package Archive::Tar;

use strict;
use Carp qw(carp);
use Cwd;
use Fcntl qw(O_RDONLY O_RDWR O_WRONLY O_CREAT O_TRUNC F_DUPFD F_GETFL);
use File::Basename;
use Symbol;
require Time::Local if $^O eq "MacOS";

use vars qw($VERSION @ISA @EXPORT_OK %EXPORT_TAGS);
$VERSION = do { my @a=q$Name: version_0_22 $ =~ /\d+/g; sprintf "%d." . ("%02d" x $#a ),@a };

require Exporter;
@ISA = qw(Exporter);

@EXPORT_OK = qw(FILE HARDLINK SYMLINK 
		CHARDEV BLOCKDEV DIR
		FIFO SOCKET INVALID);
%EXPORT_TAGS = (filetypes => \@EXPORT_OK);

# Check if symbolic links are available
my $symlinks = eval { readlink $0 or 1; };
carp "Symbolic links not available"
    unless $symlinks || !$^W;

# Check if Compress::Zlib is available
my $compression = eval { 
    local $SIG{__DIE__};
    require Compress::Zlib; 
    sub Compress::Zlib::gzFile::gzseek {
	my $tmp;

	$_[0]->gzread ($tmp, 4096), $_[1] -= 4096
	    while ($_[1] > 4096);

	$_[0]->gzread ($tmp, $_[1])
	  if $_[1];
    }
    1;
};
carp "Compression not available"
    unless $compression || !$^W;

# Check for get* (they don't exist on WinNT)
my $fake_getpwuid;
$fake_getpwuid = "unknown"
    unless eval { $_ = getpwuid (0); }; # Pointless assigment to make -w shut up

my $fake_getgrgid;
$fake_getgrgid = "unknown"
    unless eval { $_ = getgrgid (0); }; # Pointless assigment to make -w shut up

# Automagically detect gziped files if they start with this
my $gzip_magic_number = "^(?:\037\213|\037\235)";

my $tar_unpack_header 
    = 'A100 A8 A8 A8 A12 A12 A8 A1 A100 A6 A2 A32 A32 A8 A8 A155 x12';
my $tar_pack_header
    = 'a100 a8 a8 a8 a12 a12 A8 a1 a100 a6 a2 a32 a32 a8 a8 a155 x12',
my $tar_header_length = 512;

my $time_offset = ($^O eq "MacOS") ? Time::Local::timelocal(0,0,0,1,0,70) : 0;

## Subroutines to return type constants 
sub FILE() { return 0; }
sub HARDLINK() { return 1; }
sub SYMLINK() { return 2; }
sub CHARDEV() { return 3; }
sub BLOCKDEV() { return 4; }
sub DIR() { return 5; }
sub FIFO() { return 6; }
sub SOCKET() { return 8; }
sub UNKNOWN() { return 9; }

###
### Non-method functions
###

my $error;
sub _drat {
    $error = $! . '';
    return;
}

sub error {
    $error;
}

sub set_error {
    shift;
    $error = "@_";
}

## filetype -- Determine the type value for a given file
sub filetype {
    my $file = shift;

    return SYMLINK
	if (-l $file);		# Symlink

    return FILE
	if (-f _);		# Plain file

    return DIR
	if (-d _);		# Directory

    return FIFO
	if (-p _);		# Named pipe

    return SOCKET
	if (-S _);		# Socket

    return BLOCKDEV
	if (-b _);		# Block special

    return CHARDEV
	if (-c _);		# Character special

    return UNKNOWN;		# Something else (like what?)
}

sub _make_special_file_UNIX {
    # $file is the last component of $entry->{name}
    my ($entry, $file) = @_;

    if ($entry->{type} == SYMLINK) {
	symlink $entry->{linkname}, $file or
	    $^W && carp ("Making symbolic link from ", $entry->{linkname}, 
			 " to ", $entry->{name}, ", failed.\n");
    }
    elsif ($entry->{type} == HARDLINK) {
	link $entry->{linkname}, $file or
	    $^W && carp ("Hard linking ", $entry->{linkname}, 
			 " to ", $entry->{name}, ", failed.\n");
    }
    elsif ($entry->{type} == FIFO) {
	system("mknod","$file","p") or
	    $^W && carp "Making fifo ", $entry->{name}, ", failed.\n";
    }
    elsif ($entry->{type} == BLOCKDEV) {
	system("mknod","$file","b",$entry->{devmajor},$entry->{devminor}) or
	    $^W && carp ("Making block device ", $entry->{name},
			 " (maj=", $entry->{devmajor}, 
			 ", min=", $entry->{devminor}, "), failed.\n");
    }
    elsif ($entry->{type} == CHARDEV) {
	system("mknod", "$file", "c", $entry->{devmajor}, $entry->{devminor}) or
	    $^W && carp ("Making block device ", $entry->{name}, 
			 " (maj=", $entry->{devmajor}, 
			 " ,min=", $entry->{devminor}, "), failed.\n");
    }
}

sub _make_special_file_Win32 {
    # $file is the last component of $entry->{name}
    my ($entry, $file) = @_;

    if ($entry->{type} == SYMLINK) {
	$^W && carp ("Making symbolic link from ", $entry->{linkname}, 
		     " to ", $entry->{name}, ", failed.\n");
    }
    elsif ($entry->{type} == HARDLINK) {
	link $entry->{linkname}, $file or
	    $^W && carp ("Making hard link from ", $entry->{linkname}, 
			 " to ", $entry->{name}, ", failed.\n");
    }
    elsif ($entry->{type} == FIFO) {
	$^W && carp "Making fifo ", $entry->{name}, ", failed.\n";
    }
    elsif ($entry->{type} == BLOCKDEV) {
	$^W && carp ("Making block device ", $entry->{name},
		     " (maj=", $entry->{devmajor}, 
		     ", min=", $entry->{devminor}, "), failed.\n");
    }
    elsif ($entry->{type} == CHARDEV) {
	$^W && carp ("Making block device ", $entry->{name},
		     " (maj=", $entry->{devmajor}, 
		     " ,min=", $entry->{devminor}, "), failed.\n");
    }
}

*_make_special_file = $^O eq "MSWin32" ? 
    \&_make_special_file_Win32 : \&_make_special_file_UNIX;

sub _munge_file {
#
#  Mac path to the Unix like equivalent to be used in tar archives
#
    my $inpath = $_[0];
#
#  If there are no :'s in the name at all, assume it's a single item in the
#  current directory.  Return it, changing any / in the name into :
#
    if ($inpath !~ m,:,) {
	$inpath =~ s,/,:,g;
	return $inpath;
    }
#
#  If we now split on :, there will be just as many nulls in the list as
#  there should be up requests, except if it begins with a :, where there
#  will be one extra.
#
    my @names = split (/:/, $inpath);
    shift (@names)
	if ($names[0] eq "");
    my @outname = ();
#
#  Work from the end.
#
    my $i;
    for ($i = $#names; $i >= 0; --$i) {
	if ($names[$i] eq "") {
	    unshift (@outname, "..");
	} 
	else {
	    $names[$i] =~ s,/,:,g;
	    unshift (@outname, $names[$i]);
	}
    }
    my $netpath = join ("/", @outname);
    $netpath = $netpath . "/" if ($inpath =~ /:$/);
    if ($inpath !~ m,^:,) {
	return "/".$netpath;
    } 
    else {
	return $netpath;
    }
}

sub _get_handle {
    my ($fh, $flags, $mode);

    sysseek ($_[0], 0, 0)
	or goto &_drat;

    if ($^O eq "MSWin32") {
	$fh = $_[0];
    }
    else {
	$fh = fcntl ($_[0], F_DUPFD, 0)
	    or goto &_drat;
    }
    if ($compression && (@_ < 2 || $_[1] != 0)) {
	$mode = $#_ ? (int($_[1]) > 1 ?
			  "wb".int($_[1]) : "wb") : "rb";

#	$fh = Compress::Zlib::gzopen ($_[0], $mode)
#	    or &_drat;
	$fh = Compress::Zlib::gzdopen_ ($fh, $mode, 0)
	    or &_drat;
    }
    else {
	$flags = fcntl ($_[0], F_GETFL, 0) & (O_RDONLY | O_WRONLY | O_RDWR);
	$mode = ($flags == O_WRONLY) ? ">&=$fh" : 
	    ($flags == O_RDONLY) ? "<&=$fh" : "+>&=$fh";
	$fh = gensym;
	open ($fh, $mode)
	  or goto &_drat;

	$fh = bless *{$fh}{IO}, "Archive::Tar::_io";
	binmode $fh
	    or goto &_drat;
    }

    return $fh;
}

sub _read_tar {
    my ($file, $seekable, $extract) = @_;
    my $tarfile = [];
    my ($head, $offset, $size);

    $file->gzread ($head, $tar_header_length)
	or goto &_drat;

    if (substr ($head, 0, 2) =~ /$gzip_magic_number/o) {
	$error =
	    "Compression not available\n";
	return undef;
    }

    $offset = $tar_header_length
	if $seekable;

 READLOOP:
    while (length ($head) == $tar_header_length) {
	my ($name,		# string
	    $mode,		# octal number
	    $uid,		# octal number
	    $gid,		# octal number
	    $size,		# octal number
	    $mtime,		# octal number
	    $chksum,		# octal number
	    $type,		# character
	    $linkname,		# string
	    $magic,		# string
	    $version,		# two bytes
	    $uname,		# string
	    $gname,		# string
	    $devmajor,		# octal number
	    $devminor,		# octal number
	    $prefix) = unpack ($tar_unpack_header, $head);
	my ($data, $block, $entry);

	$mode = oct $mode;
	$uid = oct $uid;
	$gid = oct $gid;
	$size = oct $size;
	$mtime = oct $mtime;
	$chksum = oct $chksum;
	$devmajor = oct $devmajor;
	$devminor = oct $devminor;
	$name = $prefix."/".$name if $prefix;
	$prefix = "";
	# some broken tar-s don't set the type for directories
	# so we ass_u_me a directory if the name ends in slash
	$type = DIR
	    if $name =~ m|/$| and $type == FILE;

	last READLOOP if $head eq "\0" x 512; # End of archive
	# Apparently this should really be two blocks of 512 zeroes,
	# but GNU tar sometimes gets it wrong. See comment in the
	# source code (tar.c) to GNU cpio.

	substr ($head, 148, 8) = "        ";
	if (unpack ("%16C*", $head) != $chksum) {
	   warn "$name: checksum error.\n";
	}

	unless ($extract || $type != FILE) {
	    # Always read in full 512 byte blocks
	    $block = $size & 0x01ff ? ($size & ~0x01ff) + 512 : $size;
	    if ($seekable) {
		while ($block > 4096) {
		    $file->gzread ($data, 4096)
			or goto &_drat;
		    $block -= 4096;
		}
		$file->gzread ($data, $block)
		    or goto &_drat
			if ($block);

		# Ignore everything we've just read.
		undef $data;
	    } else {
		if ($file->gzread ($data, $block) < $block) {
		    $error = "Read error on tarfile.";
		    return undef;
		}

		# Throw away any trailing garbage
		substr ($data, $size) = "";
	    }
	}

	# Guard against tarfiles with garbage at the end
	last READLOOP if $name eq ''; 

	$entry = {name => $name,		    
		  mode => $mode,
		  uid => $uid,
		  gid => $gid,
		  size => $size,
		  mtime => $mtime,
		  chksum => $chksum,
		  type => $type,
		  linkname => $linkname,
		  magic => $magic,
		  version => $version,
		  uname => $uname,
		  gname => $gname,
		  devmajor => $devmajor,
		  devminor => $devminor,
		  prefix => $prefix,
		  offset => $offset,
		  data => $data};

	if ($extract) {
	    _extract_file ($entry, $file);
	    $file->gzread ($head, 512 - ($size & 0x1ff)) 
		or goto &_drat
		    if ($size & 0x1ff && $type == FILE);
	}
	else {
	    push @$tarfile, $entry;
	}

	if ($seekable) {
	    $offset += $tar_header_length;
	    $offset += ($size & 0x01ff) ? ($size & ~0x01ff) + 512 : $size
		if $type == FILE;
	}
	$file->gzread ($head, $tar_header_length) 
	    or goto &_drat;
    }

    $file->gzclose ()
	unless $seekable;

    return $tarfile
	unless $extract;
}

sub _format_tar_entry {
    my ($ref) = shift;
    my ($tmp,$file,$prefix,$pos);

    $file = $ref->{name};
    if (length ($file) > 99) {
	$pos = index $file, "/", (length ($file) - 100);
	next
	    if $pos == -1;	# Filename longer than 100 chars!

	$prefix = substr $file,0,$pos;
	$file = substr $file,$pos+1;
	substr ($prefix, 0, -155) = ""
	    if length($prefix)>154;
    }
    else {
	$prefix="";
    }

    $tmp = pack ($tar_pack_header,
		 $file,
		 sprintf("%06o ",$ref->{mode}),
		 sprintf("%06o ",$ref->{uid}),
		 sprintf("%06o ",$ref->{gid}),
		 sprintf("%11o ",$ref->{size}),
		 sprintf("%11o ",$ref->{mtime}),
		 "",		#checksum field - space padded by pack("A8")
		 $ref->{type},
		 $ref->{linkname},
		 $ref->{magic},
		 $ref->{version} || '00',
		 $ref->{uname},
		 $ref->{gname},
		 sprintf("%6o ",$ref->{devmajor}),
		 sprintf("%6o ",$ref->{devminor}),
		 $prefix);
    substr($tmp,148,7) = sprintf("%6o\0", unpack("%16C*",$tmp));

    return $tmp;
}

sub _format_tar_file {
    my @tarfile = @_;
    my $file = "";

    foreach (@tarfile) {
	$file .= _format_tar_entry $_;
	$file .= $_->{data};
	$file .= "\0" x (512 - ($_->{size} & 0x1ff))
	    if ($_->{size} & 0x1ff);
    }
    $file .= "\0" x 1024;

    return $file;
}

sub _write_tar {
    my $file = shift;
    my $entry;

    foreach $entry ((ref ($_[0]) eq 'ARRAY') ? @{$_[0]} : @_) {
	next
	    unless (ref ($entry) eq 'HASH');

	my $src;
        if ($^O eq "MacOS") {  #convert back from Unix to Mac path
            my @parts = split(/\//, $entry->{name});

            $src = $parts[0] ? ":" : "";
            foreach (@parts) {
		next if !$_ || $_ eq ".";  
                s,:,/,g;

		$_ = ":"
		    if ($_ eq "..");

		$src .= ($src =~ /:$/) ? $_ : ":$_";
	    }
        }
	else {
            $src = $entry->{name};
        }
	sysopen (FH, $src, O_RDONLY)
	    && binmode (FH)
		or next
		    unless $entry->{type} != FILE || $entry->{data};

	$file->gzwrite (_format_tar_entry ($entry))
	    or goto &_drat;

	if ($entry->{type} == FILE) {
	    if ($entry->{data}) {
		$file->gzwrite ($entry->{data})
		    or goto &_drat;
	    }
	    else {
		my $size = $entry->{size};
		my $data;
		while ($size >= 4096) {
		    sysread (FH, $data, 4096)
			&& $file->gzwrite ($data)
			    or goto &_drat;
		    $size -= 4096;
		}
		sysread (FH, $data, $size)
		    && $file->gzwrite ($data)
			or goto &_drat
			    if $size;
		close FH;
	    }
	    $file->gzwrite ("\0" x (512 - ($entry->{size} & 511)))
		or goto &_drat
		    if ($entry->{size} & 511);
	}
    }

    $file->gzwrite ("\0" x 1024)
	and !$file->gzclose ()
	    or goto &_drat;
}

sub _add_file {
    my $file = shift;
    my ($mode,$nlnk,$uid,$gid,$rdev,$size,$mtime,$type,$linkname);

    if (($mode,$nlnk,$uid,$gid,$rdev,$size,$mtime) = (lstat $file)[2..7,9]) {
	$linkname = "";
	$type = filetype ($file);

	$linkname = readlink $file
	    if ($type == SYMLINK) && $symlinks;

	$file = _munge_file ($file)
	    if ($^O eq "MacOS");

	return +{name => $file,		    
		 mode => $mode,
		 uid => $uid,
		 gid => $gid,
		 size => $size,
		 mtime => (($mtime - $time_offset) | 0),
		 chksum => "      ",
		 type => $type, 
		 linkname => $linkname,
		 magic => "ustar",
		 version => "00",
		 # WinNT protection
		 uname => ($fake_getpwuid || scalar getpwuid($uid)),
		 gname => ($fake_getgrgid || scalar getgrgid ($gid)),
		 devmajor => 0, # We don't handle this yet
		 devminor => 0, # We don't handle this yet
		 prefix => "",
		 data => undef,
		};
    }
}

sub _extract_file {
    my ($entry, $handle) = @_;
    my ($file, $cwd, @path);

    # For the moment, we assume that all paths in tarfiles
    # are given according to Unix standards.
    # Which they *are*, according to the tar format spec!
    @path = split(/\//,$entry->{name});
    $path[0] = '/' unless defined $path[0]; # catch absolute paths
    $file = pop @path;
    $file =~ s,:,/,g
	if $^O eq "MacOS";
    $cwd = cwd
	if @path;
    foreach (@path) {
	if ($^O eq "MacOS") {
	    s,:,/,g;
	    $_ = "::" if $_ eq "..";
	    $_ = ":" if $_ eq ".";
	}
	if (-e $_ && ! -d _) {
	    $^W && carp "$_ exists but is not a directory!\n";
	    next;
	}
	mkdir $_, 0777 unless -d _;
	chdir $_;
    }

    if ($entry->{type} == FILE) {	# Ordinary file
	sysopen (FH, $file, O_WRONLY|O_CREAT|O_TRUNC)
	    and binmode FH
		or goto &_drat;

	if ($handle) {
	    my $size = $entry->{size};
	    my $data;
	    while ($size > 4096) {
		$handle->gzread ($data, 4096)
		    and syswrite (FH, $data, length $data)
			or goto &_drat;
		$size -= 4096;
	    }
	    $handle->gzread ($data, $size)
		and syswrite (FH, $data, length $data)
		    or goto &_drat
			if ($size);
	}
	else {
	    syswrite FH, $entry->{data}, $entry->{size}
		or goto &_drat
	}
	close FH
	    or goto &_drat
    }
    elsif ($entry->{type} == DIR) { # Directory
	goto &_drat
	    if (-e $file && ! -d $file);

	mkdir $file,0777
	    unless -d $file;
    }
    elsif ($entry->{type} == UNKNOWN) {
	$error = "unknown file type: $_->{type}";
	return undef;
    }
    else {
	_make_special_file ($entry, $file);
    }
    utime time, $entry->{mtime} + $time_offset, $file;

    # We are root, and chown exists
    chown $entry->{uid}, $entry->{gid}, $file
	if ($> == 0 and $^O ne "MacOS" and $^O ne "MSWin32");

    # chmod is done last, in case it makes file readonly
    # (this accomodates DOSish OSes)
    chmod $entry->{mode}, $file;
    chdir $cwd
	if @path;
}

###
### Methods
###

##
## Class methods
##

# Perfom the equivalent of ->new()->add_files(), ->write() without the
# overhead of maintaining an Archive::Tar object.
sub create_archive {
    my ($handle, $file, $compress) = splice (@_, 0, 3);

    if ($compress && !$compression) {
	$error = "Compression not available.\n";
	return undef;
    }

    $handle = gensym;
    open $handle, ref ($file) ? ">&". fileno ($file) : ">" . $file
	and binmode ($handle)
	    or goto &_drat;

    _write_tar (_get_handle ($handle, int ($compress)),
		map {_add_file ($_)} @_);
}

# Perfom the equivalent of ->new()->list_files() without the overhead
# of maintaining an Archive::Tar object.
sub list_archive {
    my ($handle, $file, $fields) = @_;

    $handle = gensym;
    open $handle, ref ($file) ? "<&". fileno ($file) : "<" . $file
	and binmode ($handle)
	    or goto &_drat;

    my $data = _read_tar (_get_handle ($handle), 1);

    return map {my %h; @h{@$fields} = @$_{@$fields}; \%h} @$data
        if (ref $fields eq 'ARRAY'
	    && (@$fields > 1 || $fields->[0] ne 'name'));

    return map {$_->{name}} @$data;
}

# Perform the equivalen of ->new()->extract() without the overhead of
# maintaining an Archive::Tar object.
sub extract_archive {
    my ($handle, $file) = @_;

    $handle = gensym;
    open $handle, ref ($file) ? "<&". fileno ($file) : "<" . $file
	and binmode ($handle)
	    or goto &_drat;

    _read_tar (_get_handle ($handle), 0, 1);
}

# Constructor. Reads tarfile if given an argument that's the name of a
# readable file.
sub new {
    my ($class, $file) = @_;

    my $self = bless {}, $class;

    $self->read ($file)
      if defined $file;

    return $self;
}

## Return list with references to hashes representing the tar archive's
## component files.
#sub data {
#    my $self = shift;

#    return @{$self->{'_data'}};
#}

# Read a tarfile. Returns number of component files.
sub read {
    my ($self, $file) = @_;

    $self->{_data} = [];

    $self->{_handle} = gensym;
    open $self->{_handle}, ref ($file) ? "<&". fileno ($file) : "<" . $file
	and binmode ($self->{_handle})
	    or goto &_drat;

    $self->{_data} = _read_tar (_get_handle ($self->{_handle}), 
				  sysseek $self->{_handle}, 0, 1);
    return scalar @{$self->{_data}};
}

# Write a tar archive to file
sub write {
    my ($self, $file, $compress) = @_;

    return _format_tar_file (@{$self->{_data}})
	unless (@_ > 1);

    my $handle = gensym;
    open $handle, ref ($file) ? ">&". fileno ($file) : ">" . $file
	and binmode ($handle)
	    or goto &_drat;

    if ($compress && !$compression) {
	$error = "Compression not available.\n";
	return undef;
    }

    _write_tar (_get_handle ($handle, $compress || 0), $self->{_data});
}

# Add files to the archive. Returns number of successfully added files.
sub add_files {
    my $self = shift;
    my ($counter, $file, $entry);

    foreach $file (@_) {
	if ($entry = _add_file ($file)) {
	    push (@{$self->{'_data'}}, $entry);
	    ++$counter;
	}
    }

    return $counter;
}

# Add data as a file
sub add_data {
    my ($self, $file, $data, $opt) = @_;
    my $ref = {};
    my ($key);

    if($^O eq "MacOS") {
	$file = _munge_file($file);
    }
    $ref->{'data'} = $data;
    $ref->{name} = $file;
    $ref->{mode} = 0666 & (0777 - umask);
    $ref->{uid} = $>;
    $ref->{gid} = (split(/ /,$)))[0]; # Yuck
    $ref->{size} = length $data;
    $ref->{mtime} = ((time - $time_offset) | 0),
    $ref->{chksum} = "      ";	# Utterly pointless
    $ref->{type} = FILE;		# Ordinary file
    $ref->{linkname} = "";
    $ref->{magic} = "ustar";
    $ref->{version} = "00";
    # WinNT protection
    $ref->{uname} = $fake_getpwuid || getpwuid ($>);
    $ref->{gname} = $fake_getgrgid || getgrgid ($ref->{gid});
    $ref->{devmajor} = 0;
    $ref->{devminor} = 0;
    $ref->{prefix} = "";

    if ($opt) {
	foreach $key (keys %$opt) {
	    $ref->{$key} = $opt->{$key}
	}
    }

    push (@{$self->{'_data'}}, $ref);
    return 1;
}

sub rename {
    my ($self) = shift;
    my $entry;

    foreach $entry (@{$self->{_data}}) {
	@{$self->{_data}} = grep {$_->{name} ne $entry} @{$self->{'_data'}};
    }
    return $self;
}

sub remove {
    my ($self) = shift;
    my $entry;

    foreach $entry (@_) {
	@{$self->{_data}} = grep {$_->{name} ne $entry} @{$self->{'_data'}};
    }
    return $self;
}

# Get the content of a file
sub get_content {
    my ($self, $file) = @_;
    my ($entry, $data);

    foreach $entry (@{$self->{_data}}) {
	next
	    unless $entry->{name} eq $file;

	return $entry->{data}
	    unless $entry->{offset};

	my $handle = _get_handle ($self->{_handle});
	$handle->gzseek ($entry->{offset}, 0)
	    or goto &_drat;

	$handle->gzread ($data, $entry->{size}) != -1
	    or goto &_drat;

	return $data;
    }

    return;
}

# Replace the content of a file
sub replace_content {
    my ($self, $file, $content) = @_;
    my $entry;

    foreach $entry (@{$self->{_data}}) {
	next
	    unless $entry->{name} eq $file;

	$entry->{data} = $content;
	$entry->{size} = length $content;
	$entry->{offset} = undef;
	return 1;
    }
}

# Write a single (probably) file from the in-memory archive to disk
sub extract {
    my $self = shift;
    my @files = @_;
    my ($file, $entry);

    @files = list_files ($self) unless @files;
    foreach $entry (@{$self->{_data}}) {
	my $cnt = 0;
	foreach $file (@files) {
	    ++$cnt, next
		unless $entry->{name} eq $file;
	    my $handle = $entry->{offset} && _get_handle ($self->{_handle});
	    $handle->gzseek ($entry->{offset}, 0)
		or goto &_drat
		    if $handle;
	    _extract_file ($entry, $handle);
	    splice (@_, $cnt, 1);
	    last;
	}
	last
	    unless @_;
    }
    $self;
}


# Return a list names or attribute hashes for all files in the
# in-memory archive.
sub list_files {
 my ($self, $fields) = @_;

    return map {my %h; @h{@$fields} = @$_{@$fields}; \%h} @{$self->{'_data'}}
    if (ref $fields eq 'ARRAY' && (@$fields > 1 || $fields->[0] ne 'name'));

    return map {$_->{name}} @{$self->{'_data'}}
}


### Standard end of module :-)
1;

# 
# Sub-package to hide I/O differences between compressed &
# uncompressed archives.
#
# Yes, I could have used the IO::* class hierarchy here, but I'm
# trying to minimise the necessity for non-core modules on perl5
# environments < 5.004

package Archive::Tar::_io;

sub gzseek {
    sysseek $_[0], $_[1], $_[2];
}

sub gzread {
    sysread $_[0], $_[1], $_[2];
}

sub gzwrite {
    syswrite $_[0], $_[1], length $_[1];
}

sub gzclose {
    !close $_[0];
}

1;

__END__

=head1 NAME

Tar - module for manipulation of tar archives.

=head1 SYNOPSIS

  use Archive::Tar;

  Archive::Tar->create_archive ("my.tar.gz", 9, "/this/file", "/that/file");
  print join "\n", Archive::Tar->list_archive ("my.tar.gz"), "";

  $tar = Archive::Tar->new();
  $tar->read("origin.tar.gz",1);
  $tar->add_files("file/foo.c", "file/bar.c");
  $tar->add_data("file/baz.c","This is the file contents");
  $tar->write("files.tar");

=head1 DESCRIPTION

This is a module for the handling of tar archives. 

Archive::Tar provides an object oriented mechanism for handling tar
files.  It provides class methods for quick and easy files handling
while also allowing for the creation of tar file objects for custom
manipulation.  If you have the Compress::Zlib module installed,
Archive::Tar will also support compressed or gzipped tar files.

=head2 Class Methods

The class methods should be sufficient for most tar file interaction.

=over 4

=item create_archive ($file, $compression, @filelist)

Creates a tar file from the list of files provided.  The first
argument can either be the name of the tar file to create or a
reference to an open file handle (e.g. a GLOB reference).

The second argument specifies the level of compression to be used, if
any.  Compression of tar files requires the installation of the
Compress::Zlib module.  Specific levels or compression may be
requested by passing a value between 2 and 9 as the second argument.
Any other value evaluating as true will result in the default
compression level being used.

The remaining arguments list the files to be included in the tar file.
These files must all exist.  Any files which don\'t exist or can\'t be
read are silently ignored.

If the archive creation fails for any reason, C<create_archive> will
return undef.  Please use the C<error> method to find the cause of the
failure.

=item list_archive ($file, ['property', 'property',...])

=item list_archive ($file)

Returns a list of the names of all the files in the archive.  The
first argument can either be the name of the tar file to create or a
reference to an open file handle (e.g. a GLOB reference).

If C<list_archive()> is passed an array reference as its second
argument it returns a list of hash references containing the requested
properties of each file.  The following list of properties is
supported: name, size, mtime (last modified date), mode, uid, gid,
linkname, uname, gname, devmajor, devminor, prefix.

Passing an array reference containing only one element, 'name', is
special cased to return a list of names rather than a list of hash
references.

=item extract_archive ($file)

Extracts the contents of the tar file.  The first argument can either
be the name of the tar file to create or a reference to an open file
handle (e.g. a GLOB reference).  All relative paths in the tar file will
be created underneath the current working directory.

If the archive extraction fails for any reason, C<extract_archive>
will return undef.  Please use the C<error> method to find the cause
of the failure.

=item new ($file)

=item new ()

Returns a new Tar object. If given any arguments, C<new()> calls the
C<read()> method automatically, parsing on the arguments provided L<read()>.

If C<new()> is invoked with arguments and the read method fails for
any reason, C<new()> returns undef.

=back

=head2 Instance Methods

=over 4

=item read ($ref, $compressed)

Read the given tar file into memory. The first argument can either be
the name of a file or a reference to an already open file handle (e.g. a
GLOB reference).  The second argument indicates whether the file
referenced by the first argument is compressed.

The second argument is now optional as Archive::Tar will automatically
detect compressed archives.

The C<read> will I<replace> any previous content in C<$tar>!

=item add_files(@filenamelist)

Takes a list of filenames and adds them to the in-memory archive.  On
MacOS, the path to the file is automatically converted to a Unix like
equivalent for use in the archive, and the file\'s modification time
is converted from the MacOS epoch to the Unix epoch.  So tar archives
created on MacOS with B<Archive::Tar> can be read both with I<tar> on
Unix and applications like I<suntar> or I<Stuffit Expander> on MacOS.
Be aware that the file\'s type/creator and resource fork will be lost,
which is usually what you want in cross-platform archives.

=item add_data ($filename, $data, $opthashref)

Takes a filename, a scalar full of data and optionally a reference to
a hash with specific options. Will add a file to the in-memory
archive, with name C<$filename> and content C<$data>. Specific
properties can be set using C<$opthashref>, The following list of
properties is supported: name, size, mtime (last modified date), mode,
uid, gid, linkname, uname, gname, devmajor, devminor, prefix.  (On
MacOS, the file\'s path and modification times are converted to Unix
equivalents.)

=item remove (@filenamelist)

Removes any entries with names matching any of the given filenames
from the in-memory archive. String comparisons are done with C<eq>.

=item write ($file, $compressed)

Write the in-memory archive to disk.  The first argument can either be
the name of a file or a reference to an already open file handle (be a
GLOB reference).  If the second argument is true, the module will use
Compress::Zlib to write the file in a compressed format.  If
Compress:Zlib is not available, the C<write> method will fail.
Specific levels of compression can be chosen by passing the values 2
through 9 as the second parameter.

If no arguments are given, C<write> returns the entire formatted
archive as a string, which could be useful if you\'d like to stuff the
archive into a socket or a pipe to gzip or something.  This
functionality may be deprecated later, however, as you can also do
this using a GLOB reference for the first argument.

=item extract(@filenames)

Write files whose names are equivalent to any of the names in
C<@filenames> to disk, creating subdirectories as necessary. This
might not work too well under VMS.  Under MacPerl, the file\'s
modification time will be converted to the MacOS zero of time, and
appropriate conversions will be done to the path.  However, the length
of each element of the path is not inspected to see whether it\'s
longer than MacOS currently allows (32 characters).

If C<extract> is called without a list of file names, the entire
contents of the archive are extracted.

=item list_files(['property', 'property',...])

=item list_files()

Returns a list of the names of all the files in the archive.

If C<list_files()> is passed an array reference as its first argument
it returns a list of hash references containing the requested
properties of each file.  The following list of properties is
supported: name, size, mtime (last modified date), mode, uid, gid,
linkname, uname, gname, devmajor, devminor, prefix.

Passing an array reference containing only one element, 'name', is
special cased to return a list of names rather than a list of hash
references.

=item get_content($file)

Return the content of the named file.

=item replace_content($file,$content)

Make the string $content be the content for the file named $file.

=back

=head1 CHANGES

=over 4

=item Version 0.20

Added class methods for creation, extraction and listing of tar files.
No longer maintain a complete copy of the tar file in memory.  Removed
the C<data()> method.

=item Version 0.10

Numerous changes. Brought source under CVS.  All changes now recorded
in ChangeLog file in distribution.

=item Version 0.08

New developer/maintainer.  Calle has carpal-tunnel syndrome and cannot
type a great deal. Get better as soon as you can, Calle.

Added proper support for MacOS.  Thanks to Paul J. Schinder
<schinder@leprss.gsfc.nasa.gov>.

=item Version 0.071

Minor release.

Arrange to chmod() at the very end in case it makes the file read only.
Win32 is actually picky about that.

SunOS 4.x tar makes tarfiles that contain directory entries that
don\'t have typeflag set properly.  We use the trailing slash to
recognise directories in such tar files.

=item Version 0.07

Fixed (hopefully) broken portability to MacOS, reported by Paul J.
Schinder at Goddard Space Flight Center.

Fixed two bugs with symlink handling, reported in excellent detail by
an admin at teleport.com called Chris.

Primitive tar program (called ptar) included with distribution. Usage
should be pretty obvious if you\'ve used a normal tar program.

Added methods get_content and replace_content.

Added support for paths longer than 100 characters, according to
POSIX. This is compatible with just about everything except GNU tar.
Way to go, GNU tar (use a better tar, or GNU cpio).

NOTE: When adding files to an archive, files with basenames longer
      than 100 characters will be silently ignored. If the prefix part
      of a path is longer than 155 characters, only the last 155
      characters will be stored.

=item Version 0.06

Added list_files() method, as requested by Michael Wiedman.

Fixed a couple of dysfunctions when run under Windows NT. Michael
Wiedmann reported the bugs.

Changed the documentation to reflect reality a bit better.

Fixed bug in format_tar_entry. Bug reported by Michael Schilli.

=item Version 0.05

Quoted lots of barewords to make C<use strict;> stop complaining under
perl version 5.003.

Ties to L<Compress::Zlib> put in. Will warn if it isn\'t available.

$tar->write() with no argument now returns the formatted archive.

=item Version 0.04

Made changes to write_tar so that Solaris tar likes the resulting
archives better.

Protected the calls to readlink() and symlink(). AFAIK this module
should now run just fine on Windows NT.

Add method to write a single entry to disk (extract)

Added method to add entries entirely from scratch (add_data)

Changed name of add() to add_file()

All calls to croak() removed and replaced with returning undef and
setting Tar::error.

Better handling of tarfiles with garbage at the end.

=cut