#!/usr/bin/env perl

# NOTE:  Cons intentionally does not use the "perl -w" option or
# "use strict."  Because Cons "configuration files" are actually
# Perl scripts, enabling those restrictions here would force them
# on every user's config files, wanted or not.  Would users write
# "better" Construct and Conscript files if we forced "use strict"
# on them?  Probably.  But we want people to use Cons to get work
# done, not force everyone to become a Perl guru to use it, so we
# don't insist.
#
# That said, Cons' code is both "perl -w" and "use strict" clean.
# Regression tests keep the code honest by checking for warnings
# and "use strict" failures.

use vars qw( $CVS_id $CVS_ver $ver_num $ver_rev $version );

$CVS_id = 'Id';
$CVS_ver = (split(/\s+/, $CVS_id))[2];

$ver_num = "2.3";
$ver_rev = ".1";

$version = "This is Cons $ver_num$ver_rev ($CVS_id)\n";

# Cons: A Software Construction Tool.
# Copyright (c) 1996-2001 Free Software Foundation, Inc.
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; see the file COPYING.  If not, write to
# the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
# Boston, MA 02111-1307, USA.

require 5.003;
# See the NOTE above about why Cons doesn't "use strict".
use integer;
use Cwd;
use File::Copy;

use vars qw( $_WIN32 $_a $_exe $_o $_so );

#------------------------------------------------------------------
# Determine if running on win32 platform - either Windows NT or 95
#------------------------------------------------------------------

use vars qw( $PATH_SEPARATOR $iswin32 $_WIN32 $usage $indent @targets );

BEGIN {
    use Config;

    # if the version is 5.003, we can check $^O
    if ($] <  5.003) {
	eval("require Win32");
	$_WIN32 = (!$@);
    } else {
	$_WIN32 = ($^O eq "MSWin32") ? 1 : 0;
    }

    # Fetch the PATH separator from Config;
    # provide our old defaults in case it's not set.
    $PATH_SEPARATOR = $Config{path_sep};
    $PATH_SEPARATOR = $_WIN32 ? ';' : ':' if ! defined $PATH_SEPARATOR;

    # Fetch file suffixes from Config,
    # accomodating differences in the Config variables
    # used by different Perl versions.
    $_exe = $Config{_exe};
    $_exe = $Config{exe_ext} if ! defined $_exe;
    $_exe = $_WIN32 ? '.exe' : '' if ! defined $_exe;
    $_o = $Config{_o};
    $_o = $Config{obj_ext}  if ! defined $_o;
    $_o = $_WIN32 ? '.obj' : '.o' if ! defined $_o;
    $_a = $Config{_a};
    $_a = $Config{lib_ext} if ! defined $_a;
    $_a = $_WIN32 ? '.lib' : '.a' if ! defined $_a;
    $_so = ".$Config{so}";
    $_so = $_WIN32 ? '.dll' : '.so' if ! defined $_so;
}

# Flush stdout each time.
$| = 1;

# Seed random number generator.
srand(time . $$); # this works better than time ^ $$ in perlfunc manpage.

$usage = q(
Usage: cons <arguments> -- <construct-args>

Arguments can be any of the following, in any order:

  <targets>	Build the specified targets. If <target> is a directory
		recursively build everything within that directory.

  +<pattern>	Limit the cons scripts considered to just those that
		match <pattern>. Multiple + arguments are accepted.

  <name>=<val>	Sets <name> to value <val> in the ARG hash passed to the
		top-level Construct file.

  -cc           Show command that would have been executed, when
		retrieving from cache. No indication that the file
		has been retrieved is given; this is useful for
		generating build logs that can be compared with
		real build logs.

  -cd           Disable all caching. Do not retrieve from cache nor
		flush to cache.

  -cr           Build dependencies in random order. This is useful when
		building multiple similar trees with caching enabled.

  -cs           Synchronize existing build targets that are found to be
		up-to-date with cache. This is useful if caching has
		been disabled with -cc or just recently enabled with
		UseCache.

  -d            Enable dependency debugging.

  -f <file>	Use the specified file instead of "Construct" (but first
		change to containing directory of <file>).

  -h            Show a help message local to the current build if
		one such is defined,  and exit.

  -k		Keep going as far as possible after errors.

  -o <file>	Read override file <file>.

  -p		Show construction products in specified trees.
  -pa		Show construction products and associated actions.
  -pw		Show products and where they are defined.

  -q		Be quiet; multiple -q flags increase quietness level:
		1: quiet about Installing and Removing targets
		2: quiet about build commands, up-to-date targets

  -r		Remove construction products associated with <targets>

  -R <repos>	Search for files in <repos>.  Multiple -R <repos>
		directories are searched in the order specified.

  -S <pkg>	Use package sig::<pkg> to calculate file signatures.
		Currently supported values are "md5" for MD5
		signatures (the default) and "md5::debug" for MD5
		signature debug information.

  -t            Traverse up the directory hierarchy looking for a
		Construct file, if none exists in the current directory.
		(Targets will be modified to be relative to the
		Construct file.)

  -v		Show cons version and continue processing.
  -V            Show cons version and exit.

  -wf <file>    Write all filenames considered into <file>.

  -x		Show this message and exit.


   Please report any suggestions through the cons-discuss@gnu.org mailing
   list.

   To subscribe, send mail to cons-discuss-request@gnu.org with body
   'subscribe'.

   If you find a bug, please report it through the bug-cons@gnu.org
   mailing list.

   Information about CONS can be obtained from the official cons web site
   http://www.dsmit.com/cons/ or its mirrors (listed there).

   The cons maintainers can be contacted by email at cons-maintainers@gnu.org

   User documentation of cons is contained in cons and can be obtained
   by doing 'perldoc /path/to/cons'.

);

# Simplify program name, if it is a path.
{
    my ($vol, $dir, $file) = File::Spec->splitpath(File::Spec->canonpath($0));
    $0 = $file;
}

# Default parameters.
$param::topfile = 'Construct';	# Top-level construction file.
$param::install = 1;		# Show installations
$param::build = 1;		# Build targets
### $param::show = 1;		# Show building of targets.
$param::sigpro = 'md5';		# Signature protocol.
$param::depfile = '';		# Write all deps out to this file
$param::salt = '';		# Salt derived file signatures with this.
$param::sourcesig = ['*' => 'content'];# Source file signature calculation
$param::rep_sig_times_ok = 1;	# Repository .consign times are in sync
				#   w/files.
$param::conscript_chdir = 0;	# Change dir to Conscript directory
$param::quiet = 0;		# should we show the command being executed.

@param::defaults = ();

#
$indent = '';

# Display a command while executing or otherwise. This
# should be called by command builder action methods.
sub showcom {
    print($indent . $_[0] . "\n") if ($param::quiet < 2);
}

# Default environment.
# This contains only the completely platform-independent information
# we can figure out.  Platform-specific information (UNIX, Win32)
# gets added below.
@param::base = (
     'SIGNATURE'    => [ '*' => 'build' ],
     'SUFEXE'	    => $_exe,				# '' on UNIX systems
     'SUFLIB'	    => $_a,				# '.a' on UNIX systems
     'SUFLIBS'      => "$_so:$_a",			# '.so:.a' on UNIX
     'SUFOBJ'	    => $_o,				# '.o' on UNIX systems
     'SUFMAP'       => {
	 '.c'  => 'build::command::cc',
	 '.s'  => 'build::command::cc',
	 '.S'  => 'build::command::cc',
	 '.C'  => 'build::command::cxx',
	 '.cc' => 'build::command::cxx',
	 '.cxx'=> 'build::command::cxx',
	 '.cpp'=> 'build::command::cxx',
	 '.c++'=> 'build::command::cxx',
	 '.C++'=> 'build::command::cxx',
     },
     'PERL'	    => $^X,
);

%param::rulesets =
    (
     # Defaults for Win32.
     # Defined for VC++ 6.0 by Greg Spencer <greg_spencer@acm.org>
     # Your mileage may vary.
     'msvc' => [
		'CC'             => 'cl',
		'CFLAGS'         => '/nologo',
		'CCCOM'          => '%CC %CFLAGS %_IFLAGS /c %< /Fo%>',
		'CXX'            => '%CC',
		'CXXFLAGS'       => '%CFLAGS',
		'CXXCOM'         => '%CXX %CXXFLAGS %_IFLAGS /c %< /Fo%>',
		'INCDIRPREFIX'   => '/I',
		'INCDIRSUFFIX'   => '',
		'LINK'           => 'link',
		'LINKCOM'        => '%LINK %LDFLAGS /out:%> %< %_LDIRS %LIBS',
		'LINKMODULECOM'  => '%LD /r /o %> %<',
		'LIBDIRPREFIX'   => '/LIBPATH:',
		'LIBDIRSUFFIX'   => '',
		'AR'             => 'lib',
		'ARFLAGS'        => '/nologo ',
		'ARCOM'          => "%AR %ARFLAGS /out:%> %<",
		'RANLIB'         => '',
		'LD'             => 'link',
		'LDFLAGS'        => '/nologo ',
		'PREFLIB'        => '',
		],
     # Defaults for a typical (?) UNIX platform.
     # Your mileage may vary.
     'unix' => [
		'CC'             => 'cc',
		'CFLAGS'         => '',
		'CCCOM'          => '%CC %CFLAGS %_IFLAGS -c %< -o %>',
		'CXX'            => '%CC',
		'CXXFLAGS'       => '%CFLAGS',
		'CXXCOM'         => '%CXX %CXXFLAGS %_IFLAGS -c %< -o %>',
		'INCDIRPREFIX'   => '-I',
		'INCDIRSUFFIX'   => '',
		'LINK'           => '%CXX',
		'LINKCOM'        => '%LINK %LDFLAGS -o %> %< %_LDIRS %LIBS',
		'LINKMODULECOM'  => '%LD -r -o %> %<',
		'LIBDIRPREFIX'   => '-L',
		'LIBDIRSUFFIX'   => '',
		'AR'             => 'ar',
		'ARFLAGS'        => 'r', # rs?
		'ARCOM'          => ['%AR %ARFLAGS %> %<', '%RANLIB %>'],
		'RANLIB'         => 'ranlib',
		'AS'             => 'as',
		'ASFLAGS'        => '',
		'ASCOM'          => '%AS %ASFLAGS %< -o %>',
		'LD'             => 'ld',
		'LDFLAGS'        => '',
		'PREFLIB'        => 'lib',
		'ENV'            => { 'PATH' => '/bin:/usr/bin' },
		],
     );

# Set the rules based on the platform.
script::DefaultRules(script::RuleSet($_WIN32 ? 'msvc' : 'unix'));

# Handle command line arguments.
while (@ARGV) {
    $_ = shift @ARGV;
    last if /^--$/;		# Argument passing to Construct.
    &option, next			if s/^-//;
    push (@param::include, $_), next	if s/^\+//;
    &equate, next			if /=/;
    push (@targets, $_), next;
}

sub option {
    my %opt = (
		    'cc' =>   sub { $param::cachecom = 1; },
		    'cd' =>   sub { $param::cachedisable = 1; },
		    'cr' =>   sub { $param::random = 1; },
		    'cs' =>   sub { $param::cachesync = 1; },
		    'd' =>    sub { $param::depends = 1; },
		    'h' =>    sub { $param::localhelp = 1; },
		    'k' =>    sub { $param::kflag = 1; },
		    'p' =>    sub { $param::pflag = 1;
				    $param::build = 0; },
		    'pa' =>   sub { $param::pflag = 1;
				    $param::aflag = 1;
				    $indent = "... ";
				    $param::build = 0; },
		    'pw' =>   sub { $param::pflag = 1;
				    $param::wflag = 1;
				    $param::build = 0; },
		    'q' =>    sub { $param::quiet++; },
		    'r' =>    sub { $param::rflag = 1;
				    $param::build = 0; },
		    't' =>    sub { $param::traverse = 1; },
		    'v' =>    sub { print($version); },
		    'V' =>    sub { print($version), exit(0); },
		    'x' =>    sub { print($usage), exit 0; },
		);

    my %opt_arg = (
		    'f' =>    sub { $param::topfile = $_[0]; },
		    'o' =>    sub { $param::overfile = $_[0]; },
		    'R' =>    sub { script::Repository($_[0]); },
		    'S' =>    sub { $param::sigpro = $_[0]; },
		    'wf' =>   sub { $param::depfile = $_[0]; },
		);

    if (defined $opt{$_}) {
	&{$opt{$_}}();
	return;
    }
    while ($_) {
	$_  =~ m/(.)(.*)/;
	if (defined $opt{$1}) {
	    &{$opt{$1}}();
	    $_ = $2;
	    next;
	}
	if (defined $opt_arg{$1}) {
	    if (! $2) {
		$_ = shift @ARGV;
		die("$0: -$1 option requires an argument.\n") if ! $_;
	    }
	    &{$opt_arg{$1}}($2 || $_);
	    return;
	}
	$_  =~ m/(..)(.*)/;
	if (defined $opt_arg{$1}) {
	    if (! $2) {
		$_ = shift @ARGV;
		die("$0: -$1 option requires an argument.\n") if ! $_;
	    }
	    &{$opt_arg{$1}}($2 || $_);
	    return;
	}
	if ($_) {
	    die qq($0: unrecognized option "-$_".  Use -x for a usage message.\n);
	}
    }
}

# Process an equate argument (var=val).
sub equate {
    my($var, $val) = /([^=]*)=(.*)/;
    $script::ARG{$var} = $val;
}

# Define file signature protocol.
'sig'->select($param::sigpro);

# Cleanup after an interrupt.
$SIG{INT} = $SIG{QUIT} = $SIG{TERM} = sub {
    $SIG{PIPE} = $SIG{INT} = $SIG{QUIT} = $SIG{TERM} = 'IGNORE';
    $SIG{HUP} = $SIG{INT} if ! $main::_WIN32;
    warn("\n$0: killed\n");
    # Call this first, to make sure that this processing
    # occurs even if a child process does not die (and we
    # hang on the wait).
    sig::hash::END();
    wait();
    exit(1);
};
$SIG{HUP} = $SIG{INT} if ! $main::_WIN32;

# Cleanup after a broken pipe (someone piped our stdout?)
$SIG{PIPE} = sub {
    $SIG{PIPE} = $SIG{HUP} = $SIG{INT} = $SIG{QUIT} = $SIG{TERM} = 'IGNORE';
    warn("\n$0: broken pipe\n");
    sig::hash::END();
    wait();
    exit(1);
};

if ($param::depfile) {
  open (main::DEPFILE, ">".$param::depfile) ||
    die ("$0: couldn't open $param::depfile ($!)\n");
}

# If the supplied top-level Conscript file is not in the
# current directory, then change to that directory.
{
    my ($vol, $dir, $file) =
      File::Spec->splitpath(File::Spec->canonpath($param::topfile));
    if ($vol || $dir) {
	my($cd) = File::Spec->catpath($vol, $dir, undef);
	chdir($cd) || die("$0: couldn't change to directory $cd ($!)\n");
	$param::topfile = $file;
    }
}

# Walk up the directory hierarchy looking for a Conscript file (if -t set).
my($target_top);
my(@targetdir) = ();
if ($param::traverse && ! -f $param::topfile) {
    my($vol, $dirs, $file) = File::Spec->splitpath(cwd());
    my(@dirs) = (File::Spec->splitdir($dirs), $file);
    while (! -f File::Spec->catpath($vol, File::Spec->catdir(@dirs),
				    $param::topfile)) {
	die("$0: unable to find $param::topfile.\n") if ! @dirs;
	unshift(@targetdir, pop(@dirs));
    }
    my($cwd) = File::Spec->catpath($vol, File::Spec->catdir(@dirs), '');
    print "$0: Entering directory `$cwd'\n";
    chdir($cwd);
    @targets = map {File::Spec->catdir(@targetdir, $_)} @targets;
}

# Set up $dir::top and $dir::cwd, now that we are in the right directory.
dir::init();

#
if (@targetdir) {
    $target_top = $dir::top->lookupdir(File::Spec->catdir(@targetdir));
}

# Now handle override file.
package override;
if ($param::overfile) {
    my($ov) = $param::overfile;
    die qq($0: can\'t read override file "$ov" ($!)\n) if ! -f $ov; #'
    do $ov;
    if ($@) {
	chop($@);
	die qq($0: errors in override file "$ov" ($@)\n);
    }
}

# Provide this to user to setup override patterns.
sub Override {
    my($re, @env) = @_;
    return if $param::overrides{$re}; # if identical, first will win.
    $param::overrides = 1;
    $param::overrides{$re} = \@env;
    push(@param::overrides, $re);
}

package main;

use vars qw( %priority $errors );

# Check script inclusion regexps
my $re;
for $re (@param::include) {
    if (! defined eval {"" =~ /$re/}) {
	my($err) = $@;
	$err =~ s/in regexp at .*$//;
	die("$0: error in regexp $err");
    }
}

# Read the top-level construct file and its included scripts.
doscripts($param::topfile);

# Status priorities. This lets us aggregate status for directories
# and print an appropriate message (at the top-level).
%priority =
    ('none' => 1, 'handled' => 2, 'built' => 3, 'unknown' => 4, 'errors' => 5);

# If no targets were specified, supply default targets (if any).
@targets = @param::default_targets if ! @targets;

$errors = 0;

# Build the supplied target patterns.
my $tgt;
for $tgt (map($dir::top->lookup($_), @targets)) {
    if ($target_top && ! $tgt->is_under($target_top)) {
	# A -t option was used, and this target is not underneath
	# the directory where we were invoked via -t.
	# If the target is a directory and the -t directory
	# is underneath it, then build the -t directory.
	if (ref $tgt ne "dir" || ! $target_top->is_under($tgt)) {
	    next;
	}
	$tgt = $target_top;
    }
    buildtoptarget($tgt);
}

exit 0 + ($errors != 0);

sub buildtoptarget {
    my($tgt) = @_;
    return if ! $tgt;
    my($status) = buildtarget($tgt);
    if ($status ne 'built') {
	my($path) = $tgt->path;
	if ($status eq "errors") {
	    print qq($0: "$path" not remade because of errors.\n);
	    $errors++;
	} elsif ($status eq "handled") {
	    print qq($0: "$path" is up-to-date.\n) if ($param::quiet < 2);
	} elsif ($status eq "unknown") {
	    # cons error already reported.
	    $errors++;
	} elsif ($status eq "none") {
	    # search for targets that may be linked to the given path.
	    my @linked = dir::linked_targets($tgt) if $target_top;
	    if (@linked) {
		my @names = map($_->path, @linked);
		print "Linked targets: @names\n" if ($param::quiet < 1);
		map(buildtoptarget($_), @linked);
	    } else {
		print qq($0: nothing to be built in "$path".\n)
		      if $param::build && ($param::quiet < 2);
	    }
	} else {
	    print qq($0: don\'t know how to construct "$path".\n); #'
	    $errors++;
	}
    }
}

# Build the supplied target directory or files. Return aggregated status.
sub buildtarget {
    my($tgt) = @_;
    if (ref($tgt) eq "dir") {
	my($result) = "none";
	my($priority) = $priority{$result};
	if (exists $tgt->{member}) {
	    my($members) = $tgt->{member};
	    my $entry;
	    for $entry (sort keys %$members) {
		next if $entry eq $dir::CURDIR || $entry eq $dir::UPDIR;
		my($tgt) = $members->{$entry};
		next if ref($tgt) ne "dir" && !exists($tgt->{builder});
		my($stat) = buildtarget($members->{$entry});
		my($pri) = $priority{$stat};
		if ($pri > $priority) {
		    $priority = $pri;
		    $result = $stat;
		}
	    }
	}
	return $result;
    }
    if ($param::depends) {
	my($path) = $tgt->path;
	if ($tgt->{builder}) {
	    my(@dep) = (@{$tgt->{dep}}, @{$tgt->{sources}});
	    my($dep) = join(' ',map($_->path, @dep));
	    print("Target $path: $dep\n");
	} else {
	    print("Target $path: not a derived file\n");
	}
    }
    if ($param::build) {
	return build $tgt;
    } elsif ($param::pflag || $param::wflag || $param::aflag) {
	if ($tgt->{builder}) {
	    if ($param::wflag) {
		print qq(${\$tgt->path}: $tgt->{script}\n);
	    } elsif ($param::pflag) {
		print qq(${\$tgt->path}:\n) if $param::aflag;
		print qq(${\$tgt->path}\n) if !$param::aflag;
	    }
	    if ($param::aflag) {
		$tgt->{builder}->action($tgt);
	    }
	}
    } elsif ($param::rflag && $tgt->{builder}) {
	my($path) = $tgt->path;
	if (-f $path) {
	    if (unlink($path)) {
		print("Removed $path\n") if ($param::quiet < 1);
	    } else {
		warn("$0: couldn't remove $path\n");
	    }
	}
    }

    return "none";
}

package NameSpace;

# Return a hash that maps the name of symbols in a namespace to an
# array of refs for all types for which the name has a defined value.
# A list of symbols may be specified; default is all symbols in the
# name space.
sub save {
    my $package = shift;
    my(%namerefs, $var, $type);
    no strict 'refs';
    @_ = keys %{$package."::"} if ! @_;
    foreach $var (@_) {
	$namerefs{$var} = [];
	my $fqvar = $package."::".$var;
	# If the scalar for this variable name doesn't already
	# exist, *foo{SCALAR} will autovivify the reference
	# instead of returning undef, so unlike the other types,
	# we have to dereference to find out if it exists.
	push(@{$namerefs{$var}}, *{$fqvar}{SCALAR})
		    if defined ${*{$fqvar}{SCALAR}};
	foreach $type (qw(ARRAY HASH CODE IO)) {
	    push(@{$namerefs{$var}}, *{$fqvar}{$type})
			  if defined *{$fqvar}{$type};
	}
    }
    return \%namerefs;
}

# Remove the specified symbols from the namespace.
# Default is to remove all.
sub remove {
    my $package = shift;
    my(%namerefs, $var);
    no strict 'refs';
    @_ = keys %{$package."::"} if ! @_;
    foreach $var (@_) {
	delete ${$package."::"}{$var};
    }
}

# Restore values to symbols specified in a hash as returned
# by NameSpace::save.
sub restore {
    my($package, $namerefs) = @_;
    my($var, $ref);
    no strict 'refs';
    foreach $var (keys %$namerefs) {
	my $fqvar = $package."::".$var;
	foreach $ref (@{$namerefs->{$var}}) {
	    *{$fqvar} = $ref;
	}
    }
}

# Support for "building" scripts, importing and exporting variables.
# With the exception of the top-level routine here (invoked from the
# main package by cons), these are all invoked by user scripts.
package script;

use vars qw( $ARG $caller_dir_path %special_var );

BEGIN {
    # We can't Export or Import the following variables because Perl always
    # treats them as part of the "main::" package (see perlvar(1)).
    %special_var = map {$_ => 1} qw(ENV INC ARGV ARGVOUT SIG
				    STDIN STDOUT STDERR);
}

# This is called from main to interpret/run the top-level Construct
# file, passed in as the single argument.
sub main::doscripts {
    my($script) = @_;
    Build($script);
    # Now set up the includes/excludes (after the Construct file is read).
    $param::include = join('|', @param::include);

    # Save the original variable names from the script package.
    # These will stay intact, but any other "script::" variables
    # defined in a Conscript file will get saved, deleted,
    # and (when necessary) restored.
    my(%orig_script_var) = map {$_ => 1} keys %script::;
    $caller_dir_path = undef;
    my $cwd = Cwd::cwd();
    my(@scripts) = pop(@priv::scripts);
    while ($priv::self = shift(@scripts)) {
	my($path) = $priv::self->{script}->rsrcpath;
	if (-f $path) {
	    $dir::cwd = $priv::self->{script}->{dir};
	    # Handle chdir to the Conscript file directory, if necessary.
	    my ($vol, $dir, $file);
	    if ($param::conscript_chdir) {
		($vol, $dir, $file) =
		  File::Spec->splitpath(File::Spec->canonpath($path));
		if ($vol ne '' || $dir ne '') {
		    $caller_dir_path = File::Spec->catpath($vol, $dir, undef);
		    chdir($caller_dir_path) ||
			die "Could not chdir to $caller_dir_path: $!\n";
		}
	    } else {
		$file = $path;
	    }
	    # Actually process the Conscript file.
	    do $file;
	    # Save any variables defined by the Conscript file
	    # so we can restore them later, if needed;
	    # then delete them from the script:: namespace.
	    my(@del) = grep(! $orig_script_var{$_}, keys %script::);
	    if (@del) {
		$priv::self->{script}->{pkgvars} = NameSpace::save('script',
								   @del);
		NameSpace::remove('script', @del);
	    }
	    if ($caller_dir_path) {
		chdir($cwd);
		$caller_dir_path = undef;
	    }
	    if ($@) {
		chomp($@);
		my $err = ($@ =~ /\n/ms) ? ":\n$@" : " ($@)";
		print qq($0: error in file "$path"$err\n);
		$run::errors++;
	    } else {
		# Only process subsidiary scripts if no errors in parent.
		unshift(@scripts, @priv::scripts);
	    }
	    undef @priv::scripts;
	} else {
	    my $where = '';
	    my $cref = $priv::self->{script}->creator;
	    if (defined $cref) {
		my($_foo, $script, $line, $sub) = @$cref;
		$where = " ($sub in $script, line $line)";
	    }
	    warn qq(Ignoring missing script "$path"$where);
	}
    }
    die("$0: script errors encountered: construction aborted\n")
	if $run::errors;
}

# Return caller info about the method being invoked.
# This is everything from the Perl "caller" builtin function,
# including which Construct/Conscript file, line number,
# subroutine name, etc.
sub caller_info {
    my($lev) = 1;
    my(@frame);
    do {
	@frame = caller ++$lev;
	if (defined($frame[3]) && $frame[3] eq '(eval)') {
	    @frame = caller --$lev;
	    if ($caller_dir_path) {
		$frame[1] = File::Spec->catfile($caller_dir_path, $frame[1]);
	    }
	    return @frame;
	}
    } while ($frame[3]);
    return;
}

# Link a directory to another. This simply means set up the *source*
# for the directory to be the other directory.
sub Link {
    dir::link(@_);
}

# Add directories to the repository search path for files.
# Strip our current directory from the list so Repository
# (or -R options) can be used from within the repository.
sub Repository {
    my($my_dir) = Cwd::cwd();
    my $dir;
    foreach $dir (@_) {
	# The following more direct call isn't available in
	# Cwd.pm until some time after 5.003...
	#	my($d) = Cwd::abs_path($dir);
	chdir($dir);
	my($d) = Cwd::cwd();
	chdir($my_dir);
	#
	next if ! $d || ! -d $d || $d eq $my_dir;
	# We know we can get away with passing undef to lookupdir
	# as the directory because $dir is an absolute path.
	push(@param::rpath, dir::lookupdir(undef, $dir));
	push @INC, $d;
    }
}

# Return the list of Repository directories specified.
sub Repository_List {
    map($_->path, @param::rpath);
}

# Specify whether the .consign signature times in repository files are,
# in fact, consistent with the times on the files themselves.
sub Repository_Sig_Times_OK {
    $param::rep_sig_times_ok = shift;
}

sub SourceSignature {
    $param::sourcesig = [@_];
}

# Specify whether we should chdir to the containing directories
# of Conscript files.
sub Conscript_chdir {
    $param::conscript_chdir = shift;
}

# Specify files/targets that must be present and built locally,
# even if they exist already-built in a Repository.
sub Local {
    my(@files) = map($dir::cwd->lookupfile($_), @_);
    map($_->local(1), @files);
}

# Export variables to any scripts invoked from this one.
sub Export {
    my(@illegal) = grep($special_var{$_}, @_);
    if (@illegal) {
	die qq($0: cannot Export special Perl variables: @illegal\n);
    }
    @{$priv::self->{exports}} = grep(! defined $special_var{$_}, @_);
}

# Import variables from the export list of the caller
# of the current script.
sub Import {
    my(@illegal) = grep($special_var{$_}, @_);
    if (@illegal) {
	die qq($0: cannot Import special Perl variables: @illegal\n);
    }
    my($parent) = $priv::self->{parent};
    my($imports) = $priv::self->{imports};
    @{$priv::self->{exports}} = keys %$imports;
    my($var);
    foreach $var (grep(! defined $special_var{$_}, @_)) {
	if (!exists $imports->{$var}) {
	    my($path) = $parent->{script}->path;
	    die qq($0: variable "$var" not exported by file "$path"\n);
	}
	if (!defined $imports->{$var}) {
	    my $path = $parent->{script}->path;
	    my $err = "$0: variable \"$var\" exported but not " .
		      "defined by file \"$path\"\n";
	    die $err;
	}
	${"script::$var"} = $imports->{$var};
    }
}

# Build an inferior script. That is, arrange to read and execute
# the specified script, passing to it any exported variables from
# the current script.
sub Build {
    my(@files) = map($dir::cwd->lookupfile($_), @_);
    my(%imports) = map {$_ => ${"script::$_"}} @{$priv::self->{exports}};
    my $file;
    for $file (@files) {
	next if $param::include && $file->path !~ /$param::include/o;
	my($self) = {'script' => $file,
		     'parent' => $priv::self,
		     'imports' => \%imports};
	bless $self;  # may want to bless into class of parent in future
	push(@priv::scripts, $self);
    }
}

# Set up regexps dependencies to ignore. Should only be called once.
sub Ignore {
    die("Ignore called more than once\n") if $param::ignore;
    $param::ignore = join("|", map("($_)", @_)) if @_;
}

# Specification of default targets.
sub Default {
    push(@param::default_targets, map($dir::cwd->lookup($_)->path, @_));
}

# Local Help.  Should only be called once.
sub Help {
    if ($param::localhelp) {
	print "@_\n";
	exit 2;
    }
}

# For windows platforms which use unix tool sets, the msvc defaults may
# not be useful. Also, in the future, other platforms (Mac?) may have the
# same problem.
sub RuleSet {
    my $style = shift;
    my @rulesets = sort keys %param::rulesets;
    die "Unknown style for rules: $style.\n" .
	"Supported rules are: (" . join(" ", @rulesets) . ")"
	    unless eval(join("||", map("\$style eq '$_'", @rulesets)));
    return @param::base, @{$param::rulesets{$style}};
}

sub DefaultRules {
    @param::defaults = ();
    push @param::defaults, @_;
}

# Return the build name(s) of a file or file list.
sub FilePath {
    wantarray
	? map($dir::cwd->lookupfile($_)->path, @_)
	: $dir::cwd->lookupfile($_[0])->path;
}

# Return the build name(s) of a directory or directory list.
sub DirPath {
    wantarray
	? map($dir::cwd->lookupdir($_)->path, @_)
	: $dir::cwd->lookupdir($_[0])->path;
}

# Split the search path provided into components. Look each up
# relative to the current directory.
# The usual path separator problems abound; for now we'll use :
sub SplitPath {
    my($dirs) = @_;
    if (ref($dirs) ne "ARRAY") {
	$dirs = [ split(/$main::PATH_SEPARATOR/o, $dirs) ];
    }
    map { DirPath($_) } @$dirs;
}

# Return true if the supplied path is available as a source file
# or is buildable (by rules seen to-date in the build).
sub ConsPath {
    my($path) = @_;
    my($file) = $dir::cwd->lookup($path);
    return $file->accessible;
}

# Return the source path of the supplied path.
sub SourcePath {
    wantarray
	? map($dir::cwd->lookupfile($_)->rsrcpath, @_)
	: $dir::cwd->lookupfile($_[0])->rsrcpath;
}

# Search up the tree for the specified cache directory, starting with
# the current directory. Returns undef if not found, 1 otherwise.
# If the directory is found, then caching is enabled. The directory
# must be readable and writable. If the argument "mixtargets" is provided,
# then targets may be mixed in the cache (two targets may share the same
# cache file--not recommended).
sub UseCache($@) {
    my($dir, @args) = @_;
    # NOTE: it's important to process arguments here regardless of whether
    # the cache is disabled temporarily, since the mixtargets option affects
    # the salt for derived signatures.
    for (@args) {
	if ($_ eq "mixtargets") {
	    # When mixtargets is enabled, we salt the target signatures.
	    # This is done purely to avoid a scenario whereby if
	    # mixtargets is turned on or off after doing builds, and
	    # if cache synchronization with -cs is used, then
	    # cache files may be shared in the cache itself (linked
	    # under more than one name in the cache). This is not bad,
	    # per se, but simply would mean that a cache cleaning algorithm
	    # that looked for a link count of 1 would never find those
	    # particular files; they would always appear to be in use.
	    $param::salt = 'M' . $param::salt;
	    $param::mixtargets = 1;
	} else {
	    die qq($0: UseCache unrecognized option "$_"\n);
	}
    }
    if ($param::cachedisable) {
	warn("Note: caching disabled by -cd flag\n");
	return 1;
    }
    my($depth) = 15;
    while ($depth-- && ! -d $dir) {
	$dir = File::Spec->catdir($dir::UPDIR, $dir);
    }
    if (-d $dir) {
	$param::cache = $dir;
	return 1;
    }
    return undef;
}

# Salt the signature generator. The salt (a number of string) is added
# into the signature of each derived file. Changing the salt will
# force recompilation of all derived files.
sub Salt($) {
    # We append the value, so that UseCache and Salt may be used
    # in either order without changing the signature calculation.
    $param::salt .= $_[0];
}

# Mark files (or directories) to not be removed before building.
sub Precious {
    map($_->{precious} = 1, map($dir::cwd->lookup($_), @_));
}


# These methods are callable from Conscript files, via a cons
# object. Procs beginning with _ are intended for internal use.
package cons;

use vars qw( %envcache );

# This is passed the name of the base environment to instantiate.
# Overrides to the base environment may also be passed in
# as key/value pairs.
sub new {
    my($package) = shift;
    my ($env) = {@param::defaults, @_};
    @{$env->{_envcopy}} = %$env; # Note: we never change PATH
    $env->{_cwd} = $dir::cwd; # Save directory of environment for
    bless $env, $package;	# any deferred name interpretation.
}

# Clone an environment.
# Note that the working directory will be the initial directory
# of the original environment.
sub clone {
    my($env) = shift;
    my $clone = {@{$env->{_envcopy}}, @_};
    @{$clone->{_envcopy}} = %$clone; # Note: we never change PATH
    $clone->{_cwd} = $env->{_cwd};
    bless $clone, ref $env;
}

# Create a flattened hash representing the environment.
# It also contains a copy of the PATH, so that the path
# may be modified if it is converted back to a hash.
sub copy {
    my($env) = shift;
    (@{$env->{_envcopy}}, 'ENV' => {%{$env->{ENV}}}, @_)
}

# Resolve which environment to actually use for a given
# target. This is just used for simple overrides.
sub _resolve {
    return $_[0] if !$param::overrides;
    my($env, $tgt) = @_;
    my($path) = $tgt->path;
    my $re;
    for $re (@param::overrides) {
	next if $path !~ /$re/;
	# Found one. Return a combination of the original environment
	# and the override.
	my($ovr) = $param::overrides{$re};
	return $envcache{$env,$re} if $envcache{$env,$re};
	my($newenv) = {@{$env->{_envcopy}}, @$ovr};
	@{$newenv->{_envcopy}} = %$env;
	$newenv->{_cwd} = $env->{_cwd};
	return $envcache{$env,$re} = bless $newenv, ref $env;
    }
    return $env;
}

# Substitute construction environment variables into a string.
# Internal function/method.
sub _subst {
    my($env, $str) = @_;
    if (! defined $str) {
	return undef;
    } elsif (ref($str) eq "ARRAY") {
	return [ map($env->_subst($_), @$str) ];
    } else {
	# % expansion.  %% gets converted to % later, so expand any
	# %keyword construction that doesn't have a % in front of it,
	# modulo multiple %% pairs in between.
	# In Perl 5.005 and later, we could actually do this in one regex
	# using a conditional expression as follows,
	#	while ($str =~ s/($pre)\%(\{)?([_a-zA-Z]\w*)(?(2)\})/"$1".
	#                      $env->{$3}/ge) {}
	# The following two-step approach is backwards-compatible
	# to (at least) Perl5.003.
	my $pre = '^|[^\%](?:\%\%)*';
	while (($str =~ s/($pre)\%([_a-zA-Z]\w*)/$1.($env->{$2}||'')/ge) ||
	       ($str =~ s/($pre)\%\{([_a-zA-Z]\w*)\}/$1.($env->{$2}||'')/ge)) {
	}
	return $str;
    }
}

sub AfterBuild {
    my($env) = shift;
    my($perl_eval_str) = pop(@_);
    my $file;
    for $file (map($dir::cwd->lookup($_), @_)) {
	$file->{after_build_func} = $perl_eval_str;
    }
}

sub Install {
    my($env) = shift;
    my($tgtdir) = $dir::cwd->lookupdir($env->_subst(shift));
    my $file;
    for $file (map($dir::cwd->lookupfile($env->_subst($_)), @_)) {
	my($tgt) = $tgtdir->lookupfile($file->{entry});
	$tgt->bind(find build::install($env), $file);
    }
}

sub InstallAs {
    my $env = shift;
    my $tgt = shift;
    my $src = shift;
    my @sources = ();
    my @targets = ();

    if (ref $tgt) {
	die "InstallAs: Source is a file and target is a list!\n"
	    if (!ref($src));
	@sources = @$src;
	@targets = @$tgt;
    } elsif (ref $src) {
	die "InstallAs: Target is a file and source is a list!\n";
    } else {
	push @sources, $src;
	push @targets, $tgt;
    }

    if ($#sources != $#targets) {
	my $tn = $#targets+1;
	my $sn = $#sources+1;
	die "InstallAs: Source file list ($sn) and target file list ($tn) " .
	    "are inconsistent in length!\n";
    } else {
	foreach (0..$#sources) {
	    my $tfile = $dir::cwd->lookupfile($env->_subst($targets[$_]));
	    my $sfile = $dir::cwd->lookupfile($env->_subst($sources[$_]));
	    $tfile->bind(find build::install($env), $sfile);
	}
    }
}

# Installation in a local build directory,
# copying from the repository if it's already built there.
# Functionally equivalent to:
#	Install $env $dir, $file;
#	Local "$dir/$file";
sub Install_Local {
    my($env) = shift;
    my($tgtdir) = $dir::cwd->lookupdir($env->_subst(shift));
    my $file;
    for $file (map($dir::cwd->lookupfile($env->_subst($_)), @_)) {
	my($tgt) = $tgtdir->lookupfile($file->{entry});
	$tgt->bind(find build::install($env), $file);
	$tgt->local(1);
    }
}

sub Objects {
    my($env) = shift;
    map($dir::cwd->relpath($_), $env->_Objects(@_));
}

# Called with multiple source file references (or object files).
# Returns corresponding object files references.
sub _Objects {
    my($env) = shift;
    my($suffix) = $env->{SUFOBJ};
    map($env->_Object($_, $_->{dir}->lookupfile($_->base_suf($suffix))),
	map { ref $_ ? $_ : $dir::cwd->lookupfile($env->_subst($_)) }
		grep(defined $_, @_));
}

# Called with an object and source reference.  If no object reference
# is supplied, then the object file is determined implicitly from the
# source file's extension. Sets up the appropriate rules for creating
# the object from the source.  Returns the object reference.
sub _Object {
    my($env, $src, $obj) = @_;
    return $obj if $src eq $obj; # don't need to build self from self.
    my($objenv) = $env->_resolve($obj);
    my($suffix) = $src->suffix;

    my($builder) = $env->{SUFMAP}{$suffix};

    if ($builder) {
	$obj->bind((find $builder($objenv)), $src);
    } else {
	die("don't know how to construct ${\$obj->path} from " .
	    "${\$src->path}.\n");
    }
    $obj
}

sub Program {
    my($env) = shift;
    my($tgt) = $dir::cwd->lookupfile(file::addsuffix($env->_subst(shift),
						 $env->{SUFEXE}));
    my($progenv) = $env->_resolve($tgt);
    $tgt->bind(find build::command::link($progenv, $progenv->{LINKCOM}),
	       $env->_Objects(@_));
}

sub Module {
    my($env) = shift;
    my($tgt) = $dir::cwd->lookupfile($env->_subst(shift));
    my($modenv) = $env->_resolve($tgt);
    my($com) = pop(@_);
    $tgt->bind(find build::command::link($modenv, $com), $env->_Objects(@_));
}

sub LinkedModule {
    my($env) = shift;
    my($tgt) = $dir::cwd->lookupfile($env->_subst(shift));
    my($progenv) = $env->_resolve($tgt);
    $tgt->bind(find build::command::linkedmodule
	       ($progenv, $progenv->{LINKMODULECOM}),
	       $env->_Objects(@_));
}

sub Library {
    my($env) = shift;
    my($lib) = $dir::cwd->lookupfile(file::addsuffix($env->_subst(shift),
						 $env->{SUFLIB}));
    my($libenv) = $env->_resolve($lib);
    $lib->bind(find build::command::library($libenv), $env->_Objects(@_));
}

# Simple derivation: you provide target, source(s), command.
# Special variables substitute into the rule.
# Target may be a reference, in which case it is taken
# to be a multiple target (all targets built at once).
sub Command {
    my($env) = shift;
    my($tgt) = $env->_subst(shift);
    my($builder) = find build::command::user($env, pop(@_), 'script');
    my(@sources) = map($dir::cwd->lookupfile($env->_subst($_)), @_);
    if (ref($tgt)) {
	# A multi-target command.
	my(@tgts) = map($dir::cwd->lookupfile($_), @$tgt);
	die("empty target list in multi-target command\n") if !@tgts;
	$env = $env->_resolve($tgts[0]);
	my($multi) = build::multiple->new($builder, \@tgts);
	for $tgt (@tgts) {
	    $tgt->bind($multi, @sources);
	}
    } else {
	$tgt = $dir::cwd->lookupfile($tgt);
	$env = $env->_resolve($tgt);
	$tgt->bind($builder, @sources);
    }
}

sub Depends {
    my($env) = shift;
    my($tgt) = $env->_subst(shift);
    my(@deps) = map($dir::cwd->lookup($env->_subst($_)), @_);
    if (! ref($tgt)) {
	$tgt = [ $tgt ];
    }
    my($t);
    foreach $t (map($dir::cwd->lookupfile($_), @$tgt)) {
	push(@{$t->{dep}}, @deps);
    }
}

# Setup a quick scanner for the specified input file, for the
# associated environment. Any use of the input file will cause the
# scanner to be invoked, once only. The scanner sees just one line at
# a time of the file, and is expected to return a list of
# dependencies.
sub QuickScan {
    my($env, $code, $file, $path) = @_;
    $dir::cwd->lookup($env->_subst($file))->{'srcscan',$env} =
	find scan::quickscan($code, $env, $env->_subst($path));
}

# Generic builder module. Just a few default methods.  Every derivable
# file must have a builder object of some sort attached.  Usually
# builder objects are shared.
package build;

use vars qw( %builder );

# Every builder must now have at least an associated environment,
# so we can find its sigarray and calculate the proper signature.
sub find {
    my($class, $env) = @_;
    $builder{$env} || do {
	my $self = { env => $env };
	$builder{$env} = bless $self, $class;
    }
}

# Null signature for dynamic includes.
sub includes { () }

# Null signature for build script.
sub scriptsig { () }

# Not compatible with any other builder, by default.
sub compatible { 0 }


# Builder module for the Install command.
package build::install;

use vars qw( @ISA );

BEGIN { @ISA = qw(build) }

# Caching not supported for Install: generally install is trivial anyway,
# and we don't want to clutter the cache.
sub cachin { undef }
sub cachout { }

# Do the installation.
sub action {
    my($self, $tgt) = @_;
    my($src) = $tgt->{sources}[0];
    main::showcom("Install ${\$src->rpath} as ${\$tgt->path}")
	if ($param::install && $param::quiet < 1);
    return unless $param::build;
    futil::install($src->rpath, $tgt);
    return 1;
}


# Builder module for generic UNIX commands.
package build::command;

use vars qw( @ISA %com );

BEGIN { @ISA = qw(build) }

sub find {
    my($class, $env, $cmd, $package) = @_;
    my($act) = action::new($env, $cmd);
    $package ||= '';
    $com{$env,$act,$package} || do {
	my $self = { env => $env, act => $act, 'package' => $package };
	$com{$env,$act,$package} = bless $self, $class;
    }
}

# Default cache in function.
sub cachin {
    my($self, $tgt, $sig) = @_;
    if (cache::in($tgt, $sig)) {
	if ($param::cachecom) {
	    $self->{act}->show($self->{env}, $tgt);
	} else {
	    printf("Retrieved %s from cache\n", $tgt->path)
		if ($param::quiet < 1);
	}
	return 1;
    }
    return undef;
}

# Default cache out function.
sub cachout {
    my($self, $tgt, $sig) = @_;
    cache::out($tgt, $sig);
}

# Build the target using the previously specified commands.
sub action {
    my($self, $tgt) = @_;
    $self->{act}->execute($self->{env}, $tgt, $self->{'package'});
}

# Return script signature.
sub scriptsig {
    $_[0]->{act}->scriptsig
}


# Create a linked module.
package build::command::link;

use vars qw( @ISA );

BEGIN { @ISA = qw(build::command) }

# Find an appropriate linker.
sub find {
    my($class, $env, $command) = @_;
    if (!exists $env->{_LDIRS}) {
	my($ldirs) = '';
	my($wd) = $env->{_cwd};
	my($pdirs) = $env->{LIBPATH};
	if (! defined $pdirs) {
	    $pdirs = [ ];
	} elsif (ref($pdirs) ne 'ARRAY') {
	    $pdirs = [ split(/$main::PATH_SEPARATOR/o, $pdirs) ];
	}
	my($dir, $dpath);
	for $dir (map($wd->lookupdir($env->_subst($_)), @$pdirs)) {
	    $dpath = $dir->path;
	    # Add the (presumably local) directory to the -L flags
	    # if we're not using repositories, the directory exists,
	    # or it's Linked to a source directory (that is, it *will*
	    # exist by the time the link occurs).
	    $ldirs .= " ".$env->{LIBDIRPREFIX}.$dpath.$env->{LIBDIRSUFFIX}
			if ! @param::rpath || -d $dpath || $dir->is_linked;
	    next if File::Spec->file_name_is_absolute($dpath);
	    if (@param::rpath) {
		my $d;
		if ($dpath eq $dir::CURDIR) {
		    foreach $d (map($_->path, @param::rpath)) {
			$ldirs .= " " . $env->{LIBDIRPREFIX} .
				  $d . $env->{LIBDIRSUFFIX};
		    }
		} else {
		    my($rpath);
		    foreach $d (map($_->path, @param::rpath)) {
			$rpath = File::Spec->catfile($d, $dpath);
			$ldirs .= " ". $env->{LIBDIRPREFIX} .
				  $rpath . $env->{LIBDIRSUFFIX} if -d $rpath;
		    }
		}
	    }
	}
	$env->{_LDIRS} = "%($ldirs%)";
    }

    # Introduce a new magic _LIBS symbol which allows to use the
    # Unix-style -lNAME syntax for Win32 only. -lNAME will be replaced
    # with %{PREFLIB}NAME%{SUFLIB}. <schwarze@isa.de> 1998-06-18

    if ($main::_WIN32 && !exists $env->{_LIBS}) {
	my $libs;
	my $name;
	for $name (split(' ', $env->_subst($env->{LIBS} || ''))) {
	    if ($name =~ /^-l(.*)/) {
		$name = "$env->{PREFLIB}$1$env->{SUFLIB}";
	    }
	    $libs .= ' ' . $name;
	}
	$env->{_LIBS} = $libs ? "%($libs%)" : '';
    }
    bless find build::command($env, $command);
}

# Called from file::build. Make sure any libraries needed by the
# environment are built, and return the collected signatures
# of the libraries in the path.
sub includes {
    return $_[0]->{'bsig'} if exists $_[0]->{'bsig'};
    my($self, $tgt) = @_;
    my($env) = $self->{env};
    my($ewd) = $env->{_cwd};
    my $ldirs = $env->{LIBPATH};
    if (! defined $ldirs) {
	$ldirs = [ ];
    } elsif (ref($ldirs) ne 'ARRAY') {
	$ldirs = [ split(/$main::PATH_SEPARATOR/o, $ldirs) ];
    }
    my @lpath = map($ewd->lookupdir($_), @$ldirs);
    my(@sigs);
    my(@names);

    # Pass %LIBS symbol through %-substituition
    # <schwarze@isa.de> 1998-06-18
    @names = split(' ', $env->_subst($env->{LIBS} || ''));
    my $name;
    for $name (@names) {
	my ($lpath, @allnames);
	if ($name =~ /^-l(.*)/) {
	    # -l style names are looked up on LIBPATH, using all
	    # possible lib suffixes in the same search order the
	    # linker uses (according to SUFLIBS).
	    # Recognize new PREFLIB symbol, which should be 'lib' on
	    # Unix, and empty on Win32. TODO: What about shared
	    # library suffixes?  <schwarze@isa.de> 1998-05-13
	   @allnames = map("$env->{PREFLIB}$1$_",
			   split(/:/, $env->{SUFLIBS}));
	    $lpath = \@lpath;
	} else {
	    @allnames = ($name);
	    # On Win32, all library names are looked up in LIBPATH
	    # <schwarze@isa.de> 1998-05-13
	    if ($main::_WIN32) {
		$lpath = [$dir::top, @lpath];
	    }
	    else {
		$lpath = [$dir::top];
	    }
	}
	my $dir;
	DIR: for $dir (@$lpath) {
	    my $n;
	    for $n (@allnames) {
		my($lib) = $dir->lookup_accessible($n);
		if ($lib) {
		    last DIR if $lib->ignore;
		    if ((build $lib) eq 'errors') {
			$tgt->{status} = 'errors';
			return undef;
		    }
		    push(@sigs, 'sig'->signature($lib));
		    last DIR;
		}
	    }
	}
    }
    $self->{'bsig'} = 'sig'->collect(@sigs);
}

# Always compatible with other such builders, so the user
# can define a single program or module from multiple places.
sub compatible {
    my($self, $other) = @_;
    ref($other) eq "build::command::link";
}

# Link a program.
package build::command::linkedmodule;

use vars qw( @ISA );

BEGIN { @ISA = qw(build::command) }

# Always compatible with other such builders, so the user
# can define a single linked module from multiple places.
sub compatible {
    my($self, $other) = @_;
    ref($other) eq "build::command::linkedmodule";
}

# Builder for a C module
package build::command::cc;

use vars qw( @ISA );

BEGIN { @ISA = qw(build::command) }

sub find {
    $_[1]->{_cc} || do {
	my($class, $env) = @_;
	my($cpppath) = $env->_subst($env->{CPPPATH});
	my($cscanner) = find scan::cpp($env->{_cwd}, $cpppath);
	$env->{_IFLAGS} = "%(" . $cscanner->iflags($env) . "%)";
	my($self) = find build::command($env, $env->{CCCOM});
	$self->{scanner} = $cscanner;
	bless $env->{_cc} = $self;
    }
}

# Invoke the associated	 C scanner to get signature of included files.
sub includes {
    my($self, $tgt) = @_;
    $self->{scanner}->includes($tgt, $tgt->{sources}[0]);
}

# Builder for a C++ module
package build::command::cxx;

use vars qw( @ISA );

BEGIN { @ISA = qw(build::command) }

sub find {
    $_[1]->{_cxx} || do {
	my($class, $env) = @_;
	my($cpppath) = $env->_subst($env->{CPPPATH});
	my($cscanner) = find scan::cpp($env->{_cwd}, $cpppath);
	$env->{_IFLAGS} = "%(" . $cscanner->iflags($env) . "%)";
	my($self) = find build::command($env, $env->{CXXCOM});
	$self->{scanner} = $cscanner;
	bless $env->{_cxx} = $self;
    }
}

# Invoke the associated	 C scanner to get signature of included files.
sub includes {
    my($self, $tgt) = @_;
    $self->{scanner}->includes($tgt, $tgt->{sources}[0]);
}

# Builder for a user command (cons::Command).  We assume that a user
# command might be built and implement the appropriate dependencies on
# the command itself (actually, just on the first word of the command
# line).
package build::command::user;

use vars qw( @ISA );

BEGIN { @ISA = qw(build::command) }

sub includes {
    my($self, $tgt) = @_;
    my($sig) = '';

    # Check for any quick scanners attached to source files.
    my $dep;
    for $dep (@{$tgt->{dep}}, @{$tgt->{sources}}) {
	my($scanner) = $dep->{'srcscan',$self->{env}};
	if ($scanner) {
	    $sig .= $scanner->includes($tgt, $dep);
	}
    }

    # XXX Optimize this to not use ignored paths.
    if (! exists $self->{_comsig}) {
	my($env) = $self->{env};
	$self->{_comsig} = '';
	my($com, $dir);
      com:
	for $com ($self->{act}->commands) {
	    my($pdirs) = $env->{ENV}->{PATH};
	    if (! defined $pdirs) {
		$pdirs = [ ];
	    } elsif (ref($pdirs) ne 'ARRAY') {
		$pdirs = [ split(/$main::PATH_SEPARATOR/o, $pdirs) ];
	    }
	    for $dir (map($dir::top->lookupdir($_), @$pdirs)) {
		my($prog) = $dir->lookup_accessible($com);
		if ($prog) { # XXX Not checking execute permission.
		    if ((build $prog) eq 'errors') {
			$tgt->{status} = 'errors';
			return $sig;
		    }
		    next com if $prog->ignore;
		    $self->{_comsig} .= 'sig'->signature($prog);
		    next com;
		}
	    }
	}
    }

    return $self->{_comsig} . $sig
}


# Builder for a library module (archive).
# We assume that a user command might be built and implement the
# appropriate dependencies on the command itself.
package build::command::library;

use vars qw( @ISA );

BEGIN { @ISA = qw(build::command) }

sub find {
    my($class, $env) = @_;
    bless find build::command($env, $env->{ARCOM})
}

# Always compatible with other library builders, so the user
# can define a single library from multiple places.
sub compatible {
    my($self, $other) = @_;
    ref($other) eq "build::command::library";
}

# A multi-target builder.
# This allows multiple targets to be associated with a single build
# script, without forcing all the code to be aware of multiple targets.
package build::multiple;

sub new {
    my($class, $builder, $tgts) = @_;
    bless { 'builder' => $builder, 'env' => $builder->{env}, 'tgts' => $tgts };
}

sub scriptsig {
    my($self, $tgt) = @_;
    $self->{builder}->scriptsig($tgt);
}

sub includes {
    my($self, $tgt) = @_;
    $self->{builder}->includes($tgt);
}

sub compatible {
    my($self, $tgt) = @_;
    $self->{builder}->compatible($tgt);
}

sub cachin {
    my($self, $tgt, $sig) = @_;
    $self->{builder}->cachin($tgt, $sig);
}

sub cachout {
    my($self, $tgt, $sig) = @_;
    $self->{builder}->cachout($tgt, $sig);
}

sub action {
    my($self, $invoked_tgt) = @_;
    return $self->{built} if exists $self->{built};

    # Make sure all targets in the group are unlinked before building any.
    my($tgts) = $self->{tgts};
    my $tgt;
    for $tgt (@$tgts) {
	futil::mkdir($tgt->{dir});
	unlink($tgt->path) if ! $tgt->precious;
    }

    # Now do the action to build all the targets. For consistency
    # we always call the action on the first target, just so that
    # $> is deterministic.
    $self->{built} = $self->{builder}->action($tgts->[0]);

    # Now "build" all the other targets (except for the one
    # we were called with). This guarantees that the signature
    # of each target is updated appropriately. We force the
    # targets to be built even if they have been previously
    # considered and found to be OK; the only effect this
    # has is to make sure that signature files are updated
    # correctly.
    for $tgt (@$tgts) {
	if ($tgt ne $invoked_tgt) {
	    delete $tgt->{status};
	    'sig'->invalidate($tgt);
	    build $tgt;
	}
    }

    # Status of action.
    $self->{built};
}

package action;

sub new {
    my($env, $act) = @_;
    if (ref($act) eq 'CODE') {
	return action::perl->new($act);
    } else {
	return action::command->new($env, $act);
    }
}

package action::command;

use vars qw( @ISA %cmd %_varopts $_varletters );

BEGIN {
    @ISA = $main::_WIN32 ? 'action::command::win32' : 'action::command::unix';

    # Internal hash for processing variable options.
    # f: return file part
    # d: return directory part
    # F: return file part, but strip any suffix
    # b: return full path, but strip any suffix (a.k.a. return basename)
    # s: return only the suffix (or an empty string, if no suffix is there)
    # a: return the absolute path to the file
    # S: return the absolute path to a Linked source file
    %_varopts = (
	'f' => sub { return $_[0]->{entry}; },
	'd' => sub { return $_[0]->{dir}->path; },
	'F' => sub { my $subst = $_[0]->{entry};
		     $subst =~ s/\.[^\.]+$//;
		     return $subst; },
	'b' => sub { my $subst =  $_[0]->path;
		     $subst =~ s/\.[^\.]+$//;
		     return $subst; },
	's' => sub { my $subst = $_[0]->{entry};
		     $subst =~ m/(\.[^\.]+)$/;
		     return $1; },
	'a' => sub { my $path = $_[0]->path;
		     if (! File::Spec->file_name_is_absolute($path)) {
			$path = File::Spec->catfile(Cwd::cwd(), $path);
		     }
		     return $path; },
	'S' => sub { my $path = $_[0]->srcpath;
		     if (! File::Spec->file_name_is_absolute($path)) {
			my $cwd = File::Spec->canonpath(Cwd::cwd());
			$path = File::Spec->catfile($cwd, $path);
		     }
		     return $path; },
    );

    $_varletters = join('', keys %_varopts);
}

# Internal routine for processing variable options.
# Options are specified in hash in the BEGIN block above.
# no option: return path to file (relative to top,
# or absolute if it's outside)
sub _variant {
    my($opt, $file) = @_;
    $opt = '' if ! defined $opt;
    if (defined $_varopts{$opt}) {
	return &{$_varopts{$opt}}($file);
    }
    return $file->path;
}

sub new {
    my($class, $env, $cmd) = @_;
    $cmd = $env->_subst($cmd);
    $cmd{$env,$cmd} || do {
	# Remove unwanted bits from signature -- those bracketed by %( ... %)
	my $sigs = $cmd;
	my $sig = '';
	if (ref($sigs) eq 'ARRAY') {
	    # This is an array of commands..
	    my $f;
	    foreach $f (@$sigs) {
		$sig .= _strip($f);
	    }
	} else {
	    $sig = _strip($sigs);
	}
	my $self = { cmd => $cmd, cmdsig => 'sig'->cmdsig($sig) };
	$cmd{$env,$cmd} = bless $self, $class;
    }
}

sub _strip {
    my $sig = shift;
    $sig =~ s/^\@\s*//mg;
    while ($sig =~ s/%\(([^%]|%[^\(])*?%\)//g) { }
    $sig;
}

sub scriptsig {
    $_[0]->{cmdsig};
}

# Return an array of all the commands (first word on each line).
sub commands {
    my($self) = @_;
    my(@cmds) = ();
    my $com;
    my $cmd = $self->{'cmd'};
    my @allcoms;

    push @allcoms, ref $cmd ? @{$cmd} : split(/\n/, $cmd);

    for $com (@allcoms) {
	$com =~ s/^\s*//;
	$com =~ s/\s.*//;
	next if ! $com; # blank line
	push @cmds, $com;
    }
    @cmds;
}

# For the signature of a basic command, we don't bother
# including the command itself. This is not strictly correct,
# and if we wanted to be rigorous, we might want to insist
# that the command was checked for all the basic commands
# like gcc, etc. For this reason we don't have an includes
# method.

# Call this to get the command line script: an array of
# fully substituted commands.
sub getcoms {
    my($self, $env, $tgt) = @_;
    my(@coms);
    my $com;
    my @allcoms = ();
    my $cmd = $self->{'cmd'};

    push @allcoms, ref $cmd ? @{$cmd} : split(/\n/, $cmd);

    for $com (@allcoms) {
	my(@src) = (undef, @{$tgt->{sources}});
	my(@src1) = @src;

	next if $com =~ /^\s*$/;

	# NOTE: we used to have a more elegant s//.../e solution
	# for the items below, but this caused a bus error...

	# Remove %( and %) -- those are only used to bracket parts
	# of the command that we don't depend on.
	$com =~ s/%[()]//g;

	# Deal with %n, n=1,9 and variants.
	while ($com =~ /%([1-9])(:([$_varletters]?))?/o) {
	    my($match) = $&;
	    my($src) = $src1[$1];
	    my($subst) = _variant($3, $src1[$1]->rfile);
	    undef $src[$1];
	    $com =~ s/$match/$subst/;
	}

	# Deal with %0 aka %> and variants.
	while ($com =~ /%[0>](:([$_varletters]?))?/o) {
	    my($match) = $&;
	    my($subst) = _variant($2, $tgt);
	    $com =~ s/$match/$subst/;
	}

	# Deal with %< (all sources except %n's already used)
	while ($com =~ /%<(:([$_varletters]?))?/o) {
	    my($match) = $&;
	    my @list = ();
	    foreach (@src) {
		push(@list, _variant($2, $_->rfile)) if $_;
	    }
	    my($subst) = join(' ', @list);
	    $com =~ s/$match/$subst/;
	}

	# Deal with %[ %].
	$com =~ s{%\[(.*?)%\]}{
	    my($func, @args) = grep { $_ ne '' } split(/\s+/, $1);
	    die("$0: \"$func\" is not defined.\n")
		unless ($env->{$func});
	    &{$env->{$func}}(@args);
	}gex;

	# Convert left-over %% into %.
	$com =~ s/%%/%/g;

	# White space cleanup. XXX NO WAY FOR USER TO HAVE QUOTED SPACES
	$com = join(' ', split(' ', $com));
	next if $com =~ /^:/ && $com !~ /^:\S/;
	push(@coms, $com);
    }
    @coms
}

# Build the target using the previously specified commands.
sub execute {
    my($self, $env, $tgt, $package) = @_;

    if ($param::build) {
	futil::mkdir($tgt->{dir});
	unlink($tgt->path) if ! $tgt->precious;
    }

    # Set environment.
    map(delete $ENV{$_}, keys %ENV);
    %ENV = %{$env->{ENV}};

    # Handle multi-line commands.
    my $com;
    for $com ($self->getcoms($env, $tgt)) {
	if ($com !~ s/^\@\s*//) {
	    main::showcom($com);
	}
	next if ! $param::build;

	if ($com =~ /^\[perl\]\s*/) {
	    my $perlcmd = $';
	    my $status;
	    {
		# Restore the script package variables that were defined
		# in the Conscript file that defined this [perl] build,
		# so the code executes with the expected variables.
		# Then actually execute (eval) the [perl] command to build
		# the target, followed by cleaning up the name space
		# by deleting the package variables we just restored.
		my($pkgvars) = $tgt->{conscript}->{pkgvars};
		NameSpace::restore($package, $pkgvars) if $pkgvars;
		$status = eval "package $package; $perlcmd";
		NameSpace::remove($package, keys %$pkgvars) if $pkgvars;
	    }
	    if (!defined($status)) {
		warn "$0: *** Error during perl command eval: $@.\n";
		return undef;
	    } elsif ($status == 0) {
		warn "$0: *** Perl command returned $status "
		   . "(this indicates an error).\n";
		return undef;
	    }
	    next;
	}
	if (! $self->do_command($com, $tgt->path)) {
		return undef;
	}
    }

    # success.
    return 1;
}

sub show {
    my($self, $env, $tgt) = @_;
    my $com;
    for $com ($self->getcoms($env, $tgt)) {
	if ($com !~ /^\@\s*/) {
	    main::showcom($com);
	}
    }
}

package action::command::unix;

sub do_command {
    my($class, $com, $path) = @_;
    my($pid) = fork();
    die("$0: unable to fork child process ($!)\n") if !defined $pid;
    if (!$pid) {
	# This is the child.  We eval the command to suppress -w
	# warnings about not reaching the statements afterwards.
	eval 'exec($com)';
	$com =~ s/\s.*//;
	die qq($0: failed to execute "$com" ($!). )
	  . qq(Is this an executable on path "$ENV{PATH}"?\n);
    }
    for (;;) {
	do {} until wait() == $pid;
	my ($b0, $b1) = ($? & 0xFF, $? >> 8);
	# Don't actually see 0177 on stopped process; is this necessary?
	next if $b0 == 0177; # process stopped; we can wait.
	if ($b0) {
	    my($core, $sig) = ($b0 & 0200, $b0 & 0177);
	    my($coremsg) = $core ? "; core dumped" : "";
	    $com =~ s/\s.*//;
	    my $err = "$0: *** \[$path\] $com terminated by signal " .
		  "$sig$coremsg\n";
	    warn $err;
	    return undef;
	}
	if ($b1) {
	    warn qq($0: *** [$path] Error $b1\n); # trying to be like make.
	    return undef;
	}
	last;
    }
    return 1;
}

package action::command::win32;

sub do_command {
    my($class, $com, $path) = @_;
    system($com);
    if ($?) {
	my ($b0, $b1) = ($? & 0xFF, $? >> 8);
	my $err = $b1 || $?;
	my $warn = qq($0: *** [$path] Error $err);
	$warn .= " (executable not found in path?)" if $b1 == 0xFF;
	warn "$warn\n";
	return undef;
    }
    return 1;
}

package action::perl;

# THIS IS AN EXPERIMENTAL PACKAGE.  It's entirely possible that the
# interface may change as this gets completed, so use at your own risk.
#
# There are (at least) two issues that need to be solved before blessing
# this as a real, fully-supported feature:
#
#   --	We need to calculate a signature value for a Perl code ref, in
#	order to rebuild the target if there's a change to the Perl code
#	used to generate it.
#
#	This is not straightforward.  A B::Deparse package exists that
#	decompiles a coderef into text.  It's reportedly not completely
#	reliable for closures; it misses which variables are global, and
#	the values of private lexicals.  Nevertheless, it'd probably
#	be perfect for our purposes, except that it wasn't added until
#	some time between Perl 5.00502 and 5.00554, and doesn't seem to
#	really work until Perl 5.6.0, so by relying on it, we'd lose
#	support for Perl versions back to 5.003*.
#
#   --	Ideally, a code ref should be able to use something like
#	$env->_subst to fetch values from the construction environment
#	to modify its behavior without having to cut-and-paste code.
#	(Actually, since we pass the environment to the executed code
#	ref, there's no reason you can't do this with the code as it
#	stands today.)  But this REALLY complicates the signature
#	calculation, because now the actual signature would depend not
#	just on the code contents, but on the construction variables (or
#	maybe just the environment).
#
# A potentially valid workaround would be to use the contents of the
# Conscript file in which the code reference is defined as the code
# ref's signature.  This has the drawback of causing a recompilation of
# the target file even in response to unrelated changes in the Conscript
# file, but it would ensure correct builds without having to solve the
# messy issues of generating a signature directly from a code ref.
#
# Nevertheless, this seemed a useful enough skeleton of a feature that
# it made sense to release it in hopes that some practical experience
# will encourage someone to figure out how to solve the signature
# issues.  Or maybe we'll discover these aren't big issues in practice
# and end up blessing it as is.

use vars qw( %code );

sub new {
    my($class, $cref) = @_;
    $code{$cref} || do {
	my $sig = '';
	# Generating a code signature using B::Deparse doesn't really
	# work for us until Perl 5.6.0.  Here's the code in case
	# someone wants to use it.
	#use B::Deparse;
	#my $deparse = B::Deparse->new();
	#my $body = $deparse->coderef2text($cref);
	#$sig = $body;	# should be an MD5 sig
	my($self) = { cref => $cref, crefsig => $sig };
	$code{$cref} = bless $self, $class;
    }
}

sub scriptsig {
    $_[0]->{crefsig}
}

sub execute {
    my($self, $env, $tgt) = @_;
    if ($param::build) {
	futil::mkdir($tgt->{dir});
	unlink($tgt->path) if ! $tgt->precious;
	my($cref) = $self->{cref};
	&$cref($env, $tgt->path, map($_->rpath, @{$tgt->{sources}}));
    }
}

sub commands {
    return ();
}


# Generic scanning module.
package scan;

# Returns the signature of files included by the specified files on
# behalf of the associated target. Any errors in handling the included
# files are propagated to the target on whose behalf this processing
# is being done. Signatures are cached for each unique file/scanner
# pair.
sub includes {
    my($self, $tgt, @files) = @_;
    my(%files, $file);
    my($inc) = $self->{includes} || ($self->{includes} = {});
    while ($file = pop @files) {
	next if exists $files{$file};
	if ($inc->{$file}) {
	    push(@files, @{$inc->{$file}});
	    $files{$file} = 'sig'->signature($file->rfile);
	} else {
	    if ((build $file) eq 'errors') {
		$tgt->{status} = 'errors'; # tgt inherits build status
		return ();
	    }
	    $files{$file} = 'sig'->signature($file->rfile);
	    my(@includes) = $self->scan($file);
	    $inc->{$file} = \@includes;
	    push(@files, @includes);
	}
    }
    'sig'->collect(sort values %files)
}


# A simple scanner. This is used by the QuickScanfunction, to setup
# one-time target and environment-independent scanning for a source
# file. Only used for commands run by the Command method.
package scan::quickscan;

use vars qw( @ISA %scanner );

BEGIN { @ISA = qw(scan) }

sub find {
    my($class, $code, $env, $pdirs) = @_;
    if (! defined $pdirs) {
	$pdirs = [ ] ;
    } elsif (ref($pdirs) ne 'ARRAY') {
	$pdirs = [ split(/$main::PATH_SEPARATOR/o, $pdirs) ];
    }
    my(@path) = map { $dir::cwd->lookupdir($_) } @$pdirs;
    my($spath) = "@path";
    $scanner{$code,$env,$spath} || do {
	my($self) = { code => $code, env => $env, path => \@path };
	$scanner{$code,$env,$spath} = bless $self;
    }
}

# Scan the specified file for included file names.
sub scan {
    my($self, $file) = @_;
    my($code) = $self->{code};
    my(@includes);
    # File should have been built by now. If not, we'll ignore it.
    return () unless open(SCAN, $file->rpath);
    while(<SCAN>) {
	push(@includes, grep($_ ne '', &$code));
    }
    close(SCAN);
    my($wd) = $file->{dir};
    my(@files);
    my $name;
    for $name (@includes) {
	my $dir;
	for $dir ($file->{dir}, @{$self->{path}}) {
	    my($include) = $dir->lookup_accessible($name);
	    if ($include) {
		push(@files, $include) unless $include->ignore;
		last;
	    }
	}
    }
    @files
}


# CPP (C preprocessor) scanning module
package scan::cpp;

use vars qw( @ISA %scanner );

BEGIN { @ISA = qw(scan) }

# For this constructor, provide the include path argument (colon
# separated). Each path is taken relative to the provided directory.

# Note: a particular scanning object is assumed to always return the
# same result for the same input. This is why the search path is a
# parameter to the constructor for a CPP scanning object. We go to
# some pains to make sure that we return the same scanner object
# for the same path: otherwise we will unecessarily scan files.
sub find {
    my($class, $dir, $pdirs) = @_;
    if (! defined $pdirs) {
	$pdirs = [ ];
    } elsif (ref($pdirs) ne 'ARRAY') {
	$pdirs = [ split(/$main::PATH_SEPARATOR/o, $pdirs) ];
    }
    my @path = map($dir->lookupdir($_), @$pdirs);
    my($spath) = "@path";
    $scanner{$spath} || do {
	my($self) = {'path' => \@path};
	$scanner{$spath} = bless $self;
    }
}

# Scan the specified file for include lines.
sub scan {
    my($self, $file) = @_;
    my($angles, $quotes);

    if (exists $file->{angles}) {
	$angles = $file->{angles};
	$quotes = $file->{quotes};
    } else {
	my(@anglenames, @quotenames);
	return () unless open(SCAN, $file->rpath);
	while (<SCAN>) {
	    next unless /^\s*#/;
	    if (/^\s*#\s*include\s*([<"])(.*?)[>"]/) {
		if ($1 eq "<") {
		    push(@anglenames, $2);
		} else {
		    push(@quotenames, $2);
		}
	    }
	}
	close(SCAN);
	$angles = $file->{angles} = \@anglenames;
	$quotes = $file->{quotes} = \@quotenames;
    }


    my(@shortpath) = @{$self->{path}};	  # path for <> style includes
    my(@longpath) = ($file->{dir}, @shortpath); # path for "" style includes

    my(@includes);

    my $name;
    for $name (@$angles) {
	my $dir;
	for $dir (@shortpath) {
	    my($include) = $dir->lookup_accessible($name);
	    if ($include) {
		push(@includes, $include) unless $include->ignore;
		last;
	    }
	}
    }

    for $name (@$quotes) {
	my $dir;
	for $dir(@longpath) {
	    my($include) = $dir->lookup_accessible($name);
	    if ($include) {
		push(@includes, $include) unless $include->ignore;
		last;
	    }
	}
    }

    return @includes
}

# Return the include flags that would be used for a C Compile.
sub iflags {
    my($self, $env) = @_;
    my($iflags) = '';
    my($dir, $dpath);
    for $dir (@{$self->{path}}) {
	$dpath = $dir->path;
	# Add the (presumably local) directory to the -I flags
	# if we're not using repositories, the directory exists,
	# or it's Linked to a source directory (that is, it *will*
	# exist by the time the compilation occurs).
	$iflags .= " ".$env->{INCDIRPREFIX}.$dpath.$env->{INCDIRSUFFIX}
		if ! @param::rpath || -d $dpath || $dir->is_linked;
	next if File::Spec->file_name_is_absolute($dpath);
	if (@param::rpath) {
	    my $d;
	    if ($dpath eq $dir::CURDIR) {
		foreach $d (map($_->path, @param::rpath)) {
		    $iflags .= " ".$env->{INCDIRPREFIX}.$d.$env->{INCDIRSUFFIX};
		}
	    } else {
		my($rpath);
		foreach $d (map($_->path, @param::rpath)) {
		    $rpath = File::Spec->catfile($d, $dpath);
		    $iflags .= " ".$env->{INCDIRPREFIX}.$rpath.$env->{INCDIRSUFFIX}
				if -d $rpath;
		}
	    }
	}
    }
    $iflags
}

package File::Spec;

use vars qw( $_SEP $_MATCH_SEP $_MATCH_VOL );

# Cons is migrating to using File::Spec for portable path name
# manipulation.  This is the right long-term direction, but there are
# some problems with making the transition:
#
#	For multi-volume support, we need to use newer interfaces
#	(splitpath, catpath, splitdir) that are only available in
#	File::Spec 0.8.
#
#	File::Spec 0.8 doesn't work with Perl 5.00[34] due to
#	regular expression incompatibilities (use of \z).
#
#	Forcing people to use a new version of a module is painful
#	because (in the workplace) their administrators aren't
#	always going to agree to install it everywhere.
#
# As a middle ground, we provide our own versions of all the File::Spec
# methods we use, supporting both UNIX and Win32.  Some of these methods
# are home brew, some are cut-and-pasted from the real File::Spec methods.
# This way, we're not reinventing the whole wheel, at least.
#
# We can (and should) get rid of this class whenever 5.00[34] and
# versions of File::Spec prior to 0.9 (?) have faded sufficiently.
# We also may need to revisit whenever someone first wants to use
# Cons on some platform other than UNIX or Win32.

BEGIN {
    if ($main::_WIN32) {
	$_SEP = '\\';
	$_MATCH_SEP = "[\Q/$_SEP\E]";
	$_MATCH_VOL = "([a-z]:)?$_MATCH_SEP";
    } else {
	$_SEP = '/';
	$_MATCH_SEP = "\Q$_SEP\E";
	$_MATCH_VOL = $_MATCH_SEP;
    }
}

sub canonpath {
    my ($self, $path) = @_;
    if ($main::_WIN32) {
	$path =~ s/^([a-z]:)/\u$1/s;
	$path =~ s|/|\\|g;
	$path =~ s|([^\\])\\+|$1\\|g;                  # xx////xx  -> xx/xx
	$path =~ s|(\\\.)+\\|\\|g;                     # xx/././xx -> xx/xx
	$path =~ s|^(\.\\)+||s unless $path eq ".\\";  # ./xx      -> xx
	$path =~ s|\\$||
		 unless $path =~ m#^([A-Z]:)?\\$#s;   # xx/       -> xx
    } else {
	$path =~ s|/+|/|g unless($^O eq 'cygwin');     # xx////xx  -> xx/xx
	$path =~ s|(/\.)+/|/|g;                        # xx/././xx -> xx/xx
	$path =~ s|^(\./)+||s unless $path eq "./";    # ./xx      -> xx
	$path =~ s|^/(\.\./)+|/|s;                     # /../../xx -> xx
	$path =~ s|/$|| unless $path eq "/";          # xx/       -> xx
    }
    return $path;
}

sub catdir {
    my $self = shift;
    my @args = @_;
    foreach (@args) {
	# append a slash to each argument unless it has one there
	$_ .= $_SEP if $_ eq '' || substr($_,-1) ne $_SEP;
    }
    return $self->canonpath(join('', @args));
}

sub catfile {
    my $self = shift;
    my $file = pop @_;
    return $file unless @_;
    my $dir = $self->catdir(@_);
    $dir .= $_SEP unless substr($dir,-1) eq $_SEP;
    $file = '' if ! defined($file);
    return $dir.$file;
}

sub catpath {
    my $path = $_[1] . $_[0]->catfile(@_[2..$#_]);
    $path =~ s/(.)$_MATCH_SEP*$/$1/;
    $path;
}

sub curdir {
    '.'
}

sub file_name_is_absolute {
    my ($self, $file) = @_;
    return scalar($file =~ m{^$_MATCH_VOL}is);
}

sub splitdir {
    my @dirs = split(/$_MATCH_SEP/, $_[1], -1);
    push(@dirs, '') if $dirs[$#dirs];
    @dirs;
}

sub splitpath {
    my ($self, $path) = @_;
    my $vol = '';
    my $sep = $_SEP;
    if ($main::_WIN32) {
	if ($path =~ s#^([A-Za-z]:|(?:\\\\|//)[^\\/]+[\\/][^\\/]+)([\\/])#$2#) {
	    $vol = $1;
	    $sep = $2;
	}
    }
    my(@path) = split(/$_MATCH_SEP/, $path, -1);
    my $file = pop @path;
    my $dirs = join($sep, @path, '');
    return ($vol, $dirs, $file);
}

sub updir {
    '..'
}

sub case_tolerant {
    return $main::_WIN32;
}

# Directory and file handling. Files/dirs are represented by objects.
# Other packages are welcome to add component-specific attributes.
package dir;

use vars qw( $SEPARATOR $MATCH_SEPARATOR $CURDIR $UPDIR
	     $cwd_vol %root $top $cwd );

BEGIN {
    # A portable way of determing our directory separator.
    $SEPARATOR = File::Spec->catdir('', '');
    # A fast-path regular expression to match a directory separator
    # anywhere in a path name.
    if ($SEPARATOR eq '/') {
	$MATCH_SEPARATOR = "\Q$SEPARATOR\E";
    } else {
	$MATCH_SEPARATOR = "[\Q/$SEPARATOR\E]";
    }
    # Cache these values so we don't have to make a method call
    # every time we need them.
    $CURDIR = File::Spec->curdir;	# '.' on UNIX
    $UPDIR = File::Spec->updir;	# '..' on UNIX
    #
    $cwd_vol = '';
}

# Annotate a node (file or directory) with info about the
# method that created it.
sub creator {
    my($self, @frame) = @_;
    $self->{'creator'} = \@frame if @frame;
    $self->{'creator'};
}

# Handle a file|dir type exception.  We only die if we find we were
# invoked by something in a Conscript/Construct file, because
# dependencies created directly by Cons' analysis shouldn't cause
# an error.
sub _type_exception {
    my($e) = @_;
    my($line, $sub);
    (undef, undef, $line, $sub) = script::caller_info;
    if (defined $line) {
	my $err = "\"${\$e->path}\" already in use as a " . ref($e) . " before $sub on line $line";
	if ($e->{'creator'}) {
	    my $script;
	    (undef, $script, $line, $sub) = @{$e->{'creator'}};
	    $err = "\t" . $err . ",\n\t\tdefined by $sub in $script, line $line";
	}
	$err .= "\n";
	die $err;
    }
}

# This wraps up all the common File::Spec logic that we use for parsing
# directory separators in a path and turning it into individual
# subdirectories that we must create, as well as creation of root
# nodes for any new file system volumes we find.  File::Spec doesn't have
# intuitively obvious interfaces, so this is heavily commented.
#
# Note:  This is NOT an object or class method;
# it's just a utility subroutine.
sub _parse_path {
    my($dir, $path) = @_;

    # Convert all slashes to the native directory separator.
    # This allows Construct files to always be written with good
    # old POSIX path names, regardless of what we're running on.
    $path = File::Spec->canonpath($path);

    # File::Spec doesn't understand the Cons convention of
    # an initial '#' for top-relative files.  Strip it.
    my($toprel) = $path =~ s/^#//;

    # Let File::Spec do the heavy lifting of parsing the path name.
    my($vol, $directories, $entry) = File::Spec->splitpath($path);
    my @dirs = File::Spec->splitdir($directories);

    # If there was a file entry on the end of the path, then the
    # last @dirs element is '' and we don't need it.  If there
    # wasn't a file entry on the end (File::Spec->splitpath() knew
    # the last component was a directory), then the last @dirs
    # element becomes the entry we want to look up.
    my($e) = pop @dirs;
    $entry = $e if $entry eq '';

    if (File::Spec->file_name_is_absolute($path)) {
	# An absolute path name.  If no volume was supplied,
	# use the volume of our current directory.
	$vol = $cwd_vol if $vol eq '';
	$vol = uc($vol) if File::Spec->case_tolerant;
	if (! defined $root{$vol}) {
	    # This is our first time looking up a path name
	    # on this volume, so create a root node for it.
	    # (On UNIX systems, $vol is always '', so '/'
	    # always maps to the $root{''} node.)
	    $root{$vol} = {path => $vol.$SEPARATOR,
			prefix => $vol.$SEPARATOR,
			srcpath => $vol.$SEPARATOR,
			'exists' => 1 };
	    $root{$vol}->{'srcdir'} = $root{$vol};
	    bless $root{$vol};
	}
	# We're at the top, so strip the blank entry from the front of
	# the @dirs array since the initial '/' it represents will now
	# be supplied by the root node we return.
	shift @dirs;
	$dir = $root{$vol};
    } elsif ($toprel) {
	$dir = $dir::top;
    }
    ($dir, \@dirs, $entry);
}

# Common subroutine for creating directory nodes.
sub _create_dirs {
    my ($dir, @dirs) = @_;
    my $e;
    foreach $e (@dirs) {
	my $d = $dir->{member}->{$e};
	if (! defined $d) {
	    bless $d = { 'entry' => $e, 'dir' => $dir, }, 'dir';
	    $d->creator(script::caller_info);
	    $d->{member}->{$dir::CURDIR} = $d;
	    $d->{member}->{$dir::UPDIR} = $dir;
	    $dir->{member}->{$e} = $d;
	} elsif (ref $d eq 'entry') {
	    bless $d, 'dir';
	    $d->{member}->{$dir::CURDIR} = $d;
	    $d->{member}->{$dir::UPDIR} = $dir;
	} elsif (ref $d eq 'file') {
	    # This clause is to supply backwards compatibility,
	    # with a warning, for anyone that's used FilePath
	    # to refer to a directory.  After people have using
	    # 1.8 have had time to adjust (sometime in version
	    # 1.9 or later), we should remove this entire clause.
	    my($script, $line, $sub);
	    (undef, $script, $line, $sub) = @{$d->{'creator'}};
	    if ($sub eq 'script::FilePath') {
		print STDERR "$0:  Warning:  $sub used to refer to a directory\n"
			. "\tat line $line of $script.  Use DirPath instead.\n";
		bless $d, 'dir';
	    } else {
		_type_exception($d);
	    }
	} elsif (ref $d ne 'dir') {
	    _type_exception($d);
	}
	$dir = $d;
    }
    $dir;
}

# Look up an entry in a directory.  This method is for when we don't
# care whether a file or directory is returned, so if the entry already
# exists, it will simply be returned.  If not, we create it as a
# generic "entry" which can be later turned into a file or directory
# by a more-specific lookup.
#
# The file entry may be specified as relative, absolute (starts with /),
# or top-relative (starts with #).
sub lookup {
    my($dir, $entry) = @_;

    if ($entry !~ m#$MATCH_SEPARATOR#o) {
	# Fast path: simple entry name in a known directory.
	if ($entry =~ s/^#//) {
	    # Top-relative names begin with #.
	    $dir = $dir::top;
	} elsif ($entry =~ s/^!//) {
	    $dir = $dir::cwd->srcdir;
	}
    } else {
	my $dirsref;
	($dir, $dirsref, $entry) = _parse_path($dir, $entry);
	$dir = _create_dirs($dir, @$dirsref) if @$dirsref;
	return if ! defined $dir;
	return $dir if $entry eq '';
    }

    my $e = $dir->{member}->{$entry};
    if (! defined $e) {
	bless $e = { 'entry' => $entry, 'dir' => $dir, }, 'entry';
	$e->creator(script::caller_info);
	$dir->{member}->{$entry} = $e;
    }

    $e;
}

# Look up a file entry in a directory.
#
# The file entry may be specified as relative, absolute (starts with /),
# or top-relative (starts with #).
sub lookupfile {
    my($dir, $entry) = @_;

    if ($entry !~ m#$MATCH_SEPARATOR#o) {
	# Fast path: simple entry name in a known directory.
	if ($entry =~ s/^#//) {
	    # Top-relative names begin with #.
	    $dir = $dir::top;
	} elsif ($entry =~ s/^!//) {
	    $dir = $dir::cwd->srcdir;
	}
    } else {
	my $dirsref;
	($dir, $dirsref, $entry) = _parse_path($dir, $entry);
	$dir = _create_dirs($dir, @$dirsref) if @$dirsref;
	return undef if $entry eq '';
    }

    my $f = $dir->{member}->{$entry};
    if (! defined $f) {
	bless $f = { 'entry' => $entry, 'dir' => $dir, }, 'file';
	$f->creator(script::caller_info);
	$dir->{member}->{$entry} = $f;
    } elsif (ref $f eq 'entry') {
	bless $f, 'file';
    } elsif (ref $f ne 'file') {
	_type_exception($f);
    }

    $f;
}

# Look up a (sub-)directory entry in a directory.
#
# The (sub-)directory entry may be specified as relative, absolute
# (starts with /), or top-relative (starts with #).
sub lookupdir {
    my($dir, $entry) = @_;

    my $dirsref;
    if ($entry !~ m#$MATCH_SEPARATOR#o) {
	# Fast path: simple entry name in a known directory.
	if ($entry =~ s/^#//) {
	    # Top-relative names begin with #.
	    $dir = $dir::top;
	} elsif ($entry =~ s/^!//) {
	    $dir = $dir::cwd->srcdir;
	}
    } else {
	($dir, $dirsref, $entry) = _parse_path($dir, $entry);
    }
    push(@$dirsref, $entry) if $entry ne '';
    _create_dirs($dir, @$dirsref);
}

# Look up a file entry and return it if it's accessible.
sub lookup_accessible {
    my $file = $_[0]->lookupfile($_[1]);
    return ($file && $file->accessible) ? $file : undef;
}

# Return the parent directory without doing a lookupdir,
# which would create a parent if it doesn't already exist.
# A return value of undef (! $dir->up) indicates a root directory.
sub up {
    $_[0]->{member}->{$dir::UPDIR};
}

# Return whether this is an entry somewhere underneath the
# specified directory.
sub is_under {
    my $dir = $_[0];
    while ($dir) {
	return 1 if $_[1] == $dir;
	$dir = $dir->up;
    }
    return undef;
}

# Return the relative path from the calling directory ($_[1])
# to the object.  If the object is not under the directory, then
# we return it as a top-relative or absolute path name.
sub relpath {
    my ($dir, $obj) = @_;
    my @dirs;
    my $o = $obj;
    while ($o) {
	if ($dir == $o) {
	    if (@dirs < 2) {
		return $dirs[0] || '';
	    } else {
		return File::Spec->catdir(@dirs);
	    }
	}
	unshift(@dirs, $o->{entry});
	$o = $o->up;
    }
    # The object was not underneath the specified directory.
    # Use the node's cached path, which is either top-relative
    # (in which case we append '#' to the beginning) or
    # absolute.
    my $p = $obj->path;
    $p = '#' . $p if ! File::Spec->file_name_is_absolute($p);
    return $p;
}

# Return the path of the directory (file paths implemented
# separately, below).
sub path {
    $_[0]->{path} ||
	($_[0]->{path} = $_[0]->{dir}->prefix . $_[0]->{entry});
}

# Return the pathname as a prefix to be concatenated with an entry.
sub prefix {
    return $_[0]->{prefix} if exists $_[0]->{prefix};
    $_[0]->{prefix} = $_[0]->path . $SEPARATOR;
}

# Return the related source path prefix.
sub srcprefix {
    return $_[0]->{srcprefix} if exists $_[0]->{srcprefix};
    my($srcdir) = $_[0]->srcdir;
    $srcdir->{srcprefix} = $srcdir eq $_[0] ? $srcdir->prefix
					    : $srcdir->srcprefix;
}

# Return the related source directory.
sub srcdir {
    $_[0]->{'srcdir'} ||
	($_[0]->{'srcdir'} = $_[0]->{dir}->srcdir->lookupdir($_[0]->{entry}))
}

# Return if the directory is linked to a separate source directory.
sub is_linked {
    return $_[0]->{is_linked} if defined $_[0]->{is_linked};
    $_[0]->{is_linked} = $_[0]->path ne $_[0]->srcdir->path;
}

sub link {
    my(@paths) = @_;
    my($srcdir) = $dir::cwd->lookupdir(pop @paths)->srcdir;
    map($dir::cwd->lookupdir($_)->{'srcdir'} = $srcdir, @paths);

    # make a reverse lookup for the link.
    $srcdir->{links} = [] if ! $srcdir->{links};
    push @{$srcdir->{links}}, @paths;
}

use vars qw( @tail ); # TODO: Why global ????

sub linked_targets {
    my $tgt = shift;
    my @targets = ();
    my $dir;
    if (ref $tgt eq 'dir') {
	$dir = $tgt;
    } else {
	push @tail, $tgt;
	$dir = $tgt->{dir};
    }
    while ($dir) {
	if (defined $dir->{links} && @{$dir->{links}}) {
	    push @targets,
		 map(File::Spec->catdir($_, @tail), @{$dir->{links}});
	    #print STDERR "Found Link: ${\$dir->path} -> @{\$dir->{links}}\n";
	}
	unshift @tail, $dir->{entry};
	$dir = $dir->up;
    }

    return map($dir::top->lookupdir($_), @targets);
}

sub accessible {
    my $path = $_[0]->path;
    my $err = "$0: you have attempted to use path \"$path\" both as a file " .
	      "and as a directory!\n";
    die $err;
}

sub init {
    my $path = Cwd::cwd();

    # We know we can get away with passing undef to lookupdir
    # as the directory because $dir is an absolute path.
    $top = lookupdir(undef, $path);
    $top->{'path'} = $top->{srcpath} = $dir::CURDIR;
    $top->{'prefix'} = '';
    $top->{'srcdir'} = $top;

    $cwd = $top;

    ($cwd_vol, undef, undef) = File::Spec->splitpath($path);
    $cwd_vol = '' if ! defined $cwd_vol;
    $cwd_vol = uc($cwd_vol) if File::Spec->case_tolerant;
}

package file;

use vars qw( @ISA $level );

BEGIN { @ISA = qw(dir); $level = 0 }

# Return the pathname of the file.
# Define this separately from dir::path because we don't want to
# cache all file pathnames (just directory pathnames).
sub path {
    $_[0]->{dir}->prefix . $_[0]->{entry}
}

# Return the related source file path.
sub srcpath {
    $_[0]->{dir}->srcprefix . $_[0]->{entry}
}

# Return if the file is (should be) linked to a separate source file.
sub is_linked {
    $_[0]->{dir}->is_linked
}

# Repository file search.  If the local file exists, that wins.
# Otherwise, return the first existing same-named file under a
# Repository directory.  If there isn't anything with the same name
# under a Repository directory, return the local file name anyway
# so that some higher layer can try to construct it.
sub rfile {
    return $_[0]->{rfile} if exists $_[0]->{rfile};
    my($self) = @_;
    my($rfile) = $self;
    if (@param::rpath) {
	my($path) = $self->path;
	if (! File::Spec->file_name_is_absolute($path) && ! -f $path) {
	    my($dir);
	    foreach $dir (@param::rpath) {
		my($t) = $dir->prefix . $path;
		if (-f $t) {
		    $rfile = $_[0]->lookupfile($t);
		    $rfile->{'lfile'} = $self;
		    last;
		}
	    }
	}
    }
    $self->{rfile} = $rfile;
}

# Returns the local file for a repository file;
# returns self if it's already a local file.
sub lfile {
    $_[0]->{'lfile'} || $_[0]
}

# returns the "precious" status of this file.
sub precious {
    return $_[0]->{precious};
}

# "Erase" reference to a Repository file,
# making this a completely local file object
# by pointing it back to itself.
sub no_rfile {
    $_[0]->{'rfile'} = $_[0];
}

# Return a path to the first existing file under a Repository directory,
# implicitly returning the current file's path if there isn't a
# same-named file under a Repository directory.
sub rpath {
    $_[0]->{rpath} ||
	($_[0]->{rpath} = $_[0]->rfile->path)
}

# Return a path to the first linked srcpath file under a Repositoy
# directory, implicitly returning the current file's srcpath if there
# isn't a same-named file under a Repository directory.
sub rsrcpath {
    return $_[0]->{rsrcpath} if exists $_[0]->{rsrcpath};
    my($self) = @_;
    my($path) = $self->{rsrcpath} = $self->srcpath;
    if (@param::rpath && ! File::Spec->file_name_is_absolute($path) && ! -f $path) {
	my($dir);
	foreach $dir (@param::rpath) {
	    my($t) = $dir->prefix . $path;
	    if (-f $t) {
		$self->{rsrcpath} = $t;
		last;
	    }
	}
    }
    $self->{rsrcpath};
}

# Return if a same-named file source file exists.
# This handles the interaction of Link and Repository logic.
# As a side effect, it will link a source file from its Linked
# directory (preferably local, but maybe in a repository)
# into a build directory from its proper Linked directory.
sub source_exists {
    return $_[0]->{source_exists} if defined $_[0]->{source_exists};
    my($self) = @_;
    my($path) = $self->path;
    my($mtime, $ctime) = (stat($path))[9,10];
    if ($self->is_linked) {
	# Linked directory, local logic.
	my($srcpath) = $self->srcpath;
	my($src_mtime, $src_ctime) = (stat($srcpath))[9,10];
	if ($src_mtime) {
	    if (! $mtime || $src_mtime != $mtime || $src_ctime != $ctime) {
		futil::install($srcpath, $self);
	    }
	    return $self->{source_exists} = 1;
	}
	# Linked directory, repository logic.
	if (@param::rpath) {
	    if ($self != $self->rfile) {
		return $self->{source_exists} = 1;
	    }
	    my($rsrcpath) = $self->rsrcpath;
	    if ($path ne $rsrcpath) {
		my($rsrc_mtime, $rsrc_ctime) = (stat($rsrcpath))[9,10];
		if ($rsrc_mtime) {
		    if (! $mtime || $rsrc_mtime != $mtime
				 || $rsrc_ctime != $ctime) {
			futil::install($rsrcpath, $self);
		    }
		    return $self->{source_exists} = 1;
		}
	    }
	}
	# There was no source file in any Linked directory
	# under any Repository.  If there's one in the local
	# build directory, it no longer belongs there.
	if ($mtime) {
	    unlink($path) || die("$0: couldn't unlink $path ($!)\n");
	}
	return $self->{source_exists} = '';
    } else {
	if ($mtime) {
	    return $self->{source_exists} = 1;
	}
	if (@param::rpath && $self != $self->rfile) {
	    return $self->{source_exists} = 1;
	}
	return $self->{source_exists} = '';
    }
}

# Return if a same-named derived file exists under a Repository directory.
sub derived_exists {
    $_[0]->{derived_exists} ||
	($_[0]->{derived_exists} = ($_[0] != $_[0]->rfile));
}

# Return if this file is somewhere under a Repository directory.
sub is_on_rpath {
    defined $_[0]->{'lfile'};
}

sub local {
    my($self, $arg) = @_;
    if (defined $arg) {
	$self->{'local'} = $arg;
    }
    $self->{'local'};
}

# Return the entry name of the specified file with the specified
# suffix appended.  Leave it untouched if the suffix is already there.
# Differs from the addsuffix function, below, in that this strips
# the existing suffix (if any) before appending the desired one.
sub base_suf {
    my($entry) = $_[0]->{entry};
    if ($entry !~ m/$_[1]$/) {
	$entry =~ s/\.[^\.]*$//;
	$entry .= $_[1];
    }
    $entry;
}

# Return the suffix of the file; everything including and to the
# right of the last dot.
sub suffix {
    my @pieces = split(/\./, $_[0]->{entry});
    my $suffix = pop(@pieces);
    return ".$suffix";
}

# Called as a simple function file::addsuffix(name, suffix)
sub addsuffix {
    my($name, $suffix) = @_;

    if ($suffix && substr($name, -length($suffix)) ne $suffix) {
	return $name .= $suffix;
    }
    $name;
}

# Return true if the file is (or will be) accessible.
# That is, if we can build it, or if it is already present.
sub accessible {
    (exists $_[0]->{builder}) || ($_[0]->source_exists);
}

# Return true if the file should be ignored for the purpose
# of computing dependency information (should not be considered
# as a dependency and, further, should not be scanned for
# dependencies).
sub ignore {
    return 0 if !$param::ignore;
    return $_[0]->{ignore} if exists $_[0]->{ignore};
    $_[0]->{ignore} = $_[0]->path =~ /$param::ignore/o;
}

# Build the file, if necessary.
sub build {
    return $_[0]->{status} if $_[0]->{status};
    my($status) = &file::_build;
    if ($_[0]->{after_build_func}) {
	my($pkgvars) = $_[0]->{conscript}->{pkgvars};
	NameSpace::restore('script', $pkgvars) if $pkgvars;
	eval("package script; " . $_[0]->{after_build_func});
	print "Error running AfterBuild for ${\$_[0]->path}: $@\n" if ($@);
	NameSpace::remove('script', keys %$pkgvars) if $pkgvars;
    }
    return $status;
}

sub _build {
    my($self) = @_;
    print main::DEPFILE $self->path, "\n" if $param::depfile;
    print((' ' x $level), "Checking ", $self->path, "\n") if $param::depends;
    if (!exists $self->{builder}) {
	# We don't know how to build the file. This is OK, if
	# the file is present as a source file, under either the
	# local tree or a Repository.
	if ($self->source_exists) {
	    return $self->{status} = 'handled';
	} else {
	    my($name) = $self->path;
	    print("$0: don't know how to construct \"$name\"\n");
	    exit(1) unless $param::kflag;
	    return $self->{status} = 'errors'; # xxx used to be 'unknown'
	}
    }

    # An associated build object exists, so we know how to build
    # the file. We first compute the signature of the file, based
    # on its dependendencies, then only rebuild the file if the
    # signature has changed.
    my($builder) = $self->{builder};
    $level += 2;

    my(@deps) = (@{$self->{dep}}, @{$self->{sources}});
    my($rdeps) = \@deps;

    if ($param::random) {
	# If requested, build in a random order, instead of the
	# order that the dependencies were listed.
	my(%rdeps);
	map { $rdeps{$_,'*' x int(rand 10)} = $_ } @deps;
	$rdeps = [values(%rdeps)];
    }

    $self->{status} = '';

    my $dep;
    for $dep (@$rdeps) {
	if ((build $dep) eq 'errors') {
	    # Propagate dependent errors to target.
	    # but try to build all dependents regardless of errors.
	    $self->{status} = 'errors';
	}
    }

    # If any dependents had errors, then we abort.
    if ($self->{status} eq 'errors') {
	$level -= 2;
	return 'errors';
    }

    # Compute the final signature of the file, based on
    # the static dependencies (in order), dynamic dependencies,
    # output path name, and (non-substituted) build script.
    my($sig) = 'sig'->collect(map('sig'->signature($_->rfile), @deps),
			    $builder->includes($self),
			    $builder->scriptsig);

    # May have gotten errors during computation of dynamic
    # dependency signature, above.
    $level -= 2;
    return 'errors' if $self->{status} eq 'errors';

    if (@param::rpath && $self->derived_exists) {
	# There is no local file of this name, but there is one
	# under a Repository directory.

	if ('sig'->current($self->rfile, $sig)) {
	    # The Repository copy is current (its signature matches
	    # our calculated signature).
	    if ($self->local) {
		# ...but they want a local copy, so provide it.
		main::showcom("Local copy of ${\$self->path} from " .
			      "${\$self->rpath}");
		futil::install($self->rpath, $self);
		'sig'->bsig($self, $sig);
	    }
	    return $self->{status} = 'handled';
	}

	# The signatures don't match, implicitly because something
	# on which we depend exists locally.  Get rid of the reference
	# to the Repository file; we'll build this (and anything that
	# depends on it) locally.
	$self->no_rfile;
    }

    # Then check for currency.
    if (! 'sig'->current($self, $sig)) {
	# We have to build/derive the file.
	print((' ' x $level), "Rebuilding ", $self->path, ": out of date.\n")
		if $param::depends;
	# First check to see if the built file is cached.
	if ($builder->cachin($self, $sig)) {
	    'sig'->bsig($self, $sig);
	    return $self->{status} = 'built';
	} elsif ($builder->action($self)) {
	    $builder->cachout($self, $sig);
	    'sig'->bsig($self, $sig);
	    return $self->{status} = 'built';
	} else {
	    die("$0: errors constructing ${\$self->path}\n")
		unless $param::kflag;
	    return $self->{status} = 'errors';
	}
    } else {
	# Push this out to the cache if we've been asked to (-C option).
	# Don't normally do this because it slows us down.
	# In a fully built system, no accesses to the cache directory
	# are required to check any files. This is a win if cache is
	# heavily shared. Enabling this option puts the directory in the
	# loop. Useful only when you wish to recreate a cache from a build.
	if ($param::cachesync) {
	    $builder->cachout($self, $sig);
	    'sig'->bsig($self, $sig);
	}
	return $self->{status} = 'handled';
    }
}

# Bind an action to a file, with the specified sources. No return value.
sub bind {
    my($self, $builder, @sources) = @_;
    if ($self->{builder} && !$self->{builder}->compatible($builder)) {
	# Even if not "compatible", we can still check to see if the
	# derivation is identical. It should be identical if the builder is
	# the same and the sources are the same.
	if ("$self->{builder} @{$self->{sources}}" ne "$builder @sources") {
	    $main::errors++;
	    my($_foo1, $script1, $line1, $sub1) = @{$self->creator};
	    my($_foo2, $script2, $line2, $sub2) = script::caller_info;
	    my $err = "\t${\$self->path}\n" .
		      "\tbuilt (at least) two different ways:\n" .
		      "\t\t$script1, line $line1:  $sub1\n" .
		      "\t\t$script2, line $line2:  $sub2\n";
	    die $err;
	}
	return;
    }
    if ($param::wflag) {
	my($script, $line, $sub);
	(undef, $script, $line, $sub) = script::caller_info;
	$self->{script} = '' if ! defined $self->{script};
	$self->{script} .= "; " if $self->{script};
	$self->{script} .= qq($sub in "$script", line $line);
    }
    $self->{builder} = $builder;
    push(@{$self->{sources}}, @sources);
    @{$self->{dep}} = () if ! defined $self->{dep};
    $self->{conscript} = $priv::self->{script};
}

sub is_under {
    $_[0]->{dir}->is_under($_[1]);
}

sub relpath {
    my $dirpath = $_[0]->relpath($_[1]->{dir});
    if (! $dirpath) {
	return $_[1]->{entry};
    } else {
	File::Spec->catfile($dirpath, $_[1]->{entry});
    }
}

# Return the signature array for this file.
# This probably belongs in its own "sigarray" package,
# which would make it easier to optimize performance.
sub sigarray {
    if ($_[0]->{sigaref}) {
	return @{$_[0]->{sigaref}};
    }
    my $self = shift;
    # glob2pat based on The Perl Cookbook, p. 180.
    sub glob2pat {
	my $globstr = shift;
	my %patmap = (
	    '*' => '.*',
	    '?' => '.',
	    '[' => '[',
	    ']' => ']',
	    '/' => "\Q$dir::SEPARATOR",	# Cons-specific modification
	);
	$globstr =~ s{(.)} { $patmap{$1} || "\Q$1" }ge;
	return '^' . $globstr . '$';
    }
    my @sigarray;
    my $default;
    my $builder = $self->lfile->{builder};
    if (! $builder) {
	@sigarray = @$param::sourcesig;
	$default = [qw(content)];
    } else {
	if ($builder->{env} && $builder->{env}->{SIGNATURE}) {
	    @sigarray = @{$builder->{env}->{SIGNATURE}};
	} else {
	    my $class = ref $builder;
	    my $path = $self->path;
	    warn qq($0: Warning:  Builder package $class did not record\n) .
		 qq(\tthe calling environment for '$path'.\n) .
		 qq(\tUnable to use any %SIGNATURE construction variable\n) .
		 qq(\tfor signature configuration.\n);
	}
	$default = [qw(build)];
    }
    my $path = $self->path;
    while (@sigarray) {
	my($glob, $aref) = splice(@sigarray, 0, 2);
	my $re = glob2pat($glob);
	if ($path =~ /$re/) {
	    $aref = [split(/\s+/, $aref)] if ! ref $aref;
	    $self->{sigaref} = $aref;
	    return @$aref;
	}
    }
    $self->{sigaref} = $default;
    return @{$self->{sigaref}}
}

# Decide if this file's signature should be the content or build signature.
sub sigtype {
    if ($_[0]->{sigtype}) {
	return $_[0]->{sigtype};
    }
    my $self = shift;
    my @sigarray = $self->sigarray;
    my $sigtype;
    if (grep($_ eq "build", @sigarray)) {
	$sigtype = 'bsig';
    } elsif (grep($_ =~ /content$/, @sigarray)) {
	$sigtype = 'csig';
    }
    return $self->{sigtype} = $sigtype;
}

# Return whether this file is configured to use stored
# signature values from the .consign file.
sub stored {
    if (! defined $_[0]->{stored}) {
	$_[0]->{stored} = grep($_ eq "stored-content", $_[0]->sigarray);
    }
    return $_[0]->{stored};
}

# Generic entry (file or directory) handling.
# This is an empty subclass for nodes that haven't
# quite decided whether they're files or dirs.
# Use file methods until someone blesses them one way or the other.
package entry;

use vars qw( @ISA );

BEGIN { @ISA = qw(file) }

# File utilities
package futil;

# Install one file as another.
# Links them if possible (hard link), otherwise copies.
# Don't ask why, but the source is a path, the tgt is a file obj.
sub install {
    my($sp, $tgt) = @_;
    my($tp) = $tgt->path;
    return 1 if $tp eq $sp;
    return 1 if eval { link($sp, $tp) };
    unlink($tp);
    if (! futil::mkdir($tgt->{dir})) {
	return undef;
    }
    return 1 if eval { link($sp, $tp) };
    futil::copy($sp, $tp);
}

# Copy one file to another. Arguments are actual file names.
# Returns undef on failure. Preserves mtime and mode.
sub copy {
    my ($sp, $tp) = @_;
    my ($mode, $length, $atime, $mtime) = (stat($sp))[2,7,8,9];

    # Use Perl standard library module for file copying, which handles
    # binary copies. <schwarze@isa.de> 1998-06-18
    if (! File::Copy::copy($sp, $tp)) {
	warn qq($0: can\'t install "$sp" to "$tp" ($!)\n); #'
	return undef;
    }
    # The file has been created, so try both the chmod and utime,
    # first making sure the copy is writable (because permissions
    # affect the ability to modify file times on some operating
    # systems), and then changing permissions back if necessary.
    my $ret = 1;
    my $wmode = $mode | 0700;
    if (! chmod $wmode, $tp) {
	warn qq($0: can\'t set mode $wmode on file "$tp" ($!)\n); #'
	$ret = undef;
    }
    if (! utime $atime, $mtime, $tp) {
	warn qq($0: can\'t set modification time for file "$tp" ($!)\n); #'
	$ret = undef;
    }
    if ($mode !=  $wmode && ! chmod $mode, $tp) {
	warn qq($0: can\'t set mode $mode on file "$tp" ($!)\n); #'
	$ret = undef;
    }
    return $ret;
}

# Ensure that the specified directory exists.
# Aborts on failure.
sub mkdir {
    return 1 if $_[0]->{'exists'};
    if (! futil::mkdir($_[0]->{dir})) { # Recursively make parent.
	return undef;
    }
    my($path) = $_[0]->path;
    if (!-d $path && !mkdir($path, 0777)) {
	warn qq($0: can't create directory $path ($!).\n); #'
	return undef;
    }
    $_[0]->{'exists'} = 1;
}


# Signature package.
package sig::hash;

use vars qw( $called );

sub init {
    my($dir) = @_;
    my($consign) = $dir->prefix . ".consign";
    my($dhash) = $dir->{consign} = {};
    if (-f $consign) {
	open(CONSIGN, $consign) || die("$0: can't open $consign ($!)\n");
	while(<CONSIGN>) {
	    chop;
	    my ($file, $sig) = split(/:/,$_);
	    $dhash->{$file} = $sig;
	}
	close(CONSIGN);
    }
    $dhash
}

# Read the hash entry for a particular file.
sub in {
    my($dir) = $_[0]->{dir};
    ($dir->{consign} || init($dir))->{$_[0]->{entry}}
}

# Write the hash entry for a particular file.
sub out {
    my($file, $sig) = @_;
    my($dir) = $file->{dir};
    ($dir->{consign} || init($dir))->{$file->{entry}} = $sig;
    $sig::hash::dirty{$dir} = $dir;
}

# Eliminate the hash entry for a particular file.
sub clear {
    my($file) = @_;
    my($dir) = $file->{dir};
    delete $dir->{consign}->{$file->{entry}} if $dir->{consign};
    $sig::hash::dirty{$dir} = $dir;
}

# Flush hash entries. Called at end or via ^C interrupt.
sub END {
    return if $called++; # May be called twice.
    close(CONSIGN); # in case this came in via ^C.
    my $dir;
    for $dir (values %sig::hash::dirty) {
	my($consign) = $dir->prefix . ".consign";
	my($constemp) = $consign . ".$$";
	if (! open(CONSIGN, ">$constemp")) {
	    die("$0: can't create $constemp ($!)\n");
	}
	my($entry, $sig);
	while (($entry, $sig) = each %{$dir->{consign}}) {
	    if (! print CONSIGN "$entry:$sig\n") {
		die("$0: error writing to $constemp ($!)\n");
	    }
	}
	close(CONSIGN);
	if (! rename($constemp, $consign)) {
	    if (futil::copy($constemp, $consign)) {
		unlink($constemp);
	    } else {
		die("$0: couldn't rename or copy $constemp to $consign " .
		    "($!)\n");
	    }
	}
    }
}


# Derived file caching.
package cache;

# Find a file in the cache. Return non-null if the file is in the cache.
sub in {
    return undef unless $param::cache;
    my($file, $sig) = @_;
    # Add the path to the signature, to make it unique.
    $sig = 'sig'->collect($sig, $file->path) unless $param::mixtargets;
    my($dir) = substr($sig, 0, 1);
    my($cp) = File::Spec->catfile($param::cache, $dir, $sig);
    return -f $cp && futil::install($cp, $file);
}

# Try to flush a file to the cache, if not already there.
# If it doesn't make it out, due to an error, then that doesn't
# really matter.
sub out {
    return unless $param::cache;
    my($file, $sig) = @_;
    # Add the path to the signature, to make it unique.
    $sig = 'sig'->collect($sig, $file->path) unless $param::mixtargets;
    my($dir) = substr($sig, 0, 1);
    my($sp) = $file->path;
    my($cp) = File::Spec->catfile($param::cache, $dir, $sig);
    my($cdir) = File::Spec->catfile($param::cache, $dir);
    if (! -d $cdir) {
	mkdir($cdir, 0777) ||
	    die("$0: can't create cache directory $cdir ($!).\n");
    } elsif (-f $cp) {
	# Already cached: try to use that instead, to save space.
	# This can happen if the -cs option is used on a previously
	# uncached build, or if two builds occur simultaneously.
	my($lp) = ".$sig";
	unlink($lp);
	return if ! eval { link($cp, $lp) };
	rename($lp, $sp);
	# Unix98 says, "If the old argument and the new argument both
	# [refer] to the same existing file, the rename() function
	# returns successfully and performs no other action."  So, if
	# $lp and $sp are links (i.e., $cp and $sp are links), $lp is
	# left, and we must unlink it ourselves.  If the rename failed
	# for any reason, it is also good form to unlink the temporary
	# $lp.  Otherwise $lp no longer exists and, barring some race,
	# the unlink fails silently.
	unlink($lp);
	return;
    }

    return if eval { link($sp, $cp) };
    return if ! -f $sp; # if nothing to cache.
    if (futil::copy($sp, "$cp.new")) {
	rename("$cp.new", $cp);
    }
}


# Generic signature handling package.
# This handles the higher-layer distinction between content and build
# signatures, relying on an underlying calculation package like
# "sig::md5"" to provide the signature values themselves.
package sig;

use vars qw( @ISA );

# Select the underlying package to be used for signature calculation.
# We play a few namespace games here.  Specifically, we append
# "sig::" to the beginning of the subclass we're passed.  Then,
# if the package ends in "::debug", we actually subclass the
# "sig::debug" package and as a wrapper around the underlying
# (e.g.) "sig::md5" package that's doing the real calculation.
sub select {
    my($package, $subclass) = @_;
    my $p = $package . "::" . $subclass;
    my $sigpkg = $p;
    if ($p =~ /(.*)::debug$/) {
	$sigpkg = $1;
	$p = 'sig::debug';
    }
    @ISA = ($p);
    $p->init($sigpkg);
};

# Set or return the build signature of a file.
# This is computed elsewhere and passed in to us.
sub bsig {
    my($self, $file, $sig) = @_;
    if (defined $sig) {
	$file->{'bsig'} = $sig;
	$self->set($file);
    } elsif (! defined $file->{'bsig'}) {
	$file->{'bsig'} = '';
    }
    $file->{'bsig'}
}

# Determine the content signature of a file.
# This also sets the .consign entry unless the file is in a
# repository; we don't write into repositories, only read from them.
sub csig {
    my($self, $file) = @_;
    if (! $file->{'csig'}) {
	$file->{'csig'} = $self->srcsig($file->path);
	$self->set($file) if ! $file->is_on_rpath;
    }
    $_[1]->{'csig'}
}

# Determine the current signature of an already-existing or
# non-existant file.  Unless a specific signature type (bsig
# or csig) is requested, this consults the file's signature
# array to decide whether to return content or build signature,
# and whether to use a cached value from a .consign file.
sub signature {
    my($self, $file, $sigtype) = @_;
    $sigtype = $file->sigtype if ! $sigtype;
    #open(TTY, ">/dev/tty");
    #print TTY $file->path, ": $sigtype\n";
    #close(TTY);
    my($path) = $file->path;
    my($time) = (stat($path))[9];
    if ($time) {
	if ($file->{$sigtype}) {
	    return $file->{$sigtype};
	}
	if ($file->is_on_rpath || $file->stored) {
	    if ('sig'->fetch($file) && $file->{$sigtype}) {
		if ($file->{'sigtime'} == $time ||
		    ! $param::rep_sig_times_ok
		    && $file->is_on_rpath) {
		    return $file->{$sigtype};
		}
	    }
	    $file->{$sigtype} = undef;
	}
	if ($file->is_on_rpath || ! File::Spec->file_name_is_absolute($path)) {
	    my $sig = '';
	    if ($sigtype eq 'bsig') { $sig = $self->bsig($file); }
	    elsif ($sigtype eq 'csig') { $sig = $self->csig($file); }
	    return $sig;
	}
	# This file is not in a repository or under the local directory
	# structure.  In the canonical case, it's a utility that will be
	# executed by a command.  Historically, Cons has returned the
	# name of the command concatenated with the modification time.
	# Note that this is *not* the path ("cc" not "/bin/cc"), so it
	# would lose in the unlikely event that a different copy of the
	# utility was used that happened to have the same modification
	# time (due to living in a different directory on the PATH, for
	# example).  The obvious "fix" of using the path like so, however:
	#	return $path . $time;
	# is wrong.  In a multi-machine build environment, different
	# systems may have the same utility in different locations (due
	# to different NFS mount points, for example), which would
	# cause a lot of unnecessary builds if we used the full path.
	# A better solution to strengthen this signature would be to
	# also concatenate the size of the file, but that would cause
	# unnecessary rebuilds when coming from .consign files that used
	# the old scheme.  All of which is to merely explain why we're
	# leaving this as it has been, but documenting it here in case
	# there's reason to change it in the future.
	return $file->{entry} . $time;
    }
    return $file->{$sigtype} = '';
}

sub bsignature {
    my($self, $file) = @_;
    my($path) = $file->path;
    my($time) = (stat($path))[9];
    if ($time) {
	if ($file->{'bsig'}) {
	    return $file->{'bsig'};
	}
	if ('sig'->fetch($file, 'bsig') && $file->{'bsig'}) {
		if ($file->{'sigtime'} == $time ||
		    ! $param::rep_sig_times_ok
		    && $file->is_on_rpath) {
		    return $file->{'bsig'};
		}
	}
	if ($file->is_on_rpath || ! File::Spec->file_name_is_absolute($path)) {
	    return $self->bsig($file);
	}
	return $path . $time;
    }
    return $file->{'bsig'} = '';
}

# Invalidate a file's signature, also clearing its .consign entry.
sub invalidate {
    my($self, $file) = @_;
    delete $file->{'sigtime'};
    delete $file->{'bsig'};
    delete $file->{'csig'};
    sig::hash::clear($file);
}

# Store the signature for a file.
sub set {
    my($self, $file) = @_;
    my $sig = (stat($file->path))[9];
    $sig .= " " . ($file->{'bsig'} || '-');
    $sig .= " " . $file->{'csig'} if $file->{'csig'};
    sig::hash::out($file, $sig);
}

# Fetch the signature(s) for a file.
# Returns whether there was a signature to fetch.
sub fetch {
    my($self, $file, @kw) = @_;
    @kw = ('bsig', 'csig') if ! @kw;
    my $sig = sig::hash::in($file) || '';
    my($sigtime, $bsig, $csig) = split(/ /, $sig);
    $file->{'sigtime'} = $sigtime;
    $file->{'bsig'} = $bsig || '' if grep($_ eq 'bsig', @kw);
    $file->{'csig'} = $csig || '' if grep($_ eq 'csig', @kw);
    $file->{'bsig'} = '' if $file->{'bsig'} eq '-';
    return $sig ne '';
}

# MD5-based signature package.
package sig::md5;

use vars qw( $md5 );

# Initialize MD5 signature calculation by finding an appropriate
# module and creating the proper object.
sub init {
    my $self = shift;
    my @md5_modules = qw(Digest::MD5 MD5 Digest::Perl::MD5);
    # We used to find the right module more simply, using $_ as the
    # loop iterator and just doing:
    #
    #		eval "use $_";
    #		$module = $_, $last if ! $@;
    #
    # in the loop.  Empirically, though, this doesn't pass back the
    # right value in $module on some ActiveState versions.  (Maybe
    # it's something to do with the eval in a for loop, I dunno.)
    # Work around it by using $_ to pass the value out of the loop,
    # which seems to work everywhere.
    my $module;
    for $module (@md5_modules) {
	eval "use $module";
	$_ = $module, last if ! $@;
    }
    $module = $_;
    die "Cannot find any MD5 module from:  @md5_modules" if $@;

    $md5 = new $module;
}

# Is the provided signature equal to the signature of the current
# instantiation of the target (and does the target exist)?
sub current {
    my($self, $file, $sig, $sigtype) = @_;
    $self->bsignature($file) eq $sig;
}

# Return an aggregate signature for a list of signature values.
sub collect {
    my($self, @sigs) = @_;
    # The following sequence is faster than calling the hex interface.
    $md5->reset();
    $md5->add(join('', $param::salt, @sigs));
    unpack("H*", $md5->digest());
}

# Directly compute a file signature as the MD5 checksum of the
# bytes in the file.
sub srcsig {
    my($self, $path) = @_;
    $md5->reset();
    open(FILE, $path) || return '';
    binmode(FILE);
    $md5->addfile(\*FILE);
    close(FILE);
    unpack("H*", $md5->digest());
}

# Compute the signature of a command string.
# For MD5, this is just the string itself, since MD5 will condense
# the string contents into the ultimate signature.  Other signature
# schemes may need to figure this out differently.
sub cmdsig {
    my($self, $sig) = @_;
    return $sig
}

# Generic debug package for signature calculation.
# Because of the way we're called by sig::select() and then use
# the specified value to set up @ISA, this package is essentially a
# factory that creates packages like sig::md5::debug, etc., on the fly.
package sig::debug;

use vars qw( @ISA $sigpkg $outfh );

local *FH;

sub init {
    my $self = shift;
    $sigpkg = shift;
    @ISA = ($sigpkg);
    $sigpkg->init();
    my $file = $ENV{CONS_SIG_DEBUG};
    if ($file) {
	if (! open(FH, ">$file")) {
	    die "Cannot open $file: $!";
	}
	$outfh = \*FH;
    } else {
	$outfh = \*STDOUT;
    }
}

sub current {
    my($self, $file, $sig, $sigtype) = @_;
    my $fsig = $self->bsignature($file);
    my $sub = "${sigpkg}::current";
    my $sep = "\n" . ' ' x (length($sub) + 1 - 3);
    print $outfh "$sub(|$fsig|${sep}eq |$sig|)\n";
    return $fsig eq $sig;
}

sub collect {
    my($self, @sigs) = @_;
    my $sig = $sigpkg->collect(@sigs);
    my $sub = "${sigpkg}::collect";
    my $sep = ",\n" . ' ' x (length($sub) + 1);
    my $buf = join($sep, @sigs);
    $buf = $param::salt . $sep . $buf if $param::salt;
    print $outfh "$sub($buf)\n\t=> |$sig|\n";
    return $sig;
}

sub srcsig {
    my($self, $path) = @_;
    my $sig = $sigpkg->srcsig($path);
    print $outfh "${sigpkg}::srcsig($path)\n\t=> |$sig|\n";
    return $sig;
}

__END__;

=head1 NAME

Cons - A Software Construction System

=head1 DESCRIPTION

A guide and reference for version 2.3.1

Copyright (c) 1996-2001 Free Software Foundation, Inc.

This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; see the file COPYING.  If not, write to
the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA.

=head1 Introduction

B<Cons> is a system for constructing, primarily, software, but is quite
different from previous software construction systems. Cons was designed
from the ground up to deal easily with the construction of software spread
over multiple source directories. Cons makes it easy to create build scripts
that are simple, understandable and maintainable. Cons ensures that complex
software is easily and accurately reproducible.

Cons uses a number of techniques to accomplish all of this. Construction
scripts are just Perl scripts, making them both easy to comprehend and very
flexible. Global scoping of variables is replaced with an import/export
mechanism for sharing information between scripts, significantly improving
the readability and maintainability of each script. B<Construction
environments> are introduced: these are Perl objects that capture the
information required for controlling the build process. Multiple
environments are used when different semantics are required for generating
products in the build tree. Cons implements automatic dependency analysis
and uses this to globally sequence the entire build. Variant builds are
easily produced from a single source tree. Intelligent build subsetting is
possible, when working on localized changes. Overrides can be setup to
easily override build instructions without modifying any scripts. MD5
cryptographic B<signatures> are associated with derived files, and are used
to accurately determine whether a given file needs to be rebuilt.

While offering all of the above, and more, Cons remains simple and easy to
use. This will, hopefully, become clear as you read the remainder of this
document.


=head1 Why Cons? Why not Make?

Cons is a B<make> replacement. In the following paragraphs, we look at a few
of the undesirable characteristics of make--and typical build environments
based on make--that motivated the development of Cons.

=head2 Build complexity

Traditional make-based systems of any size tend to become quite complex. The
original make utility and its derivatives have contributed to this tendency
in a number of ways. Make is not good at dealing with systems that are
spread over multiple directories. Various work-arounds are used to overcome
this difficulty; the usual choice is for make to invoke itself recursively
for each sub-directory of a build. This leads to complicated code, in which
it is often unclear how a variable is set, or what effect the setting of a
variable will have on the build as a whole. The make scripting language has
gradually been extended to provide more possibilities, but these have
largely served to clutter an already overextended language. Often, builds
are done in multiple passes in order to provide appropriate products from
one directory to another directory. This represents a further increase in
build complexity.


=head2 Build reproducibility

The bane of all makes has always been the correct handling of
dependencies. Most often, an attempt is made to do a reasonable job of
dependencies within a single directory, but no serious attempt is made to do
the job between directories. Even when dependencies are working correctly,
make's reliance on a simple time stamp comparison to determine whether a
file is out of date with respect to its dependents is not, in general,
adequate for determining when a file should be rederived. If an external
library, for example, is rebuilt and then ``snapped'' into place, the
timestamps on its newly created files may well be earlier than the last
local build, since it was built before it became visible.


=head2 Variant builds

Make provides only limited facilities for handling variant builds. With the
proliferation of hardware platforms and the need for debuggable
vs. optimized code, the ability to easily create these variants is
essential. More importantly, if variants are created, it is important to
either be able to separate the variants or to be able to reproduce the
original or variant at will. With make it is very difficult to separate the
builds into multiple build directories, separate from the source. And if
this technique isn't used, it's also virtually impossible to guarantee at
any given time which variant is present in the tree, without resorting to a
complete rebuild.


=head2 Repositories

Make provides only limited support for building software from code that
exists in a central repository directory structure.  The VPATH feature of
GNU make (and some other make implementations) is intended to provide this,
but doesn't work as expected: it changes the path of target file to the
VPATH name too early in its analysis, and therefore searches for all
dependencies in the VPATH directory.  To ensure correct development builds,
it is important to be able to create a file in a local build directory and
have any files in a code repository (a VPATH directory, in make terms) that
depend on the local file get rebuilt properly.  This isn't possible with
VPATH, without coding a lot of complex repository knowledge directly into
the makefiles.


=head1 Keeping it simple

A few of the difficulties with make have been cited above. In this and
subsequent sections, we shall introduce Cons and show how these issues are
addressed.

=head2 Perl scripts

Cons is Perl-based. That is, Cons scripts--F<Conscript> and F<Construct>
files, the equivalent to F<Makefile> or F<makefile>--are all written in
Perl. This provides an immediate benefit: the language for writing scripts
is a familiar one. Even if you don't happen to be a Perl programmer, it
helps to know that Perl is basically just a simple declarative language,
with a well-defined flow of control, and familiar semantics. It has
variables that behave basically the way you would expect them to,
subroutines, flow of control, and so on. There is no special syntax
introduced for Cons. The use of Perl as a scripting language simplifies
the task of expressing the appropriate solution to the often complex
requirements of a build.


=head2 Hello, World!

To ground the following discussion, here's how you could build the B<Hello,
World!> C application with Cons:



  $env = new cons();
  Program $env 'hello', 'hello.c';

If you install this script in a directory, naming the script F<Construct>,
and create the F<hello.c> source file in the same directory, then you can
type C<cons hello> to build the application:



  % cons hello
  cc -c hello.c -o hello.o
  cc -o hello hello.o


=head2 Construction environments

A key simplification of Cons is the idea of a B<construction environment>. A
construction environment is an B<object> characterized by a set of key/value
pairs and a set of B<methods>. In order to tell Cons how to build something,
you invoke the appropriate method via an appropriate construction
environment. Consider the following example:



  $env = new cons(
	CC	=>	'gcc',
	LIBS	=>	'libworld.a'
  );

  Program $env 'hello', 'hello.c';

In this case, rather than using the default construction environment, as is,
we have overridden the value of C<CC> so that the GNU C Compiler equivalent
is used, instead. Since this version of B<Hello, World!> requires a library,
F<libworld.a>, we have specified that any program linked in this environment
should be linked with that library. If the library exists already, well and
good, but if not, then we'll also have to include the statement:



  Library $env 'libworld', 'world.c';

Now if you type C<cons hello>, the library will be built before the program
is linked, and, of course, C<gcc> will be used to compile both modules:



  % cons hello
  gcc -c hello.c -o hello.o
  gcc -c world.c -o world.o
  ar r libworld.a world.o
  ar: creating libworld.a
  ranlib libworld.a
  gcc -o hello hello.o libworld.a


=head2 Automatic and complete dependency analysis

With Cons, dependencies are handled automatically. Continuing the previous
example, note that when we modify F<world.c>, F<world.o> is recompiled,
F<libworld.a> recreated, and F<hello> relinked:



  % vi world.c
    [EDIT]
  % cons hello
  gcc -c world.c -o world.o
  ar r libworld.a world.o
  ar: creating libworld.a
  ranlib libworld.a
  gcc -o hello hello.o libworld.a

This is a relatively simple example: Cons ``knows'' F<world.o> depends upon
F<world.c>, because the dependency is explicitly set up by the C<Library>
method. It also knows that F<libworld.a> depends upon F<world.o> and that
F<hello> depends upon F<libworld.a>, all for similar reasons.

Now it turns out that F<hello.c> also includes the interface definition
file, F<world.h>:



  % emacs world.h
    [EDIT]
  % cons hello
  gcc -c hello.c -o hello.o
  gcc -o hello hello.o libworld.a

How does Cons know that F<hello.c> includes F<world.h>, and that F<hello.o>
must therefore be recompiled? For now, suffice it to say that when
considering whether or not F<hello.o> is up-to-date, Cons invokes a scanner
for its dependency, F<hello.c>. This scanner enumerates the files included
by F<hello.c> to come up with a list of further dependencies, beyond those
made explicit by the Cons script. This process is recursive: any files
included by included files will also be scanned.

Isn't this expensive? The answer is--it depends. If you do a full build of a
large system, the scanning time is insignificant. If you do a rebuild of a
large system, then Cons will spend a fair amount of time thinking about it
before it decides that nothing has to be done (although not necessarily more
time than make!). The good news is that Cons makes it very easy to
intelligently subset your build, when you are working on localized changes.


=head2 Automatic global build sequencing

Because Cons does full and accurate dependency analysis, and does this
globally, for the entire build, Cons is able to use this information to take
full control of the B<sequencing> of the build. This sequencing is evident
in the above examples, and is equivalent to what you would expect for make,
given a full set of dependencies. With Cons, this extends trivially to
larger, multi-directory builds. As a result, all of the complexity involved
in making sure that a build is organized correctly--including multi-pass
hierarchical builds--is eliminated. We'll discuss this further in the next
sections.

=head1 Building large trees--still just as simple


=head2 A hierarchy of build scripts

A larger build, in Cons, is organized by creating a hierarchy of B<build
scripts>. At the top of the tree is a script called F<Construct>. The rest
of the scripts, by convention, are each called F<Conscript>. These scripts
are connected together, very simply, by the C<Build>, C<Export>, and
C<Import> commands.


=head2 The Build command

The C<Build> command takes a list of F<Conscript> file names, and arranges
for them to be included in the build. For example:

  Build qw(
	drivers/display/Conscript
	drivers/mouse/Conscript
	parser/Conscript
	utilities/Conscript
  );

This is a simple two-level hierarchy of build scripts: all the subsidiary
F<Conscript> files are mentioned in the top-level F<Construct> file. Notice
that not all directories in the tree necessarily have build scripts
associated with them.

This could also be written as a multi-level script. For example, the
F<Construct> file might contain this command:

  Build qw(
	parser/Conscript
	drivers/Conscript
	utilities/Conscript
  );

and the F<Conscript> file in the F<drivers> directory might contain this:

  Build qw(
	display/Conscript
	mouse/Conscript
  );

Experience has shown that the former model is a little easier to understand,
since the whole construction tree is laid out in front of you, at the
top-level. Hybrid schemes are also possible. A separately maintained
component that needs to be incorporated into a build tree, for example,
might hook into the build tree in one place, but define its own construction
hierarchy.

By default, Cons does not change its working directory to the directory
containing a subsidiary F<Conscript> file it is including.  This behavior
can be enabled for a build by specifying, in the top-level F<Construct>
file:

  Conscript_chdir 1;

When enabled, Cons will change to the subsidiary F<Conscript> file's
containing directory while reading in that file, and then change back
to the top-level directory once the file has been processed.

It is expected that this behavior will become the default in some future
version of Cons.  To prepare for this transition, builds that expect
Cons to remain at the top of the build while it reads in a subsidiary
F<Conscript> file should explicitly disable this feature as follows:

  Conscript_chdir 0;


=head2 Relative, top-relative, and absolute file names

You may have noticed that the file names specified to the Build command are
relative to the location of the script it is invoked from. This is generally
true for other filename arguments to other commands, too, although we might
as well mention here that if you begin a file name with a hash mark, ``#'',
then that file is interpreted relative to the top-level directory (where the
F<Construct> file resides). And, not surprisingly, if you begin it with ``/'',
then it is considered to be an absolute pathname. This is true even on
systems which use a back slash rather than a forward slash to name absolute
paths.

(There is another file prefix, ``!'', that is interpreted specially by
Cons.  See discussion of the C<Link> command, below, for details.)


=head2 Using modules in build scripts

You may pull modules into each F<Conscript> file using the normal Perl
C<use> or C<require> statements:

  use English;
  require My::Module;

Each C<use> or C<require> only affects the one F<Conscript> file in which
it appears.  To use a module in multiple F<Conscript> files, you must
put a C<use> or C<require> statement in each one that needs the module.


=head2 Scope of variables

The top-level F<Construct> file and all F<Conscript> files begin life in
a common, separate Perl package.  B<Cons> controls the symbol table for
the package so that, the symbol table for each script is empty, except
for the F<Construct> file, which gets some of the command line arguments.
All of the variables that are set or used, therefore, are set by the
script itself--not by some external script.

Variables can be explicitly B<imported> by a script from its parent
script. To import a variable, it must have been B<exported> by the parent
and initialized (otherwise an error will occur).


=head2 The Export command

The C<Export> command is used as in the following example:

  $env = new cons();
  $INCLUDE = "#export/include";
  $LIB = "#export/lib";
  Export qw( env INCLUDE LIB );
  Build qw( util/Conscript );

The values of the simple variables mentioned in the C<Export> list will be
squirreled away by any subsequent C<Build> commands. The C<Export> command
will only export Perl B<scalar> variables, that is, variables whose name
begins with C<$>. Other variables, objects, etc. can be exported by
reference--but all scripts will refer to the same object, and this object
should be considered to be read-only by the subsidiary scripts and by the
original exporting script. It's acceptable, however, to assign a new value
to the exported scalar variable--that won't change the underlying variable
referenced. This sequence, for example, is OK:

  $env = new cons();
  Export qw( env INCLUDE LIB );
  Build qw( util/Conscript );
  $env = new cons(CFLAGS => '-O');
  Build qw( other/Conscript );

It doesn't matter whether the variable is set before or after the C<Export>
command. The important thing is the value of the variable at the time the
C<Build> command is executed. This is what gets squirreled away. Any
subsequent C<Export> commands, by the way, invalidate the first: you must
mention all the variables you wish to export on each C<Export> command.


=head2 The Import command

Variables exported by the C<Export> command can be imported into subsidiary
scripts by the C<Import> command. The subsidiary script always imports
variables directly from the superior script. Consider this example:

  Import qw( env INCLUDE );

This is only legal if the parent script exported both C<$env> and
C<$INCLUDE>. It also must have given each of these variables values. It is
OK for the subsidiary script to only import a subset of the exported
variables (in this example, C<$LIB>, which was exported by the previous
example, is not imported).

All the imported variables are automatically re-exported, so the sequence:

  Import qw ( env INCLUDE );
  Build qw ( beneath-me/Conscript );

will supply both C<$env> and C<$INCLUDE> to the subsidiary file. If only
C<$env> is to be exported, then the following will suffice:

  Import qw ( env INCLUDE );
  Export qw ( env );
  Build qw ( beneath-me/Conscript );

Needless to say, the variables may be modified locally before invoking
C<Build> on the subsidiary script.


=head2 Build script evaluation order

The only constraint on the ordering of build scripts is that superior
scripts are evaluated before their inferior scripts. The top-level
F<Construct> file, for instance, is evaluated first, followed by any
inferior scripts. This is all you really need to know about the evaluation
order, since order is generally irrelevant. Consider the following C<Build>
command:

  Build qw(
	drivers/display/Conscript
	drivers/mouse/Conscript
	parser/Conscript
	utilities/Conscript
  );

We've chosen to put the script names in alphabetical order, simply because
that's the most convenient for maintenance purposes. Changing the order will
make no difference to the build.


=head1 A Model for sharing files


=head2 Some simple conventions

In any complex software system, a method for sharing build products needs to
be established. We propose a simple set of conventions which are trivial to
implement with Cons, but very effective.

The basic rule is to require that all build products which need to be shared
between directories are shared via an intermediate directory. We have
typically called this F<export>, and, in a C environment, provided
conventional sub-directories of this directory, such as F<include>, F<lib>,
F<bin>, etc.

These directories are defined by the top-level F<Construct> file. A simple
F<Construct> file for a B<Hello, World!> application, organized using
multiple directories, might look like this:

  # Construct file for Hello, World!

  # Where to put all our shared products.
  $EXPORT = '#export';

  Export qw( CONS INCLUDE LIB BIN );

  # Standard directories for sharing products.
  $INCLUDE = "$EXPORT/include";
  $LIB = "$EXPORT/lib";
  $BIN = "$EXPORT/bin";

  # A standard construction environment.
  $CONS = new cons (
	CPPPATH => $INCLUDE,	# Include path for C Compilations
	LIBPATH => $LIB,	# Library path for linking programs
	LIBS => '-lworld',	# List of standard libraries
  );

  Build qw(
	hello/Conscript
	world/Conscript
  );

The F<world> directory's F<Conscript> file looks like this:

  # Conscript file for directory world
  Import qw( CONS INCLUDE LIB );

  # Install the products of this directory
  Install $CONS $LIB, 'libworld.a';
  Install $CONS $INCLUDE, 'world.h';

  # Internal products
  Library $CONS 'libworld.a', 'world.c';

and the F<hello> directory's F<Conscript> file looks like this:

  # Conscript file for directory hello
  Import qw( CONS BIN );

  # Exported products
  Install $CONS $BIN, 'hello';

  # Internal products
  Program $CONS 'hello', 'hello.c';

To construct a B<Hello, World!> program with this directory structure, go to
the top-level directory, and invoke C<cons> with the appropriate
arguments. In the following example, we tell Cons to build the directory
F<export>. To build a directory, Cons recursively builds all known products
within that directory (only if they need rebuilding, of course). If any of
those products depend upon other products in other directories, then those
will be built, too.

  % cons export
  Install world/world.h as export/include/world.h
  cc -Iexport/include -c hello/hello.c -o hello/hello.o
  cc -Iexport/include -c world/world.c -o world/world.o
  ar r world/libworld.a world/world.o
  ar: creating world/libworld.a
  ranlib world/libworld.a
  Install world/libworld.a as export/lib/libworld.a
  cc -o hello/hello hello/hello.o -Lexport/lib -lworld
  Install hello/hello as export/bin/hello


=head2 Clean, understandable, location-independent scripts

You'll note that the two F<Conscript> files are very clean and
to-the-point. They simply specify products of the directory and how to build
those products. The build instructions are minimal: they specify which
construction environment to use, the name of the product, and the name of
the inputs. Note also that the scripts are location-independent: if you wish
to reorganize your source tree, you are free to do so: you only have to
change the F<Construct> file (in this example), to specify the new locations
of the F<Conscript> files. The use of an export tree makes this goal easy.

Note, too, how Cons takes care of little details for you. All the F<export>
directories, for example, were made automatically. And the installed files
were really hard-linked into the respective export directories, to save
space and time. This attention to detail saves considerable work, and makes
it even easier to produce simple, maintainable scripts.


=head1 Separating source and build trees

It's often desirable to keep any derived files from the build completely
separate from the source files. This makes it much easier to keep track of
just what is a source file, and also makes it simpler to handle B<variant>
builds, especially if you want the variant builds to co-exist.


=head2 Separating build and source directories using the Link command

Cons provides a simple mechanism that handles all of these requirements. The
C<Link> command is invoked as in this example:

  Link 'build' => 'src';

The specified directories are ``linked'' to the specified source
directory. Let's suppose that you setup a source directory, F<src>, with the
sub-directories F<world> and F<hello> below it, as in the previous
example. You could then substitute for the original build lines the
following:

  Build qw(
	build/world/Conscript
	build/hello/Conscript
  );

Notice that you treat the F<Conscript> file as if it existed in the build
directory. Now if you type the same command as before, you will get the
following results:

  % cons export
  Install build/world/world.h as export/include/world.h
  cc -Iexport/include -c build/hello/hello.c -o build/hello/hello.o
  cc -Iexport/include -c build/world/world.c -o build/world/world.o
  ar r build/world/libworld.a build/world/world.o
  ar: creating build/world/libworld.a
  ranlib build/world/libworld.a
  Install build/world/libworld.a as export/lib/libworld.a
  cc -o build/hello/hello build/hello/hello.o -Lexport/lib -lworld
  Install build/hello/hello as export/bin/hello

Again, Cons has taken care of the details for you. In particular, you will
notice that all the builds are done using source files and object files from
the build directory. For example, F<build/world/world.o> is compiled from
F<build/world/world.c>, and F<export/include/world.h> is installed from
F<build/world/world.h>. This is accomplished on most systems by the simple
expedient of ``hard'' linking the required files from each source directory
into the appropriate build directory.

The links are maintained correctly by Cons, no matter what you do to the
source directory. If you modify a source file, your editor may do this ``in
place'' or it may rename it first and create a new file. In the latter case,
any hard link will be lost. Cons will detect this condition the next time
the source file is needed, and will relink it appropriately.

You'll also notice, by the way, that B<no> changes were required to the
underlying F<Conscript> files. And we can go further, as we shall see in the
next section.

=head2 Explicit references to the source directory

When using the C<Link> command on some operating systems or with some
tool chains, it's sometimes useful to have a command actually use
the path name to the source directory, not the build directory.  For
example, on systems that must copy, not "hard link," the F<src/> and
F<build/> copies of C<Linked> files, using the F<src/> path of a file
name might make an editor aware that a syntax error must be fixed in the
source directory, not the build directory.

You can tell Cons that you want to use the "source path" for a file by
preceding the file name with a ``!'' (exclamation point).  For example,
if we add a ``!'' to the beginning of a source file:

  Program $env "foo", "!foo.c";	# Notice initial ! on foo.c

Cons will compile the target as follows:

  cc -c src/foo.c -o build/foo.o
  cc -o build/foo build/foo.o

Notice that Cons has compiled the program from the the F<src/foo.c>
source file.  Without the initial ``!'', Cons would have compiled the
program using the F<build/foo.c> path name.



=head1 Variant builds


=head2 Hello, World! for baNaNa and peAcH OS's

Variant builds require just another simple extension. Let's take as an
example a requirement to allow builds for both the baNaNa and peAcH
operating systems. In this case, we are using a distributed file system,
such as NFS to access the particular system, and only one or the other of
the systems has to be compiled for any given invocation of C<cons>. Here's
one way we could set up the F<Construct> file for our B<Hello, World!>
application:

  # Construct file for Hello, World!

  die qq(OS must be specified) unless $OS = $ARG{OS};
  die qq(OS must be "peach" or "banana")
	if $OS ne "peach" && $OS ne "banana";

  # Where to put all our shared products.
  $EXPORT = "#export/$OS";

  Export qw( CONS INCLUDE LIB BIN );

  # Standard directories for sharing products.
  $INCLUDE = "$EXPORT/include";
  $LIB = "$EXPORT/lib";
  $BIN = "$EXPORT/bin";

  # A standard construction environment.
  $CONS = new cons (
	CPPPATH => $INCLUDE,	# Include path for C Compilations
	LIBPATH => $LIB,	# Library path for linking programs
	LIBS => '-lworld',	# List of standard libraries
  );

  # $BUILD is where we will derive everything.
  $BUILD = "#build/$OS";

  # Tell cons where the source files for $BUILD are.
  Link $BUILD => 'src';

  Build (
	"$BUILD/hello/Conscript",
	"$BUILD/world/Conscript",
  );

Now if we login to a peAcH system, we can build our B<Hello, World!>
application for that platform:

  % cons export OS=peach
  Install build/peach/world/world.h as export/peach/include/world.h
  cc -Iexport/peach/include -c build/peach/hello/hello.c -o build/peach/hello/hello.o
  cc -Iexport/peach/include -c build/peach/world/world.c -o build/peach/world/world.o
  ar r build/peach/world/libworld.a build/peach/world/world.o
  ar: creating build/peach/world/libworld.a
  ranlib build/peach/world/libworld.a
  Install build/peach/world/libworld.a as export/peach/lib/libworld.a
  cc -o build/peach/hello/hello build/peach/hello/hello.o -Lexport/peach/lib -lworld
  Install build/peach/hello/hello as export/peach/bin/hello


=head2 Variations on a theme

Other variations of this model are possible. For example, you might decide
that you want to separate out your include files into platform dependent and
platform independent files. In this case, you'd have to define an
alternative to C<$INCLUDE> for platform-dependent files. Most F<Conscript>
files, generating purely platform-independent include files, would not have
to change.

You might also want to be able to compile your whole system with debugging
or profiling, for example, enabled. You could do this with appropriate
command line options, such as C<DEBUG=on>. This would then be translated
into the appropriate platform-specific requirements to enable debugging
(this might include turning off optimization, for example). You could
optionally vary the name space for these different types of systems, but, as
we'll see in the next section, it's not B<essential> to do this, since Cons
is pretty smart about rebuilding things when you change options.


=head1 Signatures

Cons uses file B<signatures> to decide if a derived file is out-of-date
and needs rebuilding.  In essence, if the contents of a file change,
or the manner in which the file is built changes, the file's signature
changes as well.  This allows Cons to decide with certainty when a file
needs rebuilding, because Cons can detect, quickly and reliably, whether
any of its dependency files have been changed.


=head2 MD5 content and build signatures

Cons uses the B<MD5> (B<Message Digest 5>) algorithm to compute file
signatures.  The MD5 algorithm computes a strong cryptographic checksum
for any given input string.  Cons can, based on configuration, use two
different MD5 signatures for a given file:

The B<content signature> of a file is an MD5 checksum of the file's
contents.  Consequently, when the contents of a file change, its content
signature changes as well.

The B<build signature> of a file is a combined MD5 checksum of:

=over 4

the signatures of all the input files used to build the file

the signatures of all dependency files discovered by source scanners
(for example, C<.h> files)

the signatures of all dependency files specified explicitly via the
C<Depends> method)

the command-line string used to build the file

=back

The build signature is, in effect, a digest of all the dependency
information for the specified file.  Consequently, a file's build
signature changes whenever any part of its dependency information
changes: a new file is added, the contents of a file on which it depends
change, there's a change to the command line used to build the file (or
any of its dependency files), etc.

For example, in the previous section, the build signature of the
F<world.o> file will include:

=over 4

the signature of the F<world.c> file

the signatures of any header files that Cons detects are included,
directly or indirectly, by F<world.c>

the text of the actual command line was used to generate F<world.o>

=back

Similarly, the build signature of the F<libworld.a> file will include
all the signatures of its constituents (and hence, transitively, the
signatures of B<their> constituents), as well as the command line that
created the file.

Note that there is no need for a derived file to depend upon any
particular F<Construct> or F<Conscript> file.  If changes to these files
affect a file, then this will be automatically reflected in its build
signature, since relevant parts of the command line are included in the
signature. Unrelated F<Construct> or F<Conscript> changes will have no
effect.


=head2 Storing signatures in .consign files

Before Cons exits, it stores the calculated signatures for all of the
files it built or examined in F<.consign> files, one per directory.
Cons uses this stored information on later invocations to decide if
derived files need to be rebuilt.

After the previous example was compiled, the F<.consign> file in the
F<build/peach/world> directory looked like this:

  world.h:985533370 - d181712f2fdc07c1f05d97b16bfad904
  world.o:985533372 2a0f71e0766927c0532977b0d2158981
  world.c:985533370 - c712f77189307907f4189b5a7ab62ff3
  libworld.a:985533374 69e568fc5241d7d25be86d581e1fb6aa

After the file name and colon, the first number is a timestamp of the
file's modification time (on UNIX systems, this is typically the number
of seconds since January 1st, 1970).  The second value is the build
signature of the file (or ``-'' in the case of files with no build
signature--that is, source files).  The third value, if any, is the
content signature of the file.


=head2 Using build signatures to decide when to rebuild files

When Cons is deciding whether to build or rebuild a derived file, it
first computes the file's current build signature.  If the file doesn't
exist, it must obviously be built.

If, however, the file already exists, Cons next compares the
modification timestamp of the file against the timestamp value in
the F<.consign> file.  If the timestamps match, Cons compares the
newly-computed build signature against the build signature in the
F<.consign> file.  If the timestamps do not match or the build
signatures do not match, the derived file is rebuilt.

After the file is built or rebuilt, Cons arranges to store the
newly-computed build signature in the F<.consign> file when it exits.


=head2 Signature example

The use of these signatures is an extremely simple, efficient, and
effective method of improving--dramatically--the reproducibility of a
system.

We'll demonstrate this with a simple example:

  # Simple "Hello, World!" Construct file
  $CFLAGS = '-g' if $ARG{DEBUG} eq 'on';
  $CONS = new cons(CFLAGS => $CFLAGS);
  Program $CONS 'hello', 'hello.c';

Notice how Cons recompiles at the appropriate times:

  % cons hello
  cc -c hello.c -o hello.o
  cc -o hello hello.o
  % cons hello
  cons: "hello" is up-to-date.
  % cons DEBUG=on hello
  cc -g -c hello.c -o hello.o
  cc -o hello hello.o
  % cons DEBUG=on hello
  cons: "hello" is up-to-date.
  % cons hello
  cc -c hello.c -o hello.o
  cc -o hello hello.o


=head2 Source-file signature configuration

Cons provides a C<SourceSignature> method that allows you to configure
how the signature should be calculated for any source file when its
signature is being used to decide if a dependent file is up-to-date.
The arguments to the C<SourceSignature> method consist of one or more
pairs of strings:

  SourceSignature 'auto/*.c' => 'content',
		  '*' => 'stored-content';

The first string in each pair is a pattern to match against derived file
path names. The pattern is a file-globbing pattern, not a Perl regular
expression; the pattern <*.l> will match all Lex source files.  The C<*>
wildcard will match across directory separators; the pattern C<foo/*.c>
would match all C source files in any subdirectory underneath the C<foo>
subdirectory.

The second string in each pair contains one of the following keywords to
specify how signatures should be calculated for source files that match
the pattern.  The available keywords are:

=over 4

=item content

Use the content signature of the source file when calculating signatures
of files that depend on it.  This guarantees correct calculation of the
file's signature for all builds, by telling Cons to read the contents of
a source file to calculate its content signature each time it is run.

=item stored-content

Use the source file's content signature as stored in the F<.consign>
file, provided the file's timestamp matches the cached timestamp value
in the F<.consign> file.  This optimizes performance, with the slight
risk of an incorrect build if a source file's contents have been changed
so quickly after its previous update that the timestamp still matches
the stored timestamp in the F<.consign> file even though the contents
have changed.

=back

The Cons default behavior of always calculating a source file's
signature from the file's contents is equivalent to specifying:

  SourceSignature '*' => 'content';

The C<*> will match all source files.  The C<content> keyword
specifies that Cons will read the contents of a source file to calculate
its signature each time it is run.

A useful global performance optimization is:

  SourceSignature '*' => 'stored-content';

This specifies that Cons will use pre-computed content signatures
from F<.consign> files, when available, rather than re-calculating a
signature from the the source file's contents each time Cons is run.  In
practice, this is safe for most build situations, and only a problem
when source files are changed automatically (by scripts, for example).
The Cons default, however, errs on the side of guaranteeing a correct
build in all situations.

Cons tries to match source file path names against the patterns in the
order they are specified in the C<SourceSignature> arguments:

  SourceSignature '/usr/repository/objects/*' => 'stored-content',
		  '/usr/repository/*' => 'content',
		  '*.y' => 'content',
		  '*' => 'stored-content';

In this example, all source files under the F</usr/repository/objects>
directory will use F<.consign> file content signatures, source files
anywhere else underneath F</usr/repository> will not use F<.consign>
signature values, all Yacc source files (C<*.y>) anywhere else will not
use F<.consign> signature values, and any other source file will use
F<.consign> signature values.


=head2 Derived-file signature configuration

Cons provides a C<SIGNATURE> construction variable that allows you to
configure how signatures are calculated for any derived file when its
signature is being used to decide if a dependent file is up-to-date.
The value of the C<SIGNATURE> construction variable is a Perl array
reference that holds one or more pairs of strings, like the arguments to
the C<SourceSignature> method.

The first string in each pair is a pattern to match against derived file
path names. The pattern is a file-globbing pattern, not a Perl regular
expression; the pattern `*.obj' will match all (Win32) object files.
The C<*> wildcard will match across directory separators; the pattern
`foo/*.a' would match all (UNIX) library archives in any subdirectory
underneath the foo subdirectory.

The second string in each pair contains one of the following keywords
to specify how signatures should be calculated for derived files that
match the pattern.  The available keywords are the same as for the
C<SourceSignature> method, with an additional keyword:

=over 4

=item build

Use the build signature of the derived file when calculating signatures
of files that depend on it.  This guarantees correct builds by forcing
Cons to rebuild any and all files that depend on the derived file.

=item content

Use the content signature of the derived file when calculating signatures
of files that depend on it.  This guarantees correct calculation of the
file's signature for all builds, by telling Cons to read the contents of
a derived file to calculate its content signature each time it is run.

=item stored-content

Use the derived file's content signature as stored in the F<.consign>
file, provided the file's timestamp matches the cached timestamp value
in the F<.consign> file.  This optimizes performance, with the slight
risk of an incorrect build if a derived file's contents have been
changed so quickly after a Cons build that the file's timestamp still
matches the stored timestamp in the F<.consign> file.

=back

The Cons default behavior (as previously described) for using
derived-file signatures is equivalent to:

  $env = new cons(SIGNATURE => ['*' => 'build']);

The C<*> will match all derived files.  The C<build> keyword specifies
that all derived files' build signatures will be used when calculating
whether a dependent file is up-to-date.

A useful alternative default C<SIGNATURE> configuration for many sites:

  $env = new cons(SIGNATURE => ['*' => 'content']);

In this configuration, derived files have their signatures calculated
from the file contents.  This adds slightly to Cons' workload, but has
the useful effect of "stopping" further rebuilds if a derived file is
rebuilt to exactly the same file contents as before, which usually
outweighs the additional computation Cons must perform.

For example, changing a comment in a C file and recompiling should
generate the exact same object file (assuming the compiler doesn't
insert a timestamp in the object file's header).  In that case,
specifying C<content> or C<stored-content> for the signature calculation
will cause Cons to recognize that the object file did not actually
change as a result of being rebuilt, and libraries or programs that
include the object file will not be rebuilt.  When C<build> is
specified, however, Cons will only "know" that the object file was
rebuilt, and proceed to rebuild any additional files that include the
object file.

Note that Cons tries to match derived file path names against the
patterns in the order they are specified in the C<SIGNATURE> array
reference:

  $env = new cons(SIGNATURE => ['foo/*.o' => 'build',
				'*.o' => 'content',
				'*.a' => 'stored-content',
				'*' => 'content']);

In this example, all object files underneath the F<foo> subdirectory
will use build signatures, all other object files (including object
files underneath other subdirectories!) will use F<.consign> file
content signatures, libraries will use F<.consign> file build
signatures, and all other derived files will use content signatures.


=head2 Debugging signature calculation

Cons provides a C<-S> option that can be used to specify what internal
Perl package Cons should use to calculate signatures.  The default Cons
behavior is equivalent to specifying C<-S md5> on the command line.

The only other package (currently) available is an C<md5::debug>
package that prints out detailed information about the MD5 signature
calculations performed by Cons:

  % cons -S md5::debug hello
  sig::md5::srcsig(hello.c)
          => |52d891204c62fe93ecb95281e1571938|
  sig::md5::collect(52d891204c62fe93ecb95281e1571938)
          => |fb0660af4002c40461a2f01fbb5ffd03|
  sig::md5::collect(52d891204c62fe93ecb95281e1571938,
                    fb0660af4002c40461a2f01fbb5ffd03,
                    cc   -c %< -o %>)
          => |f7128da6c3fe3c377dc22ade70647b39|
  sig::md5::current(||
                 eq |f7128da6c3fe3c377dc22ade70647b39|)
  cc -c hello.c -o hello.o
  sig::md5::collect()
          => |d41d8cd98f00b204e9800998ecf8427e|
  sig::md5::collect(f7128da6c3fe3c377dc22ade70647b39,
                    d41d8cd98f00b204e9800998ecf8427e,
                    cc  -o %> %<  )
          => |a0bdce7fd09e0350e7efbbdb043a00b0|
  sig::md5::current(||
                 eq |a0bdce7fd09e0350e7efbbdb043a00b0|)
  cc -o hello, hello.o


=head1 Code Repositories

Many software development organizations will have one or more central
repository directory trees containing the current source code for one or
more projects, as well as the derived object files, libraries, and
executables.  In order to reduce unnecessary recompilation, it is useful to
use files from the repository to build development software--assuming, of
course, that no newer dependency file exists in the local build tree.


=head2 Repository

Cons provides a mechanism to specify a list of code repositories that will
be searched, in-order, for source files and derived files not found in the
local build directory tree.

The following lines in a F<Construct> file will instruct Cons to look first
under the F</usr/experiment/repository> directory and then under the
F</usr/product/repository> directory:

  Repository qw (
	/usr/experiment/repository
	/usr/product/repository
  );

The repository directories specified may contain source files, derived files
(objects, libraries and executables), or both.  If there is no local file
(source or derived) under the directory in which Cons is executed, then the
first copy of a same-named file found under a repository directory will be
used to build any local derived files.

Cons maintains one global list of repositories directories.  Cons will
eliminate the current directory, and any non-existent directories, from the
list.


=head2 Finding the Construct file in a Repository

Cons will also search for F<Construct> and F<Conscript> files in the
repository tree or trees.  This leads to a chicken-and-egg situation,
though: how do you look in a repository tree for a F<Construct> file if the
F<Construct> file tells you where the repository is?  To get around this,
repositories may be specified via C<-R> options on the command line:

  % cons -R /usr/experiment/repository -R /usr/product/repository .

Any repository directories specified in the F<Construct> or F<Conscript>
files will be appended to the repository directories specified by
command-line C<-R> options.

=head2 Repository source files

If the source code (include the F<Conscript> file) for the library version
of the I<Hello, World!> C application is in a repository (with no derived
files), Cons will use the repository source files to create the local object
files and executable file:

  % cons -R /usr/src_only/repository hello
  gcc -c /usr/src_only/repository/hello.c -o hello.o
  gcc -c /usr/src_only/repository/world.c -o world.o
  ar r libworld.a world.o
  ar: creating libworld.a
  ranlib libworld.a
  gcc -o hello hello.o libworld.a

Creating a local source file will cause Cons to rebuild the appropriate
derived file or files:

  % pico world.c
    [EDIT]
  % cons -R /usr/src_only/repository hello
  gcc -c world.c -o world.o
  ar r libworld.a world.o
  ar: creating libworld.a
  ranlib libworld.a
  gcc -o hello hello.o libworld.a

And removing the local source file will cause Cons to revert back to
building the derived files from the repository source:

  % rm world.c
  % cons -R /usr/src_only/repository hello
  gcc -c /usr/src_only/repository/world.c -o world.o
  ar r libworld.a world.o
  ar: creating libworld.a
  ranlib libworld.a
  gcc -o hello hello.o libworld.a


=head2 Repository derived files

If a repository tree contains derived files (usually object files,
libraries, or executables), Cons will perform its normal signature
calculation to decide whether the repository file is up-to-date or a derived
file must be built locally.  This means that, in order to ensure correct
signature calculation, a repository tree must also contain the F<.consign>
files that were created by Cons when generating the derived files.

This would usually be accomplished by building the software in the
repository (or, alternatively, in a build directory, and then copying the
result to the repository):

  % cd /usr/all/repository
  % cons hello
  gcc -c hello.c -o hello.o
  gcc -c world.c -o world.o
  ar r libworld.a world.o
  ar: creating libworld.a
  ranlib libworld.a
  gcc -o hello hello.o libworld.a

(This is safe even if the F<Construct> file lists the F</usr/all/repository>
directory in a C<Repository> command because Cons will remove the current
directory from the repository list.)

Now if we want to build a copy of the application with our own F<hello.c>
file, we only need to create the one necessary source file, and use the
C<-R> option to have Cons use other files from the repository:

  % mkdir $HOME/build1
  % cd $HOME/build1
  % ed hello.c
    [EDIT]
  % cons -R /usr/all/repository hello
  gcc -c hello.c -o hello.o
  gcc -o hello hello.o /usr/all/repository/libworld.a

Notice that Cons has not bothered to recreate a local F<libworld.a> library
(or recompile the F<world.o> module), but instead uses the already-compiled
version from the repository.

Because the MD5 signatures that Cons puts in the F<.consign> file contain
timestamps for the derived files, the signature timestamps must match the
file timestamps for a signature to be considered valid.

Some software systems may alter the timestamps on repository files (by
copying them, e.g.), in which case Cons will, by default, assume the
repository signatures are invalid and rebuild files unnecessarily.  This
behavior may be altered by specifying:

  Repository_Sig_Times_OK 0;

This tells Cons to ignore timestamps when deciding whether a signature is
valid.  (Note that avoiding this sanity check means there must be proper
control over the repository tree to ensure that the derived files cannot be
modified without updating the F<.consign> signature.)


=head2 Local copies of files

If the repository tree contains the complete results of a build, and we try
to build from the repository without any files in our local tree, something
moderately surprising happens:

  % mkdir $HOME/build2
  % cd $HOME/build2
  % cons -R /usr/all/repository hello
  cons: "hello" is up-to-date.

Why does Cons say that the F<hello> program is up-to-date when there is no
F<hello> program in the local build directory?  Because the repository (not
the local directory) contains the up-to-date F<hello> program, and Cons
correctly determines that nothing needs to be done to rebuild this
up-to-date copy of the file.

There are, however, many times in which it is appropriate to ensure that a
local copy of a file always exists.  A packaging or testing script, for
example, may assume that certain generated files exist locally.  Instead of
making these subsidiary scripts aware of the repository directory, the
C<Local> command may be added to a F<Construct> or F<Conscript> file to
specify that a certain file or files must appear in the local build
directory:

  Local qw(
	hello
  );

Then, if we re-run the same command, Cons will make a local copy of the
program from the repository copy (telling you that it is doing so):

  % cons -R /usr/all/repository hello
  Local copy of hello from /usr/all/repository/hello
  cons: "hello" is up-to-date.

Notice that, because the act of making the local copy is not considered a
"build" of the F<hello> file, Cons still reports that it is up-to-date.

Creating local copies is most useful for files that are being installed into
an intermediate directory (for sharing with other directories) via the
C<Install> command.  Accompanying the C<Install> command for a file with a
companion C<Local> command is so common that Cons provides a
C<Install_Local> command as a convenient way to do both:

  Install_Local $env, '#export', 'hello';

is exactly equivalent to:

  Install $env '#export', 'hello';
  Local '#export/hello';

Both the C<Local> and C<Install_Local> commands update the local F<.consign>
file with the appropriate file signatures, so that future builds are
performed correctly.


=head2 Repository dependency analysis

Due to its built-in scanning, Cons will search the specified repository
trees for included F<.h> files.  Unless the compiler also knows about the
repository trees, though, it will be unable to find F<.h> files that only
exist in a repository.  If, for example, the F<hello.c> file includes the
F<hello.h> file in its current directory:

  % cons -R /usr/all/repository hello
  gcc -c /usr/all/repository/hello.c -o hello.o
  /usr/all/repository/hello.c:1: hello.h: No such file or directory

Solving this problem forces some requirements onto the way construction
environments are defined and onto the way the C C<#include> preprocessor
directive is used to include files.

In order to inform the compiler about the repository trees, Cons will add
appropriate C<-I> flags to the compilation commands.  This means that the
C<CPPPATH> variable in the construction environment must explicitly specify
all subdirectories which are to be searched for included files, including the
current directory.  Consequently, we can fix the above example by changing
the environment creation in the F<Construct> file as follows:

  $env = new cons(
	CC	=> 'gcc',
	CPPPATH	=> '.',
	LIBS	=> 'libworld.a',
  );

Due to the definition of the C<CPPPATH> variable, this yields, when we
re-execute the command:

  % cons -R /usr/all/repository hello
  gcc -c -I. -I/usr/all/repository /usr/all/repository/hello.c -o hello.o
  gcc -o hello hello.o /usr/all/repository/libworld.a

The order of the C<-I> flags replicates, for the C preprocessor, the same
repository-directory search path that Cons uses for its own dependency
analysis.  If there are multiple repositories and multiple C<CPPPATH>
directories, Cons will append the repository directories to the beginning of
each C<CPPPATH> directory, rapidly multiplying the number of C<-I> flags.
As an extreme example, a F<Construct> file containing:

  Repository qw(
	/u1
	/u2
  );

  $env = new cons(
	CPPPATH	=> 'a:b:c',
  );

Would yield a compilation command of:

  cc -Ia -I/u1/a -I/u2/a -Ib -I/u1/b -I/u2/b -Ic -I/u1/c -I/u2/c -c hello.c -o hello.o

In order to shorten the command lines as much as possible, Cons will
remove C<-I> flags for any directories, locally or in the repositories,
which do not actually exist.  (Note that the C<-I> flags are not included
in the MD5 signature calculation for the target file, so the target will
not be recompiled if the compilation command changes due to a directory
coming into existence.)

Because Cons relies on the compiler's C<-I> flags to communicate the
order in which repository directories must be searched, Cons' handling
of repository directories is fundamentally incompatible with using
double-quotes on the C<#include> directives in any C source code that
you plan to modify:

  #include "file.h"	/* DON'T USE DOUBLE-QUOTES LIKE THIS */

This is because most C preprocessors, when faced with such a directive, will
always first search the directory containing the source file.  This
undermines the elaborate C<-I> options that Cons constructs to make the
preprocessor conform to its preferred search path.

Consequently, when using repository trees in Cons, B<always> use
angle-brackets for included files in any C source (.c or .h) files that
you plan to modify locally:

  #include <file.h>	/* USE ANGLE-BRACKETS INSTEAD */

Code that will not change can still safely use double quotes on #include
lines.


=head2 Repository_List

Cons provides a C<Repository_List> command to return a list of all
repository directories in their current search order.  This can be used for
debugging, or to do more complex Perl stuff:

  @list = Repository_List;
  print join(' ', @list), "\n";


=head2 Repository interaction with other Cons features

Cons' handling of repository trees interacts correctly with other Cons
features--which is to say, it generally does what you would expect.

Most notably, repository trees interact correctly, and rather powerfully,
with the 'Link' command.  A repository tree may contain one or more
subdirectories for version builds established via C<Link> to a source
subdirectory.  Cons will search for derived files in the appropriate build
subdirectories under the repository tree.


=head1 Default targets

Until now, we've demonstrated invoking Cons with an explicit target
to build:

  % cons hello

Normally, Cons does not build anything unless a target is specified,
but specifying '.' (the current directory) will build everything:

  % cons		# does not build anything

  % cons .		# builds everything under the top-level directory

Adding the C<Default> method to any F<Construct> or F<Conscript> file will add
the specified targets to a list of default targets.  Cons will build
these defaults if there are no targets specified on the command line.
So adding the following line to the top-level F<Construct> file will mimic
Make's typical behavior of building everything by default:

  Default '.';

The following would add the F<hello> and F<goodbye> commands (in the
same directory as the F<Construct> or F<Conscript> file) to the default list:

  Default qw(
	hello
	goodbye
  );

The C<Default> method may be used more than once to add targets to the
default list.

=head1 Selective builds

Cons provides two methods for reducing the size of given build. The first is
by specifying targets on the command line, and the second is a method for
pruning the build tree. We'll consider target specification first.


=head2 Selective targeting

Like make, Cons allows the specification of ``targets'' on the command
line. Cons targets may be either files or directories. When a directory is
specified, this is simply a short-hand notation for every derivable
product--that Cons knows about--in the specified directory and below. For
example:

  % cons build/hello/hello.o

means build F<hello.o> and everything that F<hello.o> might need. This is
from a previous version of the B<Hello, World!> program in which F<hello.o>
depended upon F<export/include/world.h>. If that file is not up-to-date
(because someone modified F<src/world/world.h)>, then it will be rebuilt,
even though it is in a directory remote from F<build/hello>.

In this example:

  % cons build

Everything in the F<build> directory is built, if necessary. Again, this may
cause more files to be built. In particular, both F<export/include/world.h>
and F<export/lib/libworld.a> are required by the F<build/hello> directory,
and so they will be built if they are out-of-date.

If we do, instead:

  % cons export

then only the files that should be installed in the export directory will be
rebuilt, if necessary, and then installed there. Note that C<cons build>
might build files that C<cons export> doesn't build, and vice-versa.


=head2 No ``special'' targets

With Cons, make-style ``special'' targets are not required. The simplest
analog with Cons is to use special F<export> directories, instead. Let's
suppose, for example, that you have a whole series of unit tests that are
associated with your code. The tests live in the source directory near the
code. Normally, however, you don't want to build these tests. One solution
is to provide all the build instructions for creating the tests, and then to
install the tests into a separate part of the tree. If we install the tests
in a top-level directory called F<tests>, then:

  % cons tests

will build all the tests.

  % cons export

will build the production version of the system (but not the tests), and:

  % cons build

should probably be avoided (since it will compile tests unnecessarily).

If you want to build just a single test, then you could explicitly name the
test (in either the F<tests> directory or the F<build> directory). You could
also aggregate the tests into a convenient hierarchy within the tests
directory. This hierarchy need not necessarily match the source hierarchy,
in much the same manner that the include hierarchy probably doesn't match
the source hierarchy (the include hierarchy is unlikely to be more than two
levels deep, for C programs).

If you want to build absolutely everything in the tree (subject to whatever
options you select), you can use:

  % cons .

This is not particularly efficient, since it will redundantly walk all the
trees, including the source tree. The source tree, of course, may have
buildable objects in it--nothing stops you from doing this, even if you
normally build in a separate build tree.


=head1 Build Pruning

In conjunction with target selection, B<build pruning> can be used to reduce
the scope of the build. In the previous peAcH and baNaNa example, we have
already seen how script-driven build pruning can be used to make only half
of the potential build available for any given invocation of C<cons>. Cons
also provides, as a convenience, a command line convention that allows you
to specify which F<Conscript> files actually get ``built''--that is,
incorporated into the build tree. For example:

  % cons build +world

The C<+> argument introduces a Perl regular expression. This must, of
course, be quoted at the shell level if there are any shell meta-characters
within the expression. The expression is matched against each F<Conscript>
file which has been mentioned in a C<Build> statement, and only those
scripts with matching names are actually incorporated into the build
tree. Multiple such arguments are allowed, in which case a match against any
of them is sufficient to cause a script to be included.

In the example, above, the F<hello> program will not be built, since Cons
will have no knowledge of the script F<hello/Conscript>. The F<libworld.a>
archive will be built, however, if need be.

There are a couple of uses for build pruning via the command line. Perhaps
the most useful is the ability to make local changes, and then, with
sufficient knowledge of the consequences of those changes, restrict the size
of the build tree in order to speed up the rebuild time. A second use for
build pruning is to actively prevent the recompilation of certain files that
you know will recompile due to, for example, a modified header file. You may
know that either the changes to the header file are immaterial, or that the
changes may be safely ignored for most of the tree, for testing
purposes.With Cons, the view is that it is pragmatic to admit this type of
behavior, with the understanding that on the next full build everything that
needs to be rebuilt will be. There is no equivalent to a ``make touch''
command, to mark files as permanently up-to-date. So any risk that is
incurred by build pruning is mitigated. For release quality work, obviously,
we recommend that you do not use build pruning (it's perfectly OK to use
during integration, however, for checking compilation, etc. Just be sure to
do an unconstrained build before committing the integration).


=head1 Temporary overrides

Cons provides a very simple mechanism for overriding aspects of a build. The
essence is that you write an override file containing one or more
C<Override> commands, and you specify this on the command line, when you run
C<cons>:

  % cons -o over export

will build the F<export> directory, with all derived files subject to the
overrides present in the F<over> file. If you leave out the C<-o> option,
then everything necessary to remove all overrides will be rebuilt.


=head2 Overriding environment variables

The override file can contain two types of overrides. The first is incoming
environment variables. These are normally accessible by the F<Construct>
file from the C<%ENV> hash variable. These can trivially be overridden in
the override file by setting the appropriate elements of C<%ENV> (these
could also be overridden in the user's environment, of course).


=head2 The Override command

The second type of override is accomplished with the C<Override> command,
which looks like this:

  Override <regexp>, <var1> => <value1>, <var2> => <value2>, ...;

The regular expression I<regexp> is matched against every derived file that
is a candidate for the build. If the derived file matches, then the
variable/value pairs are used to override the values in the construction
environment associated with the derived file.

Let's suppose that we have a construction environment like this:

  $CONS = new cons(
	COPT => '',
	CDBG => '-g',
	CFLAGS => '%COPT %CDBG',
  );

Then if we have an override file F<over> containing this command:

  Override '\.o$', COPT => '-O', CDBG => '';

then any C<cons> invocation with C<-o over> that creates F<.o> files via
this environment will cause them to be compiled with C<-O >and no C<-g>. The
override could, of course, be restricted to a single directory by the
appropriate selection of a regular expression.

Here's the original version of the Hello, World! program, built with this
environment. Note that Cons rebuilds the appropriate pieces when the
override is applied or removed:

  % cons hello
  cc -g -c hello.c -o hello.o
  cc -o hello hello.o
  % cons -o over hello
  cc -O -c hello.c -o hello.o
  cc -o hello hello.o
  % cons -o over hello
  cons: "hello" is up-to-date.
  % cons hello
  cc -g -c hello.c -o hello.o
  cc -o hello hello.o

It's important that the C<Override> command only be used for temporary,
on-the-fly overrides necessary for development because the overrides are not
platform independent and because they rely too much on intimate knowledge of
the workings of the scripts. For temporary use, however, they are exactly
what you want.

Note that it is still useful to provide, say, the ability to create a fully
optimized version of a system for production use--from the F<Construct> and
F<Conscript> files. This way you can tailor the optimized system to the
platform. Where optimizer trade-offs need to be made (particular files may
not be compiled with full optimization, for example), then these can be
recorded for posterity (and reproducibility) directly in the scripts.


=head1 More on construction environments

As previously mentioned, a B<construction environment> is an object that
has a set of keyword/value pairs and a set of methods, and which is used
to tell Cons how target files should be built.  This section describes
how Cons uses and expands construction environment values to control its
build behavior.

=head2 Construction variable expansion

Construction variables from a construction environment are expanded
by preceding the keyword with a C<%> (percent sign):

  Construction variables:
	XYZZY => 'abracadabra',

  The string:  "The magic word is:  %XYZZY!"
  expands to:  "The magic word is:  abracadabra!"

A construction variable name may be surrounded by C<{> and C<}> (curly
braces), which are stripped as part of the expansion.  This can
sometimes be necessary to separate a variable expansion from trailing
alphanumeric characters:

  Construction variables:
	OPT    => 'value1',
	OPTION => 'value2',

  The string:  "%OPT %{OPT}ION %OPTION %{OPTION}"
  expands to:  "value1 value1ION value2 value2"

Construction variable expansion is recursive--that is, a string
containing C<%->expansions after substitution will be re-expanded until
no further substitutions can be made:

  Construction variables:
	STRING => 'The result is:  %FOO',
	FOO    => '%BAR',
	BAR    => 'final value',

  The string:  "The string says:  %STRING"
  expands to:  "The string says:  The result is:  final value"

If a construction variable is not defined in an environment, then the
null string is substituted:

  Construction variables:
	FOO => 'value1',
	BAR => 'value2',

  The string:  "%FOO <%NO_VARIABLE> %BAR"
  expands to:  "value1 <> value2"

A doubled C<%%> will be replaced by a single C<%>:

  The string:  "Here is a percent sign:  %%"
  expands to:  "Here is a percent sign: %"

=head2 Default construction variables

When you specify no arguments when creating a new construction
environment:

  $env = new cons();

Cons creates a reference to a new, default construction
environment. This contains a number of construction variables and some
methods. At the present writing, the default construction variables on a
UNIX system are:

  CC            => 'cc',
  CFLAGS        => '',
  CCCOM         => '%CC %CFLAGS %_IFLAGS -c %< -o %>',
  CXX           => '%CC',
  CXXFLAGS      => '%CFLAGS',
  CXXCOM        => '%CXX %CXXFLAGS %_IFLAGS -c %< -o %>',
  INCDIRPREFIX  => '-I',
  INCDIRSUFFIX  => '',
  LINK          => '%CXX',
  LINKCOM       => '%LINK %LDFLAGS -o %> %< %_LDIRS %LIBS',
  LINKMODULECOM => '%LD -r -o %> %<',
  LIBDIRPREFIX  => '-L',
  LIBDIRSUFFIX  => '',
  AR		=> 'ar',
  ARFLAGS	=> 'r',
  ARCOM		=> ['%AR %ARFLAGS %> %<', '%RANLIB %>'],
  RANLIB	=> 'ranlib',
  AS		=> 'as',
  ASFLAGS	=> '',
  ASCOM		=> '%AS %ASFLAGS %< -o %>',
  LD		=> 'ld',
  LDFLAGS	=> '',
  PREFLIB	=> 'lib',
  SUFLIB	=> '.a',
  SUFLIBS	=> '.so:.a',
  SUFOBJ	=> '.o',
  SIGNATURE     => [ '*' => 'build' ],
  ENV		=> { 'PATH' => '/bin:/usr/bin' },


And on a Win32 system (Windows NT), the default construction variables
are (unless the default rule style is set using the B<DefaultRules>
method):

  CC		=> 'cl',
  CFLAGS	=> '/nologo',
  CCCOM		=> '%CC %CFLAGS %_IFLAGS /c %< /Fo%>',
  CXXCOM        => '%CXX %CXXFLAGS %_IFLAGS /c %< /Fo%>',
  INCDIRPREFIX  => '/I',
  INCDIRSUFFIX  => '',
  LINK          => 'link',
  LINKCOM       => '%LINK %LDFLAGS /out:%> %< %_LDIRS %LIBS',
  LINKMODULECOM => '%LD /r /o %> %<',
  LIBDIRPREFIX  => '/LIBPATH:',
  LIBDIRSUFFIX  => '',
  AR            => 'lib',
  ARFLAGS       => '/nologo ',
  ARCOM         => "%AR %ARFLAGS /out:%> %<",
  RANLIB        => '',
  LD            => 'link',
  LDFLAGS       => '/nologo ',
  PREFLIB       => '',
  SUFEXE	=> '.exe',
  SUFLIB	=> '.lib',
  SUFLIBS	=> '.dll:.lib',
  SUFOBJ	=> '.obj',
  SIGNATURE     => [ '*' => 'build' ],

These variables are used by the various methods associated with the
environment. In particular, any method that ultimately invokes an external
command will substitute these variables into the final command, as
appropriate. For example, the C<Objects> method takes a number of source
files and arranges to derive, if necessary, the corresponding object
files:

  Objects $env 'foo.c', 'bar.c';

This will arrange to produce, if necessary, F<foo.o> and F<bar.o>. The
command invoked is simply C<%CCCOM>, which expands, through substitution,
to the appropriate external command required to build each object. The
substitution rules will be discussed in detail in the next section.

The construction variables are also used for other purposes. For example,
C<CPPPATH> is used to specify a colon-separated path of include
directories. These are intended to be passed to the C preprocessor and are
also used by the C-file scanning machinery to determine the dependencies
involved in a C Compilation.

Variables beginning with underscore are created by various methods,
and should normally be considered ``internal'' variables. For example,
when a method is called which calls for the creation of an object from
a C source, the variable C<_IFLAGS> is created: this corresponds to the
C<-I> switches required by the C compiler to represent the directories
specified by C<CPPPATH>.

Note that, for any particular environment, the value of a variable is set
once, and then never reset (to change a variable, you must create a new
environment. Methods are provided for copying existing environments for this
purpose). Some internal variables, such as C<_IFLAGS> are created on demand,
but once set, they remain fixed for the life of the environment.

The C<CFLAGS>, C<LDFLAGS>, and C<ARFLAGS> variables all supply a place
for passing options to the compiler, loader, and archiver, respectively.

The C<INCDIRPREFIX> and C<INCDIRSUFFIX> variables specify option
strings to be appended to the beginning and end, respectively, of each
include directory so that the compiler knows where to find F<.h> files.
Similarly, the C<LIBDIRPREFIX> and C<LIBDIRSUFFIX> variables specify the
option string to be appended to the beginning of and end, respectively,
of each directory that the linker should search for libraries.

Another variable, C<ENV>, is used to determine the system environment during
the execution of an external command. By default, the only environment
variable that is set is C<PATH>, which is the execution path for a UNIX
command. For the utmost reproducibility, you should really arrange to set
your own execution path, in your top-level F<Construct> file (or perhaps by
importing an appropriate construction package with the Perl C<use>
command). The default variables are intended to get you off the ground.

=head2 Expanding variables in construction commands

Within a construction command, construction variables will be expanded
according to the rules described above.  In addition to normal variable
expansion from the construction environment, construction commands also
expand the following pseudo-variables to insert the specific input and
output files in the command line that will be executed:

=over 10

=item %>

The target file name.  In a multi-target command, this expands to the
first target mentioned.)

=item %0

Same as C<%E<gt>>.

=item %1, %2, ..., %9

These refer to the first through ninth input file, respectively.

=item %E<lt>

The full set of input file names. If any of these have been used
anywhere else in the current command line (via C<%1>, C<%2>, etc.), then
those will be deleted from the list provided by C<%E<lt>>. Consider the
following command found in a F<Conscript> file in the F<test> directory:

  Command $env 'tgt', qw(foo bar baz), qq(
	echo %< -i %1 > %>
	echo %< -i %2 >> %>
	echo %< -i %3 >> %>
  );

If F<tgt> needed to be updated, then this would result in the execution of
the following commands, assuming that no remapping has been established for
the F<test> directory:

  echo test/bar test/baz -i test/foo > test/tgt
  echo test/foo test/baz -i test/bar >> test/tgt
  echo test/foo test/bar -i test/baz >> test/tgt

=back

Any of the above pseudo-variables may be followed immediately by one of
the following suffixes to select a portion of the expanded path name:

  :a    the absolute path to the file name
  :b    the directory plus the file name stripped of any suffix
  :d    the directory
  :f    the file name
  :s    the file name suffix
  :F    the file name stripped of any suffix
  :S    the absolute path path to a Linked source file

Continuing with the above example, C<%E<lt>:f> would expand to C<foo bar baz>,
and C<%E<gt>:d> would expand to C<test>.

There are additional C<%> elements which affect the command line(s):

=over 10

=item %[ %]

It is possible to programmatically rewrite part of the command by
enclosing part of it between C<%[> and C<%]>.  This will call the
construction variable named as the first word enclosed in the brackets
as a Perl code reference; the results of this call will be used to
replace the contents of the brackets in the command line.  For example,
given an existing input file named F<tgt.in>:

  @keywords = qw(foo bar baz);
  $env = new cons(X_COMMA => sub { join(",", @_) });
  Command $env 'tgt', 'tgt.in', qq(
	echo '# Keywords: %[X_COMMA @keywords %]' > %>
	cat %< >> %>
  );

This will execute:

  echo '# Keywords: foo,bar,baz' > tgt
  cat tgt.in >> tgt

=item %( %)

Cons includes the text of the command line in the MD5 signature for a
build, so that targets get rebuilt if you change the command line (to
add or remove an option, for example).  Command-line text in between
C<%(> and C<%)>, however, will be ignored for MD5 signature calculation.

Internally, Cons uses C<%(> and C<%)> around include and library
directory options (C<-I> and C<-L> on UNIX systems, C</I> and
C</LIBPATH> on Windows NT) to avoid rebuilds just because the directory
list changes.  Rebuilds occur only if the changed directory list causes
any included I<files> to change, and a changed include file is detected
by the MD5 signature calculation on the actual file contents.

=back

=head2 Expanding construction variables in file names

Cons expands construction variables in the source and target file names
passed to the various construction methods according to the expansion
rules described above:

  $env = new cons(
	DESTDIR	=>	'programs',
	SRCDIR	=>	'src',
  );
  Program $env '%DESTDIR/hello', '%SRCDIR/hello.c';

This allows for flexible configuration, through the construction
environment, of directory names, suffixes, etc.


=head1 Build actions

Cons supports several types of B<build actions> that can be performed
to construct one or more target files.  Usually, a build action is
a construction command--that is, a command-line string that invokes
an external command.  Cons can also execute Perl code embedded in a
command-line string, and even supports an experimental ability to build
a target file by executing a Perl code reference directly.

A build action is usually specified as the value of a construction
variable:

  $env = new cons(
	CCCOM         => '%CC %CFLAGS %_IFLAGS -c %< -o %>',
	LINKCOM       => '[perl] &link_executable("%>", "%<")',
	ARCOM         => sub { my($env, $target, @sources) = @_;
				 # code to create an archive
				}
  );

A build action may be associated directly with one or more target files
via the C<Command> method; see below.

=head2 Construction commands

A construction command goes through expansion of construction variables
and C<%-> pseudo-variables, as described above, to create the actual
command line that Cons will execute to generate the target file or
files.

After substitution occurs, strings of white space are converted into
single blanks, and leading and trailing white space is eliminated. It
is therefore currently not possible to introduce variable length white
space in strings passed into a command.

If a multi-line command string is provided, the commands are executed
sequentially. If any of the commands fails, then none of the rest are
executed, and the target is not marked as updated, i.e. a new signature is
not stored for the target.

Normally, if all the commands succeed, and return a zero status (or whatever
platform-specific indication of success is required), then a new signature
is stored for the target. If a command erroneously reports success even
after a failure, then Cons will assume that the target file created by that
command is accurate and up-to-date.

The first word of each command string, after expansion, is assumed to be an
executable command looked up on the C<PATH> environment variable (which is,
in turn, specified by the C<ENV> construction variable). If this command is
found on the path, then the target will depend upon it: the command will
therefore be automatically built, as necessary. It's possible to write
multi-part commands to some shells, separated by semi-colons. Only the first
command word will be depended upon, however, so if you write your command
strings this way, you must either explicitly set up a dependency (with the
C<Depends> method), or be sure that the command you are using is a system
command which is expected to be available. If it isn't available, you will,
of course, get an error.

Cons normally prints a command before executing it.  This behavior is
suppressed if the first character of the command is C<@>.  Note that
you may need to separate the C<@> from the command name or escape it to
prevent C<@cmd> from looking like an array to Perl quote operators that
perform interpolation:

  # The first command line is incorrect,
  # because "@cp" looks like an array
  # to the Perl qq// function.
  # Use the second form instead.
  Command $env 'foo', 'foo.in', qq(
	@cp %< tempfile
	@ cp tempfile %>
  );

If there are shell meta characters anywhere in the expanded command line,
such as C<E<lt>>, C<E<gt>>, quotes, or semi-colon, then the command
will actually be executed by invoking a shell. This means that a command
such as:

  cd foo

alone will typically fail, since there is no command C<cd> on the path. But
the command string:

  cd $<:d; tar cf $>:f $<:f

when expanded will still contain the shell meta character semi-colon, and a
shell will be invoked to interpret the command. Since C<cd> is interpreted
by this sub-shell, the command will execute as expected.

=head2 Perl expressions

If any command (even one within a multi-line command) begins with
C<[perl]>, the remainder of that command line will be evaluated by the
running Perl instead of being forked by the shell.  If an error occurs
in parsing the Perl code, or if the Perl expression returns 0 or undef,
the command will be considered to have failed.  For example, here is a
simple command which creates a file C<foo> directly from Perl:

  $env = new cons();
  Command $env 'foo',
    qq([perl] open(FOO,'>foo');print FOO "hi\\n"; close(FOO); 1);

Note that when the command is executed, you are in the same package as
when the F<Construct> or F<Conscript> file was read, so you can call
Perl functions you've defined in the same F<Construct> or F<Conscript>
file in which the C<Command> appears:

  $env = new cons();
  sub create_file {
	my $file = shift;
	open(FILE, ">$file");
	print FILE "hi\n";
	close(FILE);
	return 1;
  }
  Command $env 'foo', "[perl] &create_file('%>')";

The Perl string will be used to generate the signature for the derived
file, so if you change the string, the file will be rebuilt.  The contents
of any subroutines you call, however, are not part of the signature,
so if you modify a called subroutine such as C<create_file> above,
the target will I<not> be rebuilt.  Caveat user.

=head2 Perl code references [EXPERIMENTAL]

Cons supports the ability to create a derived file by directly executing
a Perl code reference.  This feature is considered EXPERIMENTAL and
subject to change in the future.

A code reference may either be a named subroutine referenced by the
usual C<\&> syntax:

  sub build_output {
	my($env, $target, @sources) = @_;
	print "build_output building $target\n";
	open(OUT, ">$target");
	foreach $src (@sources) {
	    if (! open(IN, "<$src")) {
		print STDERR "cannot open '$src': $!\n";
		return undef;
	    }
	    print OUT, <IN>;
	}
	close(OUT);
	return 1;
  }
  Command $env 'output', \&build_output;

or the code reference may be an anonymous subroutine:

  Command $env 'output', sub {
	my($env, $target, @sources) = @_;
	print "building $target\n";
	open(FILE, ">$target");
	print FILE "hello\n";
	close(FILE);
	return 1;
  };

To build the target file, the referenced subroutine is passed, in order:
the construction environment used to generate the target; the path
name of the target itself; and the path names of all the source files
necessary to build the target file.

The code reference is expected to generate the target file, of course,
but may manipulate the source and target files in any way it chooses.
The code reference must return a false value (C<undef> or C<0>) if
the build of the file failed.  Any true value indicates a successful
build of the target.

Building target files using code references is considered EXPERIMENTAL
due to the following current limitations:

=over 4

Cons does I<not> print anything to indicate the code reference is being
called to build the file.  The only way to give the user any indication
is to have the code reference explicitly print some sort of "building"
message, as in the above examples.

Cons does not generate any signatures for code references, so if the
code in the reference changes, the target will I<not> be rebuilt.

Cons has no public method to allow a code reference to extract
construction variables.  This would be good to allow generalization of
code references based on the current construction environment, but would
also complicate the problem of generating meaningful signatures for code
references.

=back

Support for building targets via code references has been released in
this version to encourage experimentation and the seeking of possible
solutions to the above limitations.


=head1 Default construction methods

The list of default construction methods includes the following:


=head2 The C<new> constructor

The C<new> method is a Perl object constructor. That is, it is not invoked
via a reference to an existing construction environment B<reference>, but,
rather statically, using the name of the Perl B<package> where the
constructor is defined. The method is invoked like this:

  $env = new cons(<overrides>);

The environment you get back is blessed into the package C<cons>, which
means that it will have associated with it the default methods described
below. Individual construction variables can be overridden by providing
name/value pairs in an override list. Note that to override any command
environment variable (i.e. anything under C<ENV>), you will have to override
all of them. You can get around this difficulty by using the C<copy> method
on an existing construction environment.


=head2 The C<clone> method

The C<clone> method creates a clone of an existing construction environment,
and can be called as in the following example:

  $env2 = $env1->clone(<overrides>);

You can provide overrides in the usual manner to create a different
environment from the original. If you just want a new name for the same
environment (which may be helpful when exporting environments to existing
components), you can just use simple assignment.


=head2 The C<copy> method

The C<copy> method extracts the externally defined construction variables
from an environment and returns them as a list of name/value
pairs. Overrides can also be provided, in which case, the overridden values
will be returned, as appropriate. The returned list can be assigned to a
hash, as shown in the prototype, below, but it can also be manipulated in
other ways:

  %env = $env1->copy(<overrides>);

The value of C<ENV>, which is itself a hash, is also copied to a new hash,
so this may be changed without fear of affecting the original
environment. So, for example, if you really want to override just the
C<PATH> variable in the default environment, you could do the following:

  %cons = new cons()->copy();
  $cons{ENV}{PATH} = "<your path here>";
  $cons = new cons(%cons);

This will leave anything else that might be in the default execution
environment undisturbed.


=head2 The C<Install> method

The C<Install> method arranges for the specified files to be installed in
the specified directory. The installation is optimized: the file is not
copied if it can be linked. If this is not the desired behavior, you will
need to use a different method to install the file. It is called as follows:

  Install $env <directory>, <names>;

Note that, while the files to be installed may be arbitrarily named,
only the last component of each name is used for the installed target
name. So, for example, if you arrange to install F<foo/bar> in F<baz>,
this will create a F<bar> file in the F<baz> directory (not F<foo/bar>).


=head2 The C<InstallAs> method

The C<InstallAs> method arranges for the specified source file(s) to be
installed as the specified target file(s). Multiple files should be
specified as a file list. The installation is optimized: the file is not
copied if it can be linked. If this is not the desired behavior, you will
need to use a different method to install the file. It is called as follows:

C<InstallAs> works in two ways:

Single file install:

  InstallAs $env TgtFile, SrcFile;

Multiple file install:

  InstallAs $env ['tgt1', 'tgt2'], ['src1', 'src2'];

Or, even as:

  @srcs = qw(src1 src2 src3);
  @tgts = qw(tgt1 tgt2 tgt3);
  InstallAs $env [@tgts], [@srcs];

Both the target and the sources lists should be of the same length.

=head2 The C<Precious> method

The C<Precious> method asks cons not to delete the specified file or
list of files before building them again.  It is invoked as:

  Precious <files>;

This is especially useful for allowing incremental updates to libraries
or debug information files which are updated rather than rebuilt anew each
time.  Cons will still delete the files when the C<-r> flag is specified.

=head2 The C<AfterBuild> method

The C<AfterBuild> method evaluates the specified perl string after
building the given file or files (or finding that they are up to date).
The eval will happen once per specified file.  C<AfterBuild> is called
as follows:

  AfterBuild $env 'foo.o', qq(print "foo.o is up to date!\n");

The perl string is evaluated in the C<script> package, and has access
to all variables and subroutines defined in the F<Conscript> file in
which the C<AfterBuild> method is called.

=head2 The C<Command> method

The C<Command> method is a catchall method which can be used to arrange for
any build action to be executed to update the target. For this command, a
target file and list of inputs is provided. In addition, a build action
is specified as the last argument.  The build action is typically a
command line or lines, but may also contain Perl code to be executed;
see the section above on build actions for details.

The C<Command> method is called as follows:

  Command $env <target>, <inputs>, <build action>;

The target is made dependent upon the list of input files specified, and the
inputs must be built successfully or Cons will not attempt to build the
target.

To specify a command with multiple targets, you can specify a reference to a
list of targets. In Perl, a list reference can be created by enclosing a
list in square brackets. Hence the following command:

  Command $env ['foo.h', 'foo.c'], 'foo.template', q(
	gen %1
  );

could be used in a case where the command C<gen> creates two files, both
F<foo.h> and F<foo.c>.


=head2 The C<Objects> method

The C<Objects> method arranges to create the object files that correspond to
the specified source files. It is invoked as shown below:

  @files = Objects $env <source or object files>;

Under Unix, source files ending in F<.s> and F<.c> are currently
supported, and will be compiled into a name of the same file ending
in F<.o>. By default, all files are created by invoking the external
command which results from expanding the C<CCCOM> construction variable,
with C<%E<lt>> and C<%E<gt>> set to the source and object files,
respectively. (See the section above on construction variable expansion
for details).  The variable C<CPPPATH> is also used when scanning source
files for dependencies. This is a colon separated list of pathnames, and
is also used to create the construction variable C<_IFLAGS,> which will
contain the appropriate list of -C<I> options for the compilation. Any
relative pathnames in C<CPPPATH> is interpreted relative to the
directory in which the associated construction environment was created
(absolute and top-relative names may also be used). This variable is
used by C<CCCOM>. The behavior of this command can be modified by
changing any of the variables which are interpolated into C<CCCOM>, such
as C<CC>, C<CFLAGS>, and, indirectly, C<CPPPATH>. It's also possible
to replace the value of C<CCCOM>, itself. As a convenience, this file
returns the list of object filenames.


=head2 The C<Program> method

The C<Program> method arranges to link the specified program with the
specified object files. It is invoked in the following manner:

  Program $env <program name>, <source or object files>;

The program name will have the value of the C<SUFEXE> construction
variable appended (by default, C<.exe> on Win32 systems, nothing on Unix
systems) if the suffix is not already present.

Source files may be specified in place of objects files--the C<Objects>
method will be invoked to arrange the conversion of all the files into
object files, and hence all the observations about the C<Objects> method,
above, apply to this method also.

The actual linking of the program will be handled by an external command
which results from expanding the C<LINKCOM> construction variable, with
C<%E<lt>> set to the object files to be linked (in the order presented),
and C<%E<gt>> set to the target. (See the section above on construction
variable expansion for details.)  The user may set additional variables
in the construction environment, including C<LINK>, to define which
program to use for linking, C<LIBPATH>, a colon-separated list of
library search paths, for use with library specifications of the form
I<-llib>, and C<LIBS>, specifying the list of libraries to link against
(in either I<-llib> form or just as pathnames. Relative pathnames in
both C<LIBPATH> and C<LIBS> are interpreted relative to the directory
in which the associated construction environment is created (absolute
and top-relative names may also be used). Cons automatically sets up
dependencies on any libraries mentioned in C<LIBS>: those libraries will
be built before the command is linked.


=head2 The C<Library> method

The C<Library> method arranges to create the specified library from the
specified object files. It is invoked as follows:

  Library $env <library name>, <source or object files>;

The library name will have the value of the C<SUFLIB> construction
variable appended (by default, C<.lib> on Win32 systems, C<.a> on Unix
systems) if the suffix is not already present.

Source files may be specified in place of objects files--the C<Objects>
method will be invoked to arrange the conversion of all the files into
object files, and hence all the observations about the C<Objects> method,
above, apply to this method also.

The actual creation of the library will be handled by an external
command which results from expanding the C<ARCOM> construction variable,
with C<%E<lt>> set to the library members (in the order presented),
and C<%E<gt>> to the library to be created.  (See the section above
on construction variable expansion for details.)  The user may set
variables in the construction environment which will affect the
operation of the command. These include C<AR>, the archive program
to use, C<ARFLAGS>, which can be used to modify the flags given to
the program specified by C<AR>, and C<RANLIB>, the name of a archive
index generation program, if needed (if the particular need does not
require the latter functionality, then C<ARCOM> must be redefined to not
reference C<RANLIB>).

The C<Library> method allows the same library to be specified in multiple
method invocations. All of the contributing objects from all the invocations
(which may be from different directories) are combined and generated by a
single archive command. Note, however, that if you prune a build so that
only part of a library is specified, then only that part of the library will
be generated (the rest will disappear!).


=head2 The C<Module> method

The C<Module> method is a combination of the C<Program> and C<Command>
methods. Rather than generating an executable program directly, this command
allows you to specify your own command to actually generate a module. The
method is invoked as follows:

  Module $env <module name>, <source or object files>, <construction command>;

This command is useful in instances where you wish to create, for example,
dynamically loaded modules, or statically linked code libraries.


=head2 The C<Depends> method

The C<Depends> method allows you to specify additional dependencies for a
target.  It is invoked as follows:

  Depends $env <target>, <dependencies>;

This may be occasionally useful, especially in cases where no scanner exists
(or is writable) for particular types of files. Normally, dependencies are
calculated automatically from a combination of the explicit dependencies set
up by the method invocation or by scanning source files.

A set of identical dependencies for multiple targets may be specified
using a reference to a list of targets. In Perl, a list reference can
be created by enclosing a list in square brackets. Hence the following
command:

  Depends $env ['foo', 'bar'], 'input_file_1', 'input_file_2';

specifies that both the F<foo> and F<bar> files depend on the listed
input files.


=head2 The C<RuleSet> method

The C<RuleSet> method returns the construction variables for building
various components with one of the rule sets supported by Cons.  The
currently supported rule sets are:

=over 4

=item msvc

Rules for the Microsoft Visual C++ compiler suite.

=item unix

Generic rules for most UNIX-like compiler suites.

=back

On systems with more than one available compiler suite, this allows you
to easily create side-by-side environments for building software with
multiple tools:

    $msvcenv = new cons(RuleSet("msvc"));
    $cygnusenv = new cons(RuleSet("unix"));

In the future, this could also be extended to other platforms that
have different default rule sets.


=head2 The C<DefaultRules> method

The C<DefaultRules> method sets the default construction variables that
will be returned by the C<new> method to the specified arguments:

  DefaultRules(CC     => 'gcc',
	       CFLAGS => '',
	       CCCOM  => '%CC %CFLAGS %_IFLAGS -c %< -o %>');
  $env = new cons();
  # $env now contains *only* the CC, CFLAGS,
  # and CCCOM construction variables

Combined with the C<RuleSet> method, this also provides an easy way
to set explicitly the default build environment to use some supported
toolset other than the Cons defaults:

    # use a UNIX-like tool suite (like cygwin) on Win32
    DefaultRules(RuleSet('unix'));
    $env = new cons();

Note that the C<DefaultRules> method completely replaces the default
construction environment with the specified arguments, it does not
simply override the existing defaults.  To override one or more
variables in a supported C<RuleSet>, append the variables and values:

  DefaultRules(RuleSet('unix'), CFLAGS => '-O3');
  $env1 = new cons();
  $env2 = new cons();
  # both $env1 and $env2 have 'unix' defaults
  # with CFLAGS set to '-O3'


=head2 The C<Ignore> method

The C<Ignore> method allows you to ignore explicitly dependencies that
Cons infers on its own.  It is invoked as follows:

  Ignore <patterns>;

This can be used to avoid recompilations due to changes in system header
files or utilities that are known to not affect the generated targets.

If, for example, a program is built in an NFS-mounted directory on
multiple systems that have different copies of F<stdio.h>, the differences
will affect the signatures of all derived targets built from source files
that C<#include E<lt>stdio.hE<gt>>.  This will cause all those targets to
be rebuilt when changing systems.  If this is not desirable behavior, then
the following line will remove the dependencies on the F<stdio.h> file:

  Ignore '^/usr/include/stdio\.h$';

Note that the arguments to the C<Ignore> method are regular expressions,
so special characters must be escaped and you may wish to anchor the
beginning or end of the expression with C<^> or C<$> characters.


=head2 The C<Salt> method

The C<Salt> method adds a constant value to the signature calculation
for every derived file.  It is invoked as follows:

  Salt $string;

Changing the Salt value will force a complete rebuild of every derived
file.  This can be used to force rebuilds in certain desired
circumstances.  For example,

  Salt `uname -s`;

Would force a complete rebuild of every derived file whenever the
operating system on which the build is performed (as reported by C<uname
-s>) changes.


=head2 The C<UseCache> method

The C<UseCache> method instructs Cons to maintain a cache of derived
files, to be shared among separate build trees of the same project.

  UseCache("cache/<buildname>") || warn("cache directory not found");


=head2 The C<SourcePath> method

The C<SourcePath> mathod returns the real source path name of a file,
as opposed to the path name within a build directory.  It is invoked
as follows:

  $path = SourcePath <buildpath>;


=head2 The C<ConsPath> method

The C<ConsPath> method returns true if the supplied path is a derivable
file, and returns undef (false) otherwise.
It is invoked as follows:

  $result = ConsPath <path>;


=head2 The C<SplitPath> method

The C<SplitPath> method looks up multiple path names in a string separated
by the default path separator for the operating system (':' on UNIX
systems, ';' on Windows NT), and returns the fully-qualified names.
It is invoked as follows:

  @paths = SplitPath <pathlist>;

The C<SplitPath> method will convert  names prefixed '#' to the
appropriate top-level build name (without the '#') and will convert
relative names to top-level names.


=head2 The C<DirPath> method

The C<DirPath> method returns the build path name(s) of a directory or
list of directories.  It is invoked as follows:

  $cwd = DirPath <paths>;

The most common use for the C<DirPath> method is:

  $cwd = DirPath '.';

to fetch the path to the current directory of a subsidiary F<Conscript>
file.


=head2 The C<FilePath> method

The C<FilePath> method returns the build path name(s) of a file or
list of files.  It is invoked as follows:

  $file = FilePath <path>;


=head2 The C<Help> method

The C<Help> method specifies help text that will be displayed when the
user invokes C<cons -h>.  This can be used to provide documentation
of specific targets, values, build options, etc. for the build tree.
It is invoked as follows:

  Help <helptext>;

The C<Help> method may only be called once, and should typically be
specified in the top-level F<Construct> file.


=head1 Extending Cons


=head2 Overriding construction variables

There are several ways of extending Cons, which vary in degree of
difficulty. The simplest method is to define your own construction
environment, based on the default environment, but modified to reflect your
particular needs. This will often suffice for C-based applications. You can
use the C<new> constructor, and the C<clone> and C<copy> methods to create
hybrid environments. These changes can be entirely transparent to the
underlying F<Conscript> files.


=head2 Adding new methods

For slightly more demanding changes, you may wish to add new methods to the
C<cons> package. Here's an example of a very simple extension,
C<InstallScript>, which installs a tcl script in a requested location, but
edits the script first to reflect a platform-dependent path that needs to be
installed in the script:

  # cons::InstallScript - Create a platform dependent version of a shell
  # script by replacing string ``#!your-path-here'' with platform specific
  # path $BIN_DIR.

  sub cons::InstallScript {
	my ($env, $dst, $src) = @_;
	Command $env $dst, $src, qq(
		sed s+your-path-here+$BIN_DIR+ %< > %>
		chmod oug+x %>
	);
  }

Notice that this method is defined directly in the C<cons> package (by
prefixing the name with C<cons::>). A change made in this manner will be
globally visible to all environments, and could be called as in the
following example:

  InstallScript $env "$BIN/foo", "foo.tcl";

For a small improvement in generality, the C<BINDIR> variable could be
passed in as an argument or taken from the construction environment--as
C<%BINDIR>.


=head2 Overriding methods

Instead of adding the method to the C<cons> name space, you could define a
new package which inherits existing methods from the C<cons> package and
overrides or adds others. This can be done using Perl's inheritance
mechanisms.

The following example defines a new package C<cons::switch> which
overrides the standard C<Library> method. The overridden method builds
linked library modules, rather than library archives. A new
constructor is provided. Environments created with this constructor
will have the new library method; others won't.

  package cons::switch;
  BEGIN {@ISA = 'cons'}

  sub new {
	shift;
	bless new cons(@_);
  }

  sub Library {
	my($env) = shift;
	my($lib) = shift;
	my(@objs) = Objects $env @_;
	Command $env $lib, @objs, q(
		%LD -r %LDFLAGS %< -o %>
	);
  }

This functionality could be invoked as in the following example:

  $env = new cons::switch(@overrides);
  ...
  Library $env 'lib.o', 'foo.c', 'bar.c';


=head1 Invoking Cons

The C<cons> command is usually invoked from the root of the build tree. A
F<Construct> file must exist in that directory. If the C<-f> argument is
used, then an alternate F<Construct> file may be used (and, possibly, an
alternate root, since C<cons> will cd to F<Construct> file's containing
directory).

If C<cons> is invoked from a child of the root of the build tree with
the C<-t> argument, it will walk up the directory hierarchy looking for a
F<Construct> file.  (An alternate name may still be specified with C<-f>.)
The targets supplied on the command line will be modified to be relative
to the discovered F<Construct> file.  For example, from a directory
containing a top-level F<Construct> file, the following invocation:

  % cd libfoo/subdir
  % cons -t target

is exactly equivalent to:

  % cons libfoo/subdir/target

If there are any C<Default> targets specified in the directory hierarchy's
F<Construct> or F<Conscript> files, only the default targets at or below
the directory from which C<cons -t> was invoked will be built.

The command is invoked as follows:

  cons <arguments> -- <construct-args>

where I<arguments> can be any of the following, in any order:

=over 10

=item I<target>

Build the specified target. If I<target> is a directory, then recursively
build everything within that directory.

=item I<+pattern>

Limit the F<Conscript> files considered to just those that match I<pattern>,
which is a Perl regular expression. Multiple C<+> arguments are accepted.

=item I<name>=<val>

Sets I<name> to value I<val> in the C<ARG> hash passed to the top-level
F<Construct> file.

=item C<-cc>

Show command that would have been executed, when retrieving from cache. No
indication that the file has been retrieved is given; this is useful for
generating build logs that can be compared with real build logs.

=item C<-cd>

Disable all caching. Do not retrieve from cache nor flush to cache.

=item C<-cr>

Build dependencies in random order. This is useful when building multiple
similar trees with caching enabled.

=item C<-cs>

Synchronize existing build targets that are found to be up-to-date with
cache. This is useful if caching has been disabled with -cc or just recently
enabled with UseCache.

=item C<-d>

Enable dependency debugging.

=item C<-f> <file>

Use the specified file instead of F<Construct> (but first change to
containing directory of I<file>).

=item C<-h>

Show a help message local to the current build if one such is defined, and
exit.

=item C<-k>

Keep going as far as possible after errors.

=item C<-o> <file>

Read override file I<file>.

=item C<-p>

Show construction products in specified trees. No build is attempted.

=item C<-pa>

Show construction products and associated actions. No build is attempted.

=item C<-pw>

Show products and where they are defined. No build is attempted.

=item C<-q>

Make the build quiet.  Multiple C<-q> options may be specified.

A single C<-q> options suppress messages about Installing and Removing
targets.

Two C<-q> options suppress build command lines and target up-to-date
messages.

=item C<-r>

Remove construction products associated with <targets>. No build is
attempted.

=item C<-R> <repos>

Search for files in I<repos>.  Multiple B<-R> I<repos> directories are
searched in the order specified.

=item C<-S> <pkg>

Use the sig::<pkg> package to calculate.  Supported <pkg> values
include "md5" for MD5 signature calculation and "md5::debug" for debug
information about MD5 signature calculation.

If the specified package ends in <::debug>, signature debug information
will be printed to the file name specified in the C<CONS_SIG_DEBUG>
environment variable, or to standard output if the environment variable
is not set.

=item C<-t>

Traverse up the directory hierarchy looking for a F<Construct> file,
if none exists in the current directory.  Targets will be modified to
be relative to the F<Construct> file.

Internally, C<cons> will change its working directory to the directory
which contains the top-level F<Construct> file and report:

  cons: Entering directory `top-level-directory'

This message indicates to an invoking editor (such as emacs) or build
environment that Cons will now report all file names relative to the
top-level directory.  This message can not be suppressed with the C<-q>
option.

=item C<-v>

Show C<cons> version and continue processing.

=item C<-V>

Show C<cons> version and exit.

=item C<-wf> <file>

Write all filenames considered into I<file>.

=item C<-x>

Show a help message similar to this one, and exit.

=back

And I<construct-args> can be any arguments that you wish to process in the
F<Construct> file. Note that there should be a B<--> separating the arguments
to cons and the arguments that you wish to process in the F<Construct> file.

Processing of I<construct-args> can be done by any standard package like
B<Getopt> or its variants, or any user defined package. B<cons> will pass in
the I<construct-args> as B<@ARGV> and will not attempt to interpret anything
after the B<-->.

  % cons -R /usr/local/repository -d os=solaris +driver -- -c test -f DEBUG

would pass the following to cons

  -R /usr/local/repository -d os=solaris +driver

and the following, to the top level F<Construct> file as B<@ARGV>

  -c test -f DEBUG

Note that C<cons -r .> is equivalent to a full recursive C<make clean>,
but requires no support in the F<Construct> file or any F<Conscript>
files. This is most useful if you are compiling files into source
directories (if you separate the F<build> and F<export> directories,
then you can just remove the directories).

The options C<-p>, C<-pa>, and C<-pw> are extremely useful for use as an aid
in reading scripts or debugging them. If you want to know what script
installs F<export/include/foo.h>, for example, just type:

  % cons -pw export/include/foo.h


=head1 Using and writing dependency scanners

QuickScan allows simple target-independent scanners to be set up for
source files. Only one QuickScan scanner may be associated with any given
source file and environment, although the same scanner may (and should)
be used for multiple files of a given type.

A QuickScan scanner is only ever invoked once for a given source file,
and it is only invoked if the file is used by some target in the tree
(i.e., there is a dependency on the source file).

QuickScan is invoked as follows:

  QuickScan CONSENV CODEREF, FILENAME [, PATH]

The subroutine referenced by CODEREF is expected to return a list of
filenames included directly by FILE. These filenames will, in turn, be
scanned. The optional PATH argument supplies a lookup path for finding
FILENAME and/or files returned by the user-supplied subroutine.  The PATH
may be a reference to an array of lookup-directory names, or a string of
names separated by the system's separator character (':' on UNIX systems,
';' on Windows NT).

The subroutine is called once for each line in the file, with $_ set to the
current line. If the subroutine needs to look at additional lines, or, for
that matter, the entire file, then it may read them itself, from the
filehandle SCAN. It may also terminate the loop, if it knows that no further
include information is available, by closing the filehandle.

Whether or not a lookup path is provided, QuickScan first tries to lookup
the file relative to the current directory (for the top-level file
supplied directly to QuickScan), or from the directory containing the
file which referenced the file. This is not very general, but seems good
enough--especially if you have the luxury of writing your own utilities
and can control the use of the search path in a standard way.

Here's a real example, taken from a F<Construct> file here:

  sub cons::SMFgen {
      my($env, @tables) = @_;
      foreach $t (@tables) {
	  $env->QuickScan(sub { /\b\S*?\.smf\b/g }, "$t.smf",
			  $env->{SMF_INCLUDE_PATH});
	  $env->Command(["$t.smdb.cc","$t.smdb.h","$t.snmp.cc",
			 "$t.ami.cc", "$t.http.cc"], "$t.smf",
			q(smfgen %( %SMF_INCLUDE_OPT %) %<));
      }
  }

The subroutine above finds all names of the form <name>.smf in the
file. It will return the names even if they're found within comments,
but that's OK (the mechanism is forgiving of extra files; they're just
ignored on the assumption that the missing file will be noticed when
the program, in this example, smfgen, is actually invoked).

[NOTE that the form C<$env-E<gt>QuickScan ...>  and C<$env-E<gt>Command
...> should not be necessary, but, for some reason, is required
for this particular invocation. This appears to be a bug in Perl or
a misunderstanding on my part; this invocation style does not always
appear to be necessary.]

Here is another way to build the same scanner. This one uses an
explicit code reference, and also (unnecessarily, in this case) reads
the whole file itself:

  sub myscan {
      my(@includes);
      do {
	  push(@includes, /\b\S*?\.smf\b/g);
      } while <SCAN>;
      @includes
  }

Note that the order of the loop is reversed, with the loop test at the
end. This is because the first line is already read for you. This scanner
can be attached to a source file by:

  QuickScan $env \&myscan, "$_.smf";

This final example, which scans a different type of input file, takes
over the file scanning rather than being called for each input line:

  $env->QuickScan(
      sub { my(@includes) = ();
	  do {
	     push(@includes, $3)
		 if /^(#include|import)\s+(\")(.+)(\")/ && $3
	  } while <SCAN>;
	  @includes
      },
      "$idlFileName",
      "$env->{CPPPATH};$BUILD/ActiveContext/ACSCLientInterfaces"
  );

=head1 SUPPORT AND SUGGESTIONS

Cons is maintained by the user community.  To subscribe, send mail to
B<cons-discuss-request@gnu.org> with body B<subscribe>.

Please report any suggestions through the B<cons-discuss@gnu.org> mailing
list.

=head1 BUGS

Sure to be some. Please report any bugs through the B<bug-cons@gnu.org>
mailing list.

=head1 INFORMATION ABOUT CONS

Information about CONS can be obtained from the official cons web site
B<http://www.dsmit.com/cons/> or its mirrors listed there.

The cons maintainers can be contacted by email at
B<cons-maintainers@gnu.org>

=head1 AUTHORS

Originally by Bob Sidebotham. Then significantly enriched by the members
of the Cons community B<cons-discuss@gnu.org>.

The Cons community would like to thank Ulrich Pfeifer for the original pod
documentation derived from the F<cons.html> file. Cons documentation is now
a part of the program itself.

=cut