Add flex 2.5.37 from flex.sourceforge.net to contrib.

svn path=/head/; revision=250873

contrib/flex/COPYING

@@ -0,0 +1,42 @@
+Flex carries the copyright used for BSD software, slightly modified
+because it originated at the Lawrence Berkeley (not Livermore!) Laboratory,
+which operates under a contract with the Department of Energy:
+
+Copyright (c) 2001, 2002, 2003, 2004, 2005, 2006, 2007 The Flex Project.
+
+Copyright (c) 1990, 1997 The Regents of the University of California.
+All rights reserved.
+
+This code is derived from software contributed to Berkeley by
+Vern Paxson.
+
+The United States Government has rights in this work pursuant
+to contract no. DE-AC03-76SF00098 between the United States
+Department of Energy and the University of California.
+
+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.
+
+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 ``AS IS'' AND WITHOUT ANY EXPRESS OR
+IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE.
+
+This basically says "do whatever you please with this software except
+remove this notice or take advantage of the University's (or the flex
+authors') name".
+
+Note that the "flex.skl" scanner skeleton carries no copyright notice.
+You are free to do whatever you please with scanners generated using flex;
+for them, you are not even bound by the above copyright.

contrib/flex/FlexLexer.h

@@ -0,0 +1,206 @@
+// -*-C++-*-
+// FlexLexer.h -- define interfaces for lexical analyzer classes generated
+// by flex
+
+// Copyright (c) 1993 The Regents of the University of California.
+// All rights reserved.
+//
+// This code is derived from software contributed to Berkeley by
+// Kent Williams and Tom Epperly.
+//
+//  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.
+
+//  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR
+//  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+//  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+//  PURPOSE.
+
+// This file defines FlexLexer, an abstract class which specifies the
+// external interface provided to flex C++ lexer objects, and yyFlexLexer,
+// which defines a particular lexer class.
+//
+// If you want to create multiple lexer classes, you use the -P flag
+// to rename each yyFlexLexer to some other xxFlexLexer.  You then
+// include <FlexLexer.h> in your other sources once per lexer class:
+//
+//	#undef yyFlexLexer
+//	#define yyFlexLexer xxFlexLexer
+//	#include <FlexLexer.h>
+//
+//	#undef yyFlexLexer
+//	#define yyFlexLexer zzFlexLexer
+//	#include <FlexLexer.h>
+//	...
+
+#ifndef __FLEX_LEXER_H
+// Never included before - need to define base class.
+#define __FLEX_LEXER_H
+
+#include <iostream>
+#  ifndef FLEX_STD
+#    define FLEX_STD std::
+#  endif
+
+extern "C++" {
+
+struct yy_buffer_state;
+typedef int yy_state_type;
+
+class FlexLexer {
+public:
+	virtual ~FlexLexer()	{ }
+
+	const char* YYText() const	{ return yytext; }
+	int YYLeng()	const	{ return yyleng; }
+
+	virtual void
+		yy_switch_to_buffer( struct yy_buffer_state* new_buffer ) = 0;
+	virtual struct yy_buffer_state*
+		yy_create_buffer( FLEX_STD istream* s, int size ) = 0;
+	virtual void yy_delete_buffer( struct yy_buffer_state* b ) = 0;
+	virtual void yyrestart( FLEX_STD istream* s ) = 0;
+
+	virtual int yylex() = 0;
+
+	// Call yylex with new input/output sources.
+	int yylex( FLEX_STD istream* new_in, FLEX_STD ostream* new_out = 0 )
+		{
+		switch_streams( new_in, new_out );
+		return yylex();
+		}
+
+	// Switch to new input/output streams.  A nil stream pointer
+	// indicates "keep the current one".
+	virtual void switch_streams( FLEX_STD istream* new_in = 0,
+					FLEX_STD ostream* new_out = 0 ) = 0;
+
+	int lineno() const		{ return yylineno; }
+
+	int debug() const		{ return yy_flex_debug; }
+	void set_debug( int flag )	{ yy_flex_debug = flag; }
+
+protected:
+	char* yytext;
+	int yyleng;
+	int yylineno;		// only maintained if you use %option yylineno
+	int yy_flex_debug;	// only has effect with -d or "%option debug"
+};
+
+}
+#endif // FLEXLEXER_H
+
+#if defined(yyFlexLexer) || ! defined(yyFlexLexerOnce)
+// Either this is the first time through (yyFlexLexerOnce not defined),
+// or this is a repeated include to define a different flavor of
+// yyFlexLexer, as discussed in the flex manual.
+#define yyFlexLexerOnce
+
+extern "C++" {
+
+class yyFlexLexer : public FlexLexer {
+public:
+	// arg_yyin and arg_yyout default to the cin and cout, but we
+	// only make that assignment when initializing in yylex().
+	yyFlexLexer( FLEX_STD istream* arg_yyin = 0, FLEX_STD ostream* arg_yyout = 0 );
+
+	virtual ~yyFlexLexer();
+
+	void yy_switch_to_buffer( struct yy_buffer_state* new_buffer );
+	struct yy_buffer_state* yy_create_buffer( FLEX_STD istream* s, int size );
+	void yy_delete_buffer( struct yy_buffer_state* b );
+	void yyrestart( FLEX_STD istream* s );
+
+	void yypush_buffer_state( struct yy_buffer_state* new_buffer );
+	void yypop_buffer_state();
+
+	virtual int yylex();
+	virtual void switch_streams( FLEX_STD istream* new_in, FLEX_STD ostream* new_out = 0 );
+	virtual int yywrap();
+
+protected:
+	virtual int LexerInput( char* buf, int max_size );
+	virtual void LexerOutput( const char* buf, int size );
+	virtual void LexerError( const char* msg );
+
+	void yyunput( int c, char* buf_ptr );
+	int yyinput();
+
+	void yy_load_buffer_state();
+	void yy_init_buffer( struct yy_buffer_state* b, FLEX_STD istream* s );
+	void yy_flush_buffer( struct yy_buffer_state* b );
+
+	int yy_start_stack_ptr;
+	int yy_start_stack_depth;
+	int* yy_start_stack;
+
+	void yy_push_state( int new_state );
+	void yy_pop_state();
+	int yy_top_state();
+
+	yy_state_type yy_get_previous_state();
+	yy_state_type yy_try_NUL_trans( yy_state_type current_state );
+	int yy_get_next_buffer();
+
+	FLEX_STD istream* yyin;	// input source for default LexerInput
+	FLEX_STD ostream* yyout;	// output sink for default LexerOutput
+
+	// yy_hold_char holds the character lost when yytext is formed.
+	char yy_hold_char;
+
+	// Number of characters read into yy_ch_buf.
+	int yy_n_chars;
+
+	// Points to current character in buffer.
+	char* yy_c_buf_p;
+
+	int yy_init;		// whether we need to initialize
+	int yy_start;		// start state number
+
+	// Flag which is used to allow yywrap()'s to do buffer switches
+	// instead of setting up a fresh yyin.  A bit of a hack ...
+	int yy_did_buffer_switch_on_eof;
+
+
+	size_t yy_buffer_stack_top; /**< index of top of stack. */
+	size_t yy_buffer_stack_max; /**< capacity of stack. */
+	struct yy_buffer_state ** yy_buffer_stack; /**< Stack as an array. */
+	void yyensure_buffer_stack(void);
+
+	// The following are not always needed, but may be depending
+	// on use of certain flex features (like REJECT or yymore()).
+
+	yy_state_type yy_last_accepting_state;
+	char* yy_last_accepting_cpos;
+
+	yy_state_type* yy_state_buf;
+	yy_state_type* yy_state_ptr;
+
+	char* yy_full_match;
+	int* yy_full_state;
+	int yy_full_lp;
+
+	int yy_lp;
+	int yy_looking_for_trail_begin;
+
+	int yy_more_flag;
+	int yy_more_len;
+	int yy_more_offset;
+	int yy_prev_more_offset;
+};
+
+}
+
+#endif // yyFlexLexer || ! yyFlexLexerOnce
+

contrib/flex/NEWS

@@ -0,0 +1,518 @@
+This is the file NEWS for the flex package. It records user -visible
+changes between releases of flex.
+
+See the file COPYING for copying conditions.
+
+* version 2.5.37 released 2012-08-03
+
+** Import flex into git. See
+   git://flex.git.sourceforge.net/gitroot/flex/flex.
+
+** Fix make install target to not fail when the flex++ program is
+   already installed
+
+** New translations from the translation project: de, fi, pl, vi
+
+* version 2.5.36 released 2012-07-20
+
+** various portability fixes that quiet compiler warnings on 64-bit
+   hosts
+
+** various manual fixes, including correcting the name of a %option and
+   updating some simple examples to use ANSI C syntax
+
+** various bug fixes that prevent certain error conditions from
+   persisting when they should not persist
+
+** improvements to the test suite so it behaves better when linking
+   compiled files
+
+** new translations from the translation project: ca, da, es, fi, fr,
+   ga, ko, pt_br, ro, ru, sv, tr, zh_cn
+
+** the flex distribution is now built with automake 1.10.1 and automake
+   2.61
+
+* version 2.5.35 released 2008-02-26
+
+** fixed bug that prevented flex from accepting certain comments in the
+  scanner file (resolves bugs #1849809 and #1849805)
+
+** fix bug that prevented headers for all functions from being generated
+  (resolves bug #1628314)
+
+** change yy_size_t to be size_t (resolves bug #1849812)
+
+** new de, nl, pl, pt_br, vi translations from the translation project
+
+* version 2.5.34 released 2007-12-12
+
+** introduce yylex_init_extra; see the manual for details
+
+** introduce %option extra-type="your_type *" (resolves bug #1744505)
+
+** The flex program now parses multiple short concatenated options (resolves bug
+  #1619820). Thanks to Petr Machata of Red Hat on this issue.
+
+** better checking after yyalloc/yyrealloc (resolves bug #1595967)
+
+** flex now provides for a libfl_pic.a compiled with position
+   independent code. Particularly useful when including a flex scanner
+   in a shared library and with more recent versions of gcc. Thanks to the Debian project for the idea.
+
+** SourceForge feature request #1658379: Expose YY_BUF_SIZE in the
+	header file.
+
+** flex better escapes filenames with special characters in them
+   (resolves bug #1623600)
+
+** a memory leak was plugged(resolves bug #1601111)
+
+** pattern language expanded; see the manual for details on the below
+   highlights
+
+*** pattern options added to specify patterns as case-insensitive or
+    case-sensitive
+
+*** pattern options to specify whether the "." character should match
+    the newline character
+
+*** pattern options added to allow ignoring of whitespace in patterns
+
+*** POSIX character classes may be negated in patterns
+
+*** patterns may now use set difference, union operators
+
+** the manual now contains an appendix listing various common patterns
+   which may be useful when writing scanners
+
+** some memory leaks were removed from the C++ scanner (but the C++
+  scanner is still experimental and may change radically without
+  notice)
+
+** c++ scanners can now use yywrap
+
+** added new unit test for c++ and yywrap
+
+** portability fixes to some unit tests
+
+** flex man page and flex manual in pdf now distributed in the flex
+distribution
+
+** new ca, vi, ga, nl translations from the translation project
+
+** flex no longer comes with an rpm spec file
+
+** flex development now happens with automake 1.9.6
+
+* version 2.5.33 released 2006-2-20
+
+** all flex resources are now to be found from the website at
+   http://flex.sourceforge.net/
+
+** there was no release 2.5.32 published
+
+** numerous bug and security fixes
+
+** new nl, vi, sv, ro, po, ga, ca, fr, tr translations from the translation project
+
+** upgrade to use gettext 0.12 (this now makes the "pdf" and "ps"
+   targets in the build system able to be run successfully)
+
+* version 2.5.31 released 2003-4-1
+
+** remove --enable-maintainer-mode configure option; none of the
+   Makefiles were using it and it can be unduely confusing
+
+* version 2.5.30 released 2003-4-1
+
+** yylineno is per-buffer in reentrant scanners
+
+** added %top directive for placing code at the top of the generated
+   scanner; see manual for details
+
+** flex now uses m4 to generate scanners; while this means that
+   scanners are more readable, it means that flex requires m4 to be
+   installed; see manual for details
+
+* version 2.5.29 released 2003-3-5
+
+** Automatic stack management for multiple input buffers in C and C++ scanners
+
+** moved the flex documentation to a new doc/ subdirectory
+
+** cleanups to the yy namespace
+
+* version 2.5.28 released 2003-2-12
+
+** flex is now hosted at sourceforge
+
+** Fixed trailing slash bug in YY_INPUT macro def
+
+** Flex now warns if always-interactive is specified with fast or full
+
+* version 2.5.27 released 2003-1-21
+
+** flex now works with recent bison versions
+
+** new pt_br translation from the translation project
+
+* version 2.5.26 released 2003-1-14
+
+** Fixed table deserialization bug on big-endian archs. Patch sent from Bryce Nichols <bryce@bnichols.org>
+
+** yyleng has proper declarations now; this caused flex to generate
+   unusable scanners for some programs
+
+** the flex distribution now includes a spec file suitable for use
+   with rpm
+
+** some more c++ fixes
+
+** new es translation from the translation project
+
+** slight tweeks to the flex_int*_t types
+
+** flex now warns about pattern ranges that might be ambiguous when
+   generating a case-insensitive scanner
+
+
+* version 2.5.25 released 2002-12-2
+
+** flex now uses flex_int*_t types. For C99 systems, they are just the
+   int*_t types; for non-C99 systems, we just make some typedefs
+
+** new pt_br translation from the translation project
+
+* version 2.5.24 released 2002-11-25
+
+* more portability fixes
+
+** the manual continues to be updated and edited, but it's still got a
+   ways to go
+
+** it is possible to have multiple c++ scanners in the same program again
+
+** new turkish translation from the translation project
+
+* version 2.5.23 released 2002-10-21
+
+** more portability fixes
+
+** the manual includes a title page and a table-of-contents when printed
+
+** the test suite can be run with "make check" from the top-level
+   directory
+
+** configure now accepts the --enable-maintainer-mode option
+
+** gettext functionality is now only available externally
+
+** the constant FLEX_BETA is defined if flex is a beta release
+
+** the script create-test was not included in the distribution and it
+   should have been
+
+* version 2.5.22 released 2002-10-10
+
+** more portability fixes around how we get ahold of the integral
+   types; there is a constant FLEX_NEED_INTEGRAL_TYPE_DEFINITIONS
+   which you should define if you don't have the <inttypes.h> header
+   file (after you complain to your C vendor for not providing a
+   reasonable C environment)
+
+** more test suite cleanups; in particular, the test suite should run
+   correctly when build from a different directory
+
+** upgraded automake to 1.7 and consequently autoconf to 2.54; this
+   means, among other things, that there is some support for
+formatting the manual in postscript and pdf in the distributed
+   Makefile.in (and therefore in the Makefile built by configure)
+
+** the flex.1 manpage is generated by help2man; (this has been true
+   for quite a while but was not listed here)
+
+** flex now includes three defined constants to indicate which version
+  of flex generated a scanner (YY_FLEX_{MAJOR,MINOR,SUBMINOR}_VERSION)
+
+** flex tries its best to output only the relevant portions of the
+   skeleton when generating a scanner, thus avoiding as much
+   conditional compilation as possible
+
+* version 2.5.21 released 2002-9-17
+
+** one of the tests in the test suite broke the dist target
+
+* version 2.5.20 released 2002-9-16
+
+** A flex scanner has the ability to save the DFA tables to a file,
+   and load them at runtime when needed; see the manual for details
+
+** Added %option bison-bridge (--bison-bridge)
+
+** Removed %option reentrant-bison/--reentrant-bison/-Rb
+
+** yylineno is present in all scanners; Modified nasty performance
+   penalty warning with yylineno in documentation
+
+** test-table-opts is now run last in the test suite because it's so fat
+
+** flex can, to some extent, diagnose where internal problems occur
+
+** new translations from the translation project: fr, ca, de, ru, sv
+
+**Flex generates C99 defs now; see YY_TRADITIONAL_FUNC_DEFS in the
+  manual if that's not a good thing for you
+
+* version 2.5.19 released 2002-9-5
+
+** prevent segfault on input lines which are longer than the allocated
+   space (problem report from Manoj Srivastava
+   <srivasta@golden-gryphon.com>)
+
+** Changed option 'header' to 'header-file'
+
+* version 2.5.18 released 2002-9-4
+
+** portability fixes for integer constants and in the way the test
+   suite reports its results
+
+** the test for bison was reporting bison missing when it was, in
+   fact, found
+
+** if we don't find GNU indent, we're more careful when we're not
+   finding it
+
+* version 2.5.17 released 2002-8-29
+
+** more portability fixes
+
+** updated config.sub and config.guess
+
+** flex is indented by GNU indent (this was done earlier but not
+   explicitly documented)
+
+* version 2.5.16 released 2002-8-28
+
+** c++ scanners compile again
+
+** there is now an indent target in the top-level Makefile; configure
+   checks for GNU indent which is required for proper operation of the
+   indent target
+
+** some more portability fixes were made
+
+** %options and invocation sections of manual merged
+
+** a c++ test was added to the test suite
+
+** we're trying to clean up more files in the test suite's make clean
+   targets
+
+* version 2.5.15 released 2002-8-21
+
+** reject-state buffer is now dynamically allocated and REJECT buffer
+   variables are reentrant-safe
+
+** manual now discusses memory usage
+
+** skeleton now processed by m4 before mkskel.sh; (this only matters
+   if you want to change the skeleton or if you're doing flex development)
+
+** zh_cn translation added from translation project
+
+** a bug that caused a segfault has now been fixed
+
+** the test suite now respects the usual CFLAGS, etc. variables
+
+** removed some warnings which some tests trigggered with the -s option
+
+** the flex-generated header file now tries to be smarter about
+   conditionally including start conditions
+
+** tables code omitted from generated scanner when not used
+
+* version 2.5.14 released 2002-8-15
+
+** the tests using the reentrant c scanner as c++ were reworked
+   slightly to be sure that the c++ was enforced
+
+** de translation now included in the distribution
+
+** various portability fixes regarding nls support, c++ include
+   headers, etc.
+
+* version 2.5.13 released 2002-8-15
+
+** the header file output with %option header is now much smaller
+
+** Fixed type mismatch in printf in scanner skeleton
+
+** yylex_init now reports errors
+
+* version 2.5.12 released 2002-8-8
+
+** updated gettext support to 0.11.5
+
+** new fr translation from the translation project
+
+** bison is no longer needed to build flex; If you are building flex
+   from a release (i.e., not from a cvs snapshot), then you don't need
+   to have a pre-built lex around either (unless you modify scan.l, of
+   course); (This has been true for some time, but was not mentioned
+   here.)
+
+* version 2.5.11 released 2002-7-31
+
+** Fixed bug where yyless did not consider yylineno
+
+** the yylineno performance hit is now gone
+
+** fixed some typos in the manual and we now include texinfo.tex in
+   the distribution
+
+** traditional prototypes output for C scanners, controlled by a
+   preprocessor symbol; see documentation for details
+
+* version 2.5.10 released 2002-7-24
+
+** yy_globals renamed to yyscanner and yy_globals_t renamed to
+   yy_guts_t
+
+** added dist-bzip2 option to Makefile.am so we now produce a bzip2'd
+   archive in addition to the standard gzip archive
+
+*  version 2.5.9
+
+** new tests in test suite: test-mem-{nr,r}, test-posix,
+   test-posixly-correct, test-debug-{nr,r}
+
+** made changes to work with gcc-3.2 development code
+
+** ability to choose which memory functions are used in flex
+
+** new yylex_destroy() function for the non-reentrant scanner
+
+** new handling of POSIXLY_CORRECT environment variable
+
+** the test suite now has its copyrights explicitly described
+
+** new ca, de, fr, ru, sv, tr translations
+
+* version 2.5.8
+
+** a new --posix option generates scanners with posix-style abc{1,3}
+   compatible parsing, see manual for the screwy details
+
+* version 2.5.7
+
+** configure.in now includes a call to AC_PREREQ to enforce the
+   requirement for autoconf at least 2.50 (This only effects you if
+   you're doing flex development.)
+
+** configure now uses autoconf's versioning information and configure
+   --help reports the bug-reporting address for flex
+
+** test suite now only reports success versus failure; reporting
+   skipped is problematic under the current setup
+
+** compilation with --disable-nls now works
+
+** flex can now be built in a separate directory
+
+* version 2.5.6
+
+** gettext support added (from gettext 0.11)
+
+*** translations for ca, da, de, es, fr, ko, ru, sv, tr included
+
+** distribution now built under automake 1.6 and autoconf 2.53
+
+** command-line option parsing happens differently now:
+
+*** Added long option parsing
+
+*** Options -n and -c, previously deprecated, now simply do nothing
+
+*** Options are now parsed left to right
+
+** added a number of new options
+
+*** All positive %options are now accessible from the command line
+
+*** Added option -D, to define a preprocessor symbol
+
+*** Added option --header=FILE to specify a C .h file to generate
+
+*** added option --yywrap to call yywrap on EOF
+
+*** added option --yylineno to track line count in yylineno
+
+*** --yyclass=NAME name of C++ class when generating c++ scanners
+
+*** for long option names which are associated with existing short
+options, see accompanying documentation
+
+*** new %option nounistd or command-line --nounistd added to prevent
+    flex from generating #include <unistd.h> on systems that don't
+    have that include file
+
+** Support for reentrant C scanners has been added
+
+*** Updated the manual with the new reentrant API
+
+*** Two new options %option reentrant (-R) and 
+%option reentrant-bison (-Rb)
+
+*** All globals optionally placed into struct yyglobals_t
+
+*** All access to globals replaced by macro invocations
+
+*** All functions optionally take one additional
+argument, yy_globals
+
+*** New style for invoking reentrant scanner:
+yylex_init(void** scanner );
+yylex( scanner );
+yylex_destroy( scanner );
+
+*** Added get/set functions for members of struct yy_globals_t
+e.g.,  yyget_text, yyget_leng, etc
+
+*** Prefix substitution added for new functions
+
+*** Macro shortcuts to the lengthy get/set functions
+provided for use in actions, e.g.,  yytext, yyleng, etc
+
+*** Arbitrary, user-defined data, "yyextra", may be added to scanner
+
+** %option nomain no longer implies %option yywrap
+But the inverse is still true
+
+** Developer test suite added
+
+*** TESTS/ directory has been added. Users can 
+'make test' in the TESTS directory to execute the test suite
+
+** Support for bison variables yylval and yylloc added
+
+** automake support for the build process
+
+** manual is now in texinfo/info format
+
+*** flex.1 removed from distribution
+
+** flex no longer generates C-language scanners with C++-style
+   comments
+
+** flex now generates scanners in c++ which are compatible with
+   recent c++ compilers
+
+** flex input scanner now recognizes '\r' as an EOL character
+
+See the file ONEWS for changes in earlier releases.
+
+Local Variables:
+mode: text
+mode: outline-minor
+end:

contrib/flex/README

@@ -0,0 +1,79 @@
+This is flex, the fast lexical analyzer generator.
+
+flex is a tool for generating scanners: programs which recognize
+lexical patterns in text.
+
+More information about flex as well as the latest official release of
+flex can be found at:
+
+http://flex.sourceforge.net/
+
+Bug reports should be submitted using the SourceForge Bug Tracker
+facilities which can be found from flex's SourceForge project page at:
+
+http://sourceforge.net/projects/flex
+
+There are several mailing lists available as well:
+
+flex-announce@lists.sourceforge.net - where posts will be made
+announcing new releases of flex.
+
+flex-help@lists.sourceforge.net - where you can post questions about
+using flex
+
+flex-devel@lists.sourceforge.net - where you can discuss development of
+flex itself
+
+Note that flex is distributed under a copyright very similar to that of
+BSD Unix, and not under the GNU General Public License (GPL).
+
+This file is part of flex.
+
+This code is derived from software contributed to Berkeley by
+Vern Paxson.
+
+The United States Government has rights in this work pursuant
+to contract no. DE-AC03-76SF00098 between the United States
+Department of Energy and the University of California.
+
+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.
+
+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 ``AS IS'' AND WITHOUT ANY EXPRESS OR
+IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE.
+
+The flex distribution contains the following files which may be of interest:
+
+README - This file.
+
+NEWS - current version number and list of user-visible changes.
+
+INSTALL - basic installation information.
+
+ABOUT-NLS - description of internationalization support in flex.
+
+COPYING - flex's copyright and license.
+
+doc/ - user documentation.
+
+examples/ - containing examples of some possible flex scanners and a
+few other things. See the file examples/README for more details.
+
+TODO - outstanding bug reports, desired features, etc.
+
+tests/ - regression tests. See TESTS/README for details.
+
+po/ - internationalization support files.

contrib/flex/buf.c

@@ -0,0 +1,273 @@
+/* flex - tool to generate fast lexical analyzers */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "flexdef.h"
+
+/* Take note: The buffer object is sometimes used as a String buffer (one
+ * continuous string), and sometimes used as a list of strings, usually line by
+ * line.
+ * 
+ * The type is specified in buf_init by the elt_size. If the elt_size is
+ * sizeof(char), then the buffer should be treated as string buffer. If the
+ * elt_size is sizeof(char*), then the buffer should be treated as a list of
+ * strings.
+ *
+ * Certain functions are only appropriate for one type or the other. 
+ */
+
+/* global buffers. */
+struct Buf userdef_buf;		/**< for user #definitions triggered by cmd-line. */
+struct Buf defs_buf;		/**< for #define's autogenerated. List of strings. */
+struct Buf yydmap_buf;		/**< string buffer to hold yydmap elements */
+struct Buf m4defs_buf;          /**< m4 definitions. List of strings. */
+struct Buf top_buf;             /**< contains %top code. String buffer. */
+
+struct Buf *buf_print_strings(struct Buf * buf, FILE* out)
+{
+    int i;
+
+    if(!buf || !out)
+        return buf;
+
+    for (i=0; i < buf->nelts; i++){
+        const char * s = ((char**)buf->elts)[i];
+        if(s)
+            fprintf(out, "%s", s);
+    }
+    return buf;
+}
+
+/* Append a "%s" formatted string to a string buffer */
+struct Buf *buf_prints (struct Buf *buf, const char *fmt, const char *s)
+{
+	char   *t;
+        size_t tsz;
+
+	t = flex_alloc (tsz = strlen (fmt) + strlen (s) + 1);
+	if (!t)
+	    flexfatal (_("Allocation of buffer to print string failed"));
+	snprintf (t, tsz, fmt, s);
+	buf = buf_strappend (buf, t);
+	flex_free (t);
+	return buf;
+}
+
+/** Append a line directive to the string buffer.
+ * @param buf A string buffer.
+ * @param filename file name
+ * @param lineno line number
+ * @return buf
+ */
+struct Buf *buf_linedir (struct Buf *buf, const char* filename, int lineno)
+{
+    char *dst, *src, *t;
+
+    t = flex_alloc (strlen ("#line \"\"\n")          +   /* constant parts */
+                    2 * strlen (filename)            +   /* filename with possibly all backslashes escaped */
+                    (int) (1 + log10 (abs (lineno))) +   /* line number */
+                    1);                                  /* NUL */
+    if (!t)
+      flexfatal (_("Allocation of buffer for line directive failed"));
+    for (dst = t + sprintf (t, "#line %d \"", lineno), src = filename; *src; *dst++ = *src++)
+      if (*src == '\\')   /* escape backslashes */
+        *dst++ = '\\';
+    *dst++ = '"';
+    *dst++ = '\n';
+    *dst   = '\0';
+    buf = buf_strappend (buf, t);
+    flex_free (t);
+    return buf;
+}
+
+
+/** Append the contents of @a src to @a dest.
+ * @param @a dest the destination buffer
+ * @param @a dest the source buffer
+ * @return @a dest
+ */
+struct Buf *buf_concat(struct Buf* dest, const struct Buf* src)
+{
+    buf_append(dest, src->elts, src->nelts);
+    return dest;
+}
+
+
+/* Appends n characters in str to buf. */
+struct Buf *buf_strnappend (buf, str, n)
+     struct Buf *buf;
+     const char *str;
+     int n;
+{
+	buf_append (buf, str, n + 1);
+
+	/* "undo" the '\0' character that buf_append() already copied. */
+	buf->nelts--;
+
+	return buf;
+}
+
+/* Appends characters in str to buf. */
+struct Buf *buf_strappend (buf, str)
+     struct Buf *buf;
+     const char *str;
+{
+	return buf_strnappend (buf, str, strlen (str));
+}
+
+/* appends "#define str def\n" */
+struct Buf *buf_strdefine (buf, str, def)
+     struct Buf *buf;
+     const char *str;
+     const char *def;
+{
+	buf_strappend (buf, "#define ");
+	buf_strappend (buf, " ");
+	buf_strappend (buf, str);
+	buf_strappend (buf, " ");
+	buf_strappend (buf, def);
+	buf_strappend (buf, "\n");
+	return buf;
+}
+
+/** Pushes "m4_define( [[def]], [[val]])m4_dnl" to end of buffer.
+ * @param buf A buffer as a list of strings.
+ * @param def The m4 symbol to define.
+ * @param val The definition; may be NULL.
+ * @return buf
+ */
+struct Buf *buf_m4_define (struct Buf *buf, const char* def, const char* val)
+{
+    const char * fmt = "m4_define( [[%s]], [[%s]])m4_dnl\n";
+    char * str;
+    size_t strsz;
+
+    val = val?val:"";
+    str = (char*)flex_alloc(strsz = strlen(fmt) + strlen(def) + strlen(val) + 2);
+    if (!str)
+        flexfatal (_("Allocation of buffer for m4 def failed"));
+
+    snprintf(str, strsz, fmt, def, val);
+    buf_append(buf, &str, 1);
+    return buf;
+}
+
+/** Pushes "m4_undefine([[def]])m4_dnl" to end of buffer.
+ * @param buf A buffer as a list of strings.
+ * @param def The m4 symbol to undefine.
+ * @return buf
+ */
+struct Buf *buf_m4_undefine (struct Buf *buf, const char* def)
+{
+    const char * fmt = "m4_undefine( [[%s]])m4_dnl\n";
+    char * str;
+    size_t strsz;
+
+    str = (char*)flex_alloc(strsz = strlen(fmt) + strlen(def) + 2);
+    if (!str)
+        flexfatal (_("Allocation of buffer for m4 undef failed"));
+
+    snprintf(str, strsz, fmt, def);
+    buf_append(buf, &str, 1);
+    return buf;
+}
+
+/* create buf with 0 elements, each of size elem_size. */
+void buf_init (buf, elem_size)
+     struct Buf *buf;
+     size_t elem_size;
+{
+	buf->elts = (void *) 0;
+	buf->nelts = 0;
+	buf->elt_size = elem_size;
+	buf->nmax = 0;
+}
+
+/* frees memory */
+void buf_destroy (buf)
+     struct Buf *buf;
+{
+	if (buf && buf->elts)
+		flex_free (buf->elts);
+	buf->elts = (void *) 0;
+}
+
+
+/* appends ptr[] to buf, grow if necessary.
+ * n_elem is number of elements in ptr[], NOT bytes.
+ * returns buf.
+ * We grow by mod(512) boundaries.
+ */
+
+struct Buf *buf_append (buf, ptr, n_elem)
+     struct Buf *buf;
+     const void *ptr;
+     int n_elem;
+{
+	int     n_alloc = 0;
+
+	if (!ptr || n_elem == 0)
+		return buf;
+
+	/* May need to alloc more. */
+	if (n_elem + buf->nelts > buf->nmax) {
+
+		/* exact amount needed... */
+		n_alloc = (n_elem + buf->nelts) * buf->elt_size;
+
+		/* ...plus some extra */
+		if (((n_alloc * buf->elt_size) % 512) != 0
+		    && buf->elt_size < 512)
+			n_alloc +=
+				(512 -
+				 ((n_alloc * buf->elt_size) % 512)) /
+				buf->elt_size;
+
+		if (!buf->elts)
+			buf->elts =
+				allocate_array (n_alloc, buf->elt_size);
+		else
+			buf->elts =
+				reallocate_array (buf->elts, n_alloc,
+						  buf->elt_size);
+
+		buf->nmax = n_alloc;
+	}
+
+	memcpy ((char *) buf->elts + buf->nelts * buf->elt_size, ptr,
+		n_elem * buf->elt_size);
+	buf->nelts += n_elem;
+
+	return buf;
+}
+
+/* vim:set tabstop=8 softtabstop=4 shiftwidth=4: */

contrib/flex/ccl.c

@@ -0,0 +1,310 @@
+/* ccl - routines for character classes */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+ /*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "flexdef.h"
+
+/* return true if the chr is in the ccl. Takes negation into account. */
+static bool
+ccl_contains (const int cclp, const int ch)
+{
+	int     ind, len, i;
+
+	len = ccllen[cclp];
+	ind = cclmap[cclp];
+
+	for (i = 0; i < len; ++i)
+		if (ccltbl[ind + i] == ch)
+			return !cclng[cclp];
+
+    return cclng[cclp];
+}
+
+
+/* ccladd - add a single character to a ccl */
+
+void    ccladd (cclp, ch)
+     int     cclp;
+     int     ch;
+{
+	int     ind, len, newpos, i;
+
+	check_char (ch);
+
+	len = ccllen[cclp];
+	ind = cclmap[cclp];
+
+	/* check to see if the character is already in the ccl */
+
+	for (i = 0; i < len; ++i)
+		if (ccltbl[ind + i] == ch)
+			return;
+
+	/* mark newlines */
+	if (ch == nlch)
+		ccl_has_nl[cclp] = true;
+
+	newpos = ind + len;
+
+	if (newpos >= current_max_ccl_tbl_size) {
+		current_max_ccl_tbl_size += MAX_CCL_TBL_SIZE_INCREMENT;
+
+		++num_reallocs;
+
+		ccltbl = reallocate_Character_array (ccltbl,
+						     current_max_ccl_tbl_size);
+	}
+
+	ccllen[cclp] = len + 1;
+	ccltbl[newpos] = ch;
+}
+
+/* dump_cclp - same thing as list_character_set, but for cclps.  */
+
+static void    dump_cclp (FILE* file, int cclp)
+{
+	register int i;
+
+	putc ('[', file);
+
+	for (i = 0; i < csize; ++i) {
+		if (ccl_contains(cclp, i)){
+			register int start_char = i;
+
+			putc (' ', file);
+
+			fputs (readable_form (i), file);
+
+			while (++i < csize && ccl_contains(cclp,i)) ;
+
+			if (i - 1 > start_char)
+				/* this was a run */
+				fprintf (file, "-%s",
+					 readable_form (i - 1));
+
+			putc (' ', file);
+		}
+	}
+
+	putc (']', file);
+}
+
+
+
+/* ccl_set_diff - create a new ccl as the set difference of the two given ccls. */
+int
+ccl_set_diff (int a, int b)
+{
+    int  d, ch;
+
+    /* create new class  */
+    d = cclinit();
+
+    /* In order to handle negation, we spin through all possible chars,
+     * addding each char in a that is not in b.
+     * (This could be O(n^2), but n is small and bounded.)
+     */
+	for ( ch = 0; ch < csize; ++ch )
+        if (ccl_contains (a, ch) && !ccl_contains(b, ch))
+            ccladd (d, ch);
+
+    /* debug */
+    if (0){
+        fprintf(stderr, "ccl_set_diff (");
+            fprintf(stderr, "\n    ");
+            dump_cclp (stderr, a);
+            fprintf(stderr, "\n    ");
+            dump_cclp (stderr, b);
+            fprintf(stderr, "\n    ");
+            dump_cclp (stderr, d);
+        fprintf(stderr, "\n)\n");
+    }
+    return d;
+}
+
+/* ccl_set_union - create a new ccl as the set union of the two given ccls. */
+int
+ccl_set_union (int a, int b)
+{
+    int  d, i;
+
+    /* create new class  */
+    d = cclinit();
+
+    /* Add all of a */
+    for (i = 0; i < ccllen[a]; ++i)
+		ccladd (d, ccltbl[cclmap[a] + i]);
+
+    /* Add all of b */
+    for (i = 0; i < ccllen[b]; ++i)
+		ccladd (d, ccltbl[cclmap[b] + i]);
+
+    /* debug */
+    if (0){
+        fprintf(stderr, "ccl_set_union (%d + %d = %d", a, b, d);
+            fprintf(stderr, "\n    ");
+            dump_cclp (stderr, a);
+            fprintf(stderr, "\n    ");
+            dump_cclp (stderr, b);
+            fprintf(stderr, "\n    ");
+            dump_cclp (stderr, d);
+        fprintf(stderr, "\n)\n");
+    }
+    return d;
+}
+
+
+/* cclinit - return an empty ccl */
+
+int     cclinit ()
+{
+	if (++lastccl >= current_maxccls) {
+		current_maxccls += MAX_CCLS_INCREMENT;
+
+		++num_reallocs;
+
+		cclmap =
+			reallocate_integer_array (cclmap, current_maxccls);
+		ccllen =
+			reallocate_integer_array (ccllen, current_maxccls);
+		cclng = reallocate_integer_array (cclng, current_maxccls);
+		ccl_has_nl =
+			reallocate_bool_array (ccl_has_nl,
+					       current_maxccls);
+	}
+
+	if (lastccl == 1)
+		/* we're making the first ccl */
+		cclmap[lastccl] = 0;
+
+	else
+		/* The new pointer is just past the end of the last ccl.
+		 * Since the cclmap points to the \first/ character of a
+		 * ccl, adding the length of the ccl to the cclmap pointer
+		 * will produce a cursor to the first free space.
+		 */
+		cclmap[lastccl] =
+			cclmap[lastccl - 1] + ccllen[lastccl - 1];
+
+	ccllen[lastccl] = 0;
+	cclng[lastccl] = 0;	/* ccl's start out life un-negated */
+	ccl_has_nl[lastccl] = false;
+
+	return lastccl;
+}
+
+
+/* cclnegate - negate the given ccl */
+
+void    cclnegate (cclp)
+     int     cclp;
+{
+	cclng[cclp] = 1;
+	ccl_has_nl[cclp] = !ccl_has_nl[cclp];
+}
+
+
+/* list_character_set - list the members of a set of characters in CCL form
+ *
+ * Writes to the given file a character-class representation of those
+ * characters present in the given CCL.  A character is present if it
+ * has a non-zero value in the cset array.
+ */
+
+void    list_character_set (file, cset)
+     FILE   *file;
+     int     cset[];
+{
+	register int i;
+
+	putc ('[', file);
+
+	for (i = 0; i < csize; ++i) {
+		if (cset[i]) {
+			register int start_char = i;
+
+			putc (' ', file);
+
+			fputs (readable_form (i), file);
+
+			while (++i < csize && cset[i]) ;
+
+			if (i - 1 > start_char)
+				/* this was a run */
+				fprintf (file, "-%s",
+					 readable_form (i - 1));
+
+			putc (' ', file);
+		}
+	}
+
+	putc (']', file);
+}
+
+/** Determines if the range [c1-c2] is unambiguous in a case-insensitive
+ * scanner.  Specifically, if a lowercase or uppercase character, x, is in the
+ * range [c1-c2], then we require that UPPERCASE(x) and LOWERCASE(x) must also
+ * be in the range. If not, then this range is ambiguous, and the function
+ * returns false.  For example, [@-_] spans [a-z] but not [A-Z].  Beware that
+ * [a-z] will be labeled ambiguous because it does not include [A-Z].
+ *
+ * @param c1 the lower end of the range
+ * @param c2 the upper end of the range
+ * @return true if [c1-c2] is not ambiguous for a caseless scanner.
+ */
+bool range_covers_case (int c1, int c2)
+{
+	int     i, o;
+
+	for (i = c1; i <= c2; i++) {
+		if (has_case (i)) {
+			o = reverse_case (i);
+			if (o < c1 || c2 < o)
+				return false;
+		}
+	}
+	return true;
+}
+
+/** Reverse the case of a character, if possible.
+ * @return c if case-reversal does not apply.
+ */
+int reverse_case (int c)
+{
+	return isupper (c) ? tolower (c) : (islower (c) ? toupper (c) : c);
+}
+
+/** Return true if c is uppercase or lowercase. */
+bool has_case (int c)
+{
+	return (isupper (c) || islower (c)) ? true : false;
+}

contrib/flex/ecs.c

@@ -0,0 +1,219 @@
+/* ecs - equivalence class routines */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/* This file is part of flex */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+
+#include "flexdef.h"
+
+/* ccl2ecl - convert character classes to set of equivalence classes */
+
+void    ccl2ecl ()
+{
+	int     i, ich, newlen, cclp, ccls, cclmec;
+
+	for (i = 1; i <= lastccl; ++i) {
+		/* We loop through each character class, and for each character
+		 * in the class, add the character's equivalence class to the
+		 * new "character" class we are creating.  Thus when we are all
+		 * done, character classes will really consist of collections
+		 * of equivalence classes
+		 */
+
+		newlen = 0;
+		cclp = cclmap[i];
+
+		for (ccls = 0; ccls < ccllen[i]; ++ccls) {
+			ich = ccltbl[cclp + ccls];
+			cclmec = ecgroup[ich];
+
+			if (cclmec > 0) {
+				ccltbl[cclp + newlen] = cclmec;
+				++newlen;
+			}
+		}
+
+		ccllen[i] = newlen;
+	}
+}
+
+
+/* cre8ecs - associate equivalence class numbers with class members
+ *
+ * fwd is the forward linked-list of equivalence class members.  bck
+ * is the backward linked-list, and num is the number of class members.
+ *
+ * Returned is the number of classes.
+ */
+
+int     cre8ecs (fwd, bck, num)
+     int     fwd[], bck[], num;
+{
+	int     i, j, numcl;
+
+	numcl = 0;
+
+	/* Create equivalence class numbers.  From now on, ABS( bck(x) )
+	 * is the equivalence class number for object x.  If bck(x)
+	 * is positive, then x is the representative of its equivalence
+	 * class.
+	 */
+	for (i = 1; i <= num; ++i)
+		if (bck[i] == NIL) {
+			bck[i] = ++numcl;
+			for (j = fwd[i]; j != NIL; j = fwd[j])
+				bck[j] = -numcl;
+		}
+
+	return numcl;
+}
+
+
+/* mkeccl - update equivalence classes based on character class xtions
+ *
+ * synopsis
+ *    Char ccls[];
+ *    int lenccl, fwd[llsiz], bck[llsiz], llsiz, NUL_mapping;
+ *    void mkeccl( Char ccls[], int lenccl, int fwd[llsiz], int bck[llsiz],
+ *			int llsiz, int NUL_mapping );
+ *
+ * ccls contains the elements of the character class, lenccl is the
+ * number of elements in the ccl, fwd is the forward link-list of equivalent
+ * characters, bck is the backward link-list, and llsiz size of the link-list.
+ *
+ * NUL_mapping is the value which NUL (0) should be mapped to.
+ */
+
+void    mkeccl (ccls, lenccl, fwd, bck, llsiz, NUL_mapping)
+     Char    ccls[];
+     int     lenccl, fwd[], bck[], llsiz, NUL_mapping;
+{
+	int     cclp, oldec, newec;
+	int     cclm, i, j;
+	static unsigned char cclflags[CSIZE];	/* initialized to all '\0' */
+
+	/* Note that it doesn't matter whether or not the character class is
+	 * negated.  The same results will be obtained in either case.
+	 */
+
+	cclp = 0;
+
+	while (cclp < lenccl) {
+		cclm = ccls[cclp];
+
+		if (NUL_mapping && cclm == 0)
+			cclm = NUL_mapping;
+
+		oldec = bck[cclm];
+		newec = cclm;
+
+		j = cclp + 1;
+
+		for (i = fwd[cclm]; i != NIL && i <= llsiz; i = fwd[i]) {	/* look for the symbol in the character class */
+			for (; j < lenccl; ++j) {
+				register int ccl_char;
+
+				if (NUL_mapping && ccls[j] == 0)
+					ccl_char = NUL_mapping;
+				else
+					ccl_char = ccls[j];
+
+				if (ccl_char > i)
+					break;
+
+				if (ccl_char == i && !cclflags[j]) {
+					/* We found an old companion of cclm
+					 * in the ccl.  Link it into the new
+					 * equivalence class and flag it as
+					 * having been processed.
+					 */
+
+					bck[i] = newec;
+					fwd[newec] = i;
+					newec = i;
+					/* Set flag so we don't reprocess. */
+					cclflags[j] = 1;
+
+					/* Get next equivalence class member. */
+					/* continue 2 */
+					goto next_pt;
+				}
+			}
+
+			/* Symbol isn't in character class.  Put it in the old
+			 * equivalence class.
+			 */
+
+			bck[i] = oldec;
+
+			if (oldec != NIL)
+				fwd[oldec] = i;
+
+			oldec = i;
+
+		      next_pt:;
+		}
+
+		if (bck[cclm] != NIL || oldec != bck[cclm]) {
+			bck[cclm] = NIL;
+			fwd[oldec] = NIL;
+		}
+
+		fwd[newec] = NIL;
+
+		/* Find next ccl member to process. */
+
+		for (++cclp; cclflags[cclp] && cclp < lenccl; ++cclp) {
+			/* Reset "doesn't need processing" flag. */
+			cclflags[cclp] = 0;
+		}
+	}
+}
+
+
+/* mkechar - create equivalence class for single character */
+
+void    mkechar (tch, fwd, bck)
+     int     tch, fwd[], bck[];
+{
+	/* If until now the character has been a proper subset of
+	 * an equivalence class, break it away to create a new ec
+	 */
+
+	if (fwd[tch] != NIL)
+		bck[fwd[tch]] = bck[tch];
+
+	if (bck[tch] != NIL)
+		fwd[bck[tch]] = fwd[tch];
+
+	fwd[tch] = NIL;
+	bck[tch] = NIL;
+}

contrib/flex/filter.c

@@ -0,0 +1,444 @@
+/* filter - postprocessing of flex output through filters */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "flexdef.h"
+static const char * check_4_gnu_m4 =
+    "m4_dnl ifdef(`__gnu__', ,"
+    "`errprint(Flex requires GNU M4. Set the PATH or set the M4 environment variable to its path name.)"
+    " m4exit(2)')\n";
+
+
+/** global chain. */
+struct filter *output_chain = NULL;
+
+/* Allocate and initialize an external filter.
+ * @param chain the current chain or NULL for new chain
+ * @param cmd the command to execute.
+ * @param ... a NULL terminated list of (const char*) arguments to command,
+ *            not including argv[0].
+ * @return newest filter in chain
+ */
+struct filter *filter_create_ext (struct filter *chain, const char *cmd,
+				  ...)
+{
+	struct filter *f;
+	int     max_args;
+	const char *s;
+	va_list ap;
+
+	/* allocate and initialize new filter */
+	f = (struct filter *) flex_alloc (sizeof (struct filter));
+	if (!f)
+		flexerror (_("flex_alloc failed (f) in filter_create_ext"));
+	memset (f, 0, sizeof (*f));
+	f->filter_func = NULL;
+	f->extra = NULL;
+	f->next = NULL;
+	f->argc = 0;
+
+	if (chain != NULL) {
+		/* append f to end of chain */
+		while (chain->next)
+			chain = chain->next;
+		chain->next = f;
+	}
+
+
+	/* allocate argv, and populate it with the argument list. */
+	max_args = 8;
+	f->argv =
+		(const char **) flex_alloc (sizeof (char *) *
+					    (max_args + 1));
+	if (!f->argv)
+		flexerror (_("flex_alloc failed (f->argv) in filter_create_ext"));
+	f->argv[f->argc++] = cmd;
+
+	va_start (ap, cmd);
+	while ((s = va_arg (ap, const char *)) != NULL) {
+		if (f->argc >= max_args) {
+			max_args += 8;
+			f->argv =
+				(const char **) flex_realloc (f->argv,
+							      sizeof (char
+								      *) *
+							      (max_args +
+							       1));
+		}
+		f->argv[f->argc++] = s;
+	}
+	f->argv[f->argc] = NULL;
+
+	va_end (ap);
+	return f;
+}
+
+/* Allocate and initialize an internal filter.
+ * @param chain the current chain or NULL for new chain
+ * @param filter_func The function that will perform the filtering.
+ *        filter_func should return 0 if successful, and -1
+ *        if an error occurs -- or it can simply exit().
+ * @param extra optional user-defined data to pass to the filter.
+ * @return newest filter in chain
+ */
+struct filter *filter_create_int (struct filter *chain,
+				  int (*filter_func) (struct filter *),
+				  void *extra)
+{
+	struct filter *f;
+
+	/* allocate and initialize new filter */
+	f = (struct filter *) flex_alloc (sizeof (struct filter));
+	if (!f)
+		flexerror (_("flex_alloc failed in filter_create_int"));
+	memset (f, 0, sizeof (*f));
+	f->next = NULL;
+	f->argc = 0;
+	f->argv = NULL;
+
+	f->filter_func = filter_func;
+	f->extra = extra;
+
+	if (chain != NULL) {
+		/* append f to end of chain */
+		while (chain->next)
+			chain = chain->next;
+		chain->next = f;
+	}
+
+	return f;
+}
+
+/** Fork and exec entire filter chain.
+ *  @param chain The head of the chain.
+ *  @return true on success.
+ */
+bool filter_apply_chain (struct filter * chain)
+{
+	int     pid, pipes[2];
+	int     r;
+	const int readsz = 512;
+	char   *buf;
+
+
+	/* Tricky recursion, since we want to begin the chain
+	 * at the END. Why? Because we need all the forked processes
+	 * to be children of the main flex process.
+	 */
+	if (chain)
+		filter_apply_chain (chain->next);
+	else
+		return true;
+
+	/* Now we are the right-most unprocessed link in the chain.
+	 */
+
+	fflush (stdout);
+	fflush (stderr);
+
+
+	if (pipe (pipes) == -1)
+		flexerror (_("pipe failed"));
+
+	if ((pid = fork ()) == -1)
+		flexerror (_("fork failed"));
+
+	if (pid == 0) {
+		/* child */
+
+        /* We need stdin (the FILE* stdin) to connect to this new pipe.
+         * There is no portable way to set stdin to a new file descriptor,
+         * as stdin is not an lvalue on some systems (BSD).
+         * So we dup the new pipe onto the stdin descriptor and use a no-op fseek
+         * to sync the stream. This is a Hail Mary situation. It seems to work.
+         */
+		close (pipes[1]);
+clearerr(stdin);
+		if (dup2 (pipes[0], fileno (stdin)) == -1)
+			flexfatal (_("dup2(pipes[0],0)"));
+		close (pipes[0]);
+        fseek (stdin, 0, SEEK_CUR);
+
+		/* run as a filter, either internally or by exec */
+		if (chain->filter_func) {
+			int     r;
+
+			if ((r = chain->filter_func (chain)) == -1)
+				flexfatal (_("filter_func failed"));
+			exit (0);
+		}
+		else {
+			execvp (chain->argv[0],
+				(char **const) (chain->argv));
+            lerrsf_fatal ( _("exec of %s failed"),
+                    chain->argv[0]);
+		}
+
+		exit (1);
+	}
+
+	/* Parent */
+	close (pipes[0]);
+	if (dup2 (pipes[1], fileno (stdout)) == -1)
+		flexfatal (_("dup2(pipes[1],1)"));
+	close (pipes[1]);
+    fseek (stdout, 0, SEEK_CUR);
+
+	return true;
+}
+
+/** Truncate the chain to max_len number of filters.
+ * @param chain the current chain.
+ * @param max_len the maximum length of the chain.
+ * @return the resulting length of the chain.
+ */
+int filter_truncate (struct filter *chain, int max_len)
+{
+	int     len = 1;
+
+	if (!chain)
+		return 0;
+
+	while (chain->next && len < max_len) {
+		chain = chain->next;
+		++len;
+	}
+
+	chain->next = NULL;
+	return len;
+}
+
+/** Splits the chain in order to write to a header file.
+ *  Similar in spirit to the 'tee' program.
+ *  The header file name is in extra.
+ *  @return 0 (zero) on success, and -1 on failure.
+ */
+int filter_tee_header (struct filter *chain)
+{
+	/* This function reads from stdin and writes to both the C file and the
+	 * header file at the same time.
+	 */
+
+	const int readsz = 512;
+	char   *buf;
+	int     to_cfd = -1;
+	FILE   *to_c = NULL, *to_h = NULL;
+	bool    write_header;
+
+	write_header = (chain->extra != NULL);
+
+	/* Store a copy of the stdout pipe, which is already piped to C file
+	 * through the running chain. Then create a new pipe to the H file as
+	 * stdout, and fork the rest of the chain again.
+	 */
+
+	if ((to_cfd = dup (1)) == -1)
+		flexfatal (_("dup(1) failed"));
+	to_c = fdopen (to_cfd, "w");
+
+	if (write_header) {
+		if (freopen ((char *) chain->extra, "w", stdout) == NULL)
+			flexfatal (_("freopen(headerfilename) failed"));
+
+		filter_apply_chain (chain->next);
+		to_h = stdout;
+	}
+
+	/* Now to_c is a pipe to the C branch, and to_h is a pipe to the H branch.
+	 */
+
+	if (write_header) {
+        fputs (check_4_gnu_m4, to_h);
+		fputs ("m4_changecom`'m4_dnl\n", to_h);
+		fputs ("m4_changequote`'m4_dnl\n", to_h);
+		fputs ("m4_changequote([[,]])[[]]m4_dnl\n", to_h);
+	    fputs ("m4_define([[M4_YY_NOOP]])[[]]m4_dnl\n", to_h);
+		fputs ("m4_define( [[M4_YY_IN_HEADER]],[[]])m4_dnl\n",
+		       to_h);
+		fprintf (to_h, "#ifndef %sHEADER_H\n", prefix);
+		fprintf (to_h, "#define %sHEADER_H 1\n", prefix);
+		fprintf (to_h, "#define %sIN_HEADER 1\n\n", prefix);
+		fprintf (to_h,
+			 "m4_define( [[M4_YY_OUTFILE_NAME]],[[%s]])m4_dnl\n",
+			 headerfilename ? headerfilename : "<stdout>");
+
+	}
+
+    fputs (check_4_gnu_m4, to_c);
+	fputs ("m4_changecom`'m4_dnl\n", to_c);
+	fputs ("m4_changequote`'m4_dnl\n", to_c);
+	fputs ("m4_changequote([[,]])[[]]m4_dnl\n", to_c);
+	fputs ("m4_define([[M4_YY_NOOP]])[[]]m4_dnl\n", to_c);
+	fprintf (to_c, "m4_define( [[M4_YY_OUTFILE_NAME]],[[%s]])m4_dnl\n",
+		 outfilename ? outfilename : "<stdout>");
+
+	buf = (char *) flex_alloc (readsz);
+	if (!buf)
+		flexerror (_("flex_alloc failed in filter_tee_header"));
+	while (fgets (buf, readsz, stdin)) {
+		fputs (buf, to_c);
+		if (write_header)
+			fputs (buf, to_h);
+	}
+
+	if (write_header) {
+		fprintf (to_h, "\n");
+
+		/* write a fake line number. It will get fixed by the linedir filter. */
+		fprintf (to_h, "#line 4000 \"M4_YY_OUTFILE_NAME\"\n");
+
+		fprintf (to_h, "#undef %sIN_HEADER\n", prefix);
+		fprintf (to_h, "#endif /* %sHEADER_H */\n", prefix);
+		fputs ("m4_undefine( [[M4_YY_IN_HEADER]])m4_dnl\n", to_h);
+
+		fflush (to_h);
+		if (ferror (to_h))
+			lerrsf (_("error writing output file %s"),
+				(char *) chain->extra);
+
+		else if (fclose (to_h))
+			lerrsf (_("error closing output file %s"),
+				(char *) chain->extra);
+	}
+
+	fflush (to_c);
+	if (ferror (to_c))
+		lerrsf (_("error writing output file %s"),
+			outfilename ? outfilename : "<stdout>");
+
+	else if (fclose (to_c))
+		lerrsf (_("error closing output file %s"),
+			outfilename ? outfilename : "<stdout>");
+
+	while (wait (0) > 0) ;
+
+	exit (0);
+	return 0;
+}
+
+/** Adjust the line numbers in the #line directives of the generated scanner.
+ * After the m4 expansion, the line numbers are incorrect since the m4 macros
+ * can add or remove lines.  This only adjusts line numbers for generated code,
+ * not user code. This also happens to be a good place to squeeze multiple
+ * blank lines into a single blank line.
+ */
+int filter_fix_linedirs (struct filter *chain)
+{
+	char   *buf;
+	const int readsz = 512;
+	int     lineno = 1;
+	bool    in_gen = true;	/* in generated code */
+	bool    last_was_blank = false;
+
+	if (!chain)
+		return 0;
+
+	buf = (char *) flex_alloc (readsz);
+	if (!buf)
+		flexerror (_("flex_alloc failed in filter_fix_linedirs"));
+
+	while (fgets (buf, readsz, stdin)) {
+
+		regmatch_t m[10];
+
+		/* Check for #line directive. */
+		if (buf[0] == '#'
+			&& regexec (&regex_linedir, buf, 3, m, 0) == 0) {
+
+			int     num;
+			char   *fname;
+
+			/* extract the line number and filename */
+			num = regmatch_strtol (&m[1], buf, NULL, 0);
+			fname = regmatch_dup (&m[2], buf);
+
+			if (strcmp (fname,
+				outfilename ? outfilename : "<stdout>")
+					== 0
+			 || strcmp (fname,
+			 	headerfilename ? headerfilename : "<stdout>")
+					== 0) {
+
+				char    *s1, *s2;
+				char	filename[MAXLINE];
+
+				s1 = fname;
+				s2 = filename;
+
+				while ((s2 - filename) < (MAXLINE - 1) && *s1) {
+					/* Escape the backslash */
+					if (*s1 == '\\')
+						*s2++ = '\\';
+					/* Escape the double quote */
+					if (*s1 == '\"')
+						*s2++ = '\\';
+					/* Copy the character as usual */
+					*s2++ = *s1++;
+				}
+
+				*s2 = '\0';
+
+				/* Adjust the line directives. */
+				in_gen = true;
+				snprintf (buf, readsz, "#line %d \"%s\"\n",
+					  lineno + 1, filename);
+			}
+			else {
+				/* it's a #line directive for code we didn't write */
+				in_gen = false;
+			}
+
+			free (fname);
+			last_was_blank = false;
+		}
+
+		/* squeeze blank lines from generated code */
+		else if (in_gen
+			 && regexec (&regex_blank_line, buf, 0, NULL,
+				     0) == 0) {
+			if (last_was_blank)
+				continue;
+			else
+				last_was_blank = true;
+		}
+
+		else {
+			/* it's a line of normal, non-empty code. */
+			last_was_blank = false;
+		}
+
+		fputs (buf, stdout);
+		lineno++;
+	}
+	fflush (stdout);
+	if (ferror (stdout))
+		lerrsf (_("error writing output file %s"),
+			outfilename ? outfilename : "<stdout>");
+
+	else if (fclose (stdout))
+		lerrsf (_("error closing output file %s"),
+			outfilename ? outfilename : "<stdout>");
+
+	return 0;
+}
+
+/* vim:set expandtab cindent tabstop=4 softtabstop=4 shiftwidth=4 textwidth=0: */

contrib/flex/flexint.h

@@ -0,0 +1,63 @@
+/* flex integer type definitions */
+
+#ifndef FLEXINT_H
+#define FLEXINT_H
+
+/* C99 systems have <inttypes.h>. Non-C99 systems may or may not. */
+
+#if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L
+
+/* C99 says to define __STDC_LIMIT_MACROS before including stdint.h,
+ * if you want the limit (max/min) macros for int types. 
+ */
+#ifndef __STDC_LIMIT_MACROS
+#define __STDC_LIMIT_MACROS 1
+#endif
+
+#include <inttypes.h>
+typedef int8_t flex_int8_t;
+typedef uint8_t flex_uint8_t;
+typedef int16_t flex_int16_t;
+typedef uint16_t flex_uint16_t;
+typedef int32_t flex_int32_t;
+typedef uint32_t flex_uint32_t;
+#else
+typedef signed char flex_int8_t;
+typedef short int flex_int16_t;
+typedef int flex_int32_t;
+typedef unsigned char flex_uint8_t; 
+typedef unsigned short int flex_uint16_t;
+typedef unsigned int flex_uint32_t;
+
+/* Limits of integral types. */
+#ifndef INT8_MIN
+#define INT8_MIN               (-128)
+#endif
+#ifndef INT16_MIN
+#define INT16_MIN              (-32767-1)
+#endif
+#ifndef INT32_MIN
+#define INT32_MIN              (-2147483647-1)
+#endif
+#ifndef INT8_MAX
+#define INT8_MAX               (127)
+#endif
+#ifndef INT16_MAX
+#define INT16_MAX              (32767)
+#endif
+#ifndef INT32_MAX
+#define INT32_MAX              (2147483647)
+#endif
+#ifndef UINT8_MAX
+#define UINT8_MAX              (255U)
+#endif
+#ifndef UINT16_MAX
+#define UINT16_MAX             (65535U)
+#endif
+#ifndef UINT32_MAX
+#define UINT32_MAX             (4294967295U)
+#endif
+
+#endif /* ! C99 */
+
+#endif /* ! FLEXINT_H */

contrib/flex/libmain.c

@@ -0,0 +1,33 @@
+/* libmain - flex run-time support library "main" function */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+extern int yylex ();
+
+int     main (argc, argv)
+     int     argc;
+     char   *argv[];
+{
+	while (yylex () != 0) ;
+
+	return 0;
+}

contrib/flex/libyywrap.c

@@ -0,0 +1,27 @@
+/* libyywrap - flex run-time support library "yywrap" function */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+int     yywrap (void)
+{
+	return 1;
+}

contrib/flex/mkskel.sh

@@ -0,0 +1,37 @@
+#! /bin/sh
+
+#  This file is part of flex.
+
+#  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.
+
+#  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR
+#  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+#  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+#  PURPOSE.
+
+cat <<!
+/* File created from flex.skl via mkskel.sh */
+
+#include "flexdef.h"
+
+const char *skel[] = {
+!
+
+sed 's/\\/&&/g' | sed 's/"/\\"/g' | sed 's/.*/  "&",/'
+
+cat <<!
+  0
+};
+!

contrib/flex/nfa.c

@@ -0,0 +1,708 @@
+/* nfa - NFA construction routines */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "flexdef.h"
+
+
+/* declare functions that have forward references */
+
+int dupmachine PROTO ((int));
+void mkxtion PROTO ((int, int));
+
+
+/* add_accept - add an accepting state to a machine
+ *
+ * accepting_number becomes mach's accepting number.
+ */
+
+void    add_accept (mach, accepting_number)
+     int     mach, accepting_number;
+{
+	/* Hang the accepting number off an epsilon state.  if it is associated
+	 * with a state that has a non-epsilon out-transition, then the state
+	 * will accept BEFORE it makes that transition, i.e., one character
+	 * too soon.
+	 */
+
+	if (transchar[finalst[mach]] == SYM_EPSILON)
+		accptnum[finalst[mach]] = accepting_number;
+
+	else {
+		int     astate = mkstate (SYM_EPSILON);
+
+		accptnum[astate] = accepting_number;
+		(void) link_machines (mach, astate);
+	}
+}
+
+
+/* copysingl - make a given number of copies of a singleton machine
+ *
+ * synopsis
+ *
+ *   newsng = copysingl( singl, num );
+ *
+ *     newsng - a new singleton composed of num copies of singl
+ *     singl  - a singleton machine
+ *     num    - the number of copies of singl to be present in newsng
+ */
+
+int     copysingl (singl, num)
+     int     singl, num;
+{
+	int     copy, i;
+
+	copy = mkstate (SYM_EPSILON);
+
+	for (i = 1; i <= num; ++i)
+		copy = link_machines (copy, dupmachine (singl));
+
+	return copy;
+}
+
+
+/* dumpnfa - debugging routine to write out an nfa */
+
+void    dumpnfa (state1)
+     int     state1;
+
+{
+	int     sym, tsp1, tsp2, anum, ns;
+
+	fprintf (stderr,
+		 _
+		 ("\n\n********** beginning dump of nfa with start state %d\n"),
+		 state1);
+
+	/* We probably should loop starting at firstst[state1] and going to
+	 * lastst[state1], but they're not maintained properly when we "or"
+	 * all of the rules together.  So we use our knowledge that the machine
+	 * starts at state 1 and ends at lastnfa.
+	 */
+
+	/* for ( ns = firstst[state1]; ns <= lastst[state1]; ++ns ) */
+	for (ns = 1; ns <= lastnfa; ++ns) {
+		fprintf (stderr, _("state # %4d\t"), ns);
+
+		sym = transchar[ns];
+		tsp1 = trans1[ns];
+		tsp2 = trans2[ns];
+		anum = accptnum[ns];
+
+		fprintf (stderr, "%3d:  %4d, %4d", sym, tsp1, tsp2);
+
+		if (anum != NIL)
+			fprintf (stderr, "  [%d]", anum);
+
+		fprintf (stderr, "\n");
+	}
+
+	fprintf (stderr, _("********** end of dump\n"));
+}
+
+
+/* dupmachine - make a duplicate of a given machine
+ *
+ * synopsis
+ *
+ *   copy = dupmachine( mach );
+ *
+ *     copy - holds duplicate of mach
+ *     mach - machine to be duplicated
+ *
+ * note that the copy of mach is NOT an exact duplicate; rather, all the
+ * transition states values are adjusted so that the copy is self-contained,
+ * as the original should have been.
+ *
+ * also note that the original MUST be contiguous, with its low and high
+ * states accessible by the arrays firstst and lastst
+ */
+
+int     dupmachine (mach)
+     int     mach;
+{
+	int     i, init, state_offset;
+	int     state = 0;
+	int     last = lastst[mach];
+
+	for (i = firstst[mach]; i <= last; ++i) {
+		state = mkstate (transchar[i]);
+
+		if (trans1[i] != NO_TRANSITION) {
+			mkxtion (finalst[state], trans1[i] + state - i);
+
+			if (transchar[i] == SYM_EPSILON &&
+			    trans2[i] != NO_TRANSITION)
+					mkxtion (finalst[state],
+						 trans2[i] + state - i);
+		}
+
+		accptnum[state] = accptnum[i];
+	}
+
+	if (state == 0)
+		flexfatal (_("empty machine in dupmachine()"));
+
+	state_offset = state - i + 1;
+
+	init = mach + state_offset;
+	firstst[init] = firstst[mach] + state_offset;
+	finalst[init] = finalst[mach] + state_offset;
+	lastst[init] = lastst[mach] + state_offset;
+
+	return init;
+}
+
+
+/* finish_rule - finish up the processing for a rule
+ *
+ * An accepting number is added to the given machine.  If variable_trail_rule
+ * is true then the rule has trailing context and both the head and trail
+ * are variable size.  Otherwise if headcnt or trailcnt is non-zero then
+ * the machine recognizes a pattern with trailing context and headcnt is
+ * the number of characters in the matched part of the pattern, or zero
+ * if the matched part has variable length.  trailcnt is the number of
+ * trailing context characters in the pattern, or zero if the trailing
+ * context has variable length.
+ */
+
+void    finish_rule (mach, variable_trail_rule, headcnt, trailcnt,
+		     pcont_act)
+     int     mach, variable_trail_rule, headcnt, trailcnt, pcont_act;
+{
+	char    action_text[MAXLINE];
+
+	add_accept (mach, num_rules);
+
+	/* We did this in new_rule(), but it often gets the wrong
+	 * number because we do it before we start parsing the current rule.
+	 */
+	rule_linenum[num_rules] = linenum;
+
+	/* If this is a continued action, then the line-number has already
+	 * been updated, giving us the wrong number.
+	 */
+	if (continued_action)
+		--rule_linenum[num_rules];
+
+
+	/* If the previous rule was continued action, then we inherit the
+	 * previous newline flag, possibly overriding the current one.
+	 */
+	if (pcont_act && rule_has_nl[num_rules - 1])
+		rule_has_nl[num_rules] = true;
+
+	snprintf (action_text, sizeof(action_text), "case %d:\n", num_rules);
+	add_action (action_text);
+	if (rule_has_nl[num_rules]) {
+		snprintf (action_text, sizeof(action_text), "/* rule %d can match eol */\n",
+			 num_rules);
+		add_action (action_text);
+	}
+
+
+	if (variable_trail_rule) {
+		rule_type[num_rules] = RULE_VARIABLE;
+
+		if (performance_report > 0)
+			fprintf (stderr,
+				 _
+				 ("Variable trailing context rule at line %d\n"),
+				 rule_linenum[num_rules]);
+
+		variable_trailing_context_rules = true;
+	}
+
+	else {
+		rule_type[num_rules] = RULE_NORMAL;
+
+		if (headcnt > 0 || trailcnt > 0) {
+			/* Do trailing context magic to not match the trailing
+			 * characters.
+			 */
+			char   *scanner_cp = "YY_G(yy_c_buf_p) = yy_cp";
+			char   *scanner_bp = "yy_bp";
+
+			add_action
+				("*yy_cp = YY_G(yy_hold_char); /* undo effects of setting up yytext */\n");
+
+			if (headcnt > 0) {
+				snprintf (action_text, sizeof(action_text), "%s = %s + %d;\n",
+					 scanner_cp, scanner_bp, headcnt);
+				add_action (action_text);
+			}
+
+			else {
+				snprintf (action_text, sizeof(action_text), "%s -= %d;\n",
+					 scanner_cp, trailcnt);
+				add_action (action_text);
+			}
+
+			add_action
+				("YY_DO_BEFORE_ACTION; /* set up yytext again */\n");
+		}
+	}
+
+	/* Okay, in the action code at this point yytext and yyleng have
+	 * their proper final values for this rule, so here's the point
+	 * to do any user action.  But don't do it for continued actions,
+	 * as that'll result in multiple YY_RULE_SETUP's.
+	 */
+	if (!continued_action)
+		add_action ("YY_RULE_SETUP\n");
+
+	line_directive_out ((FILE *) 0, 1);
+}
+
+
+/* link_machines - connect two machines together
+ *
+ * synopsis
+ *
+ *   new = link_machines( first, last );
+ *
+ *     new    - a machine constructed by connecting first to last
+ *     first  - the machine whose successor is to be last
+ *     last   - the machine whose predecessor is to be first
+ *
+ * note: this routine concatenates the machine first with the machine
+ *  last to produce a machine new which will pattern-match first first
+ *  and then last, and will fail if either of the sub-patterns fails.
+ *  FIRST is set to new by the operation.  last is unmolested.
+ */
+
+int     link_machines (first, last)
+     int     first, last;
+{
+	if (first == NIL)
+		return last;
+
+	else if (last == NIL)
+		return first;
+
+	else {
+		mkxtion (finalst[first], last);
+		finalst[first] = finalst[last];
+		lastst[first] = MAX (lastst[first], lastst[last]);
+		firstst[first] = MIN (firstst[first], firstst[last]);
+
+		return first;
+	}
+}
+
+
+/* mark_beginning_as_normal - mark each "beginning" state in a machine
+ *                            as being a "normal" (i.e., not trailing context-
+ *                            associated) states
+ *
+ * The "beginning" states are the epsilon closure of the first state
+ */
+
+void    mark_beginning_as_normal (mach)
+     register int mach;
+{
+	switch (state_type[mach]) {
+	case STATE_NORMAL:
+		/* Oh, we've already visited here. */
+		return;
+
+	case STATE_TRAILING_CONTEXT:
+		state_type[mach] = STATE_NORMAL;
+
+		if (transchar[mach] == SYM_EPSILON) {
+			if (trans1[mach] != NO_TRANSITION)
+				mark_beginning_as_normal (trans1[mach]);
+
+			if (trans2[mach] != NO_TRANSITION)
+				mark_beginning_as_normal (trans2[mach]);
+		}
+		break;
+
+	default:
+		flexerror (_
+			   ("bad state type in mark_beginning_as_normal()"));
+		break;
+	}
+}
+
+
+/* mkbranch - make a machine that branches to two machines
+ *
+ * synopsis
+ *
+ *   branch = mkbranch( first, second );
+ *
+ *     branch - a machine which matches either first's pattern or second's
+ *     first, second - machines whose patterns are to be or'ed (the | operator)
+ *
+ * Note that first and second are NEITHER destroyed by the operation.  Also,
+ * the resulting machine CANNOT be used with any other "mk" operation except
+ * more mkbranch's.  Compare with mkor()
+ */
+
+int     mkbranch (first, second)
+     int     first, second;
+{
+	int     eps;
+
+	if (first == NO_TRANSITION)
+		return second;
+
+	else if (second == NO_TRANSITION)
+		return first;
+
+	eps = mkstate (SYM_EPSILON);
+
+	mkxtion (eps, first);
+	mkxtion (eps, second);
+
+	return eps;
+}
+
+
+/* mkclos - convert a machine into a closure
+ *
+ * synopsis
+ *   new = mkclos( state );
+ *
+ * new - a new state which matches the closure of "state"
+ */
+
+int     mkclos (state)
+     int     state;
+{
+	return mkopt (mkposcl (state));
+}
+
+
+/* mkopt - make a machine optional
+ *
+ * synopsis
+ *
+ *   new = mkopt( mach );
+ *
+ *     new  - a machine which optionally matches whatever mach matched
+ *     mach - the machine to make optional
+ *
+ * notes:
+ *     1. mach must be the last machine created
+ *     2. mach is destroyed by the call
+ */
+
+int     mkopt (mach)
+     int     mach;
+{
+	int     eps;
+
+	if (!SUPER_FREE_EPSILON (finalst[mach])) {
+		eps = mkstate (SYM_EPSILON);
+		mach = link_machines (mach, eps);
+	}
+
+	/* Can't skimp on the following if FREE_EPSILON(mach) is true because
+	 * some state interior to "mach" might point back to the beginning
+	 * for a closure.
+	 */
+	eps = mkstate (SYM_EPSILON);
+	mach = link_machines (eps, mach);
+
+	mkxtion (mach, finalst[mach]);
+
+	return mach;
+}
+
+
+/* mkor - make a machine that matches either one of two machines
+ *
+ * synopsis
+ *
+ *   new = mkor( first, second );
+ *
+ *     new - a machine which matches either first's pattern or second's
+ *     first, second - machines whose patterns are to be or'ed (the | operator)
+ *
+ * note that first and second are both destroyed by the operation
+ * the code is rather convoluted because an attempt is made to minimize
+ * the number of epsilon states needed
+ */
+
+int     mkor (first, second)
+     int     first, second;
+{
+	int     eps, orend;
+
+	if (first == NIL)
+		return second;
+
+	else if (second == NIL)
+		return first;
+
+	else {
+		/* See comment in mkopt() about why we can't use the first
+		 * state of "first" or "second" if they satisfy "FREE_EPSILON".
+		 */
+		eps = mkstate (SYM_EPSILON);
+
+		first = link_machines (eps, first);
+
+		mkxtion (first, second);
+
+		if (SUPER_FREE_EPSILON (finalst[first]) &&
+		    accptnum[finalst[first]] == NIL) {
+			orend = finalst[first];
+			mkxtion (finalst[second], orend);
+		}
+
+		else if (SUPER_FREE_EPSILON (finalst[second]) &&
+			 accptnum[finalst[second]] == NIL) {
+			orend = finalst[second];
+			mkxtion (finalst[first], orend);
+		}
+
+		else {
+			eps = mkstate (SYM_EPSILON);
+
+			first = link_machines (first, eps);
+			orend = finalst[first];
+
+			mkxtion (finalst[second], orend);
+		}
+	}
+
+	finalst[first] = orend;
+	return first;
+}
+
+
+/* mkposcl - convert a machine into a positive closure
+ *
+ * synopsis
+ *   new = mkposcl( state );
+ *
+ *    new - a machine matching the positive closure of "state"
+ */
+
+int     mkposcl (state)
+     int     state;
+{
+	int     eps;
+
+	if (SUPER_FREE_EPSILON (finalst[state])) {
+		mkxtion (finalst[state], state);
+		return state;
+	}
+
+	else {
+		eps = mkstate (SYM_EPSILON);
+		mkxtion (eps, state);
+		return link_machines (state, eps);
+	}
+}
+
+
+/* mkrep - make a replicated machine
+ *
+ * synopsis
+ *   new = mkrep( mach, lb, ub );
+ *
+ *    new - a machine that matches whatever "mach" matched from "lb"
+ *          number of times to "ub" number of times
+ *
+ * note
+ *   if "ub" is INFINITE_REPEAT then "new" matches "lb" or more occurrences of "mach"
+ */
+
+int     mkrep (mach, lb, ub)
+     int     mach, lb, ub;
+{
+	int     base_mach, tail, copy, i;
+
+	base_mach = copysingl (mach, lb - 1);
+
+	if (ub == INFINITE_REPEAT) {
+		copy = dupmachine (mach);
+		mach = link_machines (mach,
+				      link_machines (base_mach,
+						     mkclos (copy)));
+	}
+
+	else {
+		tail = mkstate (SYM_EPSILON);
+
+		for (i = lb; i < ub; ++i) {
+			copy = dupmachine (mach);
+			tail = mkopt (link_machines (copy, tail));
+		}
+
+		mach =
+			link_machines (mach,
+				       link_machines (base_mach, tail));
+	}
+
+	return mach;
+}
+
+
+/* mkstate - create a state with a transition on a given symbol
+ *
+ * synopsis
+ *
+ *   state = mkstate( sym );
+ *
+ *     state - a new state matching sym
+ *     sym   - the symbol the new state is to have an out-transition on
+ *
+ * note that this routine makes new states in ascending order through the
+ * state array (and increments LASTNFA accordingly).  The routine DUPMACHINE
+ * relies on machines being made in ascending order and that they are
+ * CONTIGUOUS.  Change it and you will have to rewrite DUPMACHINE (kludge
+ * that it admittedly is)
+ */
+
+int     mkstate (sym)
+     int     sym;
+{
+	if (++lastnfa >= current_mns) {
+		if ((current_mns += MNS_INCREMENT) >= maximum_mns)
+			lerrif (_
+				("input rules are too complicated (>= %d NFA states)"),
+current_mns);
+
+		++num_reallocs;
+
+		firstst = reallocate_integer_array (firstst, current_mns);
+		lastst = reallocate_integer_array (lastst, current_mns);
+		finalst = reallocate_integer_array (finalst, current_mns);
+		transchar =
+			reallocate_integer_array (transchar, current_mns);
+		trans1 = reallocate_integer_array (trans1, current_mns);
+		trans2 = reallocate_integer_array (trans2, current_mns);
+		accptnum =
+			reallocate_integer_array (accptnum, current_mns);
+		assoc_rule =
+			reallocate_integer_array (assoc_rule, current_mns);
+		state_type =
+			reallocate_integer_array (state_type, current_mns);
+	}
+
+	firstst[lastnfa] = lastnfa;
+	finalst[lastnfa] = lastnfa;
+	lastst[lastnfa] = lastnfa;
+	transchar[lastnfa] = sym;
+	trans1[lastnfa] = NO_TRANSITION;
+	trans2[lastnfa] = NO_TRANSITION;
+	accptnum[lastnfa] = NIL;
+	assoc_rule[lastnfa] = num_rules;
+	state_type[lastnfa] = current_state_type;
+
+	/* Fix up equivalence classes base on this transition.  Note that any
+	 * character which has its own transition gets its own equivalence
+	 * class.  Thus only characters which are only in character classes
+	 * have a chance at being in the same equivalence class.  E.g. "a|b"
+	 * puts 'a' and 'b' into two different equivalence classes.  "[ab]"
+	 * puts them in the same equivalence class (barring other differences
+	 * elsewhere in the input).
+	 */
+
+	if (sym < 0) {
+		/* We don't have to update the equivalence classes since
+		 * that was already done when the ccl was created for the
+		 * first time.
+		 */
+	}
+
+	else if (sym == SYM_EPSILON)
+		++numeps;
+
+	else {
+		check_char (sym);
+
+		if (useecs)
+			/* Map NUL's to csize. */
+			mkechar (sym ? sym : csize, nextecm, ecgroup);
+	}
+
+	return lastnfa;
+}
+
+
+/* mkxtion - make a transition from one state to another
+ *
+ * synopsis
+ *
+ *   mkxtion( statefrom, stateto );
+ *
+ *     statefrom - the state from which the transition is to be made
+ *     stateto   - the state to which the transition is to be made
+ */
+
+void    mkxtion (statefrom, stateto)
+     int     statefrom, stateto;
+{
+	if (trans1[statefrom] == NO_TRANSITION)
+		trans1[statefrom] = stateto;
+
+	else if ((transchar[statefrom] != SYM_EPSILON) ||
+		 (trans2[statefrom] != NO_TRANSITION))
+		flexfatal (_("found too many transitions in mkxtion()"));
+
+	else {			/* second out-transition for an epsilon state */
+		++eps2;
+		trans2[statefrom] = stateto;
+	}
+}
+
+/* new_rule - initialize for a new rule */
+
+void    new_rule ()
+{
+	if (++num_rules >= current_max_rules) {
+		++num_reallocs;
+		current_max_rules += MAX_RULES_INCREMENT;
+		rule_type = reallocate_integer_array (rule_type,
+						      current_max_rules);
+		rule_linenum = reallocate_integer_array (rule_linenum,
+							 current_max_rules);
+		rule_useful = reallocate_integer_array (rule_useful,
+							current_max_rules);
+		rule_has_nl = reallocate_bool_array (rule_has_nl,
+						     current_max_rules);
+	}
+
+	if (num_rules > MAX_RULE)
+		lerrif (_("too many rules (> %d)!"), MAX_RULE);
+
+	rule_linenum[num_rules] = linenum;
+	rule_useful[num_rules] = false;
+	rule_has_nl[num_rules] = false;
+}

contrib/flex/options.c

@@ -0,0 +1,280 @@
+/* flex - tool to generate fast lexical analyzers */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "options.h"
+
+/* Be sure to synchronize these options with those defined in "options.h",
+ * the giant switch() statement in "main.c", and the %option processing in
+ * "scan.l".
+ */
+
+
+/* The command-line options, passed to scanopt_init() */
+optspec_t flexopts[] = {
+
+	{"-7", OPT_7BIT, 0}
+	,
+	{"--7bit", OPT_7BIT, 0}
+	,			/* Generate 7-bit scanner. */
+	{"-8", OPT_8BIT, 0}
+	,
+	{"--8bit", OPT_8BIT, 0}
+	,			/* Generate 8-bit scanner. */
+	{"--align", OPT_ALIGN, 0}
+	,			/* Trade off larger tables for better memory alignment. */
+	{"--noalign", OPT_NO_ALIGN, 0}
+	,
+	{"--always-interactive", OPT_ALWAYS_INTERACTIVE, 0}
+	,
+	{"--array", OPT_ARRAY, 0}
+	,
+	{"-b", OPT_BACKUP, 0}
+	,
+	{"--backup", OPT_BACKUP, 0}
+	,			/* Generate backing-up information to lex.backup. */
+	{"-B", OPT_BATCH, 0}
+	,
+	{"--batch", OPT_BATCH, 0}
+	,			/* Generate batch scanner (opposite of -I). */
+	{"--bison-bridge", OPT_BISON_BRIDGE, 0}
+	,			/* Scanner to be called by a bison pure parser. */
+	{"--bison-locations", OPT_BISON_BRIDGE_LOCATIONS, 0}
+	,			/* Scanner to be called by a bison pure parser. */
+	{"-i", OPT_CASE_INSENSITIVE, 0}
+	,
+	{"--case-insensitive", OPT_CASE_INSENSITIVE, 0}
+	,			/* Generate case-insensitive scanner. */
+	
+		{"-C[aefFmr]", OPT_COMPRESSION,
+	 "Specify degree of table compression (default is -Cem)"},
+	{"-+", OPT_CPLUSPLUS, 0}
+	,
+	{"--c++", OPT_CPLUSPLUS, 0}
+	,			/* Generate C++ scanner class. */
+	{"-d", OPT_DEBUG, 0}
+	,
+	{"--debug", OPT_DEBUG, 0}
+	,			/* Turn on debug mode in generated scanner. */
+	{"--nodebug", OPT_NO_DEBUG, 0}
+	,
+	{"-s", OPT_NO_DEFAULT, 0}
+	,
+	{"--nodefault", OPT_NO_DEFAULT, 0}
+	,			/* Suppress default rule to ECHO unmatched text. */
+	{"--default", OPT_DEFAULT, 0}
+	,
+	{"-c", OPT_DONOTHING, 0}
+	,			/* For POSIX lex compatibility. */
+	{"-n", OPT_DONOTHING, 0}
+	,			/* For POSIX lex compatibility. */
+	{"--ecs", OPT_ECS, 0}
+	,			/* Construct equivalence classes. */
+	{"--noecs", OPT_NO_ECS, 0}
+	,
+	{"-F", OPT_FAST, 0}
+	,
+	{"--fast", OPT_FAST, 0}
+	,			/* Same as -CFr. */
+	{"-f", OPT_FULL, 0}
+	,
+	{"--full", OPT_FULL, 0}
+	,			/* Same as -Cfr. */
+	{"--header-file[=FILE]", OPT_HEADER_FILE, 0}
+	,
+	{"-?", OPT_HELP, 0}
+	,
+	{"-h", OPT_HELP, 0}
+	,
+	{"--help", OPT_HELP, 0}
+	,			/* Produce this help message. */
+	{"-I", OPT_INTERACTIVE, 0}
+	,
+	{"--interactive", OPT_INTERACTIVE, 0}
+	,			/* Generate interactive scanner (opposite of -B). */
+	{"-l", OPT_LEX_COMPAT, 0}
+	,
+	{"--lex-compat", OPT_LEX_COMPAT, 0}
+	,			/* Maximal compatibility with original lex. */
+	{"-X", OPT_POSIX_COMPAT, 0}
+	,
+	{"--posix-compat", OPT_POSIX_COMPAT, 0}
+	,			/* Maximal compatibility with POSIX lex. */
+        {"--preproc=NUM", OPT_PREPROC_LEVEL, 0}
+        ,
+	{"-L", OPT_NO_LINE, 0}
+	,			/* Suppress #line directives in scanner. */
+	{"--noline", OPT_NO_LINE, 0}
+	,			/* Suppress #line directives in scanner. */
+	{"--main", OPT_MAIN, 0}
+	,			/* use built-in main() function. */
+	{"--nomain", OPT_NO_MAIN, 0}
+	,
+	{"--meta-ecs", OPT_META_ECS, 0}
+	,			/* Construct meta-equivalence classes. */
+	{"--nometa-ecs", OPT_NO_META_ECS, 0}
+	,
+	{"--never-interactive", OPT_NEVER_INTERACTIVE, 0}
+	,
+	{"-o FILE", OPT_OUTFILE, 0}
+	,
+	{"--outfile=FILE", OPT_OUTFILE, 0}
+	,			/* Write to FILE (default is lex.yy.c) */
+	{"-p", OPT_PERF_REPORT, 0}
+	,
+	{"--perf-report", OPT_PERF_REPORT, 0}
+	,			/* Generate performance report to stderr. */
+	{"--pointer", OPT_POINTER, 0}
+	,
+	{"-P PREFIX", OPT_PREFIX, 0}
+	,
+	{"--prefix=PREFIX", OPT_PREFIX, 0}
+	,			/* Use PREFIX (default is yy) */
+	{"-Dmacro", OPT_PREPROCDEFINE, 0}
+	,			/* Define a preprocessor symbol. */
+	{"--read", OPT_READ, 0}
+	,			/* Use read(2) instead of stdio. */
+	{"-R", OPT_REENTRANT, 0}
+	,
+	{"--reentrant", OPT_REENTRANT, 0}
+	,			/* Generate a reentrant C scanner. */
+	{"--noreentrant", OPT_NO_REENTRANT, 0}
+	,
+	{"--reject", OPT_REJECT, 0}
+	,
+	{"--noreject", OPT_NO_REJECT, 0}
+	,
+	{"-S FILE", OPT_SKEL, 0}
+	,
+	{"--skel=FILE", OPT_SKEL, 0}
+	,			/* Use skeleton from FILE */
+	{"--stack", OPT_STACK, 0}
+	,
+	{"--stdinit", OPT_STDINIT, 0}
+	,
+	{"--nostdinit", OPT_NO_STDINIT, 0}
+	,
+	{"-t", OPT_STDOUT, 0}
+	,
+	{"--stdout", OPT_STDOUT, 0}
+	,			/* Write generated scanner to stdout. */
+	{"-T", OPT_TRACE, 0}
+	,
+	{"--trace", OPT_TRACE, 0}
+	,			/* Flex should run in trace mode. */
+	{"--tables-file[=FILE]", OPT_TABLES_FILE, 0}
+	,			/* Save tables to FILE */
+        {"--tables-verify", OPT_TABLES_VERIFY, 0}
+        ,                       /* Tables integrity check */
+	{"--nounistd", OPT_NO_UNISTD_H, 0}
+	,			/* Do not include unistd.h */
+	{"-v", OPT_VERBOSE, 0}
+	,
+	{"--verbose", OPT_VERBOSE, 0}
+	,			/* Write summary of scanner statistics to stdout. */
+	{"-V", OPT_VERSION, 0}
+	,
+	{"--version", OPT_VERSION, 0}
+	,			/* Report flex version. */
+	{"--warn", OPT_WARN, 0}
+	,
+	{"-w", OPT_NO_WARN, 0}
+	,
+	{"--nowarn", OPT_NO_WARN, 0}
+	,			/* Suppress warning messages. */
+	{"--noansi-definitions", OPT_NO_ANSI_FUNC_DEFS, 0}
+	,
+	{"--noansi-prototypes", OPT_NO_ANSI_FUNC_PROTOS, 0}
+	,
+	{"--yyclass=NAME", OPT_YYCLASS, 0}
+	,
+	{"--yylineno", OPT_YYLINENO, 0}
+	,
+	{"--noyylineno", OPT_NO_YYLINENO, 0}
+	,
+
+	{"--yymore", OPT_YYMORE, 0}
+	,
+	{"--noyymore", OPT_NO_YYMORE, 0}
+	,
+	{"--noyywrap", OPT_NO_YYWRAP, 0}
+	,
+	{"--yywrap", OPT_YYWRAP, 0}
+	,
+
+	{"--nounput", OPT_NO_UNPUT, 0}
+	,
+	{"--noyy_push_state", OPT_NO_YY_PUSH_STATE, 0}
+	,
+	{"--noyy_pop_state", OPT_NO_YY_POP_STATE, 0}
+	,
+	{"--noyy_top_state", OPT_NO_YY_TOP_STATE, 0}
+	,
+	{"--noyy_scan_buffer", OPT_NO_YY_SCAN_BUFFER, 0}
+	,
+	{"--noyy_scan_bytes", OPT_NO_YY_SCAN_BYTES, 0}
+	,
+	{"--noyy_scan_string", OPT_NO_YY_SCAN_STRING, 0}
+	,
+	{"--noyyget_extra", OPT_NO_YYGET_EXTRA, 0}
+	,
+	{"--noyyset_extra", OPT_NO_YYSET_EXTRA, 0}
+	,
+	{"--noyyget_leng", OPT_NO_YYGET_LENG, 0}
+	,
+	{"--noyyget_text", OPT_NO_YYGET_TEXT, 0}
+	,
+	{"--noyyget_lineno", OPT_NO_YYGET_LINENO, 0}
+	,
+	{"--noyyset_lineno", OPT_NO_YYSET_LINENO, 0}
+	,
+	{"--noyyget_in", OPT_NO_YYGET_IN, 0}
+	,
+	{"--noyyset_in", OPT_NO_YYSET_IN, 0}
+	,
+	{"--noyyget_out", OPT_NO_YYGET_OUT, 0}
+	,
+	{"--noyyset_out", OPT_NO_YYSET_OUT, 0}
+	,
+	{"--noyyget_lval", OPT_NO_YYGET_LVAL, 0}
+	,
+	{"--noyyset_lval", OPT_NO_YYSET_LVAL, 0}
+	,
+	{"--noyyget_lloc", OPT_NO_YYGET_LLOC, 0}
+	,
+	{"--noyyset_lloc", OPT_NO_YYSET_LLOC, 0}
+	,
+
+	{0, 0, 0}		/* required final NULL entry. */
+};
+
+/* vim:set tabstop=8 softtabstop=4 shiftwidth=4: */

contrib/flex/options.h

@@ -0,0 +1,134 @@
+/* flex - tool to generate fast lexical analyzers */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#ifndef OPTIONS_H
+#define OPTIONS_H
+#include "scanopt.h"
+
+extern optspec_t flexopts[];
+
+enum flexopt_flag_t {
+	/* Use positive integers only, since they are return codes for scanopt.
+	 * Order is not important. */
+	OPT_7BIT = 1,
+	OPT_8BIT,
+	OPT_ALIGN,
+	OPT_ALWAYS_INTERACTIVE,
+	OPT_ARRAY,
+	OPT_BACKUP,
+	OPT_BATCH,
+        OPT_BISON_BRIDGE,
+        OPT_BISON_BRIDGE_LOCATIONS,
+	OPT_CASE_INSENSITIVE,
+	OPT_COMPRESSION,
+	OPT_CPLUSPLUS,
+	OPT_DEBUG,
+	OPT_DEFAULT,
+	OPT_DONOTHING,
+	OPT_ECS,
+	OPT_FAST,
+	OPT_FULL,
+	OPT_HEADER_FILE,
+	OPT_HELP,
+	OPT_INTERACTIVE,
+	OPT_LEX_COMPAT,
+	OPT_POSIX_COMPAT,
+	OPT_MAIN,
+	OPT_META_ECS,
+	OPT_NEVER_INTERACTIVE,
+	OPT_NO_ALIGN,
+        OPT_NO_ANSI_FUNC_DEFS,
+        OPT_NO_ANSI_FUNC_PROTOS,
+	OPT_NO_DEBUG,
+	OPT_NO_DEFAULT,
+	OPT_NO_ECS,
+	OPT_NO_LINE,
+	OPT_NO_MAIN,
+	OPT_NO_META_ECS,
+	OPT_NO_REENTRANT,
+	OPT_NO_REJECT,
+	OPT_NO_STDINIT,
+	OPT_NO_UNPUT,
+	OPT_NO_WARN,
+	OPT_NO_YYGET_EXTRA,
+	OPT_NO_YYGET_IN,
+	OPT_NO_YYGET_LENG,
+	OPT_NO_YYGET_LINENO,
+	OPT_NO_YYGET_LLOC,
+	OPT_NO_YYGET_LVAL,
+	OPT_NO_YYGET_OUT,
+	OPT_NO_YYGET_TEXT,
+	OPT_NO_YYLINENO,
+	OPT_NO_YYMORE,
+	OPT_NO_YYSET_EXTRA,
+	OPT_NO_YYSET_IN,
+	OPT_NO_YYSET_LINENO,
+	OPT_NO_YYSET_LLOC,
+	OPT_NO_YYSET_LVAL,
+	OPT_NO_YYSET_OUT,
+	OPT_NO_YYWRAP,
+	OPT_NO_YY_POP_STATE,
+	OPT_NO_YY_PUSH_STATE,
+	OPT_NO_YY_SCAN_BUFFER,
+	OPT_NO_YY_SCAN_BYTES,
+	OPT_NO_YY_SCAN_STRING,
+	OPT_NO_YY_TOP_STATE,
+	OPT_OUTFILE,
+	OPT_PERF_REPORT,
+	OPT_POINTER,
+	OPT_PREFIX,
+	OPT_PREPROCDEFINE,
+	OPT_PREPROC_LEVEL,
+	OPT_READ,
+	OPT_REENTRANT,
+	OPT_REJECT,
+	OPT_SKEL,
+	OPT_STACK,
+	OPT_STDINIT,
+	OPT_STDOUT,
+	OPT_TABLES_FILE,
+	OPT_TABLES_VERIFY,
+	OPT_TRACE,
+	OPT_NO_UNISTD_H,
+	OPT_VERBOSE,
+	OPT_VERSION,
+	OPT_WARN,
+	OPT_YYCLASS,
+	OPT_YYLINENO,
+	OPT_YYMORE,
+	OPT_YYWRAP
+};
+
+#endif
+
+/* vim:set tabstop=8 softtabstop=4 shiftwidth=4 textwidth=0: */

contrib/flex/regex.c

@@ -0,0 +1,172 @@
+/** regex - regular expression functions related to POSIX regex lib. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "flexdef.h"
+
+
+static const char* REGEXP_LINEDIR = "^#line ([[:digit:]]+) \"(.*)\"";
+static const char* REGEXP_BLANK_LINE = "^[[:space:]]*$";
+
+regex_t regex_linedir; /**< matches line directives */
+regex_t regex_blank_line; /**< matches blank lines */
+
+
+/** Initialize the regular expressions.
+ * @return true upon success.
+ */
+bool flex_init_regex(void)
+{
+    flex_regcomp(&regex_linedir, REGEXP_LINEDIR, REG_EXTENDED);
+    flex_regcomp(&regex_blank_line, REGEXP_BLANK_LINE, REG_EXTENDED);
+
+    return true;
+}
+
+/** Compiles a regular expression or dies trying.
+ * @param preg  Same as for regcomp().
+ * @param regex Same as for regcomp().
+ * @param cflags Same as for regcomp().
+ */
+void flex_regcomp(regex_t *preg, const char *regex, int cflags)
+{
+    int err;
+
+	memset (preg, 0, sizeof (regex_t));
+
+	if ((err = regcomp (preg, regex, cflags)) != 0) {
+        const int errbuf_sz = 200;
+        char *errbuf, *rxerr;
+
+		errbuf = (char*)flex_alloc(errbuf_sz *sizeof(char));
+		if (!errbuf)
+			flexfatal(_("Unable to allocate buffer to report regcomp"));
+		rxerr = (char*)flex_alloc(errbuf_sz *sizeof(char));
+		if (!rxerr)
+			flexfatal(_("Unable to allocate buffer for regerror"));
+		regerror (err, preg, rxerr, errbuf_sz);
+		snprintf (errbuf, errbuf_sz, "regcomp for \"%s\" failed: %s", regex, rxerr);
+
+		flexfatal (errbuf);
+        free(errbuf);
+        free(rxerr);
+	}
+}
+
+/** Extract a copy of the match, or NULL if no match.
+ * @param m A match as returned by regexec().
+ * @param src The source string that was passed to regexec().
+ * @return The allocated string.
+ */
+char   *regmatch_dup (regmatch_t * m, const char *src)
+{
+	char   *str;
+	int     len;
+
+	if (m == NULL || m->rm_so < 0)
+		return NULL;
+	len = m->rm_eo - m->rm_so;
+	str = (char *) flex_alloc ((len + 1) * sizeof (char));
+	if (!str)
+		flexfatal(_("Unable to allocate a copy of the match"));
+	strncpy (str, src + m->rm_so, len);
+	str[len] = 0;
+	return str;
+}
+
+/** Copy the match.
+ * @param m A match as returned by regexec().
+ * @param dest The destination buffer.
+ * @param src The source string that was passed to regexec().
+ * @return dest
+ */
+char   *regmatch_cpy (regmatch_t * m, char *dest, const char *src)
+{
+	if (m == NULL || m->rm_so < 0) {
+		if (dest)
+			dest[0] = '\0';
+		return dest;
+	}
+
+	snprintf (dest, regmatch_len(m), "%s", src + m->rm_so);
+    return dest;
+}
+
+/** Get the length in characters of the match.
+ * @param m A match as returned by regexec().
+ * @param src The source string that was passed to regexec().
+ * @return The length of the match.
+ */
+int regmatch_len (regmatch_t * m)
+{
+	if (m == NULL || m->rm_so < 0) {
+		return 0;
+	}
+
+	return m->rm_eo - m->rm_so;
+}
+
+
+
+/** Convert a regmatch_t object to an integer using the strtol() function.
+ * @param m A match as returned by regexec().
+ * @param src The source string that was passed to regexec().
+ * @param endptr Same as the second argument to strtol().
+ * @param base   Same as the third argument to strtol().
+ * @return The converted integer or error (Return value is the same as for strtol()).
+ */
+int regmatch_strtol (regmatch_t * m, const char *src, char **endptr,
+		     int base)
+{
+	int     n = 0;
+
+#define bufsz 20
+	char    buf[bufsz];
+	char   *s;
+
+	if (m == NULL || m->rm_so < 0)
+		return 0;
+
+	if (regmatch_len (m) < bufsz)
+		s = regmatch_cpy (m, buf, src);
+	else
+		s = regmatch_dup (m, src);
+
+	n = strtol (s, endptr, base);
+
+	if (s != buf)
+		free (s);
+
+	return n;
+}
+
+/** Check for empty or non-existent match.
+ * @param m A match as returned by regexec().
+ * @return false if match length is non-zero.
+ * Note that reg_empty returns true even if match did not occur at all.
+ */
+bool regmatch_empty (regmatch_t * m)
+{
+	return (m == NULL || m->rm_so < 0 || m->rm_so == m->rm_eo);
+}
+
+/* vim:set expandtab cindent tabstop=4 softtabstop=4 shiftwidth=4 textwidth=0: */

contrib/flex/scanflags.c

@@ -0,0 +1,69 @@
+/* scanflags - flags used by scanning. */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "flexdef.h"
+
+scanflags_t* _sf_stk = NULL;
+size_t _sf_top_ix=0, _sf_max=0;
+
+void
+sf_push (void)
+{
+    if (_sf_top_ix + 1 >= _sf_max)
+        _sf_stk = (scanflags_t*) flex_realloc ( (void*) _sf_stk, sizeof(scanflags_t) * (_sf_max += 32));
+
+    // copy the top element
+    _sf_stk[_sf_top_ix + 1] = _sf_stk[_sf_top_ix];
+    ++_sf_top_ix;
+}
+
+void
+sf_pop (void)
+{
+    assert(_sf_top_ix > 0);
+    --_sf_top_ix;
+}
+
+/* one-time initialization. Should be called before any sf_ functions. */
+void
+sf_init (void)
+{
+    assert(_sf_stk == NULL);
+    _sf_stk = (scanflags_t*) flex_alloc ( sizeof(scanflags_t) * (_sf_max = 32));
+    if (!_sf_stk)
+        lerrsf_fatal(_("Unable to allocate %ld of stack"),
+            (long)sizeof(scanflags_t));
+    _sf_stk[_sf_top_ix] = 0;
+}
+
+/* vim:set expandtab cindent tabstop=4 softtabstop=4 shiftwidth=4 textwidth=0: */

contrib/flex/scanopt.c

@@ -0,0 +1,873 @@
+/* flex - tool to generate fast lexical analyzers */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "flexdef.h"
+#include "scanopt.h"
+
+
+/* Internal structures */
+
+#ifdef HAVE_STRCASECMP
+#define STRCASECMP(a,b) strcasecmp(a,b)
+#else
+static int STRCASECMP PROTO ((const char *, const char *));
+
+static int STRCASECMP (a, b)
+     const char *a;
+     const char *b;
+{
+	while (tolower (*a++) == tolower (*b++)) ;
+	return b - a;
+}
+#endif
+
+#define ARG_NONE 0x01
+#define ARG_REQ  0x02
+#define ARG_OPT  0x04
+#define IS_LONG  0x08
+
+struct _aux {
+	int     flags;		/* The above hex flags. */
+	int     namelen;	/* Length of the actual option word, e.g., "--file[=foo]" is 4 */
+	int     printlen;	/* Length of entire string, e.g., "--file[=foo]" is 12 */
+};
+
+
+struct _scanopt_t {
+	const optspec_t *options;	/* List of options. */
+	struct _aux *aux;	/* Auxiliary data about options. */
+	int     optc;		/* Number of options. */
+	int     argc;		/* Number of args. */
+	char  **argv;		/* Array of strings. */
+	int     index;		/* Used as: argv[index][subscript]. */
+	int     subscript;
+	char    no_err_msg;	/* If true, do not print errors. */
+	char    has_long;
+	char    has_short;
+};
+
+/* Accessor functions. These WOULD be one-liners, but portability calls. */
+static const char *NAME PROTO ((struct _scanopt_t *, int));
+static int PRINTLEN PROTO ((struct _scanopt_t *, int));
+static int RVAL PROTO ((struct _scanopt_t *, int));
+static int FLAGS PROTO ((struct _scanopt_t *, int));
+static const char *DESC PROTO ((struct _scanopt_t *, int));
+static int scanopt_err PROTO ((struct _scanopt_t *, int, int, int));
+static int matchlongopt PROTO ((char *, char **, int *, char **, int *));
+static int find_opt
+PROTO ((struct _scanopt_t *, int, char *, int, int *, int *opt_offset));
+
+static const char *NAME (s, i)
+     struct _scanopt_t *s;
+     int     i;
+{
+	return s->options[i].opt_fmt +
+		((s->aux[i].flags & IS_LONG) ? 2 : 1);
+}
+
+static int PRINTLEN (s, i)
+     struct _scanopt_t *s;
+     int     i;
+{
+	return s->aux[i].printlen;
+}
+
+static int RVAL (s, i)
+     struct _scanopt_t *s;
+     int     i;
+{
+	return s->options[i].r_val;
+}
+
+static int FLAGS (s, i)
+     struct _scanopt_t *s;
+     int     i;
+{
+	return s->aux[i].flags;
+}
+
+static const char *DESC (s, i)
+     struct _scanopt_t *s;
+     int     i;
+{
+	return s->options[i].desc ? s->options[i].desc : "";
+}
+
+#ifndef NO_SCANOPT_USAGE
+static int get_cols PROTO ((void));
+
+static int get_cols ()
+{
+	char   *env;
+	int     cols = 80;	/* default */
+
+#ifdef HAVE_NCURSES_H
+	initscr ();
+	endwin ();
+	if (COLS > 0)
+		return COLS;
+#endif
+
+	if ((env = getenv ("COLUMNS")) != NULL)
+		cols = atoi (env);
+
+	return cols;
+}
+#endif
+
+/* Macro to check for NULL before assigning a value. */
+#define SAFE_ASSIGN(ptr,val) \
+    do{                      \
+        if((ptr)!=NULL)      \
+            *(ptr) = val;    \
+    }while(0)
+
+/* Macro to assure we reset subscript whenever we adjust s->index.*/
+#define INC_INDEX(s,n)     \
+    do{                    \
+       (s)->index += (n);  \
+       (s)->subscript= 0;  \
+    }while(0)
+
+scanopt_t *scanopt_init (options, argc, argv, flags)
+     const optspec_t *options;
+     int     argc;
+     char  **argv;
+     int     flags;
+{
+	int     i;
+	struct _scanopt_t *s;
+	s = (struct _scanopt_t *) malloc (sizeof (struct _scanopt_t));
+
+	s->options = options;
+	s->optc = 0;
+	s->argc = argc;
+	s->argv = (char **) argv;
+	s->index = 1;
+	s->subscript = 0;
+	s->no_err_msg = (flags & SCANOPT_NO_ERR_MSG);
+	s->has_long = 0;
+	s->has_short = 0;
+
+	/* Determine option count. (Find entry with all zeros). */
+	s->optc = 0;
+	while (options[s->optc].opt_fmt
+	       || options[s->optc].r_val || options[s->optc].desc)
+		s->optc++;
+
+	/* Build auxiliary data */
+	s->aux = (struct _aux *) malloc (s->optc * sizeof (struct _aux));
+
+	for (i = 0; i < s->optc; i++) {
+		const Char *p, *pname;
+		const struct optspec_t *opt;
+		struct _aux *aux;
+
+		opt = s->options + i;
+		aux = s->aux + i;
+
+		aux->flags = ARG_NONE;
+
+		if (opt->opt_fmt[0] == '-' && opt->opt_fmt[1] == '-') {
+			aux->flags |= IS_LONG;
+			pname = (const Char *)(opt->opt_fmt + 2);
+			s->has_long = 1;
+		}
+		else {
+			pname = (const Char *)(opt->opt_fmt + 1);
+			s->has_short = 1;
+		}
+		aux->printlen = strlen (opt->opt_fmt);
+
+		aux->namelen = 0;
+		for (p = pname + 1; *p; p++) {
+			/* detect required arg */
+			if (*p == '=' || isspace (*p)
+			    || !(aux->flags & IS_LONG)) {
+				if (aux->namelen == 0)
+					aux->namelen = p - pname;
+				aux->flags |= ARG_REQ;
+				aux->flags &= ~ARG_NONE;
+			}
+			/* detect optional arg. This overrides required arg. */
+			if (*p == '[') {
+				if (aux->namelen == 0)
+					aux->namelen = p - pname;
+				aux->flags &= ~(ARG_REQ | ARG_NONE);
+				aux->flags |= ARG_OPT;
+				break;
+			}
+		}
+		if (aux->namelen == 0)
+			aux->namelen = p - pname;
+	}
+	return (scanopt_t *) s;
+}
+
+#ifndef NO_SCANOPT_USAGE
+/* these structs are for scanopt_usage(). */
+struct usg_elem {
+	int     idx;
+	struct usg_elem *next;
+	struct usg_elem *alias;
+};
+typedef struct usg_elem usg_elem;
+
+
+/* Prints a usage message based on contents of optlist.
+ * Parameters:
+ *   scanner  - The scanner, already initialized with scanopt_init().
+ *   fp       - The file stream to write to.
+ *   usage    - Text to be prepended to option list.
+ * Return:  Always returns 0 (zero).
+ * The output looks something like this:
+
+[indent][option, alias1, alias2...][indent][description line1
+                                            description line2...]
+ */
+int     scanopt_usage (scanner, fp, usage)
+     scanopt_t *scanner;
+     FILE   *fp;
+     const char *usage;
+{
+	struct _scanopt_t *s;
+	int     i, columns, indent = 2;
+	usg_elem *byr_val = NULL;	/* option indices sorted by r_val */
+	usg_elem *store;	/* array of preallocated elements. */
+	int     store_idx = 0;
+	usg_elem *ue;
+	int     maxlen[2];
+	int     desccol = 0;
+	int     print_run = 0;
+
+	maxlen[0] = 0;
+	maxlen[1] = 0;
+
+	s = (struct _scanopt_t *) scanner;
+
+	if (usage) {
+		fprintf (fp, "%s\n", usage);
+	}
+	else {
+		/* Find the basename of argv[0] */
+		const char *p;
+
+		p = s->argv[0] + strlen (s->argv[0]);
+		while (p != s->argv[0] && *p != '/')
+			--p;
+		if (*p == '/')
+			p++;
+
+		fprintf (fp, _("Usage: %s [OPTIONS]...\n"), p);
+	}
+	fprintf (fp, "\n");
+
+	/* Sort by r_val and string. Yes, this is O(n*n), but n is small. */
+	store = (usg_elem *) malloc (s->optc * sizeof (usg_elem));
+	for (i = 0; i < s->optc; i++) {
+
+		/* grab the next preallocate node. */
+		ue = store + store_idx++;
+		ue->idx = i;
+		ue->next = ue->alias = NULL;
+
+		/* insert into list. */
+		if (!byr_val)
+			byr_val = ue;
+		else {
+			int     found_alias = 0;
+			usg_elem **ue_curr, **ptr_if_no_alias = NULL;
+
+			ue_curr = &byr_val;
+			while (*ue_curr) {
+				if (RVAL (s, (*ue_curr)->idx) ==
+				    RVAL (s, ue->idx)) {
+					/* push onto the alias list. */
+					ue_curr = &((*ue_curr)->alias);
+					found_alias = 1;
+					break;
+				}
+				if (!ptr_if_no_alias
+				    &&
+				    STRCASECMP (NAME (s, (*ue_curr)->idx),
+						NAME (s, ue->idx)) > 0) {
+					ptr_if_no_alias = ue_curr;
+				}
+				ue_curr = &((*ue_curr)->next);
+			}
+			if (!found_alias && ptr_if_no_alias)
+				ue_curr = ptr_if_no_alias;
+			ue->next = *ue_curr;
+			*ue_curr = ue;
+		}
+	}
+
+#if 0
+	if (1) {
+		printf ("ORIGINAL:\n");
+		for (i = 0; i < s->optc; i++)
+			printf ("%2d: %s\n", i, NAME (s, i));
+		printf ("SORTED:\n");
+		ue = byr_val;
+		while (ue) {
+			usg_elem *ue2;
+
+			printf ("%2d: %s\n", ue->idx, NAME (s, ue->idx));
+			for (ue2 = ue->alias; ue2; ue2 = ue2->next)
+				printf ("  +---> %2d: %s\n", ue2->idx,
+					NAME (s, ue2->idx));
+			ue = ue->next;
+		}
+	}
+#endif
+
+	/* Now build each row of output. */
+
+	/* first pass calculate how much room we need. */
+	for (ue = byr_val; ue; ue = ue->next) {
+		usg_elem *ap;
+		int     len = 0;
+		int     nshort = 0, nlong = 0;
+
+
+#define CALC_LEN(i) do {\
+          if(FLAGS(s,i) & IS_LONG) \
+              len +=  (nlong++||nshort) ? 2+PRINTLEN(s,i) : PRINTLEN(s,i);\
+          else\
+              len +=  (nshort++||nlong)? 2+PRINTLEN(s,i) : PRINTLEN(s,i);\
+        }while(0)
+
+		if (!(FLAGS (s, ue->idx) & IS_LONG))
+			CALC_LEN (ue->idx);
+
+		/* do short aliases first. */
+		for (ap = ue->alias; ap; ap = ap->next) {
+			if (FLAGS (s, ap->idx) & IS_LONG)
+				continue;
+			CALC_LEN (ap->idx);
+		}
+
+		if (FLAGS (s, ue->idx) & IS_LONG)
+			CALC_LEN (ue->idx);
+
+		/* repeat the above loop, this time for long aliases. */
+		for (ap = ue->alias; ap; ap = ap->next) {
+			if (!(FLAGS (s, ap->idx) & IS_LONG))
+				continue;
+			CALC_LEN (ap->idx);
+		}
+
+		if (len > maxlen[0])
+			maxlen[0] = len;
+
+		/* It's much easier to calculate length for description column! */
+		len = strlen (DESC (s, ue->idx));
+		if (len > maxlen[1])
+			maxlen[1] = len;
+	}
+
+	/* Determine how much room we have, and how much we will allocate to each col.
+	 * Do not address pathological cases. Output will just be ugly. */
+	columns = get_cols () - 1;
+	if (maxlen[0] + maxlen[1] + indent * 2 > columns) {
+		/* col 0 gets whatever it wants. we'll wrap the desc col. */
+		maxlen[1] = columns - (maxlen[0] + indent * 2);
+		if (maxlen[1] < 14)	/* 14 is arbitrary lower limit on desc width. */
+			maxlen[1] = INT_MAX;
+	}
+	desccol = maxlen[0] + indent * 2;
+
+#define PRINT_SPACES(fp,n)\
+    do{\
+        int _n;\
+        _n=(n);\
+        while(_n-- > 0)\
+            fputc(' ',(fp));\
+    }while(0)
+
+
+	/* Second pass (same as above loop), this time we print. */
+	/* Sloppy hack: We iterate twice. The first time we print short and long options.
+	   The second time we print those lines that have ONLY long options. */
+	while (print_run++ < 2) {
+		for (ue = byr_val; ue; ue = ue->next) {
+			usg_elem *ap;
+			int     nwords = 0, nchars = 0, has_short = 0;
+
+/* TODO: get has_short schtick to work */
+			has_short = !(FLAGS (s, ue->idx) & IS_LONG);
+			for (ap = ue->alias; ap; ap = ap->next) {
+				if (!(FLAGS (s, ap->idx) & IS_LONG)) {
+					has_short = 1;
+					break;
+				}
+			}
+			if ((print_run == 1 && !has_short) ||
+			    (print_run == 2 && has_short))
+				continue;
+
+			PRINT_SPACES (fp, indent);
+			nchars += indent;
+
+/* Print, adding a ", " between aliases. */
+#define PRINT_IT(i) do{\
+                  if(nwords++)\
+                      nchars+=fprintf(fp,", ");\
+                  nchars+=fprintf(fp,"%s",s->options[i].opt_fmt);\
+            }while(0)
+
+			if (!(FLAGS (s, ue->idx) & IS_LONG))
+				PRINT_IT (ue->idx);
+
+			/* print short aliases first. */
+			for (ap = ue->alias; ap; ap = ap->next) {
+				if (!(FLAGS (s, ap->idx) & IS_LONG))
+					PRINT_IT (ap->idx);
+			}
+
+
+			if (FLAGS (s, ue->idx) & IS_LONG)
+				PRINT_IT (ue->idx);
+
+			/* repeat the above loop, this time for long aliases. */
+			for (ap = ue->alias; ap; ap = ap->next) {
+				if (FLAGS (s, ap->idx) & IS_LONG)
+					PRINT_IT (ap->idx);
+			}
+
+			/* pad to desccol */
+			PRINT_SPACES (fp, desccol - nchars);
+
+			/* Print description, wrapped to maxlen[1] columns. */
+			if (1) {
+				const char *pstart;
+
+				pstart = DESC (s, ue->idx);
+				while (1) {
+					int     n = 0;
+					const char *lastws = NULL, *p;
+
+					p = pstart;
+
+					while (*p && n < maxlen[1]
+					       && *p != '\n') {
+						if (isspace ((Char)(*p))
+						    || *p == '-') lastws =
+								p;
+						n++;
+						p++;
+					}
+
+					if (!*p) {	/* hit end of desc. done. */
+						fprintf (fp, "%s\n",
+							 pstart);
+						break;
+					}
+					else if (*p == '\n') {	/* print everything up to here then wrap. */
+						fprintf (fp, "%.*s\n", n,
+							 pstart);
+						PRINT_SPACES (fp, desccol);
+						pstart = p + 1;
+						continue;
+					}
+					else {	/* we hit the edge of the screen. wrap at space if possible. */
+						if (lastws) {
+							fprintf (fp,
+								 "%.*s\n",
+								 (int)(lastws - pstart),
+								 pstart);
+							pstart =
+								lastws + 1;
+						}
+						else {
+							fprintf (fp,
+								 "%.*s\n",
+								 n,
+								 pstart);
+							pstart = p + 1;
+						}
+						PRINT_SPACES (fp, desccol);
+						continue;
+					}
+				}
+			}
+		}
+	}			/* end while */
+	free (store);
+	return 0;
+}
+#endif /* no scanopt_usage */
+
+
+static int scanopt_err (s, opt_offset, is_short, err)
+     struct _scanopt_t *s;
+     int     opt_offset;
+     int     is_short;
+     int     err;
+{
+	const char *optname = "";
+	char    optchar[2];
+	const optspec_t *opt = NULL;
+
+	if (opt_offset >= 0)
+		opt = s->options + opt_offset;
+
+	if (!s->no_err_msg) {
+
+		if (s->index > 0 && s->index < s->argc) {
+			if (is_short) {
+				optchar[0] =
+					s->argv[s->index][s->subscript];
+				optchar[1] = '\0';
+				optname = optchar;
+			}
+			else {
+				optname = s->argv[s->index];
+			}
+		}
+
+		fprintf (stderr, "%s: ", s->argv[0]);
+		switch (err) {
+		case SCANOPT_ERR_ARG_NOT_ALLOWED:
+			fprintf (stderr,
+				 _
+				 ("option `%s' doesn't allow an argument\n"),
+				 optname);
+			break;
+		case SCANOPT_ERR_ARG_NOT_FOUND:
+			fprintf (stderr,
+				 _("option `%s' requires an argument\n"),
+				 optname);
+			break;
+		case SCANOPT_ERR_OPT_AMBIGUOUS:
+			fprintf (stderr, _("option `%s' is ambiguous\n"),
+				 optname);
+			break;
+		case SCANOPT_ERR_OPT_UNRECOGNIZED:
+			fprintf (stderr, _("Unrecognized option `%s'\n"),
+				 optname);
+			break;
+		default:
+			fprintf (stderr, _("Unknown error=(%d)\n"), err);
+			break;
+		}
+	}
+	return err;
+}
+
+
+/* Internal. Match str against the regex  ^--([^=]+)(=(.*))?
+ * return 1 if *looks* like a long option.
+ * 'str' is the only input argument, the rest of the arguments are output only.
+ * optname will point to str + 2
+ *
+ */
+static int matchlongopt (str, optname, optlen, arg, arglen)
+     char   *str;
+     char  **optname;
+     int    *optlen;
+     char  **arg;
+     int    *arglen;
+{
+	char   *p;
+
+	*optname = *arg = (char *) 0;
+	*optlen = *arglen = 0;
+
+	/* Match regex /--./   */
+	p = str;
+	if (p[0] != '-' || p[1] != '-' || !p[2])
+		return 0;
+
+	p += 2;
+	*optname = (char *) p;
+
+	/* find the end of optname */
+	while (*p && *p != '=')
+		++p;
+
+	*optlen = p - *optname;
+
+	if (!*p)
+		/* an option with no '=...' part. */
+		return 1;
+
+
+	/* We saw an '=' char. The rest of p is the arg. */
+	p++;
+	*arg = p;
+	while (*p)
+		++p;
+	*arglen = p - *arg;
+
+	return 1;
+}
+
+
+/* Internal. Look up long or short option by name.
+ * Long options must match a non-ambiguous prefix, or exact match.
+ * Short options must be exact.
+ * Return boolean true if found and no error.
+ * Error stored in err_code or zero if no error. */
+static int find_opt (s, lookup_long, optstart, len, err_code, opt_offset)
+     struct _scanopt_t *s;
+     int     lookup_long;
+     char   *optstart;
+     int     len;
+     int    *err_code;
+     int    *opt_offset;
+{
+	int     nmatch = 0, lastr_val = 0, i;
+
+	*err_code = 0;
+	*opt_offset = -1;
+
+	if (!optstart)
+		return 0;
+
+	for (i = 0; i < s->optc; i++) {
+		char   *optname;
+
+		optname =
+			(char *) (s->options[i].opt_fmt +
+				  (lookup_long ? 2 : 1));
+
+		if (lookup_long && (s->aux[i].flags & IS_LONG)) {
+			if (len > s->aux[i].namelen)
+				continue;
+
+			if (strncmp (optname, optstart, len) == 0) {
+				nmatch++;
+				*opt_offset = i;
+
+				/* exact match overrides all. */
+				if (len == s->aux[i].namelen) {
+					nmatch = 1;
+					break;
+				}
+
+				/* ambiguity is ok between aliases. */
+				if (lastr_val
+				    && lastr_val ==
+				    s->options[i].r_val) nmatch--;
+				lastr_val = s->options[i].r_val;
+			}
+		}
+		else if (!lookup_long && !(s->aux[i].flags & IS_LONG)) {
+			if (optname[0] == optstart[0]) {
+				nmatch++;
+				*opt_offset = i;
+			}
+		}
+	}
+
+	if (nmatch == 0) {
+		*err_code = SCANOPT_ERR_OPT_UNRECOGNIZED;
+		*opt_offset = -1;
+	}
+	else if (nmatch > 1) {
+		*err_code = SCANOPT_ERR_OPT_AMBIGUOUS;
+		*opt_offset = -1;
+	}
+
+	return *err_code ? 0 : 1;
+}
+
+
+int     scanopt (svoid, arg, optindex)
+     scanopt_t *svoid;
+     char  **arg;
+     int    *optindex;
+{
+	char   *optname = NULL, *optarg = NULL, *pstart;
+	int     namelen = 0, arglen = 0;
+	int     errcode = 0, has_next;
+	const optspec_t *optp;
+	struct _scanopt_t *s;
+	struct _aux *auxp;
+	int     is_short;
+	int     opt_offset = -1;
+
+	s = (struct _scanopt_t *) svoid;
+
+	/* Normalize return-parameters. */
+	SAFE_ASSIGN (arg, NULL);
+	SAFE_ASSIGN (optindex, s->index);
+
+	if (s->index >= s->argc)
+		return 0;
+
+	/* pstart always points to the start of our current scan. */
+	pstart = s->argv[s->index] + s->subscript;
+	if (!pstart)
+		return 0;
+
+	if (s->subscript == 0) {
+
+		/* test for exact match of "--" */
+		if (pstart[0] == '-' && pstart[1] == '-' && !pstart[2]) {
+			SAFE_ASSIGN (optindex, s->index + 1);
+			INC_INDEX (s, 1);
+			return 0;
+		}
+
+		/* Match an opt. */
+		if (matchlongopt
+		    (pstart, &optname, &namelen, &optarg, &arglen)) {
+
+			/* it LOOKS like an opt, but is it one?! */
+			if (!find_opt
+			    (s, 1, optname, namelen, &errcode,
+			     &opt_offset)) {
+				scanopt_err (s, opt_offset, 0, errcode);
+				return errcode;
+			}
+			/* We handle this below. */
+			is_short = 0;
+
+			/* Check for short opt.  */
+		}
+		else if (pstart[0] == '-' && pstart[1]) {
+			/* Pass through to below. */
+			is_short = 1;
+			s->subscript++;
+			pstart++;
+		}
+
+		else {
+			/* It's not an option. We're done. */
+			return 0;
+		}
+	}
+
+	/* We have to re-check the subscript status because it
+	 * may have changed above. */
+
+	if (s->subscript != 0) {
+
+		/* we are somewhere in a run of short opts,
+		 * e.g., at the 'z' in `tar -xzf` */
+
+		optname = pstart;
+		namelen = 1;
+		is_short = 1;
+
+		if (!find_opt
+		    (s, 0, pstart, namelen, &errcode, &opt_offset)) {
+			return scanopt_err (s, opt_offset, 1, errcode);
+		}
+
+		optarg = pstart + 1;
+		if (!*optarg) {
+			optarg = NULL;
+			arglen = 0;
+		}
+		else
+			arglen = strlen (optarg);
+	}
+
+	/* At this point, we have a long or short option matched at opt_offset into
+	 * the s->options array (and corresponding aux array).
+	 * A trailing argument is in {optarg,arglen}, if any.
+	 */
+
+	/* Look ahead in argv[] to see if there is something
+	 * that we can use as an argument (if needed). */
+	has_next = s->index + 1 < s->argc
+		&& strcmp ("--", s->argv[s->index + 1]) != 0;
+
+	optp = s->options + opt_offset;
+	auxp = s->aux + opt_offset;
+
+	/* case: no args allowed */
+	if (auxp->flags & ARG_NONE) {
+		if (optarg && !is_short) {
+			scanopt_err (s, opt_offset, is_short, errcode =
+				     SCANOPT_ERR_ARG_NOT_ALLOWED);
+			INC_INDEX (s, 1);
+			return errcode;
+		}
+		else if (!optarg)
+			INC_INDEX (s, 1);
+		else
+			s->subscript++;
+		return optp->r_val;
+	}
+
+	/* case: required */
+	if (auxp->flags & ARG_REQ) {
+		if (!optarg && !has_next)
+			return scanopt_err (s, opt_offset, is_short,
+					    SCANOPT_ERR_ARG_NOT_FOUND);
+
+		if (!optarg) {
+			/* Let the next argv element become the argument. */
+			SAFE_ASSIGN (arg, s->argv[s->index + 1]);
+			INC_INDEX (s, 2);
+		}
+		else {
+			SAFE_ASSIGN (arg, (char *) optarg);
+			INC_INDEX (s, 1);
+		}
+		return optp->r_val;
+	}
+
+	/* case: optional */
+	if (auxp->flags & ARG_OPT) {
+		SAFE_ASSIGN (arg, optarg);
+		INC_INDEX (s, 1);
+		return optp->r_val;
+	}
+
+
+	/* Should not reach here. */
+	return 0;
+}
+
+
+int     scanopt_destroy (svoid)
+     scanopt_t *svoid;
+{
+	struct _scanopt_t *s;
+
+	s = (struct _scanopt_t *) svoid;
+	if (s) {
+		if (s->aux)
+			free (s->aux);
+		free (s);
+	}
+	return 0;
+}
+
+
+/* vim:set tabstop=8 softtabstop=4 shiftwidth=4: */

contrib/flex/scanopt.h

@@ -0,0 +1,132 @@
+/* flex - tool to generate fast lexical analyzers */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#ifndef SCANOPT_H
+#define SCANOPT_H
+
+#include "flexdef.h"
+
+
+#ifndef NO_SCANOPT_USAGE
+/* Used by scanopt_usage for pretty-printing. */
+#ifdef HAVE_NCURSES_H
+#include <ncurses.h>
+#endif
+#endif
+
+#ifdef __cplusplus
+extern  "C" {
+#endif
+#ifndef PROTO
+#define PROTO(args) args
+#endif
+/* Error codes. */ enum scanopt_err_t {
+		SCANOPT_ERR_OPT_UNRECOGNIZED = -1,	/* Unrecognized option. */
+		SCANOPT_ERR_OPT_AMBIGUOUS = -2,	/* It matched more than one option name. */
+		SCANOPT_ERR_ARG_NOT_FOUND = -3,	/* The required arg was not found. */
+		SCANOPT_ERR_ARG_NOT_ALLOWED = -4	/* Option does not take an argument. */
+	};
+
+
+/* flags passed to scanopt_init */
+	enum scanopt_flag_t {
+		SCANOPT_NO_ERR_MSG = 0x01	/* Suppress printing to stderr. */
+	};
+
+/* Specification for a single option. */
+	struct optspec_t {
+		const char *opt_fmt;	/* e.g., "--foo=FILE", "-f FILE", "-n [NUM]" */
+		int     r_val;	/* Value to be returned by scanopt_ex(). */
+		const char *desc;	/* Brief description of this option, or NULL. */
+	};
+	typedef struct optspec_t optspec_t;
+
+
+/* Used internally by scanopt() to maintain state. */
+/* Never modify these value directly. */
+	typedef void *scanopt_t;
+
+
+/* Initializes scanner and checks option list for errors.
+ * Parameters:
+ *   options - Array of options.
+ *   argc    - Same as passed to main().
+ *   argv    - Same as passed to main(). First element is skipped.
+ *   flags   - Control behavior.
+ * Return:  A malloc'd pointer .
+ */
+	scanopt_t *scanopt_init PROTO ((const optspec_t * options,
+					int argc, char **argv, int flags));
+
+/* Frees memory used by scanner.
+ * Always returns 0. */
+	int scanopt_destroy PROTO ((scanopt_t * scanner));
+
+#ifndef NO_SCANOPT_USAGE
+/* Prints a usage message based on contents of optlist.
+ * Parameters:
+ *   scanner  - The scanner, already initialized with scanopt_init().
+ *   fp       - The file stream to write to.
+ *   usage    - Text to be prepended to option list. May be NULL.
+ * Return:  Always returns 0 (zero).
+ */
+	int scanopt_usage
+		PROTO (
+		       (scanopt_t * scanner, FILE * fp,
+			const char *usage));
+#endif
+
+/* Scans command-line options in argv[].
+ * Parameters:
+ *   scanner  - The scanner, already initialized with scanopt_init().
+ *   optarg   - Return argument, may be NULL.
+ *              On success, it points to start of an argument.
+ *   optindex - Return argument, may be NULL.
+ *              On success or failure, it is the index of this option.
+ *              If return is zero, then optindex is the NEXT valid option index.
+ *
+ * Return:  > 0 on success. Return value is from optspec_t->rval.
+ *         == 0 if at end of options.
+ *          < 0 on error (return value is an error code).
+ *
+ */
+	int scanopt
+		PROTO (
+		       (scanopt_t * scanner, char **optarg,
+			int *optindex));
+
+#ifdef __cplusplus
+}
+#endif
+#endif
+/* vim:set tabstop=8 softtabstop=4 shiftwidth=4: */

contrib/flex/sym.c

@@ -0,0 +1,272 @@
+/* sym - symbol table routines */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "flexdef.h"
+
+/* Variables for symbol tables:
+ * sctbl - start-condition symbol table
+ * ndtbl - name-definition symbol table
+ * ccltab - character class text symbol table
+ */
+
+struct hash_entry {
+	struct hash_entry *prev, *next;
+	char   *name;
+	char   *str_val;
+	int     int_val;
+};
+
+typedef struct hash_entry **hash_table;
+
+#define NAME_TABLE_HASH_SIZE 101
+#define START_COND_HASH_SIZE 101
+#define CCL_HASH_SIZE 101
+
+static struct hash_entry *ndtbl[NAME_TABLE_HASH_SIZE];
+static struct hash_entry *sctbl[START_COND_HASH_SIZE];
+static struct hash_entry *ccltab[CCL_HASH_SIZE];
+
+
+/* declare functions that have forward references */
+
+static int addsym PROTO ((register char[], char *, int, hash_table, int));
+static struct hash_entry *findsym PROTO ((register const char *sym,
+					  hash_table table,
+
+					  int table_size));
+static int hashfunct PROTO ((register const char *, int));
+
+
+/* addsym - add symbol and definitions to symbol table
+ *
+ * -1 is returned if the symbol already exists, and the change not made.
+ */
+
+static int addsym (sym, str_def, int_def, table, table_size)
+     register char sym[];
+     char   *str_def;
+     int     int_def;
+     hash_table table;
+     int     table_size;
+{
+	int     hash_val = hashfunct (sym, table_size);
+	register struct hash_entry *sym_entry = table[hash_val];
+	register struct hash_entry *new_entry;
+	register struct hash_entry *successor;
+
+	while (sym_entry) {
+		if (!strcmp (sym, sym_entry->name)) {	/* entry already exists */
+			return -1;
+		}
+
+		sym_entry = sym_entry->next;
+	}
+
+	/* create new entry */
+	new_entry = (struct hash_entry *)
+		flex_alloc (sizeof (struct hash_entry));
+
+	if (new_entry == NULL)
+		flexfatal (_("symbol table memory allocation failed"));
+
+	if ((successor = table[hash_val]) != 0) {
+		new_entry->next = successor;
+		successor->prev = new_entry;
+	}
+	else
+		new_entry->next = NULL;
+
+	new_entry->prev = NULL;
+	new_entry->name = sym;
+	new_entry->str_val = str_def;
+	new_entry->int_val = int_def;
+
+	table[hash_val] = new_entry;
+
+	return 0;
+}
+
+
+/* cclinstal - save the text of a character class */
+
+void    cclinstal (ccltxt, cclnum)
+     Char    ccltxt[];
+     int     cclnum;
+{
+	/* We don't bother checking the return status because we are not
+	 * called unless the symbol is new.
+	 */
+
+	(void) addsym ((char *) copy_unsigned_string (ccltxt),
+		       (char *) 0, cclnum, ccltab, CCL_HASH_SIZE);
+}
+
+
+/* ccllookup - lookup the number associated with character class text
+ *
+ * Returns 0 if there's no CCL associated with the text.
+ */
+
+int     ccllookup (ccltxt)
+     Char    ccltxt[];
+{
+	return findsym ((char *) ccltxt, ccltab, CCL_HASH_SIZE)->int_val;
+}
+
+
+/* findsym - find symbol in symbol table */
+
+static struct hash_entry *findsym (sym, table, table_size)
+     register const char *sym;
+     hash_table table;
+     int     table_size;
+{
+	static struct hash_entry empty_entry = {
+		(struct hash_entry *) 0, (struct hash_entry *) 0,
+		(char *) 0, (char *) 0, 0,
+	};
+	register struct hash_entry *sym_entry =
+
+		table[hashfunct (sym, table_size)];
+
+	while (sym_entry) {
+		if (!strcmp (sym, sym_entry->name))
+			return sym_entry;
+		sym_entry = sym_entry->next;
+	}
+
+	return &empty_entry;
+}
+
+/* hashfunct - compute the hash value for "str" and hash size "hash_size" */
+
+static int hashfunct (str, hash_size)
+     register const char *str;
+     int     hash_size;
+{
+	register int hashval;
+	register int locstr;
+
+	hashval = 0;
+	locstr = 0;
+
+	while (str[locstr]) {
+		hashval = (hashval << 1) + (unsigned char) str[locstr++];
+		hashval %= hash_size;
+	}
+
+	return hashval;
+}
+
+
+/* ndinstal - install a name definition */
+
+void    ndinstal (name, definition)
+     const char *name;
+     Char    definition[];
+{
+
+	if (addsym (copy_string (name),
+		    (char *) copy_unsigned_string (definition), 0,
+		    ndtbl, NAME_TABLE_HASH_SIZE))
+			synerr (_("name defined twice"));
+}
+
+
+/* ndlookup - lookup a name definition
+ *
+ * Returns a nil pointer if the name definition does not exist.
+ */
+
+Char   *ndlookup (nd)
+     const char *nd;
+{
+	return (Char *) findsym (nd, ndtbl, NAME_TABLE_HASH_SIZE)->str_val;
+}
+
+
+/* scextend - increase the maximum number of start conditions */
+
+void    scextend ()
+{
+	current_max_scs += MAX_SCS_INCREMENT;
+
+	++num_reallocs;
+
+	scset = reallocate_integer_array (scset, current_max_scs);
+	scbol = reallocate_integer_array (scbol, current_max_scs);
+	scxclu = reallocate_integer_array (scxclu, current_max_scs);
+	sceof = reallocate_integer_array (sceof, current_max_scs);
+	scname = reallocate_char_ptr_array (scname, current_max_scs);
+}
+
+
+/* scinstal - make a start condition
+ *
+ * NOTE
+ *    The start condition is "exclusive" if xcluflg is true.
+ */
+
+void    scinstal (str, xcluflg)
+     const char *str;
+     int     xcluflg;
+{
+
+	if (++lastsc >= current_max_scs)
+		scextend ();
+
+	scname[lastsc] = copy_string (str);
+
+	if (addsym (scname[lastsc], (char *) 0, lastsc,
+		    sctbl, START_COND_HASH_SIZE))
+			format_pinpoint_message (_
+						 ("start condition %s declared twice"),
+str);
+
+	scset[lastsc] = mkstate (SYM_EPSILON);
+	scbol[lastsc] = mkstate (SYM_EPSILON);
+	scxclu[lastsc] = xcluflg;
+	sceof[lastsc] = false;
+}
+
+
+/* sclookup - lookup the number associated with a start condition
+ *
+ * Returns 0 if no such start condition.
+ */
+
+int     sclookup (str)
+     const char *str;
+{
+	return findsym (str, sctbl, START_COND_HASH_SIZE)->int_val;
+}

contrib/flex/tables.c

@@ -0,0 +1,502 @@
+/*  tables.c - tables serialization code
+ *
+ *  Copyright (c) 1990 The Regents of the University of California.
+ *  All rights reserved.
+ *
+ *  This code is derived from software contributed to Berkeley by
+ *  Vern Paxson.
+ *
+ *  The United States Government has rights in this work pursuant
+ *  to contract no. DE-AC03-76SF00098 between the United States
+ *  Department of Energy and the University of California.
+ *
+ *  This file is part of flex.
+ *
+ *  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.
+ *
+ *  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR
+ *  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+ *  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ *  PURPOSE.
+ */
+
+
+#include "flexdef.h"
+#include "tables.h"
+
+/** Convert size_t to t_flag.
+ *  @param n in {1,2,4}
+ *  @return YYTD_DATA*. 
+ */
+#define BYTES2TFLAG(n)\
+    (((n) == sizeof(flex_int8_t))\
+        ? YYTD_DATA8\
+        :(((n)== sizeof(flex_int16_t))\
+            ? YYTD_DATA16\
+            : YYTD_DATA32))
+
+/** Clear YYTD_DATA* bit flags
+ * @return the flag with the YYTD_DATA* bits cleared
+ */
+#define TFLAGS_CLRDATA(flg) ((flg) & ~(YYTD_DATA8 | YYTD_DATA16 | YYTD_DATA32))
+
+int     yytbl_write32 (struct yytbl_writer *wr, flex_uint32_t v);
+int     yytbl_write16 (struct yytbl_writer *wr, flex_uint16_t v);
+int     yytbl_write8 (struct yytbl_writer *wr, flex_uint8_t v);
+int     yytbl_writen (struct yytbl_writer *wr, void *v, flex_int32_t len);
+static flex_int32_t yytbl_data_geti (const struct yytbl_data *tbl, int i);
+/* XXX Not used
+static flex_int32_t yytbl_data_getijk (const struct yytbl_data *tbl, int i,
+				  int j, int k);
+ */
+
+
+/** Initialize the table writer.
+ *  @param wr an uninitialized writer
+ *  @param the output file
+ *  @return 0 on success
+ */
+int yytbl_writer_init (struct yytbl_writer *wr, FILE * out)
+{
+	wr->out = out;
+	wr->total_written = 0;
+	return 0;
+}
+
+/** Initialize a table header.
+ *  @param th  The uninitialized structure
+ *  @param version_str the  version string
+ *  @param name the name of this table set
+ */
+int yytbl_hdr_init (struct yytbl_hdr *th, const char *version_str,
+		    const char *name)
+{
+	memset (th, 0, sizeof (struct yytbl_hdr));
+
+	th->th_magic = YYTBL_MAGIC;
+	th->th_hsize = 14 + strlen (version_str) + 1 + strlen (name) + 1;
+	th->th_hsize += yypad64 (th->th_hsize);
+	th->th_ssize = 0;	// Not known at this point.
+	th->th_flags = 0;
+	th->th_version = copy_string (version_str);
+	th->th_name = copy_string (name);
+	return 0;
+}
+
+/** Allocate and initialize a table data structure.
+ *  @param tbl a pointer to an uninitialized table
+ *  @param id  the table identifier
+ *  @return 0 on success
+ */
+int yytbl_data_init (struct yytbl_data *td, enum yytbl_id id)
+{
+
+	memset (td, 0, sizeof (struct yytbl_data));
+	td->td_id = id;
+	td->td_flags = YYTD_DATA32;
+	return 0;
+}
+
+/** Clean up table and data array.
+ *  @param td will be destroyed
+ *  @return 0 on success
+ */
+int yytbl_data_destroy (struct yytbl_data *td)
+{
+	if (td->td_data)
+		free (td->td_data);
+	td->td_data = 0;
+	free (td);
+	return 0;
+}
+
+/** Write enough padding to bring the file pointer to a 64-bit boundary. */
+static int yytbl_write_pad64 (struct yytbl_writer *wr)
+{
+	int     pad, bwritten = 0;
+
+	pad = yypad64 (wr->total_written);
+	while (pad-- > 0)
+		if (yytbl_write8 (wr, 0) < 0)
+			return -1;
+		else
+			bwritten++;
+	return bwritten;
+}
+
+/** write the header.
+ *  @param out the output stream
+ *  @param th table header to be written
+ *  @return -1 on error, or bytes written on success.
+ */
+int yytbl_hdr_fwrite (struct yytbl_writer *wr, const struct yytbl_hdr *th)
+{
+	int  sz, rv;
+	int     bwritten = 0;
+
+	if (yytbl_write32 (wr, th->th_magic) < 0
+	    || yytbl_write32 (wr, th->th_hsize) < 0)
+		flex_die (_("th_magic|th_hsize write32 failed"));
+	bwritten += 8;
+
+	if (fgetpos (wr->out, &(wr->th_ssize_pos)) != 0)
+		flex_die (_("fgetpos failed"));
+
+	if (yytbl_write32 (wr, th->th_ssize) < 0
+	    || yytbl_write16 (wr, th->th_flags) < 0)
+		flex_die (_("th_ssize|th_flags write failed"));
+	bwritten += 6;
+
+	sz = strlen (th->th_version) + 1;
+	if ((rv = yytbl_writen (wr, th->th_version, sz)) != sz)
+		flex_die (_("th_version writen failed"));
+	bwritten += rv;
+
+	sz = strlen (th->th_name) + 1;
+	if ((rv = yytbl_writen (wr, th->th_name, sz)) != sz)
+		flex_die (_("th_name writen failed"));
+	bwritten += rv;
+
+	/* add padding */
+	if ((rv = yytbl_write_pad64 (wr)) < 0)
+		flex_die (_("pad64 failed"));
+	bwritten += rv;
+
+	/* Sanity check */
+	if (bwritten != (int) th->th_hsize)
+		flex_die (_("pad64 failed"));
+
+	return bwritten;
+}
+
+
+/** Write this table.
+ *  @param out the file writer
+ *  @param td table data to be written
+ *  @return -1 on error, or bytes written on success.
+ */
+int yytbl_data_fwrite (struct yytbl_writer *wr, struct yytbl_data *td)
+{
+	int  rv;
+	flex_int32_t bwritten = 0;
+	flex_int32_t i, total_len;
+	fpos_t  pos;
+
+	if ((rv = yytbl_write16 (wr, td->td_id)) < 0)
+		return -1;
+	bwritten += rv;
+
+	if ((rv = yytbl_write16 (wr, td->td_flags)) < 0)
+		return -1;
+	bwritten += rv;
+
+	if ((rv = yytbl_write32 (wr, td->td_hilen)) < 0)
+		return -1;
+	bwritten += rv;
+
+	if ((rv = yytbl_write32 (wr, td->td_lolen)) < 0)
+		return -1;
+	bwritten += rv;
+
+	total_len = yytbl_calc_total_len (td);
+	for (i = 0; i < total_len; i++) {
+		switch (YYTDFLAGS2BYTES (td->td_flags)) {
+		case sizeof (flex_int8_t):
+			rv = yytbl_write8 (wr, yytbl_data_geti (td, i));
+			break;
+		case sizeof (flex_int16_t):
+			rv = yytbl_write16 (wr, yytbl_data_geti (td, i));
+			break;
+		case sizeof (flex_int32_t):
+			rv = yytbl_write32 (wr, yytbl_data_geti (td, i));
+			break;
+		default:
+			flex_die (_("invalid td_flags detected"));
+		}
+		if (rv < 0) {
+			flex_die (_("error while writing tables"));
+			return -1;
+		}
+		bwritten += rv;
+	}
+
+	/* Sanity check */
+	if (bwritten != (int) (12 + total_len * YYTDFLAGS2BYTES (td->td_flags))) {
+		flex_die (_("insanity detected"));
+		return -1;
+	}
+
+	/* add padding */
+	if ((rv = yytbl_write_pad64 (wr)) < 0) {
+		flex_die (_("pad64 failed"));
+		return -1;
+	}
+	bwritten += rv;
+
+	/* Now go back and update the th_hsize member */
+	if (fgetpos (wr->out, &pos) != 0
+	    || fsetpos (wr->out, &(wr->th_ssize_pos)) != 0
+	    || yytbl_write32 (wr, wr->total_written) < 0
+	    || fsetpos (wr->out, &pos)) {
+		flex_die (_("get|set|fwrite32 failed"));
+		return -1;
+	}
+	else
+		/* Don't count the int we just wrote. */
+		wr->total_written -= sizeof (flex_int32_t);
+	return bwritten;
+}
+
+/** Write n bytes.
+ *  @param  wr   the table writer
+ *  @param  v    data to be written
+ *  @param  len  number of bytes
+ *  @return  -1 on error. number of bytes written on success.
+ */
+int yytbl_writen (struct yytbl_writer *wr, void *v, flex_int32_t len)
+{
+	int  rv;
+
+	rv = fwrite (v, 1, len, wr->out);
+	if (rv != len)
+		return -1;
+	wr->total_written += len;
+	return len;
+}
+
+/** Write four bytes in network byte order
+ *  @param  wr  the table writer
+ *  @param  v    a dword in host byte order
+ *  @return  -1 on error. number of bytes written on success.
+ */
+int yytbl_write32 (struct yytbl_writer *wr, flex_uint32_t v)
+{
+	flex_uint32_t vnet;
+	size_t  bytes, rv;
+
+	vnet = htonl (v);
+	bytes = sizeof (flex_uint32_t);
+	rv = fwrite (&vnet, bytes, 1, wr->out);
+	if (rv != 1)
+		return -1;
+	wr->total_written += bytes;
+	return bytes;
+}
+
+/** Write two bytes in network byte order.
+ *  @param  wr  the table writer
+ *  @param  v    a word in host byte order
+ *  @return  -1 on error. number of bytes written on success.
+ */
+int yytbl_write16 (struct yytbl_writer *wr, flex_uint16_t v)
+{
+	flex_uint16_t vnet;
+	size_t  bytes, rv;
+
+	vnet = htons (v);
+	bytes = sizeof (flex_uint16_t);
+	rv = fwrite (&vnet, bytes, 1, wr->out);
+	if (rv != 1)
+		return -1;
+	wr->total_written += bytes;
+	return bytes;
+}
+
+/** Write a byte.
+ *  @param  wr  the table writer
+ *  @param  v    the value to be written
+ *  @return  -1 on error. number of bytes written on success.
+ */
+int yytbl_write8 (struct yytbl_writer *wr, flex_uint8_t v)
+{
+	size_t  bytes, rv;
+
+	bytes = sizeof (flex_uint8_t);
+	rv = fwrite (&v, bytes, 1, wr->out);
+	if (rv != 1)
+		return -1;
+	wr->total_written += bytes;
+	return bytes;
+}
+
+
+/* XXX Not Used */
+#if 0
+/** Extract data element [i][j] from array data tables. 
+ * @param tbl data table
+ * @param i index into higher dimension array. i should be zero for one-dimensional arrays.
+ * @param j index into lower dimension array.
+ * @param k index into struct, must be 0 or 1. Only valid for YYTD_ID_TRANSITION table
+ * @return data[i][j + k]
+ */
+static flex_int32_t yytbl_data_getijk (const struct yytbl_data *tbl, int i,
+				  int j, int k)
+{
+	flex_int32_t lo;
+
+	k %= 2;
+	lo = tbl->td_lolen;
+
+	switch (YYTDFLAGS2BYTES (tbl->td_flags)) {
+	case sizeof (flex_int8_t):
+		return ((flex_int8_t *) (tbl->td_data))[(i * lo + j) * (k + 1) +
+						   k];
+	case sizeof (flex_int16_t):
+		return ((flex_int16_t *) (tbl->td_data))[(i * lo + j) * (k +
+								    1) +
+						    k];
+	case sizeof (flex_int32_t):
+		return ((flex_int32_t *) (tbl->td_data))[(i * lo + j) * (k +
+								    1) +
+						    k];
+	default:
+		flex_die (_("invalid td_flags detected"));
+		break;
+	}
+
+	return 0;
+}
+#endif /* Not used */
+
+/** Extract data element [i] from array data tables treated as a single flat array of integers.
+ * Be careful for 2-dimensional arrays or for YYTD_ID_TRANSITION, which is an array
+ * of structs. 
+ * @param tbl data table
+ * @param i index into array.
+ * @return data[i]
+ */
+static flex_int32_t yytbl_data_geti (const struct yytbl_data *tbl, int i)
+{
+
+	switch (YYTDFLAGS2BYTES (tbl->td_flags)) {
+	case sizeof (flex_int8_t):
+		return ((flex_int8_t *) (tbl->td_data))[i];
+	case sizeof (flex_int16_t):
+		return ((flex_int16_t *) (tbl->td_data))[i];
+	case sizeof (flex_int32_t):
+		return ((flex_int32_t *) (tbl->td_data))[i];
+	default:
+		flex_die (_("invalid td_flags detected"));
+		break;
+	}
+	return 0;
+}
+
+/** Set data element [i] in array data tables treated as a single flat array of integers.
+ * Be careful for 2-dimensional arrays or for YYTD_ID_TRANSITION, which is an array
+ * of structs. 
+ * @param tbl data table
+ * @param i index into array.
+ * @param newval new value for data[i]
+ */
+static void yytbl_data_seti (const struct yytbl_data *tbl, int i,
+			     flex_int32_t newval)
+{
+
+	switch (YYTDFLAGS2BYTES (tbl->td_flags)) {
+	case sizeof (flex_int8_t):
+		((flex_int8_t *) (tbl->td_data))[i] = (flex_int8_t) newval;
+		break;
+	case sizeof (flex_int16_t):
+		((flex_int16_t *) (tbl->td_data))[i] = (flex_int16_t) newval;
+		break;
+	case sizeof (flex_int32_t):
+		((flex_int32_t *) (tbl->td_data))[i] = (flex_int32_t) newval;
+		break;
+	default:
+		flex_die (_("invalid td_flags detected"));
+		break;
+	}
+}
+
+/** Calculate the number of bytes  needed to hold the largest
+ *  absolute value in this data array.
+ *  @param tbl  the data table
+ *  @return sizeof(n) where n in {flex_int8_t, flex_int16_t, flex_int32_t}
+ */
+static size_t min_int_size (struct yytbl_data *tbl)
+{
+	flex_uint32_t i, total_len;
+	flex_int32_t max = 0;
+
+	total_len = yytbl_calc_total_len (tbl);
+
+	for (i = 0; i < total_len; i++) {
+		flex_int32_t n;
+
+		n = abs (yytbl_data_geti (tbl, i));
+
+		if (n > max)
+			max = n;
+	}
+
+	if (max <= INT8_MAX)
+		return sizeof (flex_int8_t);
+	else if (max <= INT16_MAX)
+		return sizeof (flex_int16_t);
+	else
+		return sizeof (flex_int32_t);
+}
+
+/** Transform data to smallest possible of (int32, int16, int8).
+ * For example, we may have generated an int32 array due to user options
+ * (e.g., %option align), but if the maximum value in that array
+ * is 80 (for example), then we can serialize it with only 1 byte per int.
+ * This is NOT the same as compressed DFA tables. We're just trying
+ * to save storage space here.
+ *
+ * @param tbl the table to be compressed
+ */
+void yytbl_data_compress (struct yytbl_data *tbl)
+{
+	flex_int32_t i, newsz, total_len;
+	struct yytbl_data newtbl;
+
+	yytbl_data_init (&newtbl, tbl->td_id);
+	newtbl.td_hilen = tbl->td_hilen;
+	newtbl.td_lolen = tbl->td_lolen;
+	newtbl.td_flags = tbl->td_flags;
+
+	newsz = min_int_size (tbl);
+
+
+	if (newsz == (int) YYTDFLAGS2BYTES (tbl->td_flags))
+		/* No change in this table needed. */
+		return;
+
+	if (newsz > (int) YYTDFLAGS2BYTES (tbl->td_flags)) {
+		flex_die (_("detected negative compression"));
+		return;
+	}
+
+	total_len = yytbl_calc_total_len (tbl);
+	newtbl.td_data = calloc (total_len, newsz);
+	newtbl.td_flags =
+		TFLAGS_CLRDATA (newtbl.td_flags) | BYTES2TFLAG (newsz);
+
+	for (i = 0; i < total_len; i++) {
+		flex_int32_t g;
+
+		g = yytbl_data_geti (tbl, i);
+		yytbl_data_seti (&newtbl, i, g);
+	}
+
+
+	/* Now copy over the old table */
+	free (tbl->td_data);
+	*tbl = newtbl;
+}
+
+/* vim:set noexpandtab cindent tabstop=8 softtabstop=0 shiftwidth=8 textwidth=0: */

contrib/flex/tables.h

@@ -0,0 +1,85 @@
+/*  tables.h - tables serialization code
+ *
+ *  Copyright (c) 1990 The Regents of the University of California.
+ *  All rights reserved.
+ *
+ *  This code is derived from software contributed to Berkeley by
+ *  Vern Paxson.
+ *
+ *  The United States Government has rights in this work pursuant
+ *  to contract no. DE-AC03-76SF00098 between the United States
+ *  Department of Energy and the University of California.
+ *
+ *  This file is part of flex.
+ *
+ *  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.
+ *
+ *  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR
+ *  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+ *  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ *  PURPOSE.
+ */
+
+#ifndef TABLES_H
+#define TABLES_H
+
+#ifdef __cplusplus
+/* *INDENT-OFF* */
+extern  "C" {
+/* *INDENT-ON* */
+#endif
+
+/* Tables serialization API declarations. */
+#include "tables_shared.h"
+struct yytbl_writer {
+	FILE   *out;
+	flex_uint32_t total_written;
+			    /**< bytes written so far */
+	fpos_t  th_ssize_pos;
+			    /**< position of th_ssize */
+};
+
+/* These are used by main.c, gen.c, etc.
+ * tablesext - if true, create external tables
+ * tablesfilename - filename for external tables
+ * tablesname - name that goes in serialized data, e.g., "yytables"
+ * tableswr -  writer for external tables
+ * tablesverify - true if tables-verify option specified
+ * gentables - true if we should spit out the normal C tables
+ */
+extern bool tablesext, tablesverify,gentables;
+extern char *tablesfilename, *tablesname;
+extern struct yytbl_writer tableswr;
+
+int     yytbl_writer_init (struct yytbl_writer *, FILE *);
+int     yytbl_hdr_init (struct yytbl_hdr *th, const char *version_str,
+			const char *name);
+int     yytbl_data_init (struct yytbl_data *tbl, enum yytbl_id id);
+int     yytbl_data_destroy (struct yytbl_data *td);
+int     yytbl_hdr_fwrite (struct yytbl_writer *wr,
+			  const struct yytbl_hdr *th);
+int     yytbl_data_fwrite (struct yytbl_writer *wr, struct yytbl_data *td);
+void    yytbl_data_compress (struct yytbl_data *tbl);
+struct yytbl_data *mkftbl (void);
+
+
+#ifdef __cplusplus
+/* *INDENT-OFF* */
+}
+/* *INDENT-ON* */
+#endif
+#endif
+
+/* vim:set expandtab cindent tabstop=4 softtabstop=4 shiftwidth=4 textwidth=0: */

contrib/flex/tables_shared.c

@@ -0,0 +1,70 @@
+#ifdef FLEX_SCANNER
+/*
+dnl   tables_shared.c - tables serialization code
+dnl 
+dnl   Copyright (c) 1990 The Regents of the University of California.
+dnl   All rights reserved.
+dnl 
+dnl   This code is derived from software contributed to Berkeley by
+dnl   Vern Paxson.
+dnl 
+dnl   The United States Government has rights in this work pursuant
+dnl   to contract no. DE-AC03-76SF00098 between the United States
+dnl   Department of Energy and the University of California.
+dnl 
+dnl   This file is part of flex.
+dnl 
+dnl   Redistribution and use in source and binary forms, with or without
+dnl   modification, are permitted provided that the following conditions
+dnl   are met:
+dnl 
+dnl   1. Redistributions of source code must retain the above copyright
+dnl      notice, this list of conditions and the following disclaimer.
+dnl   2. Redistributions in binary form must reproduce the above copyright
+dnl      notice, this list of conditions and the following disclaimer in the
+dnl      documentation and/or other materials provided with the distribution.
+dnl 
+dnl   Neither the name of the University nor the names of its contributors
+dnl   may be used to endorse or promote products derived from this software
+dnl   without specific prior written permission.
+dnl 
+dnl   THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR
+dnl   IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+dnl   WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+dnl   PURPOSE.
+dnl 
+*/
+
+/* This file is meant to be included in both the skeleton and the actual
+ * flex code (hence the name "_shared"). 
+ */
+#ifndef yyskel_static
+#define yyskel_static static
+#endif
+#else
+#include "flexdef.h"
+#include "tables.h"
+#ifndef yyskel_static
+#define yyskel_static
+#endif
+#endif
+
+
+/** Get the number of integers in this table. This is NOT the
+ *  same thing as the number of elements.
+ *  @param td the table 
+ *  @return the number of integers in the table
+ */
+yyskel_static flex_int32_t yytbl_calc_total_len (const struct yytbl_data *tbl)
+{
+	flex_int32_t n;
+
+	/* total number of ints */
+	n = tbl->td_lolen;
+	if (tbl->td_hilen > 0)
+		n *= tbl->td_hilen;
+
+	if (tbl->td_id == YYTD_ID_TRANSITION)
+		n *= 2;
+	return n;
+}

contrib/flex/tables_shared.h

@@ -0,0 +1,144 @@
+#ifdef FLEX_SCANNER
+/*
+dnl  tables_shared.h - tables serialization header
+dnl
+dnl  Copyright (c) 1990 The Regents of the University of California.
+dnl  All rights reserved.
+dnl
+dnl  This code is derived from software contributed to Berkeley by
+dnl  Vern Paxson.
+dnl
+dnl  The United States Government has rights in this work pursuant
+dnl  to contract no. DE-AC03-76SF00098 between the United States
+dnl  Department of Energy and the University of California.
+dnl
+dnl  This file is part of flex.
+dnl
+dnl  Redistribution and use in source and binary forms, with or without
+dnl  modification, are permitted provided that the following conditions
+dnl  are met:
+dnl
+dnl  1. Redistributions of source code must retain the above copyright
+dnl     notice, this list of conditions and the following disclaimer.
+dnl  2. Redistributions in binary form must reproduce the above copyright
+dnl     notice, this list of conditions and the following disclaimer in the
+dnl     documentation and/or other materials provided with the distribution.
+dnl
+dnl  Neither the name of the University nor the names of its contributors
+dnl  may be used to endorse or promote products derived from this software
+dnl  without specific prior written permission.
+dnl
+dnl  THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR
+dnl  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+dnl  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+dnl  PURPOSE.
+
+dnl
+dnl  This file is meant to be included in both the skeleton and the actual
+dnl  flex code (hence the name "_shared").
+*/
+#ifndef yyskel_static
+#define yyskel_static static
+#endif
+#else
+#ifndef yyskel_static
+#define yyskel_static
+#endif
+#endif
+
+/* Structures and prototypes for serializing flex tables.  The
+ * binary format is documented in the manual.
+ *
+ * Design considerations:
+ *
+ *  -  The format allows many tables per file.
+ *  -  The tables can be streamed.
+ *  -  All data is stored in network byte order.
+ *  -  We do not hinder future unicode support.
+ *  -  We can lookup tables by name.
+ */
+
+/** Magic number for serialized format. */
+#ifndef YYTBL_MAGIC
+#define YYTBL_MAGIC 0xF13C57B1
+#endif
+
+/** Calculate (0-7) = number bytes needed to pad n to next 64-bit boundary. */
+#ifndef yypad64
+#define yypad64(n) ((8-((n)%8))%8)
+#endif
+
+
+#ifndef YYTABLES_TYPES
+#define YYTABLES_TYPES
+/** Possible values for td_id field. Each one corresponds to a
+ *  scanner table of the same name.
+ */
+enum yytbl_id {
+	YYTD_ID_ACCEPT = 0x01,		/**< 1-dim ints */
+	YYTD_ID_BASE = 0x02,		/**< 1-dim ints */
+	YYTD_ID_CHK = 0x03,		/**< 1-dim ints */
+	YYTD_ID_DEF = 0x04,		/**< 1-dim ints */
+	YYTD_ID_EC = 0x05,		/**< 1-dim ints */
+	YYTD_ID_META = 0x06,		/**< 1-dim ints */
+	YYTD_ID_NUL_TRANS = 0x07,	/**< 1-dim ints, maybe indices */
+	YYTD_ID_NXT = 0x08,		/**< may be 2 dimensional ints */
+	YYTD_ID_RULE_CAN_MATCH_EOL = 0x09, /**< 1-dim ints */
+	YYTD_ID_START_STATE_LIST = 0x0A,	/**< 1-dim indices into trans tbl  */
+	YYTD_ID_TRANSITION = 0x0B,	/**< structs */
+	YYTD_ID_ACCLIST = 0x0C		/**< 1-dim ints */
+};
+
+/** bit flags for t_flags field of struct yytbl_data */
+enum yytbl_flags {
+	/* These first three are mutually exclusive */
+	YYTD_DATA8 = 0x01,   /**< data is an array of type flex_int8_t */
+	YYTD_DATA16 = 0x02,  /**< data is an array of type flex_int16_t */
+	YYTD_DATA32 = 0x04,  /**< data is an array of type flex_int32_t */
+
+	/* These two are mutually exclusive. */
+	YYTD_PTRANS = 0x08,  /**< data is a list of indexes of entries
+                                 into the expanded `yy_transition'
+                                 array. See notes in manual. */
+	YYTD_STRUCT = 0x10  /**< data consists of yy_trans_info structs */
+};
+
+/* The serialized tables header. */
+struct yytbl_hdr {
+	flex_uint32_t th_magic;  /**< Must be 0xF13C57B1 (comes from "Flex Table") */
+	flex_uint32_t th_hsize;  /**< Size of this header in bytes. */
+	flex_uint32_t th_ssize;  /**< Size of this dataset, in bytes, including header. */
+	flex_uint16_t th_flags;  /**< Currently unused, must be 0 */
+	char   *th_version; /**< Flex version string. NUL terminated. */
+	char   *th_name;    /**< The name of this table set. NUL terminated. */
+};
+
+/** A single serialized table */
+struct yytbl_data {
+	flex_uint16_t td_id;      /**< enum yytbl_id table identifier */
+	flex_uint16_t td_flags;   /**< how to interpret this data */
+	flex_uint32_t td_hilen;   /**< num elements in highest dimension array */
+	flex_uint32_t td_lolen;   /**< num elements in lowest dimension array */
+	void   *td_data;     /**< table data */
+};
+#endif
+
+/** Extract corresponding data size_t from td_flags */
+#ifndef YYTDFLAGS2BYTES
+#define YYTDFLAGS2BYTES(td_flags)\
+        (((td_flags) & YYTD_DATA8)\
+            ? sizeof(flex_int8_t)\
+            :(((td_flags) & YYTD_DATA16)\
+                ? sizeof(flex_int16_t)\
+                :sizeof(flex_int32_t)))
+#endif
+
+#ifdef FLEX_SCANNER
+%not-for-header
+#endif
+yyskel_static flex_int32_t yytbl_calc_total_len (const struct yytbl_data *tbl);
+#ifdef FLEX_SCANNER
+%ok-for-header
+#endif
+
+/* vim:set noexpandtab cindent tabstop=8 softtabstop=0 shiftwidth=8 textwidth=0: */

contrib/flex/tblcmp.c

@@ -0,0 +1,853 @@
+/* tblcmp - table compression routines */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include "flexdef.h"
+
+
+/* declarations for functions that have forward references */
+
+void mkentry PROTO ((register int *, int, int, int, int));
+void mkprot PROTO ((int[], int, int));
+void mktemplate PROTO ((int[], int, int));
+void mv2front PROTO ((int));
+int tbldiff PROTO ((int[], int, int[]));
+
+
+/* bldtbl - build table entries for dfa state
+ *
+ * synopsis
+ *   int state[numecs], statenum, totaltrans, comstate, comfreq;
+ *   bldtbl( state, statenum, totaltrans, comstate, comfreq );
+ *
+ * State is the statenum'th dfa state.  It is indexed by equivalence class and
+ * gives the number of the state to enter for a given equivalence class.
+ * totaltrans is the total number of transitions out of the state.  Comstate
+ * is that state which is the destination of the most transitions out of State.
+ * Comfreq is how many transitions there are out of State to Comstate.
+ *
+ * A note on terminology:
+ *    "protos" are transition tables which have a high probability of
+ * either being redundant (a state processed later will have an identical
+ * transition table) or nearly redundant (a state processed later will have
+ * many of the same out-transitions).  A "most recently used" queue of
+ * protos is kept around with the hope that most states will find a proto
+ * which is similar enough to be usable, and therefore compacting the
+ * output tables.
+ *    "templates" are a special type of proto.  If a transition table is
+ * homogeneous or nearly homogeneous (all transitions go to the same
+ * destination) then the odds are good that future states will also go
+ * to the same destination state on basically the same character set.
+ * These homogeneous states are so common when dealing with large rule
+ * sets that they merit special attention.  If the transition table were
+ * simply made into a proto, then (typically) each subsequent, similar
+ * state will differ from the proto for two out-transitions.  One of these
+ * out-transitions will be that character on which the proto does not go
+ * to the common destination, and one will be that character on which the
+ * state does not go to the common destination.  Templates, on the other
+ * hand, go to the common state on EVERY transition character, and therefore
+ * cost only one difference.
+ */
+
+void    bldtbl (state, statenum, totaltrans, comstate, comfreq)
+     int     state[], statenum, totaltrans, comstate, comfreq;
+{
+	int     extptr, extrct[2][CSIZE + 1];
+	int     mindiff, minprot, i, d;
+
+	/* If extptr is 0 then the first array of extrct holds the result
+	 * of the "best difference" to date, which is those transitions
+	 * which occur in "state" but not in the proto which, to date,
+	 * has the fewest differences between itself and "state".  If
+	 * extptr is 1 then the second array of extrct hold the best
+	 * difference.  The two arrays are toggled between so that the
+	 * best difference to date can be kept around and also a difference
+	 * just created by checking against a candidate "best" proto.
+	 */
+
+	extptr = 0;
+
+	/* If the state has too few out-transitions, don't bother trying to
+	 * compact its tables.
+	 */
+
+	if ((totaltrans * 100) < (numecs * PROTO_SIZE_PERCENTAGE))
+		mkentry (state, numecs, statenum, JAMSTATE, totaltrans);
+
+	else {
+		/* "checkcom" is true if we should only check "state" against
+		 * protos which have the same "comstate" value.
+		 */
+		int     checkcom =
+
+			comfreq * 100 > totaltrans * CHECK_COM_PERCENTAGE;
+
+		minprot = firstprot;
+		mindiff = totaltrans;
+
+		if (checkcom) {
+			/* Find first proto which has the same "comstate". */
+			for (i = firstprot; i != NIL; i = protnext[i])
+				if (protcomst[i] == comstate) {
+					minprot = i;
+					mindiff = tbldiff (state, minprot,
+							   extrct[extptr]);
+					break;
+				}
+		}
+
+		else {
+			/* Since we've decided that the most common destination
+			 * out of "state" does not occur with a high enough
+			 * frequency, we set the "comstate" to zero, assuring
+			 * that if this state is entered into the proto list,
+			 * it will not be considered a template.
+			 */
+			comstate = 0;
+
+			if (firstprot != NIL) {
+				minprot = firstprot;
+				mindiff = tbldiff (state, minprot,
+						   extrct[extptr]);
+			}
+		}
+
+		/* We now have the first interesting proto in "minprot".  If
+		 * it matches within the tolerances set for the first proto,
+		 * we don't want to bother scanning the rest of the proto list
+		 * to see if we have any other reasonable matches.
+		 */
+
+		if (mindiff * 100 >
+		    totaltrans * FIRST_MATCH_DIFF_PERCENTAGE) {
+			/* Not a good enough match.  Scan the rest of the
+			 * protos.
+			 */
+			for (i = minprot; i != NIL; i = protnext[i]) {
+				d = tbldiff (state, i, extrct[1 - extptr]);
+				if (d < mindiff) {
+					extptr = 1 - extptr;
+					mindiff = d;
+					minprot = i;
+				}
+			}
+		}
+
+		/* Check if the proto we've decided on as our best bet is close
+		 * enough to the state we want to match to be usable.
+		 */
+
+		if (mindiff * 100 >
+		    totaltrans * ACCEPTABLE_DIFF_PERCENTAGE) {
+			/* No good.  If the state is homogeneous enough,
+			 * we make a template out of it.  Otherwise, we
+			 * make a proto.
+			 */
+
+			if (comfreq * 100 >=
+			    totaltrans * TEMPLATE_SAME_PERCENTAGE)
+					mktemplate (state, statenum,
+						    comstate);
+
+			else {
+				mkprot (state, statenum, comstate);
+				mkentry (state, numecs, statenum,
+					 JAMSTATE, totaltrans);
+			}
+		}
+
+		else {		/* use the proto */
+			mkentry (extrct[extptr], numecs, statenum,
+				 prottbl[minprot], mindiff);
+
+			/* If this state was sufficiently different from the
+			 * proto we built it from, make it, too, a proto.
+			 */
+
+			if (mindiff * 100 >=
+			    totaltrans * NEW_PROTO_DIFF_PERCENTAGE)
+					mkprot (state, statenum, comstate);
+
+			/* Since mkprot added a new proto to the proto queue,
+			 * it's possible that "minprot" is no longer on the
+			 * proto queue (if it happened to have been the last
+			 * entry, it would have been bumped off).  If it's
+			 * not there, then the new proto took its physical
+			 * place (though logically the new proto is at the
+			 * beginning of the queue), so in that case the
+			 * following call will do nothing.
+			 */
+
+			mv2front (minprot);
+		}
+	}
+}
+
+
+/* cmptmps - compress template table entries
+ *
+ * Template tables are compressed by using the 'template equivalence
+ * classes', which are collections of transition character equivalence
+ * classes which always appear together in templates - really meta-equivalence
+ * classes.
+ */
+
+void    cmptmps ()
+{
+	int     tmpstorage[CSIZE + 1];
+	register int *tmp = tmpstorage, i, j;
+	int     totaltrans, trans;
+
+	peakpairs = numtemps * numecs + tblend;
+
+	if (usemecs) {
+		/* Create equivalence classes based on data gathered on
+		 * template transitions.
+		 */
+		nummecs = cre8ecs (tecfwd, tecbck, numecs);
+	}
+
+	else
+		nummecs = numecs;
+
+	while (lastdfa + numtemps + 1 >= current_max_dfas)
+		increase_max_dfas ();
+
+	/* Loop through each template. */
+
+	for (i = 1; i <= numtemps; ++i) {
+		/* Number of non-jam transitions out of this template. */
+		totaltrans = 0;
+
+		for (j = 1; j <= numecs; ++j) {
+			trans = tnxt[numecs * i + j];
+
+			if (usemecs) {
+				/* The absolute value of tecbck is the
+				 * meta-equivalence class of a given
+				 * equivalence class, as set up by cre8ecs().
+				 */
+				if (tecbck[j] > 0) {
+					tmp[tecbck[j]] = trans;
+
+					if (trans > 0)
+						++totaltrans;
+				}
+			}
+
+			else {
+				tmp[j] = trans;
+
+				if (trans > 0)
+					++totaltrans;
+			}
+		}
+
+		/* It is assumed (in a rather subtle way) in the skeleton
+		 * that if we're using meta-equivalence classes, the def[]
+		 * entry for all templates is the jam template, i.e.,
+		 * templates never default to other non-jam table entries
+		 * (e.g., another template)
+		 */
+
+		/* Leave room for the jam-state after the last real state. */
+		mkentry (tmp, nummecs, lastdfa + i + 1, JAMSTATE,
+			 totaltrans);
+	}
+}
+
+
+
+/* expand_nxt_chk - expand the next check arrays */
+
+void    expand_nxt_chk ()
+{
+	register int old_max = current_max_xpairs;
+
+	current_max_xpairs += MAX_XPAIRS_INCREMENT;
+
+	++num_reallocs;
+
+	nxt = reallocate_integer_array (nxt, current_max_xpairs);
+	chk = reallocate_integer_array (chk, current_max_xpairs);
+
+	zero_out ((char *) (chk + old_max),
+		  (size_t) (MAX_XPAIRS_INCREMENT * sizeof (int)));
+}
+
+
+/* find_table_space - finds a space in the table for a state to be placed
+ *
+ * synopsis
+ *     int *state, numtrans, block_start;
+ *     int find_table_space();
+ *
+ *     block_start = find_table_space( state, numtrans );
+ *
+ * State is the state to be added to the full speed transition table.
+ * Numtrans is the number of out-transitions for the state.
+ *
+ * find_table_space() returns the position of the start of the first block (in
+ * chk) able to accommodate the state
+ *
+ * In determining if a state will or will not fit, find_table_space() must take
+ * into account the fact that an end-of-buffer state will be added at [0],
+ * and an action number will be added in [-1].
+ */
+
+int     find_table_space (state, numtrans)
+     int    *state, numtrans;
+{
+	/* Firstfree is the position of the first possible occurrence of two
+	 * consecutive unused records in the chk and nxt arrays.
+	 */
+	register int i;
+	register int *state_ptr, *chk_ptr;
+	register int *ptr_to_last_entry_in_state;
+
+	/* If there are too many out-transitions, put the state at the end of
+	 * nxt and chk.
+	 */
+	if (numtrans > MAX_XTIONS_FULL_INTERIOR_FIT) {
+		/* If table is empty, return the first available spot in
+		 * chk/nxt, which should be 1.
+		 */
+		if (tblend < 2)
+			return 1;
+
+		/* Start searching for table space near the end of
+		 * chk/nxt arrays.
+		 */
+		i = tblend - numecs;
+	}
+
+	else
+		/* Start searching for table space from the beginning
+		 * (skipping only the elements which will definitely not
+		 * hold the new state).
+		 */
+		i = firstfree;
+
+	while (1) {		/* loops until a space is found */
+		while (i + numecs >= current_max_xpairs)
+			expand_nxt_chk ();
+
+		/* Loops until space for end-of-buffer and action number
+		 * are found.
+		 */
+		while (1) {
+			/* Check for action number space. */
+			if (chk[i - 1] == 0) {
+				/* Check for end-of-buffer space. */
+				if (chk[i] == 0)
+					break;
+
+				else
+					/* Since i != 0, there is no use
+					 * checking to see if (++i) - 1 == 0,
+					 * because that's the same as i == 0,
+					 * so we skip a space.
+					 */
+					i += 2;
+			}
+
+			else
+				++i;
+
+			while (i + numecs >= current_max_xpairs)
+				expand_nxt_chk ();
+		}
+
+		/* If we started search from the beginning, store the new
+		 * firstfree for the next call of find_table_space().
+		 */
+		if (numtrans <= MAX_XTIONS_FULL_INTERIOR_FIT)
+			firstfree = i + 1;
+
+		/* Check to see if all elements in chk (and therefore nxt)
+		 * that are needed for the new state have not yet been taken.
+		 */
+
+		state_ptr = &state[1];
+		ptr_to_last_entry_in_state = &chk[i + numecs + 1];
+
+		for (chk_ptr = &chk[i + 1];
+		     chk_ptr != ptr_to_last_entry_in_state; ++chk_ptr)
+			if (*(state_ptr++) != 0 && *chk_ptr != 0)
+				break;
+
+		if (chk_ptr == ptr_to_last_entry_in_state)
+			return i;
+
+		else
+			++i;
+	}
+}
+
+
+/* inittbl - initialize transition tables
+ *
+ * Initializes "firstfree" to be one beyond the end of the table.  Initializes
+ * all "chk" entries to be zero.
+ */
+void    inittbl ()
+{
+	register int i;
+
+	zero_out ((char *) chk,
+
+		  (size_t) (current_max_xpairs * sizeof (int)));
+
+	tblend = 0;
+	firstfree = tblend + 1;
+	numtemps = 0;
+
+	if (usemecs) {
+		/* Set up doubly-linked meta-equivalence classes; these
+		 * are sets of equivalence classes which all have identical
+		 * transitions out of TEMPLATES.
+		 */
+
+		tecbck[1] = NIL;
+
+		for (i = 2; i <= numecs; ++i) {
+			tecbck[i] = i - 1;
+			tecfwd[i - 1] = i;
+		}
+
+		tecfwd[numecs] = NIL;
+	}
+}
+
+
+/* mkdeftbl - make the default, "jam" table entries */
+
+void    mkdeftbl ()
+{
+	int     i;
+
+	jamstate = lastdfa + 1;
+
+	++tblend;		/* room for transition on end-of-buffer character */
+
+	while (tblend + numecs >= current_max_xpairs)
+		expand_nxt_chk ();
+
+	/* Add in default end-of-buffer transition. */
+	nxt[tblend] = end_of_buffer_state;
+	chk[tblend] = jamstate;
+
+	for (i = 1; i <= numecs; ++i) {
+		nxt[tblend + i] = 0;
+		chk[tblend + i] = jamstate;
+	}
+
+	jambase = tblend;
+
+	base[jamstate] = jambase;
+	def[jamstate] = 0;
+
+	tblend += numecs;
+	++numtemps;
+}
+
+
+/* mkentry - create base/def and nxt/chk entries for transition array
+ *
+ * synopsis
+ *   int state[numchars + 1], numchars, statenum, deflink, totaltrans;
+ *   mkentry( state, numchars, statenum, deflink, totaltrans );
+ *
+ * "state" is a transition array "numchars" characters in size, "statenum"
+ * is the offset to be used into the base/def tables, and "deflink" is the
+ * entry to put in the "def" table entry.  If "deflink" is equal to
+ * "JAMSTATE", then no attempt will be made to fit zero entries of "state"
+ * (i.e., jam entries) into the table.  It is assumed that by linking to
+ * "JAMSTATE" they will be taken care of.  In any case, entries in "state"
+ * marking transitions to "SAME_TRANS" are treated as though they will be
+ * taken care of by whereever "deflink" points.  "totaltrans" is the total
+ * number of transitions out of the state.  If it is below a certain threshold,
+ * the tables are searched for an interior spot that will accommodate the
+ * state array.
+ */
+
+void    mkentry (state, numchars, statenum, deflink, totaltrans)
+     register int *state;
+     int     numchars, statenum, deflink, totaltrans;
+{
+	register int minec, maxec, i, baseaddr;
+	int     tblbase, tbllast;
+
+	if (totaltrans == 0) {	/* there are no out-transitions */
+		if (deflink == JAMSTATE)
+			base[statenum] = JAMSTATE;
+		else
+			base[statenum] = 0;
+
+		def[statenum] = deflink;
+		return;
+	}
+
+	for (minec = 1; minec <= numchars; ++minec) {
+		if (state[minec] != SAME_TRANS)
+			if (state[minec] != 0 || deflink != JAMSTATE)
+				break;
+	}
+
+	if (totaltrans == 1) {
+		/* There's only one out-transition.  Save it for later to fill
+		 * in holes in the tables.
+		 */
+		stack1 (statenum, minec, state[minec], deflink);
+		return;
+	}
+
+	for (maxec = numchars; maxec > 0; --maxec) {
+		if (state[maxec] != SAME_TRANS)
+			if (state[maxec] != 0 || deflink != JAMSTATE)
+				break;
+	}
+
+	/* Whether we try to fit the state table in the middle of the table
+	 * entries we have already generated, or if we just take the state
+	 * table at the end of the nxt/chk tables, we must make sure that we
+	 * have a valid base address (i.e., non-negative).  Note that
+	 * negative base addresses dangerous at run-time (because indexing
+	 * the nxt array with one and a low-valued character will access
+	 * memory before the start of the array.
+	 */
+
+	/* Find the first transition of state that we need to worry about. */
+	if (totaltrans * 100 <= numchars * INTERIOR_FIT_PERCENTAGE) {
+		/* Attempt to squeeze it into the middle of the tables. */
+		baseaddr = firstfree;
+
+		while (baseaddr < minec) {
+			/* Using baseaddr would result in a negative base
+			 * address below; find the next free slot.
+			 */
+			for (++baseaddr; chk[baseaddr] != 0; ++baseaddr) ;
+		}
+
+		while (baseaddr + maxec - minec + 1 >= current_max_xpairs)
+			expand_nxt_chk ();
+
+		for (i = minec; i <= maxec; ++i)
+			if (state[i] != SAME_TRANS &&
+			    (state[i] != 0 || deflink != JAMSTATE) &&
+			    chk[baseaddr + i - minec] != 0) {	/* baseaddr unsuitable - find another */
+				for (++baseaddr;
+				     baseaddr < current_max_xpairs &&
+				     chk[baseaddr] != 0; ++baseaddr) ;
+
+				while (baseaddr + maxec - minec + 1 >=
+				       current_max_xpairs)
+						expand_nxt_chk ();
+
+				/* Reset the loop counter so we'll start all
+				 * over again next time it's incremented.
+				 */
+
+				i = minec - 1;
+			}
+	}
+
+	else {
+		/* Ensure that the base address we eventually generate is
+		 * non-negative.
+		 */
+		baseaddr = MAX (tblend + 1, minec);
+	}
+
+	tblbase = baseaddr - minec;
+	tbllast = tblbase + maxec;
+
+	while (tbllast + 1 >= current_max_xpairs)
+		expand_nxt_chk ();
+
+	base[statenum] = tblbase;
+	def[statenum] = deflink;
+
+	for (i = minec; i <= maxec; ++i)
+		if (state[i] != SAME_TRANS)
+			if (state[i] != 0 || deflink != JAMSTATE) {
+				nxt[tblbase + i] = state[i];
+				chk[tblbase + i] = statenum;
+			}
+
+	if (baseaddr == firstfree)
+		/* Find next free slot in tables. */
+		for (++firstfree; chk[firstfree] != 0; ++firstfree) ;
+
+	tblend = MAX (tblend, tbllast);
+}
+
+
+/* mk1tbl - create table entries for a state (or state fragment) which
+ *            has only one out-transition
+ */
+
+void    mk1tbl (state, sym, onenxt, onedef)
+     int     state, sym, onenxt, onedef;
+{
+	if (firstfree < sym)
+		firstfree = sym;
+
+	while (chk[firstfree] != 0)
+		if (++firstfree >= current_max_xpairs)
+			expand_nxt_chk ();
+
+	base[state] = firstfree - sym;
+	def[state] = onedef;
+	chk[firstfree] = state;
+	nxt[firstfree] = onenxt;
+
+	if (firstfree > tblend) {
+		tblend = firstfree++;
+
+		if (firstfree >= current_max_xpairs)
+			expand_nxt_chk ();
+	}
+}
+
+
+/* mkprot - create new proto entry */
+
+void    mkprot (state, statenum, comstate)
+     int     state[], statenum, comstate;
+{
+	int     i, slot, tblbase;
+
+	if (++numprots >= MSP || numecs * numprots >= PROT_SAVE_SIZE) {
+		/* Gotta make room for the new proto by dropping last entry in
+		 * the queue.
+		 */
+		slot = lastprot;
+		lastprot = protprev[lastprot];
+		protnext[lastprot] = NIL;
+	}
+
+	else
+		slot = numprots;
+
+	protnext[slot] = firstprot;
+
+	if (firstprot != NIL)
+		protprev[firstprot] = slot;
+
+	firstprot = slot;
+	prottbl[slot] = statenum;
+	protcomst[slot] = comstate;
+
+	/* Copy state into save area so it can be compared with rapidly. */
+	tblbase = numecs * (slot - 1);
+
+	for (i = 1; i <= numecs; ++i)
+		protsave[tblbase + i] = state[i];
+}
+
+
+/* mktemplate - create a template entry based on a state, and connect the state
+ *              to it
+ */
+
+void    mktemplate (state, statenum, comstate)
+     int     state[], statenum, comstate;
+{
+	int     i, numdiff, tmpbase, tmp[CSIZE + 1];
+	Char    transset[CSIZE + 1];
+	int     tsptr;
+
+	++numtemps;
+
+	tsptr = 0;
+
+	/* Calculate where we will temporarily store the transition table
+	 * of the template in the tnxt[] array.  The final transition table
+	 * gets created by cmptmps().
+	 */
+
+	tmpbase = numtemps * numecs;
+
+	if (tmpbase + numecs >= current_max_template_xpairs) {
+		current_max_template_xpairs +=
+			MAX_TEMPLATE_XPAIRS_INCREMENT;
+
+		++num_reallocs;
+
+		tnxt = reallocate_integer_array (tnxt,
+						 current_max_template_xpairs);
+	}
+
+	for (i = 1; i <= numecs; ++i)
+		if (state[i] == 0)
+			tnxt[tmpbase + i] = 0;
+		else {
+			transset[tsptr++] = i;
+			tnxt[tmpbase + i] = comstate;
+		}
+
+	if (usemecs)
+		mkeccl (transset, tsptr, tecfwd, tecbck, numecs, 0);
+
+	mkprot (tnxt + tmpbase, -numtemps, comstate);
+
+	/* We rely on the fact that mkprot adds things to the beginning
+	 * of the proto queue.
+	 */
+
+	numdiff = tbldiff (state, firstprot, tmp);
+	mkentry (tmp, numecs, statenum, -numtemps, numdiff);
+}
+
+
+/* mv2front - move proto queue element to front of queue */
+
+void    mv2front (qelm)
+     int     qelm;
+{
+	if (firstprot != qelm) {
+		if (qelm == lastprot)
+			lastprot = protprev[lastprot];
+
+		protnext[protprev[qelm]] = protnext[qelm];
+
+		if (protnext[qelm] != NIL)
+			protprev[protnext[qelm]] = protprev[qelm];
+
+		protprev[qelm] = NIL;
+		protnext[qelm] = firstprot;
+		protprev[firstprot] = qelm;
+		firstprot = qelm;
+	}
+}
+
+
+/* place_state - place a state into full speed transition table
+ *
+ * State is the statenum'th state.  It is indexed by equivalence class and
+ * gives the number of the state to enter for a given equivalence class.
+ * Transnum is the number of out-transitions for the state.
+ */
+
+void    place_state (state, statenum, transnum)
+     int    *state, statenum, transnum;
+{
+	register int i;
+	register int *state_ptr;
+	int     position = find_table_space (state, transnum);
+
+	/* "base" is the table of start positions. */
+	base[statenum] = position;
+
+	/* Put in action number marker; this non-zero number makes sure that
+	 * find_table_space() knows that this position in chk/nxt is taken
+	 * and should not be used for another accepting number in another
+	 * state.
+	 */
+	chk[position - 1] = 1;
+
+	/* Put in end-of-buffer marker; this is for the same purposes as
+	 * above.
+	 */
+	chk[position] = 1;
+
+	/* Place the state into chk and nxt. */
+	state_ptr = &state[1];
+
+	for (i = 1; i <= numecs; ++i, ++state_ptr)
+		if (*state_ptr != 0) {
+			chk[position + i] = i;
+			nxt[position + i] = *state_ptr;
+		}
+
+	if (position + numecs > tblend)
+		tblend = position + numecs;
+}
+
+
+/* stack1 - save states with only one out-transition to be processed later
+ *
+ * If there's room for another state on the "one-transition" stack, the
+ * state is pushed onto it, to be processed later by mk1tbl.  If there's
+ * no room, we process the sucker right now.
+ */
+
+void    stack1 (statenum, sym, nextstate, deflink)
+     int     statenum, sym, nextstate, deflink;
+{
+	if (onesp >= ONE_STACK_SIZE - 1)
+		mk1tbl (statenum, sym, nextstate, deflink);
+
+	else {
+		++onesp;
+		onestate[onesp] = statenum;
+		onesym[onesp] = sym;
+		onenext[onesp] = nextstate;
+		onedef[onesp] = deflink;
+	}
+}
+
+
+/* tbldiff - compute differences between two state tables
+ *
+ * "state" is the state array which is to be extracted from the pr'th
+ * proto.  "pr" is both the number of the proto we are extracting from
+ * and an index into the save area where we can find the proto's complete
+ * state table.  Each entry in "state" which differs from the corresponding
+ * entry of "pr" will appear in "ext".
+ *
+ * Entries which are the same in both "state" and "pr" will be marked
+ * as transitions to "SAME_TRANS" in "ext".  The total number of differences
+ * between "state" and "pr" is returned as function value.  Note that this
+ * number is "numecs" minus the number of "SAME_TRANS" entries in "ext".
+ */
+
+int     tbldiff (state, pr, ext)
+     int     state[], pr, ext[];
+{
+	register int i, *sp = state, *ep = ext, *protp;
+	register int numdiff = 0;
+
+	protp = &protsave[numecs * (pr - 1)];
+
+	for (i = numecs; i > 0; --i) {
+		if (*++protp == *++sp)
+			*++ep = SAME_TRANS;
+		else {
+			*++ep = *sp;
+			++numdiff;
+		}
+	}
+
+	return numdiff;
+}

contrib/flex/version.h

@@ -0,0 +1 @@
+#define FLEX_VERSION VERSION

contrib/flex/yylex.c

@@ -0,0 +1,211 @@
+/* yylex - scanner front-end for flex */
+
+/*  Copyright (c) 1990 The Regents of the University of California. */
+/*  All rights reserved. */
+
+/*  This code is derived from software contributed to Berkeley by */
+/*  Vern Paxson. */
+
+/*  The United States Government has rights in this work pursuant */
+/*  to contract no. DE-AC03-76SF00098 between the United States */
+/*  Department of Energy and the University of California. */
+
+/*  This file is part of flex. */
+
+/*  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. */
+
+/*  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 ``AS IS'' AND WITHOUT ANY EXPRESS OR */
+/*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
+/*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
+/*  PURPOSE. */
+
+#include <ctype.h>
+#include "flexdef.h"
+#include "parse.h"
+
+
+/* yylex - scan for a regular expression token */
+
+int     yylex ()
+{
+	int     toktype;
+	static int beglin = false;
+	extern char *yytext;
+
+	if (eofseen)
+		toktype = EOF;
+	else
+		toktype = flexscan ();
+
+	if (toktype == EOF || toktype == 0) {
+		eofseen = 1;
+
+		if (sectnum == 1) {
+			synerr (_("premature EOF"));
+			sectnum = 2;
+			toktype = SECTEND;
+		}
+
+		else
+			toktype = 0;
+	}
+
+	if (trace) {
+		if (beglin) {
+			fprintf (stderr, "%d\t", num_rules + 1);
+			beglin = 0;
+		}
+
+		switch (toktype) {
+		case '<':
+		case '>':
+		case '^':
+		case '$':
+		case '"':
+		case '[':
+		case ']':
+		case '{':
+		case '}':
+		case '|':
+		case '(':
+		case ')':
+		case '-':
+		case '/':
+		case '\\':
+		case '?':
+		case '.':
+		case '*':
+		case '+':
+		case ',':
+			(void) putc (toktype, stderr);
+			break;
+
+		case '\n':
+			(void) putc ('\n', stderr);
+
+			if (sectnum == 2)
+				beglin = 1;
+
+			break;
+
+		case SCDECL:
+			fputs ("%s", stderr);
+			break;
+
+		case XSCDECL:
+			fputs ("%x", stderr);
+			break;
+
+		case SECTEND:
+			fputs ("%%\n", stderr);
+
+			/* We set beglin to be true so we'll start
+			 * writing out numbers as we echo rules.
+			 * flexscan() has already assigned sectnum.
+			 */
+			if (sectnum == 2)
+				beglin = 1;
+
+			break;
+
+		case NAME:
+			fprintf (stderr, "'%s'", nmstr);
+			break;
+
+		case CHAR:
+			switch (yylval) {
+			case '<':
+			case '>':
+			case '^':
+			case '$':
+			case '"':
+			case '[':
+			case ']':
+			case '{':
+			case '}':
+			case '|':
+			case '(':
+			case ')':
+			case '-':
+			case '/':
+			case '\\':
+			case '?':
+			case '.':
+			case '*':
+			case '+':
+			case ',':
+				fprintf (stderr, "\\%c", yylval);
+				break;
+
+			default:
+				if (!isascii (yylval) || !isprint (yylval))
+					fprintf (stderr,
+						 "\\%.3o",
+						 (unsigned int) yylval);
+				else
+					(void) putc (yylval, stderr);
+				break;
+			}
+
+			break;
+
+		case NUMBER:
+			fprintf (stderr, "%d", yylval);
+			break;
+
+		case PREVCCL:
+			fprintf (stderr, "[%d]", yylval);
+			break;
+
+		case EOF_OP:
+			fprintf (stderr, "<<EOF>>");
+			break;
+
+		case OPTION_OP:
+			fprintf (stderr, "%s ", yytext);
+			break;
+
+		case OPT_OUTFILE:
+		case OPT_PREFIX:
+		case CCE_ALNUM:
+		case CCE_ALPHA:
+		case CCE_BLANK:
+		case CCE_CNTRL:
+		case CCE_DIGIT:
+		case CCE_GRAPH:
+		case CCE_LOWER:
+		case CCE_PRINT:
+		case CCE_PUNCT:
+		case CCE_SPACE:
+		case CCE_UPPER:
+		case CCE_XDIGIT:
+			fprintf (stderr, "%s", yytext);
+			break;
+
+		case 0:
+			fprintf (stderr, _("End Marker\n"));
+			break;
+
+		default:
+			fprintf (stderr,
+				 _
+				 ("*Something Weird* - tok: %d val: %d\n"),
+				 toktype, yylval);
+			break;
+		}
+	}
+
+	return toktype;
+}