap *cm,
+		   union tree *tree,
+		   int level)			/* level number (top == 0) of this block */
+{
+	int i;
+	union tree *t;
+	union tree *fillt = &cm->tree[level+1];
+	union tree *cb;
+
+	assert(level < NBYTS-1);	/* this level has pointers */
+	for (i = BYTTAB-1; i >= 0; i--) {
+		t = tree->tptr[i];
+		assert(t != NULL);
+		if (t != fillt) {
+			if (level < NBYTS-2) {	/* more pointer blocks below */
+				cmtreefree(cm, t, level+1);
+				FREE(t);
+			} else {		/* color block below */
+				cb = cm->cd[t->tcolor[0]].block;
+				if (t != cb)	/* not a solid block */
+					FREE(t);
+			}
+		}
+	}
+}
+
+/*
+ * setcolor - set the color of a character in a colormap
+ */
+static color			/* previous color */
+setcolor(struct colormap *cm,
+		 chr c,
+		 pcolor co)
+{
+	uchr uc = c;
+	int shift;
+	int level;
+	int b;
+	int bottom;
+	union tree *t;
+	union tree *newt;
+	union tree *fillt;
+	union tree *lastt;
+	union tree *cb;
+	color prev;
+
+	assert(cm->magic == CMMAGIC);
+	if (CISERR() || co == COLORLESS)
+		return COLORLESS;
+
+	t = cm->tree;
+	for (level = 0, shift = BYTBITS * (NBYTS - 1); shift > 0;
+						level++, shift -= BYTBITS) {
+		b = (uc >> shift) & BYTMASK;
+		lastt = t;
+		t = lastt->tptr[b];
+		assert(t != NULL);
+		fillt = &cm->tree[level+1];
+		bottom = (shift <= BYTBITS) ? 1 : 0;
+		cb = (bottom) ? cm->cd[t->tcolor[0]].block : fillt;
+		if (t == fillt || t == cb) {	/* must allocate a new block */
+			newt = (union tree *)MALLOC((bottom) ?
+				sizeof(struct colors) : sizeof(struct ptrs));
+			if (newt == NULL) {
+				CERR(REG_ESPACE);
+				return COLORLESS;
+			}
+			if (bottom)
+				memcpy(VS(newt->tcolor), VS(t->tcolor),
+							BYTTAB*sizeof(color));
+			else
+				memcpy(VS(newt->tptr), VS(t->tptr),
+						BYTTAB*sizeof(union tree *));
+			t = newt;
+			lastt->tptr[b] = t;
+		}
+	}
+
+	b = uc & BYTMASK;
+	prev = t->tcolor[b];
+	t->tcolor[b] = (color)co;
+	return prev;
+}
+
+/*
+ * maxcolor - report largest color number in use
+ */
+static color
+maxcolor(struct colormap *cm)
+{
+	if (CISERR())
+		return COLORLESS;
+
+	return (color)cm->max;
+}
+
+/*
+ * newcolor - find a new color (must be subject of setcolor at once)
+ * Beware:  may relocate the colordescs.
+ */
+static color			/* COLORLESS for error */
+newcolor(struct colormap *cm)
+{
+	struct colordesc *cd;
+	struct colordesc *new;
+	size_t n;
+
+	if (CISERR())
+		return COLORLESS;
+
+	if (cm->free != 0) {
+		assert(cm->free > 0);
+		assert((size_t)cm->free < cm->ncds);
+		cd = &cm->cd[cm->free];
+		assert(UNUSEDCOLOR(cd));
+		assert(cd->arcs == NULL);
+		cm->free = cd->sub;
+	} else if (cm->max < cm->ncds - 1) {
+		cm->max++;
+		cd = &cm->cd[cm->max];
+	} else {
+		/* oops, must allocate more */
+		n = cm->ncds * 2;
+		if (cm->cd == cm->cdspace) {
+			new = (struct colordesc *)MALLOC(n *
+						sizeof(struct colordesc));
+			if (new != NULL)
+				memcpy(VS(new), VS(cm->cdspace), cm->ncds *
+						sizeof(struct colordesc));
+		} else
+			new = (struct colordesc *)REALLOC(cm->cd,
+						n * sizeof(struct colordesc));
+		if (new == NULL) {
+			CERR(REG_ESPACE);
+			return COLORLESS;
+		}
+		cm->cd = new;
+		cm->ncds = n;
+		assert(cm->max < cm->ncds - 1);
+		cm->max++;
+		cd = &cm->cd[cm->max];
+	}
+
+	cd->nchrs = 0;
+	cd->sub = NOSUB;
+	cd->arcs = NULL;
+	cd->flags = 0;
+	cd->block = NULL;
+
+	return (color)(cd - cm->cd);
+}
+
+/*
+ * freecolor - free a color (must have no arcs or subcolor)
+ */
+static void
+freecolor(struct colormap *cm,
+		  pcolor co)
+{
+	struct colordesc *cd = &cm->cd[co];
+	color pco, nco;			/* for freelist scan */
+
+	assert(co >= 0);
+	if (co == WHITE)
+		return;
+
+	assert(cd->arcs == NULL);
+	assert(cd->sub == NOSUB);
+	assert(cd->nchrs == 0);
+	cd->flags = FREECOL;
+	if (cd->block != NULL) {
+		FREE(cd->block);
+		cd->block = NULL;	/* just paranoia */
+	}
+
+	if ((size_t)co == cm->max) {
+		while (cm->max > WHITE && UNUSEDCOLOR(&cm->cd[cm->max]))
+			cm->max--;
+		assert(cm->free >= 0);
+		while ((size_t)cm->free > cm->max)
+			cm->free = cm->cd[cm->free].sub;
+		if (cm->free > 0) {
+			assert(cm->free < cm->max);
+			pco = cm->free;
+			nco = cm->cd[pco].sub;
+			while (nco > 0)
+				if ((size_t)nco > cm->max) {
+					/* take this one out of freelist */
+					nco = cm->cd[nco].sub;
+					cm->cd[pco].sub = nco;
+				} else {
+					assert(nco < cm->max);
+					pco = nco;
+					nco = cm->cd[pco].sub;
+				}
+		}
+	} else {
+		cd->sub = cm->free;
+		cm->free = (color)(cd - cm->cd);
+	}
+}
+
+/*
+ * pseudocolor - allocate a false color, to be managed by other means
+ */
+static color
+pseudocolor(struct colormap *cm)
+{
+	color co;
+
+	co = newcolor(cm);
+	if (CISERR())
+		return COLORLESS;
+	cm->cd[co].nchrs = 1;
+	cm->cd[co].flags = PSEUDO;
+	return co;
+}
+
+/*
+ * subcolor - allocate a new subcolor (if necessary) to this chr
+ */
+static color
+subcolor(struct colormap *cm, chr c)
+{
+	color co;			/* current color of c */
+	color sco;			/* new subcolor */
+
+	co = GETCOLOR(cm, c);
+	sco = newsub(cm, co);
+	if (CISERR())
+		return COLORLESS;
+	assert(sco != COLORLESS);
+
+	if (co == sco)		/* already in an open subcolor */
+		return co;	/* rest is redundant */
+	cm->cd[co].nchrs--;
+	cm->cd[sco].nchrs++;
+	setcolor(cm, c, sco);
+	return sco;
+}
+
+/*
+ * newsub - allocate a new subcolor (if necessary) for a color
+ */
+static color
+newsub(struct colormap *cm,
+	   pcolor co)
+{
+	color sco;			/* new subcolor */
+
+	sco = cm->cd[co].sub;
+	if (sco == NOSUB) {		/* color has no open subcolor */
+		if (cm->cd[co].nchrs == 1)	/* optimization */
+			return co;
+		sco = newcolor(cm);	/* must create subcolor */
+		if (sco == COLORLESS) {
+			assert(CISERR());
+			return COLORLESS;
+		}
+		cm->cd[co].sub = sco;
+		cm->cd[sco].sub = sco;	/* open subcolor points to self */
+	}
+	assert(sco != NOSUB);
+
+	return sco;
+}
+
+/*
+ * subrange - allocate new subcolors to this range of chrs, fill in arcs
+ */
+static void
+subrange(struct vars *v,
+		 chr from,
+		 chr to,
+		 struct state *lp,
+		 struct state *rp)
+{
+	uchr uf;
+	int i;
+
+	assert(from <= to);
+
+	/* first, align "from" on a tree-block boundary */
+	uf = (uchr)from;
+	i = (int)( ((uf + BYTTAB-1) & (uchr)~BYTMASK) - uf );
+	for (; from <= to && i > 0; i--, from++)
+		newarc(v->nfa, PLAIN, subcolor(v->cm, from), lp, rp);
+	if (from > to)			/* didn't reach a boundary */
+		return;
+
+	/* deal with whole blocks */
+	for (; to - from >= BYTTAB; from += BYTTAB)
+		subblock(v, from, lp, rp);
+
+	/* clean up any remaining partial table */
+	for (; from <= to; from++)
+		newarc(v->nfa, PLAIN, subcolor(v->cm, from), lp, rp);
+}
+
+/*
+ * subblock - allocate new subcolors for one tree block of chrs, fill in arcs
+ */
+static void
+subblock(struct vars *v,
+		 chr start,				/* first of BYTTAB chrs */
+		 struct state *lp,
+		 struct state *rp)
+{
+	uchr uc = start;
+	struct colormap *cm = v->cm;
+	int shift;
+	int level;
+	int i;
+	int b;
+	union tree *t;
+	union tree *cb;
+	union tree *fillt;
+	union tree *lastt;
+	int previ;
+	int ndone;
+	color co;
+	color sco;
+
+	assert((uc % BYTTAB) == 0);
+
+	/* find its color block, making new pointer blocks as needed */
+	t = cm->tree;
+	fillt = NULL;
+	for (level = 0, shift = BYTBITS * (NBYTS - 1); shift > 0;
+						level++, shift -= BYTBITS) {
+		b = (uc >> shift) & BYTMASK;
+		lastt = t;
+		t = lastt->tptr[b];
+		assert(t != NULL);
+		fillt = &cm->tree[level+1];
+		if (t == fillt && shift > BYTBITS) {	/* need new ptr block */
+			t = (union tree *)MALLOC(sizeof(struct ptrs));
+			if (t == NULL) {
+				CERR(REG_ESPACE);
+				return;
+			}
+			memcpy(VS(t->tptr), VS(fillt->tptr),
+						BYTTAB*sizeof(union tree *));
+			lastt->tptr[b] = t;
+		}
+	}
+
+	/* special cases:  fill block or solid block */
+	co = t->tcolor[0];
+	cb = cm->cd[co].block;
+	if (t == fillt || t == cb) {
+		/* either way, we want a subcolor solid block */
+		sco = newsub(cm, co);
+		t = cm->cd[sco].block;
+		if (t == NULL) {	/* must set it up */
+			t = (union tree *)MALLOC(sizeof(struct colors));
+			if (t == NULL) {
+				CERR(REG_ESPACE);
+				return;
+			}
+			for (i = 0; i < BYTTAB; i++)
+				t->tcolor[i] = sco;
+			cm->cd[sco].block = t;
+		}
+		/* find loop must have run at least once */
+		lastt->tptr[b] = t;
+		newarc(v->nfa, PLAIN, sco, lp, rp);
+		cm->cd[co].nchrs -= BYTTAB;
+		cm->cd[sco].nchrs += BYTTAB;
+		return;
+	}
+
+	/* general case, a mixed block to be altered */
+	i = 0;
+	while (i < BYTTAB) {
+		co = t->tcolor[i];
+		sco = newsub(cm, co);
+		newarc(v->nfa, PLAIN, sco, lp, rp);
+		previ = i;
+		do {
+			t->tcolor[i++] = sco;
+		} while (i < BYTTAB && t->tcolor[i] == co);
+		ndone = i - previ;
+		cm->cd[co].nchrs -= ndone;
+		cm->cd[sco].nchrs += ndone;
+	}
+}
+
+/*
+ * okcolors - promote subcolors to full colors
+ */
+static void
+okcolors(struct nfa *nfa,
+		 struct colormap *cm)
+{
+	struct colordesc *cd;
+	struct colordesc *end = CDEND(cm);
+	struct colordesc *scd;
+	struct arc *a;
+	color co;
+	color sco;
+
+	for (cd = cm->cd, co = 0; cd < end; cd++, co++) {
+		sco = cd->sub;
+		if (UNUSEDCOLOR(cd) || sco == NOSUB) {
+			/* has no subcolor, no further action */
+		} else if (sco == co) {
+			/* is subcolor, let parent deal with it */
+		} else if (cd->nchrs == 0) {
+			/* parent empty, its arcs change color to subcolor */
+			cd->sub = NOSUB;
+			scd = &cm->cd[sco];
+			assert(scd->nchrs > 0);
+			assert(scd->sub == sco);
+			scd->sub = NOSUB;
+			while ((a = cd->arcs) != NULL) {
+				assert(a->co == co);
+				/* uncolorchain(cm, a); */
+				cd->arcs = a->colorchain;
+				a->co = sco;
+				/* colorchain(cm, a); */
+				a->colorchain = scd->arcs;
+				scd->arcs = a;
+			}
+			freecolor(cm, co);
+		} else {
+			/* parent's arcs must gain parallel subcolor arcs */
+			cd->sub = NOSUB;
+			scd = &cm->cd[sco];
+			assert(scd->nchrs > 0);
+			assert(scd->sub == sco);
+			scd->sub = NOSUB;
+			for (a = cd->arcs; a != NULL; a = a->colorchain) {
+				assert(a->co == co);
+				newarc(nfa, a->type, sco, a->from, a->to);
+			}
+		}
+	}
+}
+
+/*
+ * colorchain - add this arc to the color chain of its color
+ */
+static void
+colorchain(struct colormap *cm,
+		   struct arc *a)
+{
+	struct colordesc *cd = &cm->cd[a->co];
+
+	a->colorchain = cd->arcs;
+	cd->arcs = a;
+}
+
+/*
+ * uncolorchain - delete this arc from the color chain of its color
+ */
+static void
+uncolorchain(struct colormap *cm,
+			 struct arc *a)
+{
+	struct colordesc *cd = &cm->cd[a->co];
+	struct arc *aa;
+
+	aa = cd->arcs;
+	if (aa == a)		/* easy case */
+		cd->arcs = a->colorchain;
+	else {
+		for (; aa != NULL && aa->colorchain != a; aa = aa->colorchain)
+			continue;
+		assert(aa != NULL);
+		aa->colorchain = a->colorchain;
+	}
+	a->colorchain = NULL;	/* paranoia */
+}
+
+/*
+ * singleton - is this character in its own color?
+ */
+static int			/* predicate */
+singleton(struct colormap *cm,
+		  chr c)
+{
+	color co;			/* color of c */
+
+	co = GETCOLOR(cm, c);
+	if (cm->cd[co].nchrs == 1 && cm->cd[co].sub == NOSUB)
+		return 1;
+	return 0;
+}
+
+/*
+ * rainbow - add arcs of all full colors (but one) between specified states
+ */
+static void
+rainbow(struct nfa *nfa,
+		struct colormap *cm,
+		int type,
+		pcolor but,				/* COLORLESS if no exceptions */
+		struct state *from,
+		struct state *to)
+{
+	struct colordesc *cd;
+	struct colordesc *end = CDEND(cm);
+	color co;
+
+	for (cd = cm->cd, co = 0; cd < end && !CISERR(); cd++, co++)
+		if (!UNUSEDCOLOR(cd) && cd->sub != co && co != but &&
+							!(cd->flags&PSEUDO))
+			newarc(nfa, type, co, from, to);
+}
+
+/*
+ * colorcomplement - add arcs of complementary colors
+ *
+ * The calling sequence ought to be reconciled with cloneouts().
+ */
+static void
+colorcomplement(struct nfa *nfa,
+				struct colormap *cm,
+				int type,
+				struct state *of, /* complements of this guy's PLAIN outarcs */
+				struct state *from,
+				struct state *to)
+{
+	struct colordesc *cd;
+	struct colordesc *end = CDEND(cm);
+	color co;
+
+	assert(of != from);
+	for (cd = cm->cd, co = 0; cd < end && !CISERR(); cd++, co++)
+		if (!UNUSEDCOLOR(cd) && !(cd->flags&PSEUDO))
+			if (findarc(of, PLAIN, co) == NULL)
+				newarc(nfa, type, co, from, to);
+}
+
+
+#ifdef REG_DEBUG
+
+/*
+ * dumpcolors - debugging output
+ */
+static void
+dumpcolors(struct colormap *cm,
+		   FILE *f)
+{
+	struct colordesc *cd;
+	struct colordesc *end;
+	color co;
+	chr c;
+	char *has;
+
+	fprintf(f, "max %ld\n", (long)cm->max);
+	if (NBYTS > 1)
+		fillcheck(cm, cm->tree, 0, f);
+	end = CDEND(cm);
+	for (cd = cm->cd + 1, co = 1; cd < end; cd++, co++)	/* skip 0 */
+		if (!UNUSEDCOLOR(cd)) {
+			assert(cd->nchrs > 0);
+			has = (cd->block != NULL) ? "#" : "";
+			if (cd->flags&PSEUDO)
+				fprintf(f, "#%2ld%s(ps): ", (long)co, has);
+			else
+				fprintf(f, "#%2ld%s(%2d): ", (long)co,
+							has, cd->nchrs);
+			/* it's hard to do this more efficiently */
+			for (c = CHR_MIN; c < CHR_MAX; c++)
+				if (GETCOLOR(cm, c) == co)
+					dumpchr(c, f);
+			assert(c == CHR_MAX);
+			if (GETCOLOR(cm, c) == co)
+				dumpchr(c, f);
+			fprintf(f, "\n");
+		}
+}
+
+/*
+ * fillcheck - check proper filling of a tree
+ */
+static void
+fillcheck(struct colormap *cm,
+		  union tree *tree,
+		  int level,			/* level number (top == 0) of this block */
+		  FILE *f)
+{
+	int i;
+	union tree *t;
+	union tree *fillt = &cm->tree[level+1];
+
+	assert(level < NBYTS-1);	/* this level has pointers */
+	for (i = BYTTAB-1; i >= 0; i--) {
+		t = tree->tptr[i];
+		if (t == NULL)
+			fprintf(f, "NULL found in filled tree!\n");
+		else if (t == fillt)
+			{}
+		else if (level < NBYTS-2)	/* more pointer blocks below */
+			fillcheck(cm, t, level+1, f);
+	}
+}
+
+/*
+ * dumpchr - print a chr
+ *
+ * Kind of char-centric but works well enough for debug use.
+ */
+static void
+dumpchr(chr c,
+		FILE *f)
+{
+	if (c == '\\')
+		fprintf(f, "\\\\");
+	else if (c > ' ' && c <= '~')
+		putc((char)c, f);
+	else
+		fprintf(f, "\\u%04lx", (long)c);
+}
+
+#endif /* REG_DEBUG */

src/backend/regex/regc_cvec.c

@@ -0,0 +1,194 @@
+/*
+ * Utility functions for handling cvecs
+ * This file is #included by regcomp.c.
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ * 
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them. 
+ * 
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ * 
+ * I'd appreciate being given credit for this package in the documentation
+ * of software which uses it, but that is not a requirement.
+ * 
+ * THIS SOFTWARE IS PROVIDED ``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
+ * HENRY SPENCER 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.
+ *
+ * $Header: /cvsroot/pgsql/src/backend/regex/regc_cvec.c,v 1.1 2003/02/05 17:41:32 tgl Exp $
+ *
+ */
+
+/*
+ * newcvec - allocate a new cvec
+ */
+static struct cvec *
+newcvec(int nchrs,				/* to hold this many chrs... */
+		int nranges,			/* ... and this many ranges... */
+		int nmcces)				/* ... and this many MCCEs */
+{
+    size_t n;
+    size_t nc;
+    struct cvec *cv;
+
+    nc = (size_t)nchrs + (size_t)nmcces*(MAXMCCE+1) + (size_t)nranges*2;
+    n = sizeof(struct cvec) + (size_t)(nmcces-1)*sizeof(chr *)
+	    + nc*sizeof(chr);
+    cv = (struct cvec *)MALLOC(n);
+    if (cv == NULL) {
+	return NULL;
+    }
+    cv->chrspace = nchrs;
+    cv->chrs = (chr *)&cv->mcces[nmcces];	/* chrs just after MCCE ptrs */
+    cv->mccespace = nmcces;
+    cv->ranges = cv->chrs + nchrs + nmcces*(MAXMCCE+1);
+    cv->rangespace = nranges;
+    return clearcvec(cv);
+}
+
+/*
+ * clearcvec - clear a possibly-new cvec
+ * Returns pointer as convenience.
+ */
+static struct cvec *
+clearcvec(struct cvec *cv)
+{
+    int i;
+
+    assert(cv != NULL);
+    cv->nchrs = 0;
+    assert(cv->chrs == (chr *)&cv->mcces[cv->mccespace]);
+    cv->nmcces = 0;
+    cv->nmccechrs = 0;
+    cv->nranges = 0;
+    for (i = 0; i < cv->mccespace; i++) {
+	cv->mcces[i] = NULL;
+    }
+
+    return cv;
+}
+
+/*
+ * addchr - add a chr to a cvec
+ */
+static void
+addchr(struct cvec *cv,			/* character vector */
+	   chr c)				/* character to add */
+{
+    assert(cv->nchrs < cv->chrspace - cv->nmccechrs);
+    cv->chrs[cv->nchrs++] = (chr)c;
+}
+
+/*
+ * addrange - add a range to a cvec
+ */
+static void
+addrange(struct cvec *cv,			/* character vector */
+		 chr from,				/* first character of range */
+		 chr to)				/* last character of range */
+{
+    assert(cv->nranges < cv->rangespace);
+    cv->ranges[cv->nranges*2] = (chr)from;
+    cv->ranges[cv->nranges*2 + 1] = (chr)to;
+    cv->nranges++;
+}
+
+/*
+ * addmcce - add an MCCE to a cvec
+ */
+static void
+addmcce(struct cvec *cv,			/* character vector */
+		chr *startp,			/* beginning of text */
+		chr *endp)				/* just past end of text */
+{
+    int len;
+    int i;
+    chr *s;
+    chr *d;
+
+    if (startp == NULL && endp == NULL) {
+	return;
+    }
+    len = endp - startp;
+    assert(len > 0);
+    assert(cv->nchrs + len < cv->chrspace - cv->nmccechrs);
+    assert(cv->nmcces < cv->mccespace);
+    d = &cv->chrs[cv->chrspace - cv->nmccechrs - len - 1];
+    cv->mcces[cv->nmcces++] = d;
+    for (s = startp, i = len; i > 0; s++, i--) {
+	*d++ = *s;
+    }
+    *d++ = 0;				/* endmarker */
+    assert(d == &cv->chrs[cv->chrspace - cv->nmccechrs]);
+    cv->nmccechrs += len + 1;
+}
+
+/*
+ * haschr - does a cvec contain this chr?
+ */
+static int				/* predicate */
+haschr(struct cvec *cv,			/* character vector */
+	   chr c)				/* character to test for */
+{
+    int i;
+    chr *p;
+
+    for (p = cv->chrs, i = cv->nchrs; i > 0; p++, i--) {
+	if (*p == c) {
+	    return 1;
+	}
+    }
+    for (p = cv->ranges, i = cv->nranges; i > 0; p += 2, i--) {
+	if ((*p <= c) && (c <= *(p+1))) {
+	    return 1;
+	}
+    }
+    return 0;
+}
+
+/*
+ * getcvec - get a cvec, remembering it as v->cv
+ */
+static struct cvec *
+getcvec(struct vars *v,			/* context */
+		int nchrs,				/* to hold this many chrs... */
+		int nranges,			/* ... and this many ranges... */
+		int nmcces)				/* ... and this many MCCEs */
+{
+    if (v->cv != NULL && nchrs <= v->cv->chrspace &&
+	    nranges <= v->cv->rangespace && nmcces <= v->cv->mccespace) {
+	return clearcvec(v->cv);
+    }
+
+    if (v->cv != NULL) {
+	freecvec(v->cv);
+    }
+    v->cv = newcvec(nchrs, nranges, nmcces);
+    if (v->cv == NULL) {
+	ERR(REG_ESPACE);
+    }
+
+    return v->cv;
+}
+
+/*
+ * freecvec - free a cvec
+ */
+static void
+freecvec(struct cvec *cv)
+{
+    FREE(cv);
+}

src/backend/regex/regc_locale.c

@@ -0,0 +1,615 @@
+/* 
+ * regc_locale.c --
+ *
+ *	This file contains locale-specific regexp routines.
+ *	This file is #included by regcomp.c.
+ *
+ * Copyright (c) 1998 by Scriptics Corporation.
+ *
+ * This software is copyrighted by the Regents of the University of
+ * California, Sun Microsystems, Inc., Scriptics Corporation, ActiveState
+ * Corporation and other parties.  The following terms apply to all files
+ * associated with the software unless explicitly disclaimed in
+ * individual files.
+ * 
+ * The authors hereby grant permission to use, copy, modify, distribute,
+ * and license this software and its documentation for any purpose, provided
+ * that existing copyright notices are retained in all copies and that this
+ * notice is included verbatim in any distributions. No written agreement,
+ * license, or royalty fee is required for any of the authorized uses.
+ * Modifications to this software may be copyrighted by their authors
+ * and need not follow the licensing terms described here, provided that
+ * the new terms are clearly indicated on the first page of each file where
+ * they apply.
+ * 
+ * IN NO EVENT SHALL THE AUTHORS OR DISTRIBUTORS BE LIABLE TO ANY PARTY
+ * FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE, ITS DOCUMENTATION, OR ANY
+ * DERIVATIVES THEREOF, EVEN IF THE AUTHORS HAVE BEEN ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ * 
+ * THE AUTHORS AND DISTRIBUTORS SPECIFICALLY DISCLAIM ANY WARRANTIES,
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE, AND NON-INFRINGEMENT.  THIS SOFTWARE
+ * IS PROVIDED ON AN "AS IS" BASIS, AND THE AUTHORS AND DISTRIBUTORS HAVE
+ * NO OBLIGATION TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR
+ * MODIFICATIONS.
+ * 
+ * GOVERNMENT USE: If you are acquiring this software on behalf of the
+ * U.S. government, the Government shall have only "Restricted Rights"
+ * in the software and related documentation as defined in the Federal 
+ * Acquisition Regulations (FARs) in Clause 52.227.19 (c) (2).  If you
+ * are acquiring the software on behalf of the Department of Defense, the
+ * software shall be classified as "Commercial Computer Software" and the
+ * Government shall have only "Restricted Rights" as defined in Clause
+ * 252.227-7013 (c) (1) of DFARs.  Notwithstanding the foregoing, the
+ * authors grant the U.S. Government and others acting in its behalf
+ * permission to use and distribute the software in accordance with the
+ * terms specified in this license. 
+ *
+ * $Header: /cvsroot/pgsql/src/backend/regex/regc_locale.c,v 1.1 2003/02/05 17:41:32 tgl Exp $
+ */
+
+/* ASCII character-name table */
+
+static struct cname {
+    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}
+};
+
+/*
+ * some ctype functions with non-ascii-char guard
+ */
+static int
+pg_isdigit(pg_wchar c)
+{
+	return (c >= 0 && c <= UCHAR_MAX && isdigit((unsigned char) c));
+}
+
+static int
+pg_isalpha(pg_wchar c)
+{
+	return (c >= 0 && c <= UCHAR_MAX && isalpha((unsigned char) c));
+}
+
+static int
+pg_isalnum(pg_wchar c)
+{
+	return (c >= 0 && c <= UCHAR_MAX && isalnum((unsigned char) c));
+}
+
+static int
+pg_isupper(pg_wchar c)
+{
+	return (c >= 0 && c <= UCHAR_MAX && isupper((unsigned char) c));
+}
+
+static int
+pg_islower(pg_wchar c)
+{
+	return (c >= 0 && c <= UCHAR_MAX && islower((unsigned char) c));
+}
+
+static int
+pg_isgraph(pg_wchar c)
+{
+	return (c >= 0 && c <= UCHAR_MAX && isgraph((unsigned char) c));
+}
+
+static int
+pg_ispunct(pg_wchar c)
+{
+	return (c >= 0 && c <= UCHAR_MAX && ispunct((unsigned char) c));
+}
+
+static int
+pg_isspace(pg_wchar c)
+{
+	return (c >= 0 && c <= UCHAR_MAX && isspace((unsigned char) c));
+}
+
+static pg_wchar
+pg_toupper(pg_wchar c)
+{
+	if (c >= 0 && c <= UCHAR_MAX)
+		return toupper((unsigned char) c);
+	return c;
+}
+
+static pg_wchar
+pg_tolower(pg_wchar c)
+{
+	if (c >= 0 && c <= UCHAR_MAX)
+		return tolower((unsigned char) c);
+	return c;
+}
+
+
+/*
+ * nmcces - how many distinct MCCEs are there?
+ */
+static int
+nmcces(struct vars *v)
+{
+    /*
+     * No multi-character collating elements defined at the moment.
+     */
+    return 0;
+}
+
+/*
+ * nleaders - how many chrs can be first chrs of MCCEs?
+ */
+static int
+nleaders(struct vars *v)
+{
+    return 0;
+}
+
+/*
+ * allmcces - return a cvec with all the MCCEs of the locale
+ */
+static struct cvec *
+allmcces(struct vars *v,			/* context */
+		 struct cvec *cv)			/* this is supposed to have enough room */
+{
+    return clearcvec(cv);
+}
+
+/*
+ * element - map collating-element name to celt
+ */
+static celt
+element(struct vars *v,			/* context */
+		chr *startp,			/* points to start of name */
+		chr *endp)				/* points just past end of name */
+{
+    struct cname *cn;
+    size_t len;
+
+    /* generic:  one-chr names stand for themselves */
+    assert(startp < endp);
+    len = endp - startp;
+    if (len == 1) {
+	return *startp;
+    }
+
+    NOTE(REG_ULOCALE);
+
+    /* search table */
+    for (cn=cnames; cn->name!=NULL; cn++) {
+	if (strlen(cn->name)==len &&
+		pg_char_and_wchar_strncmp(cn->name, startp, len)==0) {
+	    break;			/* NOTE BREAK OUT */
+	}
+    }
+    if (cn->name != NULL) {
+	return CHR(cn->code);
+    }
+
+    /* couldn't find it */
+    ERR(REG_ECOLLATE);
+    return 0;
+}
+
+/*
+ * range - supply cvec for a range, including legality check
+ */
+static struct cvec *
+range(struct vars *v,			/* context */
+	  celt a,				/* range start */
+	  celt b,				/* range end, might equal a */
+	  int cases)				/* case-independent? */
+{
+    int nchrs;
+    struct cvec *cv;
+    celt c, lc, uc;
+
+    if (a != b && !before(a, b)) {
+	ERR(REG_ERANGE);
+	return NULL;
+    }
+
+    if (!cases) {			/* easy version */
+	cv = getcvec(v, 0, 1, 0);
+	NOERRN();
+	addrange(cv, a, b);
+	return cv;
+    }
+
+    /*
+     * When case-independent, it's hard to decide when cvec ranges are
+     * usable, so for now at least, we won't try.  We allocate enough
+     * space for two case variants plus a little extra for the two
+     * title case variants.
+     */
+
+    nchrs = (b - a + 1)*2 + 4;
+
+    cv = getcvec(v, nchrs, 0, 0);
+    NOERRN();
+
+    for (c=a; c<=b; c++) {
+	addchr(cv, c);
+	lc = pg_tolower((chr)c);
+	if (c != lc) {
+	    addchr(cv, lc);
+	}
+	uc = pg_toupper((chr)c);
+	if (c != uc) {
+	    addchr(cv, uc);
+	}
+    }
+
+    return cv;
+}
+
+/*
+ * before - is celt x before celt y, for purposes of range legality?
+ */
+static int				/* predicate */
+before(celt x, celt y)
+{
+    /* trivial because no MCCEs */
+    if (x < y) {
+	return 1;
+    }
+    return 0;
+}
+
+/*
+ * eclass - supply cvec for an equivalence class
+ * Must include case counterparts on request.
+ */
+static struct cvec *
+eclass(struct vars *v,			/* context */
+	   celt c,					/* Collating element representing
+								 * the equivalence class. */
+	   int cases)				/* all cases? */
+{
+    struct cvec *cv;
+
+    /* crude fake equivalence class for testing */
+    if ((v->cflags&REG_FAKE) && c == 'x') {
+	cv = getcvec(v, 4, 0, 0);
+	addchr(cv, (chr)'x');
+	addchr(cv, (chr)'y');
+	if (cases) {
+	    addchr(cv, (chr)'X');
+	    addchr(cv, (chr)'Y');
+	}
+	return cv;
+    }
+
+    /* otherwise, none */
+    if (cases) {
+	return allcases(v, c);
+    }
+    cv = getcvec(v, 1, 0, 0);
+    assert(cv != NULL);
+    addchr(cv, (chr)c);
+    return cv;
+}
+
+/*
+ * cclass - supply cvec for a character class
+ *
+ * Must include case counterparts on request.
+ */
+static struct cvec *
+cclass(struct vars *v,			/* context */
+	   chr *startp,			/* where the name starts */
+	   chr *endp,				/* just past the end of the name */
+	   int cases)				/* case-independent? */
+{
+    size_t len;
+    struct cvec *cv = NULL;
+    char **namePtr;
+    int i, index;
+
+    /*
+     * The following arrays define the valid character class names.
+     */
+
+    static char *classNames[] = {
+	"alnum", "alpha", "ascii", "blank", "cntrl", "digit", "graph",
+	"lower", "print", "punct", "space", "upper", "xdigit", NULL
+    };
+
+    enum classes {
+	CC_ALNUM, CC_ALPHA, CC_ASCII, CC_BLANK, CC_CNTRL, CC_DIGIT, CC_GRAPH,
+	CC_LOWER, CC_PRINT, CC_PUNCT, CC_SPACE, CC_UPPER, CC_XDIGIT
+    };
+
+    /*
+     * Map the name to the corresponding enumerated value.
+     */
+    len = endp - startp;
+    index = -1;
+    for (namePtr=classNames,i=0 ; *namePtr!=NULL ; namePtr++,i++) {
+	if (strlen(*namePtr) == len &&
+		pg_char_and_wchar_strncmp(*namePtr, startp, len) == 0) {
+	    index = i;
+	    break;
+	}
+    }
+    if (index == -1) {
+	ERR(REG_ECTYPE);
+	return NULL;
+    }
+
+    /*
+     * Remap lower and upper to alpha if the match is case insensitive.
+     */
+
+    if (cases &&
+		((enum classes) index == CC_LOWER ||
+		 (enum classes) index == CC_UPPER))
+		index = (int) CC_ALPHA;
+    
+    /*
+     * Now compute the character class contents.
+	 *
+	 * For the moment, assume that only char codes < 256 can be in these
+	 * classes.
+     */
+
+    switch((enum classes) index) {
+    case CC_PRINT:
+    case CC_ALNUM:
+	cv = getcvec(v, UCHAR_MAX, 1, 0);
+	if (cv) {
+	    for (i=0 ; i<= UCHAR_MAX ; i++) {
+			if (pg_isalpha((chr) i))
+				addchr(cv, (chr) i);
+	    }
+		addrange(cv, (chr) '0', (chr) '9');
+	}
+	break;
+    case CC_ALPHA:
+	cv = getcvec(v, UCHAR_MAX, 0, 0);
+	if (cv) {
+	    for (i=0 ; i<= UCHAR_MAX ; i++) {
+			if (pg_isalpha((chr) i))
+				addchr(cv, (chr) i);
+	    }
+	}
+	break;
+    case CC_ASCII:
+	cv = getcvec(v, 0, 1, 0);
+	if (cv) {
+	    addrange(cv, 0, 0x7f);
+	}
+	break;
+    case CC_BLANK:
+	cv = getcvec(v, 2, 0, 0);
+	addchr(cv, '\t');
+	addchr(cv, ' ');
+	break;
+    case CC_CNTRL:
+	cv = getcvec(v, 0, 2, 0);
+	addrange(cv, 0x0, 0x1f);
+	addrange(cv, 0x7f, 0x9f);
+	break;
+    case CC_DIGIT:
+	cv = getcvec(v, 0, 1, 0);
+	if (cv) {	
+		addrange(cv, (chr) '0', (chr) '9');
+	}
+	break;
+    case CC_PUNCT:
+	cv = getcvec(v, UCHAR_MAX, 0, 0);
+	if (cv) {
+	    for (i=0 ; i<= UCHAR_MAX ; i++) {
+			if (pg_ispunct((chr) i))
+				addchr(cv, (chr) i);
+	    }
+	}
+	break;
+    case CC_XDIGIT:
+	cv = getcvec(v, 0, 3, 0);
+	if (cv) {	
+	    addrange(cv, '0', '9');
+	    addrange(cv, 'a', 'f');
+	    addrange(cv, 'A', 'F');
+	}
+	break;
+    case CC_SPACE:
+	cv = getcvec(v, UCHAR_MAX, 0, 0);
+	if (cv) {
+	    for (i=0 ; i<= UCHAR_MAX ; i++) {
+			if (pg_isspace((chr) i))
+				addchr(cv, (chr) i);
+	    }
+	}
+	break;
+    case CC_LOWER:
+	cv = getcvec(v, UCHAR_MAX, 0, 0);
+	if (cv) {
+	    for (i=0 ; i<= UCHAR_MAX ; i++) {
+			if (pg_islower((chr) i))
+				addchr(cv, (chr) i);
+	    }
+	}
+	break;
+    case CC_UPPER:
+	cv = getcvec(v, UCHAR_MAX, 0, 0);
+	if (cv) {
+	    for (i=0 ; i<= UCHAR_MAX ; i++) {
+			if (pg_isupper((chr) i))
+				addchr(cv, (chr) i);
+	    }
+	}
+	break;
+    case CC_GRAPH:
+	cv = getcvec(v, UCHAR_MAX, 0, 0);
+	if (cv) {
+	    for (i=0 ; i<= UCHAR_MAX ; i++) {
+			if (pg_isgraph((chr) i))
+				addchr(cv, (chr) i);
+	    }
+	}
+	break;
+    }
+    if (cv == NULL) {
+	ERR(REG_ESPACE);
+    }
+    return cv;
+}
+
+/*
+ * allcases - supply cvec for all case counterparts of a chr (including itself)
+ *
+ * This is a shortcut, preferably an efficient one, for simple characters;
+ * messy cases are done via range().
+ */
+static struct cvec *
+allcases(struct vars *v,			/* context */
+		 chr pc)				/* character to get case equivs of */
+{
+    struct cvec *cv;
+    chr c = (chr)pc;
+    chr lc, uc;
+
+    lc = pg_tolower((chr)c);
+    uc = pg_toupper((chr)c);
+
+	cv = getcvec(v, 2, 0, 0);
+    addchr(cv, lc);
+    if (lc != uc) {
+	addchr(cv, uc);
+    }
+    return cv;
+}
+
+/*
+ * cmp - chr-substring compare
+ *
+ * Backrefs need this.  It should preferably be efficient.
+ * Note that it does not need to report anything except equal/unequal.
+ * Note also that the length is exact, and the comparison should not
+ * stop at embedded NULs!
+ */
+static int				/* 0 for equal, nonzero for unequal */
+cmp(const chr *x, const chr *y,			/* strings to compare */
+    size_t len)				/* exact length of comparison */
+{
+    return memcmp(VS(x), VS(y), len*sizeof(chr));
+}
+
+/*
+ * casecmp - case-independent chr-substring compare
+ *
+ * REG_ICASE backrefs need this.  It should preferably be efficient.
+ * Note that it does not need to report anything except equal/unequal.
+ * Note also that the length is exact, and the comparison should not
+ * stop at embedded NULs!
+ */
+static int				/* 0 for equal, nonzero for unequal */
+casecmp(const chr *x, const chr *y,			/* strings to compare */
+		size_t len)				/* exact length of comparison */
+{
+    for (; len > 0; len--, x++, y++) {
+	if ((*x!=*y) && (pg_tolower(*x) != pg_tolower(*y))) {
+	    return 1;
+	}
+    }
+    return 0;
+}

src/backend/regex/rege_dfa.c

@@ -0,0 +1,655 @@
+/*
+ * DFA routines
+ * This file is #included by regexec.c.
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ * 
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them. 
+ * 
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ * 
+ * I'd appreciate being given credit for this package in the documentation
+ * of software which uses it, but that is not a requirement.
+ * 
+ * THIS SOFTWARE IS PROVIDED ``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
+ * HENRY SPENCER 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.
+ *
+ * $Header: /cvsroot/pgsql/src/backend/regex/rege_dfa.c,v 1.1 2003/02/05 17:41:33 tgl Exp $
+ *
+ */
+
+/*
+ * longest - longest-preferred matching engine
+ */
+static chr *			/* endpoint, or NULL */
+longest(struct vars *v,			/* used only for debug and exec flags */
+		struct dfa *d,
+		chr *start,			/* where the match should start */
+		chr *stop,			/* match must end at or before here */
+		int *hitstopp)			/* record whether hit v->stop, if non-NULL */
+{
+	chr *cp;
+	chr *realstop = (stop == v->stop) ? stop : stop + 1;
+	color co;
+	struct sset *css;
+	struct sset *ss;
+	chr *post;
+	int i;
+	struct colormap *cm = d->cm;
+
+	/* initialize */
+	css = initialize(v, d, start);
+	cp = start;
+	if (hitstopp != NULL)
+		*hitstopp = 0;
+
+	/* startup */
+	FDEBUG(("+++ startup +++\n"));
+	if (cp == v->start) {
+		co = d->cnfa->bos[(v->eflags&REG_NOTBOL) ? 0 : 1];
+		FDEBUG(("color %ld\n", (long)co));
+	} else {
+		co = GETCOLOR(cm, *(cp - 1));
+		FDEBUG(("char %c, color %ld\n", (char)*(cp-1), (long)co));
+	}
+	css = miss(v, d, css, co, cp, start);
+	if (css == NULL)
+		return NULL;
+	css->lastseen = cp;
+
+	/* main loop */
+	if (v->eflags&REG_FTRACE)
+		while (cp < realstop) {
+			FDEBUG(("+++ at c%d +++\n", css - d->ssets));
+			co = GETCOLOR(cm, *cp);
+			FDEBUG(("char %c, color %ld\n", (char)*cp, (long)co));
+			ss = css->outs[co];
+			if (ss == NULL) {
+				ss = miss(v, d, css, co, cp+1, start);
+				if (ss == NULL)
+					break;	/* NOTE BREAK OUT */
+			}
+			cp++;
+			ss->lastseen = cp;
+			css = ss;
+		}
+	else
+		while (cp < realstop) {
+			co = GETCOLOR(cm, *cp);
+			ss = css->outs[co];
+			if (ss == NULL) {
+				ss = miss(v, d, css, co, cp+1, start);
+				if (ss == NULL)
+					break;	/* NOTE BREAK OUT */
+			}
+			cp++;
+			ss->lastseen = cp;
+			css = ss;
+		}
+
+	/* shutdown */
+	FDEBUG(("+++ shutdown at c%d +++\n", css - d->ssets));
+	if (cp == v->stop && stop == v->stop) {
+		if (hitstopp != NULL)
+			*hitstopp = 1;
+		co = d->cnfa->eos[(v->eflags&REG_NOTEOL) ? 0 : 1];
+		FDEBUG(("color %ld\n", (long)co));
+		ss = miss(v, d, css, co, cp, start);
+		/* special case:  match ended at eol? */
+		if (ss != NULL && (ss->flags&POSTSTATE))
+			return cp;
+		else if (ss != NULL)
+			ss->lastseen = cp;	/* to be tidy */
+	}
+
+	/* find last match, if any */
+	post = d->lastpost;
+	for (ss = d->ssets, i = d->nssused; i > 0; ss++, i--)
+		if ((ss->flags&POSTSTATE) && post != ss->lastseen &&
+					(post == NULL || post < ss->lastseen))
+			post = ss->lastseen;
+	if (post != NULL)		/* found one */
+		return post - 1;
+
+	return NULL;
+}
+
+/*
+ * shortest - shortest-preferred matching engine
+ */
+static chr *			/* endpoint, or NULL */
+shortest(struct vars *v,
+		 struct dfa *d,
+		 chr *start,			/* where the match should start */
+		 chr *min,			/* match must end at or after here */
+		 chr *max,			/* match must end at or before here */
+		 chr **coldp,			/* store coldstart pointer here, if nonNULL */
+		 int *hitstopp)			/* record whether hit v->stop, if non-NULL */
+{
+	chr *cp;
+	chr *realmin = (min == v->stop) ? min : min + 1;
+	chr *realmax = (max == v->stop) ? max : max + 1;
+	color co;
+	struct sset *css;
+	struct sset *ss;
+	struct colormap *cm = d->cm;
+
+	/* initialize */
+	css = initialize(v, d, start);
+	cp = start;
+	if (hitstopp != NULL)
+		*hitstopp = 0;
+
+	/* startup */
+	FDEBUG(("--- startup ---\n"));
+	if (cp == v->start) {
+		co = d->cnfa->bos[(v->eflags&REG_NOTBOL) ? 0 : 1];
+		FDEBUG(("color %ld\n", (long)co));
+	} else {
+		co = GETCOLOR(cm, *(cp - 1));
+		FDEBUG(("char %c, color %ld\n", (char)*(cp-1), (long)co));
+	}
+	css = miss(v, d, css, co, cp, start);
+	if (css == NULL)
+		return NULL;
+	css->lastseen = cp;
+	ss = css;
+
+	/* main loop */
+	if (v->eflags&REG_FTRACE)
+		while (cp < realmax) {
+			FDEBUG(("--- at c%d ---\n", css - d->ssets));
+			co = GETCOLOR(cm, *cp);
+			FDEBUG(("char %c, color %ld\n", (char)*cp, (long)co));
+			ss = css->outs[co];
+			if (ss == NULL) {
+				ss = miss(v, d, css, co, cp+1, start);
+				if (ss == NULL)
+					break;	/* NOTE BREAK OUT */
+			}
+			cp++;
+			ss->lastseen = cp;
+			css = ss;
+			if ((ss->flags&POSTSTATE) && cp >= realmin)
+				break;		/* NOTE BREAK OUT */
+		}
+	else
+		while (cp < realmax) {
+			co = GETCOLOR(cm, *cp);
+			ss = css->outs[co];
+			if (ss == NULL) {
+				ss = miss(v, d, css, co, cp+1, start);
+				if (ss == NULL)
+					break;	/* NOTE BREAK OUT */
+			}
+			cp++;
+			ss->lastseen = cp;
+			css = ss;
+			if ((ss->flags&POSTSTATE) && cp >= realmin)
+				break;		/* NOTE BREAK OUT */
+		}
+
+	if (ss == NULL)
+		return NULL;
+
+	if (coldp != NULL)	/* report last no-progress state set, if any */
+		*coldp = lastcold(v, d);
+
+	if ((ss->flags&POSTSTATE) && cp > min) {
+		assert(cp >= realmin);
+		cp--;
+	} else if (cp == v->stop && max == v->stop) {
+		co = d->cnfa->eos[(v->eflags&REG_NOTEOL) ? 0 : 1];
+		FDEBUG(("color %ld\n", (long)co));
+		ss = miss(v, d, css, co, cp, start);
+		/* match might have ended at eol */
+		if ((ss == NULL || !(ss->flags&POSTSTATE)) && hitstopp != NULL)
+			*hitstopp = 1;
+	}
+
+	if (ss == NULL || !(ss->flags&POSTSTATE))
+		return NULL;
+
+	return cp;
+}
+
+/*
+ * lastcold - determine last point at which no progress had been made
+ */
+static chr *			/* endpoint, or NULL */
+lastcold(struct vars *v,
+		 struct dfa *d)
+{
+	struct sset *ss;
+	chr *nopr;
+	int i;
+
+	nopr = d->lastnopr;
+	if (nopr == NULL)
+		nopr = v->start;
+	for (ss = d->ssets, i = d->nssused; i > 0; ss++, i--)
+		if ((ss->flags&NOPROGRESS) && nopr < ss->lastseen)
+			nopr = ss->lastseen;
+	return nopr;
+}
+
+/*
+ * newdfa - set up a fresh DFA
+ */
+static struct dfa *
+newdfa(struct vars *v,
+	   struct cnfa *cnfa,
+	   struct colormap *cm,
+	   struct smalldfa *small)		/* preallocated space, may be NULL */
+{
+	struct dfa *d;
+	size_t nss = cnfa->nstates * 2;
+	int wordsper = (cnfa->nstates + UBITS - 1) / UBITS;
+	struct smalldfa *smallwas = small;
+
+	assert(cnfa != NULL && cnfa->nstates != 0);
+
+	if (nss <= FEWSTATES && cnfa->ncolors <= FEWCOLORS) {
+		assert(wordsper == 1);
+		if (small == NULL) {
+			small = (struct smalldfa *)MALLOC(
+						sizeof(struct smalldfa));
+			if (small == NULL) {
+				ERR(REG_ESPACE);
+				return NULL;
+			}
+		}
+		d = &small->dfa;
+		d->ssets = small->ssets;
+		d->statesarea = small->statesarea;
+		d->work = &d->statesarea[nss];
+		d->outsarea = small->outsarea;
+		d->incarea = small->incarea;
+		d->cptsmalloced = 0;
+		d->mallocarea = (smallwas == NULL) ? (char *)small : NULL;
+	} else {
+		d = (struct dfa *)MALLOC(sizeof(struct dfa));
+		if (d == NULL) {
+			ERR(REG_ESPACE);
+			return NULL;
+		}
+		d->ssets = (struct sset *)MALLOC(nss * sizeof(struct sset));
+		d->statesarea = (unsigned *)MALLOC((nss+WORK) * wordsper *
+							sizeof(unsigned));
+		d->work = &d->statesarea[nss * wordsper];
+		d->outsarea = (struct sset **)MALLOC(nss * cnfa->ncolors *
+							sizeof(struct sset *));
+		d->incarea = (struct arcp *)MALLOC(nss * cnfa->ncolors *
+							sizeof(struct arcp));
+		d->cptsmalloced = 1;
+		d->mallocarea = (char *)d;
+		if (d->ssets == NULL || d->statesarea == NULL ||
+				d->outsarea == NULL || d->incarea == NULL) {
+			freedfa(d);
+			ERR(REG_ESPACE);
+			return NULL;
+		}
+	}
+
+	d->nssets = (v->eflags&REG_SMALL) ? 7 : nss;
+	d->nssused = 0;
+	d->nstates = cnfa->nstates;
+	d->ncolors = cnfa->ncolors;
+	d->wordsper = wordsper;
+	d->cnfa = cnfa;
+	d->cm = cm;
+	d->lastpost = NULL;
+	d->lastnopr = NULL;
+	d->search = d->ssets;
+
+	/* initialization of sset fields is done as needed */
+
+	return d;
+}
+
+/*
+ * freedfa - free a DFA
+ */
+static void
+freedfa(struct dfa *d)
+{
+	if (d->cptsmalloced) {
+		if (d->ssets != NULL)
+			FREE(d->ssets);
+		if (d->statesarea != NULL)
+			FREE(d->statesarea);
+		if (d->outsarea != NULL)
+			FREE(d->outsarea);
+		if (d->incarea != NULL)
+			FREE(d->incarea);
+	}
+
+	if (d->mallocarea != NULL)
+		FREE(d->mallocarea);
+}
+
+/*
+ * hash - construct a hash code for a bitvector
+ *
+ * There are probably better ways, but they're more expensive.
+ */
+static unsigned
+hash(unsigned *uv,
+	 int n)
+{
+	int i;
+	unsigned h;
+
+	h = 0;
+	for (i = 0; i < n; i++)
+		h ^= uv[i];
+	return h;
+}
+
+/*
+ * initialize - hand-craft a cache entry for startup, otherwise get ready
+ */
+static struct sset *
+initialize(struct vars *v,			/* used only for debug flags */
+		   struct dfa *d,
+		   chr *start)
+{
+	struct sset *ss;
+	int i;
+
+	/* is previous one still there? */
+	if (d->nssused > 0 && (d->ssets[0].flags&STARTER))
+		ss = &d->ssets[0];
+	else {				/* no, must (re)build it */
+		ss = getvacant(v, d, start, start);
+		for (i = 0; i < d->wordsper; i++)
+			ss->states[i] = 0;
+		BSET(ss->states, d->cnfa->pre);
+		ss->hash = HASH(ss->states, d->wordsper);
+		assert(d->cnfa->pre != d->cnfa->post);
+		ss->flags = STARTER|LOCKED|NOPROGRESS;
+		/* lastseen dealt with below */
+	}
+
+	for (i = 0; i < d->nssused; i++)
+		d->ssets[i].lastseen = NULL;
+	ss->lastseen = start;		/* maybe untrue, but harmless */
+	d->lastpost = NULL;
+	d->lastnopr = NULL;
+	return ss;
+}
+
+/*
+ * miss - handle a cache miss
+ */
+static struct sset *		/* NULL if goes to empty set */
+miss(struct vars *v,			/* used only for debug flags */
+	 struct dfa *d,
+	 struct sset *css,
+	 pcolor co,
+	 chr *cp,			/* next chr */
+	 chr *start)			/* where the attempt got started */
+{
+	struct cnfa *cnfa = d->cnfa;
+	int i;
+	unsigned h;
+	struct carc *ca;
+	struct sset *p;
+	int ispost;
+	int noprogress;
+	int gotstate;
+	int dolacons;
+	int sawlacons;
+
+	/* for convenience, we can be called even if it might not be a miss */
+	if (css->outs[co] != NULL) {
+		FDEBUG(("hit\n"));
+		return css->outs[co];
+	}
+	FDEBUG(("miss\n"));
+
+	/* first, what set of states would we end up in? */
+	for (i = 0; i < d->wordsper; i++)
+		d->work[i] = 0;
+	ispost = 0;
+	noprogress = 1;
+	gotstate = 0;
+	for (i = 0; i < d->nstates; i++)
+		if (ISBSET(css->states, i))
+			for (ca = cnfa->states[i]+1; ca->co != COLORLESS; ca++)
+				if (ca->co == co) {
+					BSET(d->work, ca->to);
+					gotstate = 1;
+					if (ca->to == cnfa->post)
+						ispost = 1;
+					if (!cnfa->states[ca->to]->co)
+						noprogress = 0;
+					FDEBUG(("%d -> %d\n", i, ca->to));
+				}
+	dolacons = (gotstate) ? (cnfa->flags&HASLACONS) : 0;
+	sawlacons = 0;
+	while (dolacons) {		/* transitive closure */
+		dolacons = 0;
+		for (i = 0; i < d->nstates; i++)
+			if (ISBSET(d->work, i))
+				for (ca = cnfa->states[i]+1; ca->co != COLORLESS;
+									ca++) {
+					if (ca->co <= cnfa->ncolors)
+						continue; /* NOTE CONTINUE */
+					sawlacons = 1;
+					if (ISBSET(d->work, ca->to))
+						continue; /* NOTE CONTINUE */
+					if (!lacon(v, cnfa, cp, ca->co))
+						continue; /* NOTE CONTINUE */
+					BSET(d->work, ca->to);
+					dolacons = 1;
+					if (ca->to == cnfa->post)
+						ispost = 1;
+					if (!cnfa->states[ca->to]->co)
+						noprogress = 0;
+					FDEBUG(("%d :> %d\n", i, ca->to));
+				}
+	}
+	if (!gotstate)
+		return NULL;
+	h = HASH(d->work, d->wordsper);
+
+	/* next, is that in the cache? */
+	for (p = d->ssets, i = d->nssused; i > 0; p++, i--)
+		if (HIT(h, d->work, p, d->wordsper)) {
+			FDEBUG(("cached c%d\n", p - d->ssets));
+			break;			/* NOTE BREAK OUT */
+		}
+	if (i == 0) {		/* nope, need a new cache entry */
+		p = getvacant(v, d, cp, start);
+		assert(p != css);
+		for (i = 0; i < d->wordsper; i++)
+			p->states[i] = d->work[i];
+		p->hash = h;
+		p->flags = (ispost) ? POSTSTATE : 0;
+		if (noprogress)
+			p->flags |= NOPROGRESS;
+		/* lastseen to be dealt with by caller */
+	}
+
+	if (!sawlacons) {		/* lookahead conds. always cache miss */
+		FDEBUG(("c%d[%d]->c%d\n", css - d->ssets, co, p - d->ssets));
+		css->outs[co] = p;
+		css->inchain[co] = p->ins;
+		p->ins.ss = css;
+		p->ins.co = (color)co;
+	}
+	return p;
+}
+
+/*
+ * lacon - lookahead-constraint checker for miss()
+ */
+static int			/* predicate:  constraint satisfied? */
+lacon(struct vars *v,
+	  struct cnfa *pcnfa,		/* parent cnfa */
+	  chr *cp,
+	  pcolor co)			/* "color" of the lookahead constraint */
+{
+	int n;
+	struct subre *sub;
+	struct dfa *d;
+	struct smalldfa sd;
+	chr *end;
+
+	n = co - pcnfa->ncolors;
+	assert(n < v->g->nlacons && v->g->lacons != NULL);
+	FDEBUG(("=== testing lacon %d\n", n));
+	sub = &v->g->lacons[n];
+	d = newdfa(v, &sub->cnfa, &v->g->cmap, &sd);
+	if (d == NULL) {
+		ERR(REG_ESPACE);
+		return 0;
+	}
+	end = longest(v, d, cp, v->stop, (int *)NULL);
+	freedfa(d);
+	FDEBUG(("=== lacon %d match %d\n", n, (end != NULL)));
+	return (sub->subno) ? (end != NULL) : (end == NULL);
+}
+
+/*
+ * getvacant - get a vacant state set
+ * This routine clears out the inarcs and outarcs, but does not otherwise
+ * clear the innards of the state set -- that's up to the caller.
+ */
+static struct sset *
+getvacant(struct vars *v,			/* used only for debug flags */
+		  struct dfa *d,
+		  chr *cp,
+		  chr *start)
+{
+	int i;
+	struct sset *ss;
+	struct sset *p;
+	struct arcp ap;
+	struct arcp lastap;
+	color co;
+
+	ss = pickss(v, d, cp, start);
+	assert(!(ss->flags&LOCKED));
+
+	/* clear out its inarcs, including self-referential ones */
+	ap = ss->ins;
+	while ((p = ap.ss) != NULL) {
+		co = ap.co;
+		FDEBUG(("zapping c%d's %ld outarc\n", p - d->ssets, (long)co));
+		p->outs[co] = NULL;
+		ap = p->inchain[co];
+		p->inchain[co].ss = NULL;	/* paranoia */
+	}
+	ss->ins.ss = NULL;
+
+	/* take it off the inarc chains of the ssets reached by its outarcs */
+	for (i = 0; i < d->ncolors; i++) {
+		p = ss->outs[i];
+		assert(p != ss);		/* not self-referential */
+		if (p == NULL)
+			continue;		/* NOTE CONTINUE */
+		FDEBUG(("del outarc %d from c%d's in chn\n", i, p - d->ssets));
+		if (p->ins.ss == ss && p->ins.co == i)
+			p->ins = ss->inchain[i];
+		else {
+			assert(p->ins.ss != NULL);
+			for (ap = p->ins; ap.ss != NULL &&
+						!(ap.ss == ss && ap.co == i);
+						ap = ap.ss->inchain[ap.co])
+				lastap = ap;
+			assert(ap.ss != NULL);
+			lastap.ss->inchain[lastap.co] = ss->inchain[i];
+		}
+		ss->outs[i] = NULL;
+		ss->inchain[i].ss = NULL;
+	}
+
+	/* if ss was a success state, may need to remember location */
+	if ((ss->flags&POSTSTATE) && ss->lastseen != d->lastpost &&
+			(d->lastpost == NULL || d->lastpost < ss->lastseen))
+		d->lastpost = ss->lastseen;
+
+	/* likewise for a no-progress state */
+	if ((ss->flags&NOPROGRESS) && ss->lastseen != d->lastnopr &&
+			(d->lastnopr == NULL || d->lastnopr < ss->lastseen))
+		d->lastnopr = ss->lastseen;
+
+	return ss;
+}
+
+/*
+ * pickss - pick the next stateset to be used
+ */
+static struct sset *
+pickss(struct vars *v,			/* used only for debug flags */
+	   struct dfa *d,
+	   chr *cp,
+	   chr *start)
+{
+	int i;
+	struct sset *ss;
+	struct sset *end;
+	chr *ancient;
+
+	/* shortcut for cases where cache isn't full */
+	if (d->nssused < d->nssets) {
+		i = d->nssused;
+		d->nssused++;
+		ss = &d->ssets[i];
+		FDEBUG(("new c%d\n", i));
+		/* set up innards */
+		ss->states = &d->statesarea[i * d->wordsper];
+		ss->flags = 0;
+		ss->ins.ss = NULL;
+		ss->ins.co = WHITE;		/* give it some value */
+		ss->outs = &d->outsarea[i * d->ncolors];
+		ss->inchain = &d->incarea[i * d->ncolors];
+		for (i = 0; i < d->ncolors; i++) {
+			ss->outs[i] = NULL;
+			ss->inchain[i].ss = NULL;
+		}
+		return ss;
+	}
+
+	/* look for oldest, or old enough anyway */
+	if (cp - start > d->nssets*2/3)		/* oldest 33% are expendable */
+		ancient = cp - d->nssets*2/3;
+	else
+		ancient = start;
+	for (ss = d->search, end = &d->ssets[d->nssets]; ss < end; ss++)
+		if ((ss->lastseen == NULL || ss->lastseen < ancient) &&
+							!(ss->flags&LOCKED)) {
+			d->search = ss + 1;
+			FDEBUG(("replacing c%d\n", ss - d->ssets));
+			return ss;
+		}
+	for (ss = d->ssets, end = d->search; ss < end; ss++)
+		if ((ss->lastseen == NULL || ss->lastseen < ancient) &&
+							!(ss->flags&LOCKED)) {
+			d->search = ss + 1;
+			FDEBUG(("replacing c%d\n", ss - d->ssets));
+			return ss;
+		}
+
+	/* nobody's old enough?!? -- something's really wrong */
+	FDEBUG(("can't find victim to replace!\n"));
+	assert(NOTREACHED);
+	ERR(REG_ASSERT);
+	return d->ssets;
+}

src/backend/regex/regerror.c

@@ -1,181 +1,110 @@
-/*-
+/*
- * Copyright (c) 1992, 1993, 1994 Henry Spencer.
+ * regerror - error-code expansion
- * 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
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
- * modification, are permitted provided that the following conditions
+ * 
- * are met:
+ * Development of this software was funded, in part, by Cray Research Inc.,
- * 1. Redistributions of source code must retain the above copyright
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
- *	  notice, this list of conditions and the following disclaimer.
+ * Corporation, none of whom are responsible for the results.  The author
- * 2. Redistributions in binary form must reproduce the above copyright
+ * thanks all of them. 
- *	  notice, this list of conditions and the following disclaimer in the
+ * 
- *	  documentation and/or other materials provided with the distribution.
+ * Redistribution and use in source and binary forms -- with or without
- * 3. All advertising materials mentioning features or use of this software
+ * modification -- are permitted for any purpose, provided that
- *	  must display the following acknowledgement:
+ * redistributions in source form retain this entire copyright notice and
- *		This product includes software developed by the University of
+ * indicate the origin and nature of any modifications.
- *		California, Berkeley and its contributors.
+ * 
- * 4. Neither the name of the University nor the names of its contributors
+ * I'd appreciate being given credit for this package in the documentation
- *	  may be used to endorse or promote products derived from this software
+ * of software which uses it, but that is not a requirement.
- *	  without specific prior written permission.
+ * 
+ * THIS SOFTWARE IS PROVIDED ``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
+ * HENRY SPENCER 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.
  *
- * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * $Header: /cvsroot/pgsql/src/backend/regex/regerror.c,v 1.25 2003/02/05 17:41:33 tgl Exp $
- * 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
  */
 
-#include "postgres.h"
+#include "regex/regguts.h"
-
-#include <ctype.h>
-#include <limits.h>
-#include <assert.h>
 
-#include "regex/regex.h"
+/* unknown-error explanation */
-#include "regex/utils.h"
+static char unk[] = "*** unknown regex error code 0x%x ***";
-#include "regex/regex2.h"
-
-static char *regatoi(const regex_t *preg, char *localbuf);
-
-static struct rerr
-{
-	int			code;
-	char	   *name;
-	char	   *explain;
-}	rerrs[] =
 
-{
+/* struct to map among codes, code names, and explanations */
-	{
+static struct rerr {
-		/* NOMATCH is not really an error condition, it just says no match */
+	int code;
-		REG_NOMATCH, "REG_NOMATCH", "no pattern match found"
+	char *name;
-	},
+	char *explain;
-	{
+} rerrs[] = {
-		REG_BADPAT, "REG_BADPAT", "invalid regex struct"
+	/* the actual table is built from regex.h */
-	},
+#include "regex/regerrs.h"
-	{
+	{ -1,	"",	"oops" },	/* explanation special-cased in code */
-		REG_ECOLLATE, "REG_ECOLLATE", "invalid collating element"
-	},
-	{
-		REG_ECTYPE, "REG_ECTYPE", "invalid character class"
-	},
-	{
-		REG_EESCAPE, "REG_EESCAPE", "trailing backslash (\\)"
-	},
-	{
-		REG_ESUBREG, "REG_ESUBREG", "invalid backreference number"
-	},
-	{
-		REG_EBRACK, "REG_EBRACK", "brackets [ ] not balanced"
-	},
-	{
-		REG_EPAREN, "REG_EPAREN", "parentheses ( ) not balanced"
-	},
-	{
-		REG_EBRACE, "REG_EBRACE", "braces { } not balanced"
-	},
-	{
-		REG_BADBR, "REG_BADBR", "invalid repetition count(s) in { }"
-	},
-	{
-		REG_ERANGE, "REG_ERANGE", "invalid character range in [ ]"
-	},
-	{
-		REG_ESPACE, "REG_ESPACE", "ran out of memory"
-	},
-	{
-		REG_BADRPT, "REG_BADRPT", "?, *, or + operand invalid"
-	},
-	{
-		REG_EMPTY, "REG_EMPTY", "empty expression or subexpression"
-	},
-	{
-		REG_ASSERT, "REG_ASSERT", "\"can't happen\" -- you found a bug"
-	},
-	{
-		REG_INVARG, "REG_INVARG", "invalid argument to regex routine"
-	},
-	{
-		0, "", "*** unknown regexp error code ***"
-	}
 };
 
 /*
- * regerror - the interface to error numbers
+ * pg_regerror - the interface to error numbers
  */
 /* ARGSUSED */
-size_t
+size_t				/* actual space needed (including NUL) */
-pg_regerror(int errcode, const regex_t *preg,
+pg_regerror(int errcode,		/* error code, or REG_ATOI or REG_ITOA */
-			char *errbuf, size_t errbuf_size)
+			const regex_t *preg, /* associated regex_t (unused at present) */
+			char *errbuf,		/* result buffer (unless errbuf_size==0) */
+			size_t errbuf_size)	/* available space in errbuf, can be 0 */
 {
 	struct rerr *r;
-	size_t		len;
+	char *msg;
-	int			target = errcode & ~REG_ITOA;
+	char convbuf[sizeof(unk)+50];	/* 50 = plenty for int */
-	char	   *s;
+	size_t len;
-	char		convbuf[50];
+	int icode;
 
-	if (errcode == REG_ATOI)
+	switch (errcode) {
-		s = regatoi(preg, convbuf);
+	case REG_ATOI:		/* convert name to number */
-	else
+		for (r = rerrs; r->code >= 0; r++)
-	{
+			if (strcmp(r->name, errbuf) == 0)
-		for (r = rerrs; r->code != 0; r++)
-			if (r->code == target)
 				break;
-
+		sprintf(convbuf, "%d", r->code);	/* -1 for unknown */
-		if (errcode & REG_ITOA)
+		msg = convbuf;
-		{
+		break;
-			if (r->code != 0)
+	case REG_ITOA:		/* convert number to name */
-				strcpy(convbuf, r->name);
+		icode = atoi(errbuf);	/* not our problem if this fails */
-			else
+		for (r = rerrs; r->code >= 0; r++)
-				sprintf(convbuf, "REG_0x%x", target);
+			if (r->code == icode)
-			assert(strlen(convbuf) < sizeof(convbuf));
+				break;
-			s = convbuf;
+		if (r->code >= 0)
+			msg = r->name;
+		else {			/* unknown; tell him the number */
+			sprintf(convbuf, "REG_%u", (unsigned)icode);
+			msg = convbuf;
 		}
-		else
+		break;
-			s = r->explain;
+	default:		/* a real, normal error code */
+		for (r = rerrs; r->code >= 0; r++)
+			if (r->code == errcode)
+				break;
+		if (r->code >= 0)
+			msg = r->explain;
+		else {			/* unknown; say so */
+			sprintf(convbuf, unk, errcode);
+			msg = convbuf;
+		}
+		break;
 	}
 
-	len = strlen(s) + 1;
+	len = strlen(msg) + 1;		/* space needed, including NUL */
-	if (errbuf_size > 0)
+	if (errbuf_size > 0) {
-	{
 		if (errbuf_size > len)
-			strcpy(errbuf, s);
+			strcpy(errbuf, msg);
-		else
+		else {			/* truncate to fit */
-		{
+			strncpy(errbuf, msg, errbuf_size-1);
-			strncpy(errbuf, s, errbuf_size - 1);
+			errbuf[errbuf_size-1] = '\0';
-			errbuf[errbuf_size - 1] = '\0';
 		}
 	}
 
 	return len;
 }
-
-/*
- * regatoi - internal routine to implement REG_ATOI
- */
-static char *
-regatoi(const regex_t *preg, char *localbuf)
-{
-	struct rerr *r;
-
-	for (r = rerrs; r->code != 0; r++)
-		if (pg_char_and_wchar_strcmp(r->name, preg->re_endp) == 0)
-			break;
-
-	if (r->code == 0)
-		return "0";
-
-	sprintf(localbuf, "%d", r->code);
-	return localbuf;
-}

src/backend/regex/regex.3

@@ -1,538 +0,0 @@
-.\" 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. All advertising materials mentioning features or use of this software
-.\"    must display the following acknowledgement:
-.\"	This product includes software developed by the University of
-.\"	California, Berkeley and its contributors.
-.\" 4. Neither the name of the University nor the names of its contributors
-.\"    may be used to endorse or promote products derived from this software
-.\"    without specific prior written permission.
-.\"
-.\" 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.3	8.4 (Berkeley) 3/20/94
-.\"
-.TH REGEX 3 "March 20, 1994"
-.de ZR
-.\" one other place knows this name:  the SEE ALSO section
-.IR re_format (7) \\$1
-..
-.SH NAME
-regcomp, regexec, regerror, regfree \- regular-expression library
-.SH SYNOPSIS
-.ft B
-.\".na
-#include <sys/types.h>
-.br
-#include <regex.h>
-.HP 10
-int regcomp(regex_t\ *preg, const\ char\ *pattern, int\ cflags);
-.HP
-int\ regexec(const\ regex_t\ *preg, const\ char\ *string,
-size_t\ nmatch, regmatch_t\ pmatch[], int\ eflags);
-.HP
-size_t\ regerror(int\ errcode, const\ regex_t\ *preg,
-char\ *errbuf, size_t\ errbuf_size);
-.HP
-void\ regfree(regex_t\ *preg);
-.\".ad
-.ft
-.SH DESCRIPTION
-These routines implement POSIX 1003.2 regular expressions (``RE''s);
-see
-.ZR .
-.I Regcomp
-compiles an RE written as a string into an internal form,
-.I regexec
-matches that internal form against a string and reports results,
-.I regerror
-transforms error codes from either into human-readable messages,
-and
-.I regfree
-frees any dynamically-allocated storage used by the internal form
-of an RE.
-.PP
-The header
-.I <regex.h>
-declares two structure types,
-.I regex_t
-and
-.IR regmatch_t ,
-the former for compiled internal forms and the latter for match reporting.
-It also declares the four functions,
-a type
-.IR regoff_t ,
-and a number of constants with names starting with ``REG_''.
-.PP
-.I Regcomp
-compiles the regular expression contained in the
-.I pattern
-string,
-subject to the flags in
-.IR cflags ,
-and places the results in the
-.I regex_t
-structure pointed to by
-.IR preg .
-.I Cflags
-is the bitwise OR of zero or more of the following flags:
-.IP REG_EXTENDED \w'REG_EXTENDED'u+2n
-Compile modern (``extended'') REs,
-rather than the obsolete (``basic'') REs that
-are the default.
-.IP REG_BASIC
-This is a synonym for 0,
-provided as a counterpart to REG_EXTENDED to improve readability.
-.IP REG_NOSPEC
-Compile with recognition of all special characters turned off.
-All characters are thus considered ordinary,
-so the ``RE'' is a literal string.
-This is an extension,
-compatible with but not specified by POSIX 1003.2,
-and should be used with
-caution in software intended to be portable to other systems.
-REG_EXTENDED and REG_NOSPEC may not be used
-in the same call to
-.IR regcomp .
-.IP REG_ICASE
-Compile for matching that ignores upper/lower case distinctions.
-See
-.ZR .
-.IP REG_NOSUB
-Compile for matching that need only report success or failure,
-not what was matched.
-.IP REG_NEWLINE
-Compile for newline-sensitive matching.
-By default, newline is a completely ordinary character with no special
-meaning in either REs or strings.
-With this flag,
-`[^' bracket expressions and `.' never match newline,
-a `^' anchor matches the null string after any newline in the string
-in addition to its normal function,
-and the `$' anchor matches the null string before any newline in the
-string in addition to its normal function.
-.IP REG_PEND
-The regular expression ends,
-not at the first NUL,
-but just before the character pointed to by the
-.I re_endp
-member of the structure pointed to by
-.IR preg .
-The
-.I re_endp
-member is of type
-.IR const\ char\ * .
-This flag permits inclusion of NULs in the RE;
-they are considered ordinary characters.
-This is an extension,
-compatible with but not specified by POSIX 1003.2,
-and should be used with
-caution in software intended to be portable to other systems.
-.PP
-When successful,
-.I regcomp
-returns 0 and fills in the structure pointed to by
-.IR preg .
-One member of that structure
-(other than
-.IR re_endp )
-is publicized:
-.IR re_nsub ,
-of type
-.IR size_t ,
-contains the number of parenthesized subexpressions within the RE
-(except that the value of this member is undefined if the
-REG_NOSUB flag was used).
-If
-.I regcomp
-fails, it returns a non-zero error code;
-see DIAGNOSTICS.
-.PP
-.I Regexec
-matches the compiled RE pointed to by
-.I preg
-against the
-.IR string ,
-subject to the flags in
-.IR eflags ,
-and reports results using
-.IR nmatch ,
-.IR pmatch ,
-and the returned value.
-The RE must have been compiled by a previous invocation of
-.IR regcomp .
-The compiled form is not altered during execution of
-.IR regexec ,
-so a single compiled RE can be used simultaneously by multiple threads.
-.PP
-By default,
-the NUL-terminated string pointed to by
-.I string
-is considered to be the text of an entire line, minus any terminating
-newline.
-The
-.I eflags
-argument is the bitwise OR of zero or more of the following flags:
-.IP REG_NOTBOL \w'REG_STARTEND'u+2n
-The first character of
-the string
-is not the beginning of a line, so the `^' anchor should not match before it.
-This does not affect the behavior of newlines under REG_NEWLINE.
-.IP REG_NOTEOL
-The NUL terminating
-the string
-does not end a line, so the `$' anchor should not match before it.
-This does not affect the behavior of newlines under REG_NEWLINE.
-.IP REG_STARTEND
-The string is considered to start at
-\fIstring\fR\ + \fIpmatch\fR[0].\fIrm_so\fR
-and to have a terminating NUL located at
-\fIstring\fR\ + \fIpmatch\fR[0].\fIrm_eo\fR
-(there need not actually be a NUL at that location),
-regardless of the value of
-.IR nmatch .
-See below for the definition of
-.IR pmatch
-and
-.IR nmatch .
-This is an extension,
-compatible with but not specified by POSIX 1003.2,
-and should be used with
-caution in software intended to be portable to other systems.
-Note that a non-zero \fIrm_so\fR does not imply REG_NOTBOL;
-REG_STARTEND affects only the location of the string,
-not how it is matched.
-.PP
-See
-.ZR
-for a discussion of what is matched in situations where an RE or a
-portion thereof could match any of several substrings of
-.IR string .
-.PP
-Normally,
-.I regexec
-returns 0 for success and the non-zero code REG_NOMATCH for failure.
-Other non-zero error codes may be returned in exceptional situations;
-see DIAGNOSTICS.
-.PP
-If REG_NOSUB was specified in the compilation of the RE,
-or if
-.I nmatch
-is 0,
-.I regexec
-ignores the
-.I pmatch
-argument (but see below for the case where REG_STARTEND is specified).
-Otherwise,
-.I pmatch
-points to an array of
-.I nmatch
-structures of type
-.IR regmatch_t .
-Such a structure has at least the members
-.I rm_so
-and
-.IR rm_eo ,
-both of type
-.I regoff_t
-(a signed arithmetic type at least as large as an
-.I off_t
-and a
-.IR ssize_t ),
-containing respectively the offset of the first character of a substring
-and the offset of the first character after the end of the substring.
-Offsets are measured from the beginning of the
-.I string
-argument given to
-.IR regexec .
-An empty substring is denoted by equal offsets,
-both indicating the character following the empty substring.
-.PP
-The 0th member of the
-.I pmatch
-array is filled in to indicate what substring of
-.I string
-was matched by the entire RE.
-Remaining members report what substring was matched by parenthesized
-subexpressions within the RE;
-member
-.I i
-reports subexpression
-.IR i ,
-with subexpressions counted (starting at 1) by the order of their opening
-parentheses in the RE, left to right.
-Unused entries in the array\(emcorresponding either to subexpressions that
-did not participate in the match at all, or to subexpressions that do not
-exist in the RE (that is, \fIi\fR\ > \fIpreg\fR\->\fIre_nsub\fR)\(emhave both
-.I rm_so
-and
-.I rm_eo
-set to \-1.
-If a subexpression participated in the match several times,
-the reported substring is the last one it matched.
-(Note, as an example in particular, that when the RE `(b*)+' matches `bbb',
-the parenthesized subexpression matches each of the three `b's and then
-an infinite number of empty strings following the last `b',
-so the reported substring is one of the empties.)
-.PP
-If REG_STARTEND is specified,
-.I pmatch
-must point to at least one
-.I regmatch_t
-(even if
-.I nmatch
-is 0 or REG_NOSUB was specified),
-to hold the input offsets for REG_STARTEND.
-Use for output is still entirely controlled by
-.IR nmatch ;
-if
-.I nmatch
-is 0 or REG_NOSUB was specified,
-the value of
-.IR pmatch [0]
-will not be changed by a successful
-.IR regexec .
-.PP
-.I Regerror
-maps a non-zero
-.I errcode
-from either
-.I regcomp
-or
-.I regexec
-to a human-readable, printable message.
-If
-.I preg
-is non-NULL,
-the error code should have arisen from use of
-the
-.I regex_t
-pointed to by
-.IR preg ,
-and if the error code came from
-.IR regcomp ,
-it should have been the result from the most recent
-.I regcomp
-using that
-.IR regex_t .
-.RI ( Regerror
-may be able to supply a more detailed message using information
-from the
-.IR regex_t .)
-.I Regerror
-places the NUL-terminated message into the buffer pointed to by
-.IR errbuf ,
-limiting the length (including the NUL) to at most
-.I errbuf_size
-bytes.
-If the whole message won't fit,
-as much of it as will fit before the terminating NUL is supplied.
-In any case,
-the returned value is the size of buffer needed to hold the whole
-message (including terminating NUL).
-If
-.I errbuf_size
-is 0,
-.I errbuf
-is ignored but the return value is still correct.
-.PP
-If the
-.I errcode
-given to
-.I regerror
-is first ORed with REG_ITOA,
-the ``message'' that results is the printable name of the error code,
-e.g. ``REG_NOMATCH'',
-rather than an explanation thereof.
-If
-.I errcode
-is REG_ATOI,
-then
-.I preg
-shall be non-NULL and the
-.I re_endp
-member of the structure it points to
-must point to the printable name of an error code;
-in this case, the result in
-.I errbuf
-is the decimal digits of
-the numeric value of the error code
-(0 if the name is not recognized).
-REG_ITOA and REG_ATOI are intended primarily as debugging facilities;
-they are extensions,
-compatible with but not specified by POSIX 1003.2,
-and should be used with
-caution in software intended to be portable to other systems.
-Be warned also that they are considered experimental and changes are possible.
-.PP
-.I Regfree
-frees any dynamically-allocated storage associated with the compiled RE
-pointed to by
-.IR preg .
-The remaining
-.I regex_t
-is no longer a valid compiled RE
-and the effect of supplying it to
-.I regexec
-or
-.I regerror
-is undefined.
-.PP
-None of these functions references global variables except for tables
-of constants;
-all are safe for use from multiple threads if the arguments are safe.
-.SH IMPLEMENTATION CHOICES
-There are a number of decisions that 1003.2 leaves up to the implementor,
-either by explicitly saying ``undefined'' or by virtue of them being
-forbidden by the RE grammar.
-This implementation treats them as follows.
-.PP
-See
-.ZR
-for a discussion of the definition of case-independent matching.
-.PP
-There is no particular limit on the length of REs,
-except insofar as memory is limited.
-Memory usage is approximately linear in RE size, and largely insensitive
-to RE complexity, except for bounded repetitions.
-See BUGS for one short RE using them
-that will run almost any system out of memory.
-.PP
-A backslashed character other than one specifically given a magic meaning
-by 1003.2 (such magic meanings occur only in obsolete [``basic''] REs)
-is taken as an ordinary character.
-.PP
-Any unmatched [ is a REG_EBRACK error.
-.PP
-Equivalence classes cannot begin or end bracket-expression ranges.
-The endpoint of one range cannot begin another.
-.PP
-RE_DUP_MAX, the limit on repetition counts in bounded repetitions, is 255.
-.PP
-A repetition operator (?, *, +, or bounds) cannot follow another
-repetition operator.
-A repetition operator cannot begin an expression or subexpression
-or follow `^' or `|'.
-.PP
-`|' cannot appear first or last in a (sub)expression or after another `|',
-i.e. an operand of `|' cannot be an empty subexpression.
-An empty parenthesized subexpression, `()', is legal and matches an
-empty (sub)string.
-An empty string is not a legal RE.
-.PP
-A `{' followed by a digit is considered the beginning of bounds for a
-bounded repetition, which must then follow the syntax for bounds.
-A `{' \fInot\fR followed by a digit is considered an ordinary character.
-.PP
-`^' and `$' beginning and ending subexpressions in obsolete (``basic'')
-REs are anchors, not ordinary characters.
-.SH SEE ALSO
-grep(1), re_format(7)
-.PP
-POSIX 1003.2, sections 2.8 (Regular Expression Notation)
-and
-B.5 (C Binding for Regular Expression Matching).
-.SH DIAGNOSTICS
-Non-zero error codes from
-.I regcomp
-and
-.I regexec
-include the following:
-.PP
-.nf
-.ta \w'REG_ECOLLATE'u+3n
-REG_NOMATCH	no pattern match found
-REG_BADPAT	invalid regex struct
-REG_ECOLLATE	invalid collating element
-REG_ECTYPE	invalid character class
-REG_EESCAPE	trailing backslash (\e)
-REG_ESUBREG	invalid backreference number
-REG_EBRACK	brackets [ ] not balanced
-REG_EPAREN	parentheses ( ) not balanced
-REG_EBRACE	braces { } not balanced
-REG_BADBR	invalid repetition count(s) in { }
-REG_ERANGE	invalid character range in [ ]
-REG_ESPACE	ran out of memory
-REG_BADRPT	?, *, or + operand invalid
-REG_EMPTY	empty expression or subexpression
-REG_ASSERT	``can't happen''\(emyou found a bug
-REG_INVARG	invalid argument, e.g. negative-length string
-.fi
-.SH HISTORY
-Originally written by Henry Spencer.
-Altered for inclusion in the 4.4BSD distribution.
-.SH BUGS
-This is an alpha release with known defects.
-Please report problems.
-.PP
-There is one known functionality bug.
-The implementation of internationalization is incomplete:
-the locale is always assumed to be the default one of 1003.2,
-and only the collating elements etc. of that locale are available.
-.PP
-The back-reference code is subtle and doubts linger about its correctness
-in complex cases.
-.PP
-.I Regexec
-performance is poor.
-This will improve with later releases.
-.I Nmatch
-exceeding 0 is expensive;
-.I nmatch
-exceeding 1 is worse.
-.I Regexec
-is largely insensitive to RE complexity \fIexcept\fR that back
-references are massively expensive.
-RE length does matter; in particular, there is a strong speed bonus
-for keeping RE length under about 30 characters,
-with most special characters counting roughly double.
-.PP
-.I Regcomp
-implements bounded repetitions by macro expansion,
-which is costly in time and space if counts are large
-or bounded repetitions are nested.
-An RE like, say,
-`((((a{1,100}){1,100}){1,100}){1,100}){1,100}'
-will (eventually) run almost any existing machine out of swap space.
-.PP
-There are suspected problems with response to obscure error conditions.
-Notably,
-certain kinds of internal overflow,
-produced only by truly enormous REs or by multiply nested bounded repetitions,
-are probably not handled well.
-.PP
-Due to a mistake in 1003.2, things like `a)b' are legal REs because `)' is
-a special character only in the presence of a previous unmatched `('.
-This can't be fixed until the spec is fixed.
-.PP
-The standard's definition of back references is vague.
-For example, does
-`a\e(\e(b\e)*\e2\e)*d' match `abbbd'?
-Until the standard is clarified,
-behavior in such cases should not be relied on.
-.PP
-The implementation of word-boundary matching is a bit of a kludge,
-and bugs may lurk in combinations of word-boundary matching and anchoring.

src/backend/regex/regfree.c

@@ -1,75 +1,54 @@
-/*-
+/*
- * Copyright (c) 1992, 1993, 1994 Henry Spencer.
+ * regfree - free an RE
- * 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
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
- * Henry Spencer.
+ * 
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them. 
+ * 
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ * 
+ * I'd appreciate being given credit for this package in the documentation
+ * of software which uses it, but that is not a requirement.
+ * 
+ * THIS SOFTWARE IS PROVIDED ``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
+ * HENRY SPENCER 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.
  *
- * Redistribution and use in source and binary forms, with or without
+ * $Header: /cvsroot/pgsql/src/backend/regex/regfree.c,v 1.16 2003/02/05 17:41:33 tgl Exp $
- * 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. All advertising materials mentioning features or use of this software
- *	  must display the following acknowledgement:
- *		This product includes software developed by the University of
- *		California, Berkeley and its contributors.
- * 4. Neither the name of the University nor the names of its contributors
- *	  may be used to endorse or promote products derived from this software
- *	  without specific prior written permission.
  *
- * 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
+ * You might think that this could be incorporated into regcomp.c, and
+ * that would be a reasonable idea... except that this is a generic
+ * function (with a generic name), applicable to all compiled REs
+ * regardless of the size of their characters, whereas the stuff in
+ * regcomp.c gets compiled once per character size.
  */
 
-#include "postgres.h"
+#include "regex/regguts.h"
-
 
-#include "regex/regex.h"
-#include "regex/utils.h"
-#include "regex/regex2.h"
 
 /*
- * regfree - free everything
+ * pg_regfree - free an RE (generic function, punts to RE-specific function)
+ *
+ * Ignoring invocation with NULL is a convenience.
  */
 void
-pg_regfree(regex_t *preg)
+pg_regfree(regex_t *re)
 {
-	struct re_guts *g;
+	if (re == NULL)
-
-	if (preg->re_magic != MAGIC1)		/* oops */
-		return;					/* nice to complain, but hard */
-
-	g = preg->re_g;
-	if (g == NULL || g->magic != MAGIC2)		/* oops again */
 		return;
-	preg->re_magic = 0;			/* mark it invalid */
+	(*((struct fns *)re->re_fns)->free)(re);
-	g->magic = 0;				/* mark it invalid */
-
-	if (preg->patsave != NULL)
-		free((char *) preg->patsave);
-	if (g->strip != NULL)
-		free((char *) g->strip);
-	if (g->sets != NULL)
-		free((char *) g->sets);
-	if (g->setbits != NULL)
-		free((char *) g->setbits);
-	if (g->must != NULL)
-		free(g->must);
-	free((char *) g);
 }

src/backend/regex/retest.c

@@ -1,44 +0,0 @@
-/*
- * a simple regexp debug program
- *
- * $Header: /cvsroot/pgsql/src/backend/regex/Attic/retest.c,v 1.5 2002/06/11 15:41:37 thomas Exp $
- */
-
-#include "postgres.h"
-#include "regex/regex.h"
-
-int
-main()
-{
-	int			sts;
-	regex_t		re;
-	char		buf[1024];
-	char	   *p;
-
-	printf("type in regexp string: ");
-	if (!fgets(buf, sizeof(buf), stdin))
-		exit(0);
-	p = strchr(buf, '\n');
-	if (p)
-		*p = '\0';
-
-	sts = pg_regcomp(&re, buf, 1);
-	printf("regcomp: parses \"%s\" and returns %d\n", buf, sts);
-	for (;;)
-	{
-		printf("type in target string: ");
-		if (!fgets(buf, sizeof(buf), stdin))
-			exit(0);
-		p = strchr(buf, '\n');
-		if (p)
-			*p = '\0';
-
-		sts = pg_regexec(&re, buf, 0, 0, 0);
-		printf("regexec: returns %d\n", sts);
-	}
-}
-
-void
-elog(int lev, const char *fmt,...)
-{
-}

src/backend/utils/adt/regexp.c

@@ -1,14 +1,14 @@
 /*-------------------------------------------------------------------------
  *
  * regexp.c
- *	  regular expression handling code.
+ *	  Postgres' interface to the regular expression package.
  *
  * Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
  * Portions Copyright (c) 1994, Regents of the University of California
  *
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/utils/adt/regexp.c,v 1.43 2002/09/22 17:27:23 tgl Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/utils/adt/regexp.c,v 1.44 2003/02/05 17:41:32 tgl Exp $
  *
  *		Alistair Crooks added the code for the regex caching
  *		agc - cached the regular expressions used - there's a good chance
@@ -30,171 +30,189 @@
 #include "postgres.h"
 
 #include "regex/regex.h"
+#include "mb/pg_wchar.h"
 #include "utils/builtins.h"
 
-#if defined(DISABLE_XOPEN_NLS)
-#undef _XOPEN_SOURCE
-#endif   /* DISABLE_XOPEN_NLS */
 
-/* this is the number of cached regular expressions held. */
+/*
+ * We cache precompiled regular expressions using a "self organizing list"
+ * structure, in which recently-used items tend to be near the front.
+ * Whenever we use an entry, it's moved up to the front of the list.
+ * Over time, an item's average position corresponds to its frequency of use.
+ *
+ * When we first create an entry, it's inserted at the front of
+ * the array, dropping the entry at the end of the array if necessary to
+ * make room.  (This might seem to be weighting the new entry too heavily,
+ * but if we insert new entries further back, we'll be unable to adjust to
+ * a sudden shift in the query mix where we are presented with MAX_CACHED_RES
+ * never-before-seen items used circularly.  We ought to be able to handle
+ * that case, so we have to insert at the front.)
+ *
+ * Knuth mentions a variant strategy in which a used item is moved up just
+ * one place in the list.  Although he says this uses fewer comparisons on
+ * average, it seems not to adapt very well to the situation where you have
+ * both some reusable patterns and a steady stream of non-reusable patterns.
+ * A reusable pattern that isn't used at least as often as non-reusable
+ * patterns are seen will "fail to keep up" and will drop off the end of the
+ * cache.  With move-to-front, a reusable pattern is guaranteed to stay in
+ * the cache as long as it's used at least once in every MAX_CACHED_RES uses.
+ */
+
+/* this is the maximum number of cached regular expressions */
 #ifndef MAX_CACHED_RES
 #define MAX_CACHED_RES	32
 #endif
 
-/* this structure describes a cached regular expression */
+/* this structure describes one cached regular expression */
-struct cached_re_str
+typedef struct cached_re_str
 {
-	char	   *cre_s;			/* pattern as null-terminated string */
+	text	   *cre_pat;		/* original RE (untoasted TEXT form) */
-	int			cre_type;		/* compiled-type: extended,icase etc */
+	int			cre_flags;		/* compile flags: extended,icase etc */
 	regex_t		cre_re;			/* the compiled regular expression */
-	unsigned long cre_lru;		/* lru tag */
+} cached_re_str;
-};
 
-static int	rec = 0;			/* # of cached re's */
+static int	num_res = 0;		/* # of cached re's */
-static struct cached_re_str rev[MAX_CACHED_RES];		/* cached re's */
+static cached_re_str re_array[MAX_CACHED_RES]; /* cached re's */
-static unsigned long lru;		/* system lru tag */
-static int	pg_lastrec = 0;
 
-/* attempt to compile `re' as an re, then match it against text */
+
-/* cflags - flag to regcomp indicates case sensitivity */
+/*
+ * RE_compile_and_execute - compile and execute a RE, caching if possible
+ *
+ * Returns TRUE on match, FALSE on no match
+ *
+ *	text_re --- the pattern, expressed as an *untoasted* TEXT object
+ *	dat --- the data to match against (need not be null-terminated)
+ *	dat_len --- the length of the data string
+ *	cflags --- compile options for the pattern
+ *	nmatch, pmatch  --- optional return area for match details
+ *
+ * Both pattern and data are given in the database encoding.  We internally
+ * convert to array of pg_wchar which is what Spencer's regex package wants.
+ */
 static bool
-RE_compile_and_execute(text *text_re, char *text, int cflags,
+RE_compile_and_execute(text *text_re, unsigned char *dat, int dat_len,
-					   int nmatch, regmatch_t *pmatch)
+					   int cflags, int nmatch, regmatch_t *pmatch)
 {
-	char	   *re;
+	int			text_re_len = VARSIZE(text_re);
-	int			oldest;
+	pg_wchar   *data;
+	size_t		data_len;
+	pg_wchar   *pattern;
+	size_t		pattern_len;
 	int			i;
 	int			regcomp_result;
+	int			regexec_result;
+	cached_re_str	re_temp;
 
-	/* Convert 'text' pattern to null-terminated string */
+	/* Convert data string to wide characters */
-	re = DatumGetCString(DirectFunctionCall1(textout,
+	data = (pg_wchar *) palloc((dat_len + 1) * sizeof(pg_wchar));
-											 PointerGetDatum(text_re)));
+	data_len = pg_mb2wchar_with_len(dat, data, dat_len);
 
 	/*
-	 * Find a previously compiled regular expression. Run the cache as a
+	 * Look for a match among previously compiled REs.  Since the data
-	 * ring buffer, starting the search from the previous match if any.
+	 * structure is self-organizing with most-used entries at the front,
+	 * our search strategy can just be to scan from the front.
 	 */
-	i = pg_lastrec;
+	for (i = 0; i < num_res; i++)
-	while (i < rec)
 	{
-		if (rev[i].cre_s != NULL)
+		if (memcmp(re_array[i].cre_pat, text_re, text_re_len) == 0 &&
+			re_array[i].cre_flags == cflags)
 		{
-			if (strcmp(rev[i].cre_s, re) == 0 &&
+			/*
-				rev[i].cre_type == cflags)
+			 * Found a match; move it to front if not there already.
+			 */
+			if (i > 0)
 			{
-				pg_lastrec = i;
+				re_temp = re_array[i];
-				rev[i].cre_lru = ++lru;
+				memmove(&re_array[1], &re_array[0], i * sizeof(cached_re_str));
-				pfree(re);
+				re_array[0] = re_temp;
-				return (pg_regexec(&rev[i].cre_re,
-								   text, nmatch,
-								   pmatch, 0) == 0);
 			}
-		}
-		i++;
 
-		/*
+			/* Perform RE match and return result */
-		 * If we were not at the first slot to start, then think about
+			regexec_result = pg_regexec(&re_array[0].cre_re,
-		 * wrapping if necessary.
+										data,
-		 */
+										data_len,
-		if (pg_lastrec != 0)
+										NULL, /* no details */
-		{
+										nmatch,
-			if (i >= rec)
+										pmatch,
-				i = 0;
+										0);
-			else if (i == pg_lastrec)
-				break;
-		}
-	}
 
-	/* we didn't find it - make room in the cache for it */
+			pfree(data);
-	if (rec >= MAX_CACHED_RES)
-	{
-		/* cache is full - find the oldest entry */
-		for (oldest = 0, i = 1; i < rec; i++)
-		{
-			if (rev[i].cre_lru < rev[oldest].cre_lru)
-				oldest = i;
-		}
-	}
-	else
-		oldest = rec++;
 
-	/* if there was an old re, then de-allocate the space it used */
+			return (regexec_result == 0);
-	if (rev[oldest].cre_s != (char *) NULL)
-	{
-		for (lru = i = 0; i < rec; i++)
-		{
-			/* downweight all of the other cached entries */
-			rev[i].cre_lru = (rev[i].cre_lru - rev[oldest].cre_lru) / 2;
-			if (rev[i].cre_lru > lru)
-				lru = rev[i].cre_lru;
 		}
-		pg_regfree(&rev[oldest].cre_re);
-
-		/*
-		 * use malloc/free for the cre_s field because the storage has to
-		 * persist across transactions
-		 */
-		free(rev[oldest].cre_s);
-		rev[oldest].cre_s = (char *) NULL;
 	}
 
-	/* compile the re */
+	/*
-	regcomp_result = pg_regcomp(&rev[oldest].cre_re, re, cflags);
+	 * Couldn't find it, so try to compile the new RE.  To avoid leaking
-	if (regcomp_result == 0)
+	 * resources on failure, we build into the re_temp local.
-	{
+	 */
-		pg_lastrec = oldest;
+
-
+	/* Convert pattern string to wide characters */
-		/*
+	pattern = (pg_wchar *) palloc((text_re_len - VARHDRSZ + 1) * sizeof(pg_wchar));
-		 * use malloc/free for the cre_s field because the storage has to
+	pattern_len = pg_mb2wchar_with_len((unsigned char *) VARDATA(text_re),
-		 * persist across transactions
+									   pattern,
-		 */
+									   text_re_len - VARHDRSZ);
-		rev[oldest].cre_s = strdup(re);
-		rev[oldest].cre_lru = ++lru;
-		rev[oldest].cre_type = cflags;
-		pfree(re);
-		/* agc - fixed an old typo here */
-		return (pg_regexec(&rev[oldest].cre_re, text,
-						   nmatch, pmatch, 0) == 0);
-	}
-	else
-	{
-		char		errMsg[1000];
 
+	regcomp_result = pg_regcomp(&re_temp.cre_re,
+								pattern,
+								pattern_len,
+								cflags);
+
+	pfree(pattern);
+
+	if (regcomp_result != 0)
+	{
 		/* re didn't compile */
-		pg_regerror(regcomp_result, &rev[oldest].cre_re, errMsg,
+		char		errMsg[100];
-					sizeof(errMsg));
+
+		pg_regerror(regcomp_result, &re_temp.cre_re, errMsg, sizeof(errMsg));
+		/* XXX should we pg_regfree here? */
 		elog(ERROR, "Invalid regular expression: %s", errMsg);
 	}
 
-	/* not reached */
+	/*
-	return false;
+	 * use malloc/free for the cre_pat field because the storage has to
-}
+	 * persist across transactions
-
+	 */
+	re_temp.cre_pat = malloc(text_re_len);
+	if (re_temp.cre_pat == NULL)
+	{
+		pg_regfree(&re_temp.cre_re);
+		elog(ERROR, "Out of memory");
+	}
+	memcpy(re_temp.cre_pat, text_re, text_re_len);
+	re_temp.cre_flags = cflags;
 
-/*
+	/*
-   fixedlen_regexeq:
+	 * Okay, we have a valid new item in re_temp; insert it into the
+	 * storage array.  Discard last entry if needed.
+	 */
+	if (num_res >= MAX_CACHED_RES)
+	{
+		--num_res;
+		Assert(num_res < MAX_CACHED_RES);
+		pg_regfree(&re_array[num_res].cre_re);
+		free(re_array[num_res].cre_pat);
+	}
 
-   a generic fixed length regexp routine
+	if (num_res > 0)
-		 s		- the string to match against (not necessarily null-terminated)
+		memmove(&re_array[1], &re_array[0], num_res * sizeof(cached_re_str));
-		 p		- the pattern (as a text*)
-		 charlen   - the length of the string
-*/
-static bool
-fixedlen_regexeq(char *s, text *p, int charlen, int cflags)
-{
-	char	   *sterm;
-	bool		result;
 
-	/* be sure sterm is null-terminated */
+	re_array[0] = re_temp;
-	sterm = (char *) palloc(charlen + 1);
+	num_res++;
-	memcpy(sterm, s, charlen);
-	sterm[charlen] = '\0';
 
-	result = RE_compile_and_execute(p, sterm, cflags, 0, NULL);
+	/* Perform RE match and return result */
+	regexec_result = pg_regexec(&re_array[0].cre_re,
+								data,
+								data_len,
+								NULL, /* no details */
+								nmatch,
+								pmatch,
+								0);
 
-	pfree(sterm);
+	pfree(data);
 
-	return result;
+	return (regexec_result == 0);
 }
 
 
@@ -208,10 +226,11 @@ nameregexeq(PG_FUNCTION_ARGS)
 	Name		n = PG_GETARG_NAME(0);
 	text	   *p = PG_GETARG_TEXT_P(1);
 
-	PG_RETURN_BOOL(fixedlen_regexeq(NameStr(*n),
+	PG_RETURN_BOOL(RE_compile_and_execute(p,
-									p,
+										  (unsigned char *) NameStr(*n),
-									strlen(NameStr(*n)),
+										  strlen(NameStr(*n)),
-									REG_EXTENDED));
+										  REG_ADVANCED,
+										  0, NULL));
 }
 
 Datum
@@ -220,10 +239,11 @@ nameregexne(PG_FUNCTION_ARGS)
 	Name		n = PG_GETARG_NAME(0);
 	text	   *p = PG_GETARG_TEXT_P(1);
 
-	PG_RETURN_BOOL(!fixedlen_regexeq(NameStr(*n),
+	PG_RETURN_BOOL(!RE_compile_and_execute(p,
-									 p,
+										   (unsigned char *) NameStr(*n),
-									 strlen(NameStr(*n)),
+										   strlen(NameStr(*n)),
-									 REG_EXTENDED));
+										   REG_ADVANCED,
+										   0, NULL));
 }
 
 Datum
@@ -232,10 +252,11 @@ textregexeq(PG_FUNCTION_ARGS)
 	text	   *s = PG_GETARG_TEXT_P(0);
 	text	   *p = PG_GETARG_TEXT_P(1);
 
-	PG_RETURN_BOOL(fixedlen_regexeq(VARDATA(s),
+	PG_RETURN_BOOL(RE_compile_and_execute(p,
-									p,
+										  (unsigned char *) VARDATA(s),
-									VARSIZE(s) - VARHDRSZ,
+										  VARSIZE(s) - VARHDRSZ,
-									REG_EXTENDED));
+										  REG_ADVANCED,
+										  0, NULL));
 }
 
 Datum
@@ -244,10 +265,11 @@ textregexne(PG_FUNCTION_ARGS)
 	text	   *s = PG_GETARG_TEXT_P(0);
 	text	   *p = PG_GETARG_TEXT_P(1);
 
-	PG_RETURN_BOOL(!fixedlen_regexeq(VARDATA(s),
+	PG_RETURN_BOOL(!RE_compile_and_execute(p,
-									 p,
+										   (unsigned char *) VARDATA(s),
-									 VARSIZE(s) - VARHDRSZ,
+										   VARSIZE(s) - VARHDRSZ,
-									 REG_EXTENDED));
+										   REG_ADVANCED,
+										   0, NULL));
 }
 
 
@@ -258,82 +280,81 @@ textregexne(PG_FUNCTION_ARGS)
 
 
 Datum
-texticregexeq(PG_FUNCTION_ARGS)
+nameicregexeq(PG_FUNCTION_ARGS)
 {
-	text	   *s = PG_GETARG_TEXT_P(0);
+	Name		n = PG_GETARG_NAME(0);
 	text	   *p = PG_GETARG_TEXT_P(1);
 
-	PG_RETURN_BOOL(fixedlen_regexeq(VARDATA(s),
+	PG_RETURN_BOOL(RE_compile_and_execute(p,
-									p,
+										  (unsigned char *) NameStr(*n),
-									VARSIZE(s) - VARHDRSZ,
+										  strlen(NameStr(*n)),
-									REG_ICASE | REG_EXTENDED));
+										  REG_ICASE | REG_ADVANCED,
+										  0, NULL));
 }
 
 Datum
-texticregexne(PG_FUNCTION_ARGS)
+nameicregexne(PG_FUNCTION_ARGS)
 {
-	text	   *s = PG_GETARG_TEXT_P(0);
+	Name		n = PG_GETARG_NAME(0);
 	text	   *p = PG_GETARG_TEXT_P(1);
 
-	PG_RETURN_BOOL(!fixedlen_regexeq(VARDATA(s),
+	PG_RETURN_BOOL(!RE_compile_and_execute(p,
-									 p,
+										   (unsigned char *) NameStr(*n),
-									 VARSIZE(s) - VARHDRSZ,
+										   strlen(NameStr(*n)),
-									 REG_ICASE | REG_EXTENDED));
+										   REG_ICASE | REG_ADVANCED,
+										   0, NULL));
 }
 
 Datum
-nameicregexeq(PG_FUNCTION_ARGS)
+texticregexeq(PG_FUNCTION_ARGS)
 {
-	Name		n = PG_GETARG_NAME(0);
+	text	   *s = PG_GETARG_TEXT_P(0);
 	text	   *p = PG_GETARG_TEXT_P(1);
 
-	PG_RETURN_BOOL(fixedlen_regexeq(NameStr(*n),
+	PG_RETURN_BOOL(RE_compile_and_execute(p,
-									p,
+										  (unsigned char *) VARDATA(s),
-									strlen(NameStr(*n)),
+										  VARSIZE(s) - VARHDRSZ,
-									REG_ICASE | REG_EXTENDED));
+										  REG_ICASE | REG_ADVANCED,
+										  0, NULL));
 }
 
 Datum
-nameicregexne(PG_FUNCTION_ARGS)
+texticregexne(PG_FUNCTION_ARGS)
 {
-	Name		n = PG_GETARG_NAME(0);
+	text	   *s = PG_GETARG_TEXT_P(0);
 	text	   *p = PG_GETARG_TEXT_P(1);
 
-	PG_RETURN_BOOL(!fixedlen_regexeq(NameStr(*n),
+	PG_RETURN_BOOL(!RE_compile_and_execute(p,
-									 p,
+										   (unsigned char *) VARDATA(s),
-									 strlen(NameStr(*n)),
+										   VARSIZE(s) - VARHDRSZ,
-									 REG_ICASE | REG_EXTENDED));
+										   REG_ICASE | REG_ADVANCED,
+										   0, NULL));
 }
 
 
-/* textregexsubstr()
+/*
- * Return a substring matched by a regular expression.
+ * textregexsubstr()
+ *		Return a substring matched by a regular expression.
  */
 Datum
 textregexsubstr(PG_FUNCTION_ARGS)
 {
 	text	   *s = PG_GETARG_TEXT_P(0);
 	text	   *p = PG_GETARG_TEXT_P(1);
-	char	   *sterm;
-	int			len;
 	bool		match;
 	regmatch_t	pmatch[2];
 
-	/* be sure sterm is null-terminated */
-	len = VARSIZE(s) - VARHDRSZ;
-	sterm = (char *) palloc(len + 1);
-	memcpy(sterm, VARDATA(s), len);
-	sterm[len] = '\0';
-
 	/*
 	 * We pass two regmatch_t structs to get info about the overall match
 	 * and the match for the first parenthesized subexpression (if any).
 	 * If there is a parenthesized subexpression, we return what it matched;
 	 * else return what the whole regexp matched.
 	 */
-	match = RE_compile_and_execute(p, sterm, REG_EXTENDED, 2, pmatch);
+	match = RE_compile_and_execute(p,
-
+								   (unsigned char *) VARDATA(s),
-	pfree(sterm);
+								   VARSIZE(s) - VARHDRSZ,
+								   REG_ADVANCED,
+								   2, pmatch);
 
 	/* match? then return the substring matching the pattern */
 	if (match)

src/include/regex/cclass.h

@@ -1,99 +0,0 @@
-/*-
- * 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. All advertising materials mentioning features or use of this software
- *	  must display the following acknowledgement:
- *		This product includes software developed by the University of
- *		California, Berkeley and its contributors.
- * 4. Neither the name of the University nor the names of its contributors
- *	  may be used to endorse or promote products derived from this software
- *	  without specific prior written permission.
- *
- * 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
-{
-	char	   *name;
-	char	   *chars;
-	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, NULL, ""
-	}
-};

src/include/regex/cname.h

@@ -1,336 +0,0 @@
-/*-
- * 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. All advertising materials mentioning features or use of this software
- *	  must display the following acknowledgement:
- *		This product includes software developed by the University of
- *		California, Berkeley and its contributors.
- * 4. Neither the name of the University nor the names of its contributors
- *	  may be used to endorse or promote products derived from this software
- *	  without specific prior written permission.
- *
- * 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
-{
-	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
-	}
-};

src/include/regex/regcustom.h

@@ -0,0 +1,64 @@
+/*
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ * 
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them. 
+ * 
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ * 
+ * I'd appreciate being given credit for this package in the documentation
+ * of software which uses it, but that is not a requirement.
+ * 
+ * THIS SOFTWARE IS PROVIDED ``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
+ * HENRY SPENCER 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.
+ *
+ * $Id: regcustom.h,v 1.1 2003/02/05 17:41:32 tgl Exp $
+ */
+
+/* headers if any */
+#include "postgres.h"
+
+#include <ctype.h>
+#include <limits.h>
+
+#include "mb/pg_wchar.h"
+
+
+/* overrides for regguts.h definitions, if any */
+#define	FUNCPTR(name, args)	(*name) args
+#define	MALLOC(n)		malloc(n)
+#define	FREE(p)			free(VS(p))
+#define	REALLOC(p,n)		realloc(VS(p),n)
+
+/* internal character type and related */
+typedef pg_wchar chr;	/* the type itself */
+typedef unsigned uchr;		/* unsigned type that will hold a chr */
+typedef int celt;		/* type to hold chr, MCCE number, or NOCELT */
+#define	NOCELT	(-1)		/* celt value which is not valid chr or MCCE */
+#define	CHR(c)	((unsigned char) (c)) /* turn char literal into chr literal */
+#define	DIGITVAL(c)	((c)-'0')	/* turn chr digit into its value */
+#define	CHRBITS	32		/* bits in a chr; must not use sizeof */
+#define	CHR_MIN	0x00000000		/* smallest and largest chr; the value */
+#define	CHR_MAX	0xfffffffe		/*  CHR_MAX-CHR_MIN+1 should fit in uchr */
+
+/* functions operating on chr */
+#define	iscalnum(x)	pg_isalnum(x)
+#define	iscalpha(x)	pg_isalpha(x)
+#define	iscdigit(x)	pg_isdigit(x)
+#define	iscspace(x)	pg_isspace(x)
+
+/* and pick up the standard header */
+#include "regex.h"

src/include/regex/regerrs.h

@@ -0,0 +1,22 @@
+/*
+ * $Id: regerrs.h,v 1.1 2003/02/05 17:41:32 tgl Exp $
+ */
+
+{ REG_OKAY,	"REG_OKAY",	"no errors detected" },
+{ REG_NOMATCH,	"REG_NOMATCH",	"failed to match" },
+{ REG_BADPAT,	"REG_BADPAT",	"invalid regexp (reg version 0.8)" },
+{ REG_ECOLLATE,	"REG_ECOLLATE",	"invalid collating element" },
+{ REG_ECTYPE,	"REG_ECTYPE",	"invalid character class" },
+{ REG_EESCAPE,	"REG_EESCAPE",	"invalid escape \\ sequence" },
+{ REG_ESUBREG,	"REG_ESUBREG",	"invalid backreference number" },
+{ REG_EBRACK,	"REG_EBRACK",	"brackets [] not balanced" },
+{ REG_EPAREN,	"REG_EPAREN",	"parentheses () not balanced" },
+{ REG_EBRACE,	"REG_EBRACE",	"braces {} not balanced" },
+{ REG_BADBR,	"REG_BADBR",	"invalid repetition count(s)" },
+{ REG_ERANGE,	"REG_ERANGE",	"invalid character range" },
+{ REG_ESPACE,	"REG_ESPACE",	"out of memory" },
+{ REG_BADRPT,	"REG_BADRPT",	"quantifier operand invalid" },
+{ REG_ASSERT,	"REG_ASSERT",	"\"can't happen\" -- you found a bug" },
+{ REG_INVARG,	"REG_INVARG",	"invalid argument to regex function" },
+{ REG_MIXED,	"REG_MIXED",	"character widths of regex and string differ" },
+{ REG_BADOPT,	"REG_BADOPT",	"invalid embedded option" },

src/include/regex/regex.h

@@ -1,110 +1,165 @@
-/*-
+#ifndef _REGEX_H_
- * Copyright (c) 1992 Henry Spencer.
+#define	_REGEX_H_	/* never again */
- * Copyright (c) 1992, 1993
+/*
- *		The Regents of the University of California.  All rights reserved.
+ * regular expressions
- *
- * 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
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
- * modification, are permitted provided that the following conditions
+ * 
- * are met:
+ * Development of this software was funded, in part, by Cray Research Inc.,
- * 1. Redistributions of source code must retain the above copyright
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
- *	  notice, this list of conditions and the following disclaimer.
+ * Corporation, none of whom are responsible for the results.  The author
- * 2. Redistributions in binary form must reproduce the above copyright
+ * thanks all of them. 
- *	  notice, this list of conditions and the following disclaimer in the
+ * 
- *	  documentation and/or other materials provided with the distribution.
+ * Redistribution and use in source and binary forms -- with or without
- * 3. All advertising materials mentioning features or use of this software
+ * modification -- are permitted for any purpose, provided that
- *	  must display the following acknowledgement:
+ * redistributions in source form retain this entire copyright notice and
- *		This product includes software developed by the University of
+ * indicate the origin and nature of any modifications.
- *		California, Berkeley and its contributors.
- * 4. Neither the name of the University nor the names of its contributors
- *	  may be used to endorse or promote products derived from this software
- *	  without specific prior written permission.
  *
- * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * I'd appreciate being given credit for this package in the documentation
- * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * of software which uses it, but that is not a requirement.
- * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * 
- * ARE DISCLAIMED.	IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
- * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
- * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL
- * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HENRY SPENCER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
- * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
- * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
- * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
- * SUCH DAMAGE.
+ * 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
+ * $Id: regex.h,v 1.23 2003/02/05 17:41:32 tgl Exp $
  */
 
-#ifndef _REGEX_H_
+/*
-#define _REGEX_H_
+ * Add your own defines, if needed, here.
-
+ */
-#include <sys/types.h>
 #include "mb/pg_wchar.h"
 
-/* types */
+/*
-typedef off_t regoff_t;
+ * interface types etc.
+ */
 
-typedef struct
+/*
-{
+ * regoff_t has to be large enough to hold either off_t or ssize_t,
-	int			re_magic;
+ * and must be signed; it's only a guess that long is suitable.
-	size_t		re_nsub;		/* number of parenthesized subexpressions */
+ */
-	const pg_wchar *re_endp;	/* end pointer for REG_PEND */
+typedef long regoff_t;
-	struct re_guts *re_g;		/* none of your business :-) */
+
-	pg_wchar   *patsave;		/* me too :-) */
+/*
+ * other interface types
+ */
+
+/* the biggie, a compiled RE (or rather, a front end to same) */
+typedef struct {
+	int re_magic;		/* magic number */
+	size_t re_nsub;		/* number of subexpressions */
+	long re_info;		/* information about RE */
+#		define	REG_UBACKREF		000001
+#		define	REG_ULOOKAHEAD		000002
+#		define	REG_UBOUNDS		000004
+#		define	REG_UBRACES		000010
+#		define	REG_UBSALNUM		000020
+#		define	REG_UPBOTCH		000040
+#		define	REG_UBBS		000100
+#		define	REG_UNONPOSIX		000200
+#		define	REG_UUNSPEC		000400
+#		define	REG_UUNPORT		001000
+#		define	REG_ULOCALE		002000
+#		define	REG_UEMPTYMATCH		004000
+#		define	REG_UIMPOSSIBLE		010000
+#		define	REG_USHORTEST		020000
+	int re_csize;		/* sizeof(character) */
+	char *re_endp;		/* backward compatibility kludge */
+	/* the rest is opaque pointers to hidden innards */
+	char *re_guts;		/* `char *' is more portable than `void *' */
+	char *re_fns;
 } regex_t;
 
-typedef struct
+/* result reporting (may acquire more fields later) */
-{
+typedef struct {
-	regoff_t	rm_so;			/* start of match */
+	regoff_t rm_so;		/* start of substring */
-	regoff_t	rm_eo;			/* end of match */
+	regoff_t rm_eo;		/* end of substring */
 } regmatch_t;
 
-/* regcomp() flags */
+/* supplementary control and reporting */
-#define REG_BASIC		0000
+typedef struct {
-#define REG_EXTENDED	0001
+	regmatch_t rm_extend;	/* see REG_EXPECT */
-#define REG_ICASE		0002
+} rm_detail_t;
-#define REG_NOSUB		0004
+
-#define REG_NEWLINE		0010
+
-#define REG_NOSPEC		0020
+
-#define REG_PEND		0040
+/*
-#define REG_DUMP		0200
+ * regex compilation flags
-
+ */
-/* regerror() flags */
+#define	REG_BASIC	000000	/* BREs (convenience) */
-#define REG_NOMATCH		 1
+#define	REG_EXTENDED	000001	/* EREs */
-#define REG_BADPAT		 2
+#define	REG_ADVF	000002	/* advanced features in EREs */
-#define REG_ECOLLATE	 3
+#define	REG_ADVANCED	000003	/* AREs (which are also EREs) */
-#define REG_ECTYPE		 4
+#define	REG_QUOTE	000004	/* no special characters, none */
-#define REG_EESCAPE		 5
+#define	REG_NOSPEC	REG_QUOTE	/* historical synonym */
-#define REG_ESUBREG		 6
+#define	REG_ICASE	000010	/* ignore case */
-#define REG_EBRACK		 7
+#define	REG_NOSUB	000020	/* don't care about subexpressions */
-#define REG_EPAREN		 8
+#define	REG_EXPANDED	000040	/* expanded format, white space & comments */
-#define REG_EBRACE		 9
+#define	REG_NLSTOP	000100	/* \n doesn't match . or [^ ] */
-#define REG_BADBR		10
+#define	REG_NLANCH	000200	/* ^ matches after \n, $ before */
-#define REG_ERANGE		11
+#define	REG_NEWLINE	000300	/* newlines are line terminators */
-#define REG_ESPACE		12
+#define	REG_PEND	000400	/* ugh -- backward-compatibility hack */
-#define REG_BADRPT		13
+#define	REG_EXPECT	001000	/* report details on partial/limited matches */
-#define REG_EMPTY		14
+#define	REG_BOSONLY	002000	/* temporary kludge for BOS-only matches */
-#define REG_ASSERT		15
+#define	REG_DUMP	004000	/* none of your business :-) */
-#define REG_INVARG		16
+#define	REG_FAKE	010000	/* none of your business :-) */
-#define REG_ATOI		255		/* convert name to number (!) */
+#define	REG_PROGRESS	020000	/* none of your business :-) */
-#define REG_ITOA		0400	/* convert number to name (!) */
+
-
+
-/* regexec() flags */
+
-#define REG_NOTBOL		00001
+/*
-#define REG_NOTEOL		00002
+ * regex execution flags
-#define REG_STARTEND	00004
+ */
-#define REG_TRACE		00400	/* tracing of execution */
+#define	REG_NOTBOL	0001	/* BOS is not BOL */
-#define REG_LARGE		01000	/* force large representation */
+#define	REG_NOTEOL	0002	/* EOS is not EOL */
-#define REG_BACKR		02000	/* force use of backref code */
+#define	REG_STARTEND	0004	/* backward compatibility kludge */
-
+#define	REG_FTRACE	0010	/* none of your business */
-extern int	pg_regcomp(regex_t *preg, const char *pattern, int cflags);
+#define	REG_MTRACE	0020	/* none of your business */
-extern size_t pg_regerror(int errcode, const regex_t *preg,
+#define	REG_SMALL	0040	/* none of your business */
-			char *errbuf, size_t errbuf_size);
+
-extern int pg_regexec(const regex_t *preg, const char *string,
+
-		   size_t nmatch,
+/*
-		   regmatch_t *pmatch, int eflags);
+ * error reporting
-extern void pg_regfree(regex_t *preg);
+ * Be careful if modifying the list of error codes -- the table used by
-
+ * regerror() is generated automatically from this file!
-#endif   /* !_REGEX_H_ */
+ */
+#define	REG_OKAY	 0	/* no errors detected */
+#define	REG_NOMATCH	 1	/* failed to match */
+#define	REG_BADPAT	 2	/* invalid regexp */
+#define	REG_ECOLLATE	 3	/* invalid collating element */
+#define	REG_ECTYPE	 4	/* invalid character class */
+#define	REG_EESCAPE	 5	/* invalid escape \ sequence */
+#define	REG_ESUBREG	 6	/* invalid backreference number */
+#define	REG_EBRACK	 7	/* brackets [] not balanced */
+#define	REG_EPAREN	 8	/* parentheses () not balanced */
+#define	REG_EBRACE	 9	/* braces {} not balanced */
+#define	REG_BADBR	10	/* invalid repetition count(s) */
+#define	REG_ERANGE	11	/* invalid character range */
+#define	REG_ESPACE	12	/* out of memory */
+#define	REG_BADRPT	13	/* quantifier operand invalid */
+#define	REG_ASSERT	15	/* "can't happen" -- you found a bug */
+#define	REG_INVARG	16	/* invalid argument to regex function */
+#define	REG_MIXED	17	/* character widths of regex and string differ */
+#define	REG_BADOPT	18	/* invalid embedded option */
+/* two specials for debugging and testing */
+#define	REG_ATOI	101	/* convert error-code name to number */
+#define	REG_ITOA	102	/* convert error-code number to name */
+
+
+
+/*
+ * the prototypes for exported functions
+ */
+extern int pg_regcomp(regex_t *, const pg_wchar *, size_t, int);
+extern int pg_regexec(regex_t *, const pg_wchar *, size_t, rm_detail_t *, size_t, regmatch_t [], int);
+extern void pg_regfree(regex_t *);
+extern size_t pg_regerror(int, const regex_t *, char *, size_t);
+
+#endif /* _REGEX_H_ */

src/include/regex/regex2.h

@@ -1,174 +0,0 @@
-/*-
- * 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. All advertising materials mentioning features or use of this software
- *	  must display the following acknowledgement:
- *		This product includes software developed by the University of
- *		California, Berkeley and its contributors.
- * 4. Neither the name of the University nor the names of its contributors
- *	  may be used to endorse or promote products derived from this software
- *	  without specific prior written permission.
- *
- * 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
- */
-
-#include <limits.h>
-
-/*
- * internals of regex_t
- */
-#define MAGIC1	((('r'^0200)<<8) | 'e')
-
-/*
- * The internal representation is a *strip*, a sequence of
- * operators ending with an endmarker.	(Some terminology etc. is a
- * historical relic of earlier versions which used multiple strips.)
- * Certain oddities in the representation are there to permit running
- * the machinery backwards; in particular, any deviation from sequential
- * flow must be marked at both its source and its destination.	Some
- * fine points:
- *
- * - OPLUS_ and O_PLUS are *inside* the loop they create.
- * - OQUEST_ and O_QUEST are *outside* the bypass they create.
- * - OCH_ and O_CH are *outside* the multi-way branch they create, while
- *	 OOR1 and OOR2 are respectively the end and the beginning of one of
- *	 the branches.	Note that there is an implicit OOR2 following OCH_
- *	 and an implicit OOR1 preceding O_CH.
- *
- * In state representations, an operator's bit is on to signify a state
- * immediately *preceding* "execution" of that operator.
- */
-typedef unsigned long sop;		/* strip operator */
-typedef long sopno;
-
-#define OPRMASK ((sop) 0xf8000000)
-#define OPDMASK ((sop) 0x07ffffff)
-#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	((size_t)1<<OPSHIFT)	/* endmarker	-						*/
-#define OCHAR	((size_t)2<<OPSHIFT)	/* character	unsigned char			*/
-#define OBOL	((size_t)3<<OPSHIFT)	/* left anchor	-						*/
-#define OEOL	((size_t)4<<OPSHIFT)	/* right anchor -						*/
-#define OANY	((size_t)5<<OPSHIFT)	/* .			-						*/
-#define OANYOF	((size_t)6<<OPSHIFT)	/* [...]		set number				*/
-#define OBACK_	((size_t)7<<OPSHIFT)	/* begin \d		paren number			*/
-#define O_BACK	((size_t)8<<OPSHIFT)	/* end \d		paren number			*/
-#define OPLUS_	((size_t)9<<OPSHIFT)	/* + prefix		fwd to suffix			*/
-#define O_PLUS	((size_t)10<<OPSHIFT)	/* + suffix		back to prefix			*/
-#define OQUEST_ ((size_t)11<<OPSHIFT)	/* ? prefix		fwd to suffix			*/
-#define O_QUEST ((size_t)12<<OPSHIFT)	/* ? suffix		back to prefix			*/
-#define OLPAREN ((size_t)13<<OPSHIFT)	/* (			fwd to )				*/
-#define ORPAREN ((size_t)14<<OPSHIFT)	/* )			back to (				*/
-#define OCH_	((size_t)15<<OPSHIFT)	/* begin choice fwd to OOR2				*/
-#define OOR1	((size_t)16<<OPSHIFT)	/* | pt. 1		back to OOR1 or
-										 * OCH_    */
-#define OOR2	((size_t)17<<OPSHIFT)	/* | pt. 2		fwd to OOR2 or
-										 * O_CH		*/
-#define O_CH	((size_t)18<<OPSHIFT)	/* end choice	back to OOR1			*/
-#define OBOW	((size_t)19<<OPSHIFT)	/* begin word	-						*/
-#define OEOW	((size_t)20<<OPSHIFT)	/* end word		-						*/
-
-/*
- * Structure for [] character-set representation.  Character sets are
- * 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] */
-	uch			mask;			/* bit within array */
-	pg_wchar	hash;			/* hash code */
-	unsigned int lc;			/* leading character (character-set) */
-	size_t		smultis;
-	char	   *multis;			/* -> char[smulti]	ab\0cd\0ef\0\0 */
-} cset;
-
-/* note that CHadd and CHsub are unsafe, and CHIN doesn't yield 0/1 */
-#define CHlc(c) (((unsigned)(c)&0xff0000)>>16)
-#define CHadd(cs, c)	((cs)->ptr[(unsigned)(c)&0xffff] |= (cs)->mask, (cs)->hash += (unsigned)(c)&0xffff,\
-			 (cs)->lc = CHlc(c))
-#define CHsub(cs, c)	((cs)->ptr[(unsigned)(c)&0xffff] &= ~(cs)->mask, (cs)->hash -= (unsigned)(c)&0xffff)
-#define CHIN(cs, c) ((cs)->ptr[(unsigned)(c)&0xffff] & (cs)->mask && \
-			 ((cs)->lc == CHlc(c)))
-#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 */
-typedef unsigned char cat_t;
-
-/*
- * main compiled-expression structure
- */
-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) */
-	sopno		laststate;		/* the final OEND */
-	int			iflags;			/* internal flags */
-#define  USEBOL  01				/* used ^ */
-#define  USEEOL  02				/* used $ */
-#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] */
-	pg_wchar   *must;			/* match must contain this string */
-	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		  (16777216+1)	/* 16777216 == 2^24 == 3 bytes */
-
-#define ISWORD(c)	(((c) >= 0 && (c) <= UCHAR_MAX) && \
-			 (isalnum((unsigned char) (c)) || (c) == '_'))

src/include/regex/regguts.h

@@ -0,0 +1,393 @@
+/*
+ * Internal interface definitions, etc., for the reg package
+ *
+ * Copyright (c) 1998, 1999 Henry Spencer.  All rights reserved.
+ * 
+ * Development of this software was funded, in part, by Cray Research Inc.,
+ * UUNET Communications Services Inc., Sun Microsystems Inc., and Scriptics
+ * Corporation, none of whom are responsible for the results.  The author
+ * thanks all of them. 
+ * 
+ * Redistribution and use in source and binary forms -- with or without
+ * modification -- are permitted for any purpose, provided that
+ * redistributions in source form retain this entire copyright notice and
+ * indicate the origin and nature of any modifications.
+ * 
+ * I'd appreciate being given credit for this package in the documentation
+ * of software which uses it, but that is not a requirement.
+ * 
+ * THIS SOFTWARE IS PROVIDED ``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
+ * HENRY SPENCER 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.
+ *
+ * $Id: regguts.h,v 1.1 2003/02/05 17:41:32 tgl Exp $
+ */
+
+
+
+/*
+ * Environmental customization.  It should not (I hope) be necessary to
+ * alter the file you are now reading -- regcustom.h should handle it all,
+ * given care here and elsewhere.
+ */
+#include "regcustom.h"
+
+
+
+/*
+ * Things that regcustom.h might override.
+ */
+
+/* assertions */
+#ifndef assert
+#	ifndef REG_DEBUG
+#	define	NDEBUG		/* no assertions */
+#	endif
+#include <assert.h>
+#endif
+
+/* voids */
+#ifndef DISCARD
+#define	DISCARD	void			/* for throwing values away */
+#endif
+#ifndef VS
+#define	VS(x)	((void *)(x))		/* cast something to generic ptr */
+#endif
+
+/* function-pointer declarator */
+#ifndef FUNCPTR
+#define	FUNCPTR(name, args)	(*name) args
+#endif
+
+/* memory allocation */
+#ifndef MALLOC
+#define	MALLOC(n)	malloc(n)
+#endif
+#ifndef REALLOC
+#define	REALLOC(p, n)	realloc(VS(p), n)
+#endif
+#ifndef FREE
+#define	FREE(p)		free(VS(p))
+#endif
+
+/* want size of a char in bits, and max value in bounded quantifiers */
+#ifndef CHAR_BIT
+#include <limits.h>
+#endif
+#ifndef _POSIX2_RE_DUP_MAX
+#define	_POSIX2_RE_DUP_MAX	255	/* normally from <limits.h> */
+#endif
+
+
+
+/*
+ * misc
+ */
+
+#define	NOTREACHED	0
+#define	xxx		1
+
+#define	DUPMAX	_POSIX2_RE_DUP_MAX
+#define	INFINITY	(DUPMAX+1)
+
+#define	REMAGIC	0xfed7		/* magic number for main struct */
+
+
+
+/*
+ * debugging facilities
+ */
+#ifdef REG_DEBUG
+/* FDEBUG does finite-state tracing */
+#define	FDEBUG(arglist)	{ if (v->eflags&REG_FTRACE) printf arglist; }
+/* MDEBUG does higher-level tracing */
+#define	MDEBUG(arglist)	{ if (v->eflags&REG_MTRACE) printf arglist; }
+#else
+#define	FDEBUG(arglist)	{}
+#define	MDEBUG(arglist)	{}
+#endif
+
+
+
+/*
+ * bitmap manipulation
+ */
+#define	UBITS	(CHAR_BIT * sizeof(unsigned))
+#define	BSET(uv, sn)	((uv)[(sn)/UBITS] |= (unsigned)1 << ((sn)%UBITS))
+#define	ISBSET(uv, sn)	((uv)[(sn)/UBITS] & ((unsigned)1 << ((sn)%UBITS)))
+
+
+
+/*
+ * We dissect a chr into byts for colormap table indexing.  Here we define
+ * a byt, which will be the same as a byte on most machines...  The exact
+ * size of a byt is not critical, but about 8 bits is good, and extraction
+ * of 8-bit chunks is sometimes especially fast.
+ */
+#ifndef BYTBITS
+#define	BYTBITS	8		/* bits in a byt */
+#endif
+#define	BYTTAB	(1<<BYTBITS)	/* size of table with one entry per byt value */
+#define	BYTMASK	(BYTTAB-1)	/* bit mask for byt */
+#define	NBYTS	((CHRBITS+BYTBITS-1)/BYTBITS)
+/* the definition of GETCOLOR(), below, assumes NBYTS <= 4 */
+
+
+
+/*
+ * As soon as possible, we map chrs into equivalence classes -- "colors" --
+ * which are of much more manageable number.
+ */
+typedef short color;		/* colors of characters */
+typedef int pcolor;		/* what color promotes to */
+#define	COLORLESS	(-1)	/* impossible color */
+#define	WHITE		0	/* default color, parent of all others */
+
+
+
+/*
+ * A colormap is a tree -- more precisely, a DAG -- indexed at each level
+ * by a byt of the chr, to map the chr to a color efficiently.  Because
+ * lower sections of the tree can be shared, it can exploit the usual
+ * sparseness of such a mapping table.  The tree is always NBYTS levels
+ * deep (in the past it was shallower during construction but was "filled"
+ * to full depth at the end of that); areas that are unaltered as yet point
+ * to "fill blocks" which are entirely WHITE in color.
+ */
+
+/* the tree itself */
+struct colors {
+	color ccolor[BYTTAB];
+};
+struct ptrs {
+	union tree *pptr[BYTTAB];
+};
+union tree {
+	struct colors colors;
+	struct ptrs ptrs;
+};
+#define	tcolor	colors.ccolor
+#define	tptr	ptrs.pptr
+
+/* internal per-color structure for the color machinery */
+struct colordesc {
+	uchr nchrs;		/* number of chars of this color */
+	color sub;		/* open subcolor (if any); free chain ptr */
+#		define	NOSUB	COLORLESS
+	struct arc *arcs;	/* color chain */
+	int flags;
+#		define	FREECOL	01	/* currently free */
+#		define	PSEUDO	02	/* pseudocolor, no real chars */
+#	define	UNUSEDCOLOR(cd)	((cd)->flags&FREECOL)
+	union tree *block;	/* block of solid color, if any */
+};
+
+/* the color map itself */
+struct colormap {
+	int magic;
+#		define	CMMAGIC	0x876
+	struct vars *v;			/* for compile error reporting */
+	size_t ncds;			/* number of colordescs */
+	size_t max;			/* highest in use */
+	color free;			/* beginning of free chain (if non-0) */
+	struct colordesc *cd;
+#	define	CDEND(cm)	(&(cm)->cd[(cm)->max + 1])
+#		define	NINLINECDS	((size_t)10)
+	struct colordesc cdspace[NINLINECDS];
+	union tree tree[NBYTS];		/* tree top, plus fill blocks */
+};
+
+/* optimization magic to do fast chr->color mapping */
+#define	B0(c)	((c) & BYTMASK)
+#define	B1(c)	(((c)>>BYTBITS) & BYTMASK)
+#define	B2(c)	(((c)>>(2*BYTBITS)) & BYTMASK)
+#define	B3(c)	(((c)>>(3*BYTBITS)) & BYTMASK)
+#if NBYTS == 1
+#define	GETCOLOR(cm, c)	((cm)->tree->tcolor[B0(c)])
+#endif
+/* beware, for NBYTS>1, GETCOLOR() is unsafe -- 2nd arg used repeatedly */
+#if NBYTS == 2
+#define	GETCOLOR(cm, c)	((cm)->tree->tptr[B1(c)]->tcolor[B0(c)])
+#endif
+#if NBYTS == 4
+#define	GETCOLOR(cm, c)	((cm)->tree->tptr[B3(c)]->tptr[B2(c)]->tptr[B1(c)]->tcolor[B0(c)])
+#endif
+
+
+
+/*
+ * Interface definitions for locale-interface functions in locale.c.
+ * Multi-character collating elements (MCCEs) cause most of the trouble.
+ */
+struct cvec {
+	int nchrs;		/* number of chrs */
+	int chrspace;		/* number of chrs possible */
+	chr *chrs;		/* pointer to vector of chrs */
+	int nranges;		/* number of ranges (chr pairs) */
+	int rangespace;		/* number of chrs possible */
+	chr *ranges;		/* pointer to vector of chr pairs */
+	int nmcces;		/* number of MCCEs */
+	int mccespace;		/* number of MCCEs possible */
+	int nmccechrs;		/* number of chrs used for MCCEs */
+	chr *mcces[1];		/* pointers to 0-terminated MCCEs */
+				/* and both batches of chrs are on the end */
+};
+
+/* caution:  this value cannot be changed easily */
+#define	MAXMCCE	2		/* length of longest MCCE */
+
+
+
+/*
+ * definitions for NFA internal representation
+ *
+ * Having a "from" pointer within each arc may seem redundant, but it
+ * saves a lot of hassle.
+ */
+struct state;
+
+struct arc {
+	int type;
+#		define	ARCFREE	'\0'
+	color co;
+	struct state *from;	/* where it's from (and contained within) */
+	struct state *to;	/* where it's to */
+	struct arc *outchain;	/* *from's outs chain or free chain */
+#	define	freechain	outchain
+	struct arc *inchain;	/* *to's ins chain */
+	struct arc *colorchain;	/* color's arc chain */
+};
+
+struct arcbatch {		/* for bulk allocation of arcs */
+	struct arcbatch *next;
+#	define	ABSIZE	10
+	struct arc a[ABSIZE];
+};
+
+struct state {
+	int no;
+#		define	FREESTATE	(-1)
+	char flag;		/* marks special states */
+	int nins;		/* number of inarcs */
+	struct arc *ins;	/* chain of inarcs */
+	int nouts;		/* number of outarcs */
+	struct arc *outs;	/* chain of outarcs */
+	struct arc *free;	/* chain of free arcs */
+	struct state *tmp;	/* temporary for traversal algorithms */
+	struct state *next;	/* chain for traversing all */
+	struct state *prev;	/* back chain */
+	struct arcbatch oas;	/* first arcbatch, avoid malloc in easy case */
+	int noas;		/* number of arcs used in first arcbatch */
+};
+
+struct nfa {
+	struct state *pre;	/* pre-initial state */
+	struct state *init;	/* initial state */
+	struct state *final;	/* final state */
+	struct state *post;	/* post-final state */
+	int nstates;		/* for numbering states */
+	struct state *states;	/* state-chain header */
+	struct state *slast;	/* tail of the chain */
+	struct state *free;	/* free list */
+	struct colormap *cm;	/* the color map */
+	color bos[2];		/* colors, if any, assigned to BOS and BOL */
+	color eos[2];		/* colors, if any, assigned to EOS and EOL */
+	struct vars *v;		/* simplifies compile error reporting */
+	struct nfa *parent;	/* parent NFA, if any */
+};
+
+
+
+/*
+ * definitions for compacted NFA
+ */
+struct carc {
+	color co;		/* COLORLESS is list terminator */
+	int to;			/* state number */
+};
+
+struct cnfa {
+	int nstates;		/* number of states */
+	int ncolors;		/* number of colors */
+	int flags;
+#		define	HASLACONS	01	/* uses lookahead constraints */
+	int pre;		/* setup state number */
+	int post;		/* teardown state number */
+	color bos[2];		/* colors, if any, assigned to BOS and BOL */
+	color eos[2];		/* colors, if any, assigned to EOS and EOL */
+	struct carc **states;	/* vector of pointers to outarc lists */
+	struct carc *arcs;	/* the area for the lists */
+};
+#define	ZAPCNFA(cnfa)	((cnfa).nstates = 0)
+#define	NULLCNFA(cnfa)	((cnfa).nstates == 0)
+
+
+
+/*
+ * subexpression tree
+ */
+struct subre {
+	char op;		/* '|', '.' (concat), 'b' (backref), '(', '=' */
+	char flags;
+#		define	LONGER	01	/* prefers longer match */
+#		define	SHORTER	02	/* prefers shorter match */
+#		define	MIXED	04	/* mixed preference below */
+#		define	CAP	010	/* capturing parens below */
+#		define	BACKR	020	/* back reference below */
+#		define	INUSE	0100	/* in use in final tree */
+#		define	LOCAL	03	/* bits which may not propagate up */
+#		define	LMIX(f)	((f)<<2)	/* LONGER -> MIXED */
+#		define	SMIX(f)	((f)<<1)	/* SHORTER -> MIXED */
+#		define	UP(f)	(((f)&~LOCAL) | (LMIX(f) & SMIX(f) & MIXED))
+#		define	MESSY(f)	((f)&(MIXED|CAP|BACKR))
+#		define	PREF(f)	((f)&LOCAL)
+#		define	PREF2(f1, f2)	((PREF(f1) != 0) ? PREF(f1) : PREF(f2))
+#		define	COMBINE(f1, f2)	(UP((f1)|(f2)) | PREF2(f1, f2))
+	short retry;		/* index into retry memory */
+	int subno;		/* subexpression number (for 'b' and '(') */
+	short min;		/* min repetitions, for backref only */
+	short max;		/* max repetitions, for backref only */
+	struct subre *left;	/* left child, if any (also freelist chain) */
+	struct subre *right;	/* right child, if any */
+	struct state *begin;	/* outarcs from here... */
+	struct state *end;	/* ...ending in inarcs here */
+	struct cnfa cnfa;	/* compacted NFA, if any */
+	struct subre *chain;	/* for bookkeeping and error cleanup */
+};
+
+
+
+/*
+ * table of function pointers for generic manipulation functions
+ * A regex_t's re_fns points to one of these.
+ */
+struct fns {
+	void FUNCPTR(free, (regex_t *));
+};
+
+
+
+/*
+ * the insides of a regex_t, hidden behind a void *
+ */
+struct guts {
+	int magic;
+#		define	GUTSMAGIC	0xfed9
+	int cflags;		/* copy of compile flags */
+	long info;		/* copy of re_info */
+	size_t nsub;		/* copy of re_nsub */
+	struct subre *tree;
+	struct cnfa search;	/* for fast preliminary search */
+	int ntree;
+	struct colormap cmap;
+	int FUNCPTR(compare, (const chr *, const chr *, size_t));
+	struct subre *lacons;	/* lookahead-constraint vector */
+	int nlacons;		/* size of lacons */
+};

src/include/regex/utils.h

@@ -1,60 +0,0 @@
-/*-
- * 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. All advertising materials mentioning features or use of this software
- *	  must display the following acknowledgement:
- *		This product includes software developed by the University of
- *		California, Berkeley and its contributors.
- * 4. Neither the name of the University nor the names of its contributors
- *	  may be used to endorse or promote products derived from this software
- *	  without specific prior written permission.
- *
- * 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
- */
-
-#ifndef _REGEX_UTILS_H
-#define _REGEX_UTILS_H
-
-#include <limits.h>
-
-/* utility definitions */
-#define DUPMAX			100000000		/* xxx is this right? */
-#define INFINITY		(DUPMAX + 1)
-
-#define NC				(SHRT_MAX - SHRT_MIN + 1)
-
-typedef unsigned char uch;
-
-/* switch off assertions (if not already off) if no REDEBUG */
-#ifndef REDEBUG
-#ifndef NDEBUG
-#define NDEBUG					/* no assertions please */
-#endif
-#endif
-
-#endif   /* _REGEX_UTILS_H */