1 ### the gnu tar specification:
2 ### http://www.gnu.org/software/tar/manual/tar.html
4 ### and the pax format spec, which tar derives from:
5 ### http://www.opengroup.org/onlinepubs/007904975/utilities/pax.html
13 use Carp qw(carp croak);
15 use File::Spec::Unix ();
18 use Archive::Tar::File;
19 use Archive::Tar::Constant;
24 use vars qw[$DEBUG $error $VERSION $WARN $FOLLOW_SYMLINK $CHOWN $CHMOD
25 $DO_NOT_USE_PREFIX $HAS_PERLIO $HAS_IO_STRING $SAME_PERMISSIONS
26 $INSECURE_EXTRACT_MODE @ISA @EXPORT
30 @EXPORT = qw[ COMPRESS_GZIP COMPRESS_BZIP ];
37 $SAME_PERMISSIONS = $> == 0 ? 1 : 0;
38 $DO_NOT_USE_PREFIX = 0;
39 $INSECURE_EXTRACT_MODE = 0;
43 $HAS_PERLIO = $Config::Config{useperlio};
45 ### try and load IO::String anyway, so you can dynamically
46 ### switch between perlio and IO::String
47 $HAS_IO_STRING = eval {
56 Archive::Tar - module for manipulations of tar archives
61 my $tar = Archive::Tar->new;
63 $tar->read('origin.tgz');
66 $tar->add_files('file/foo.pl', 'docs/README');
67 $tar->add_data('file/baz.txt', 'This is the contents now');
69 $tar->rename('oldname', 'new/file/name');
71 $tar->write('files.tar'); # plain tar
72 $tar->write('files.tgz', COMPRESS_GZIP); # gzip compressed
73 $tar->write('files.tbz', COMPRESS_BZIP); # bzip2 compressed
77 Archive::Tar provides an object oriented mechanism for handling tar
78 files. It provides class methods for quick and easy files handling
79 while also allowing for the creation of tar file objects for custom
80 manipulation. If you have the IO::Zlib module installed,
81 Archive::Tar will also support compressed or gzipped tar files.
83 An object of class Archive::Tar represents a .tar(.gz) archive full
88 =head2 Archive::Tar->new( [$file, $compressed] )
90 Returns a new Tar object. If given any arguments, C<new()> calls the
91 C<read()> method automatically, passing on the arguments provided to
94 If C<new()> is invoked with arguments and the C<read()> method fails
95 for any reason, C<new()> returns undef.
104 ### install get/set accessors for this object.
105 for my $key ( keys %$tmpl ) {
107 *{__PACKAGE__."::$key"} = sub {
109 $self->{$key} = $_[0] if @_;
110 return $self->{$key};
116 $class = ref $class if ref $class;
118 ### copying $tmpl here since a shallow copy makes it use the
119 ### same aref, causing for files to remain in memory always.
120 my $obj = bless { _data => [ ], _file => 'Unknown' }, $class;
123 unless ( $obj->read( @_ ) ) {
124 $obj->_error(qq[No data could be read from file]);
132 =head2 $tar->read ( $filename|$handle, [$compressed, {opt => 'val'}] )
134 Read the given tar file into memory.
135 The first argument can either be the name of a file or a reference to
136 an already open filehandle (or an IO::Zlib object if it's compressed)
138 The C<read> will I<replace> any previous content in C<$tar>!
140 The second argument may be considered optional, but remains for
141 backwards compatibility. Archive::Tar now looks at the file
142 magic to determine what class should be used to open the file
143 and will transparently Do The Right Thing.
145 Archive::Tar will warn if you try to pass a bzip2 compressed file and the
146 IO::Zlib / IO::Uncompress::Bunzip2 modules are not available and simply return.
148 Note that you can currently B<not> pass a C<gzip> compressed
149 filehandle, which is not opened with C<IO::Zlib>, a C<bzip2> compressed
150 filehandle, which is not opened with C<IO::Uncompress::Bunzip2>, nor a string
151 containing the full archive information (either compressed or
152 uncompressed). These are worth while features, but not currently
153 implemented. See the C<TODO> section.
155 The third argument can be a hash reference with options. Note that
156 all options are case-sensitive.
162 Do not read more than C<limit> files. This is useful if you have
163 very big archives, and are only interested in the first few files.
167 Can be set to a regular expression. Only files with names that match
168 the expression will be read.
172 If set to true, immediately extract entries when reading them. This
173 gives you the same memory break as the C<extract_archive> function.
174 Note however that entries will not be read into memory, but written
175 straight to disk. This means no C<Archive::Tar::File> objects are
176 created for you to inspect.
180 All files are stored internally as C<Archive::Tar::File> objects.
181 Please consult the L<Archive::Tar::File> documentation for details.
183 Returns the number of files read in scalar context, and a list of
184 C<Archive::Tar::File> objects in list context.
191 my $gzip = shift || 0;
192 my $opts = shift || {};
194 unless( defined $file ) {
195 $self->_error( qq[No file to read from!] );
198 $self->_file( $file );
201 my $handle = $self->_get_handle($file, $gzip, READ_ONLY->( ZLIB ) )
204 my $data = $self->_read_tar( $handle, $opts ) or return;
206 $self->_data( $data );
208 return wantarray ? @$data : scalar @$data;
213 my $file = shift; return unless defined $file;
214 return $file if ref $file;
215 my $compress = shift || 0;
216 my $mode = shift || READ_ONLY->( ZLIB ); # default to read only
219 ### get a FH opened to the right class, so we can use it transparently
220 ### throughout the program
222 { ### reading magic only makes sense if we're opening a file for
223 ### reading. otherwise, just use what the user requested.
225 if( MODE_READ->($mode) ) {
226 open my $tmp, $file or do {
227 $self->_error( qq[Could not open '$file' for reading: $!] );
231 ### read the first 4 bites of the file to figure out which class to
232 ### use to open the file.
233 sysread( $tmp, $magic, 4 );
238 ### if you asked specifically for bzip compression, or if we're in
239 ### read mode and the magic numbers add up, use bzip
241 ($compress eq COMPRESS_BZIP) or
242 ( MODE_READ->($mode) and $magic =~ BZIP_MAGIC_NUM )
246 ### different reader/writer modules, different error vars... sigh
247 if( MODE_READ->($mode) ) {
248 $fh = IO::Uncompress::Bunzip2->new( $file ) or do {
249 $self->_error( qq[Could not read '$file': ] .
250 $IO::Uncompress::Bunzip2::Bunzip2Error
256 $fh = IO::Compress::Bzip2->new( $file ) or do {
257 $self->_error( qq[Could not write to '$file': ] .
258 $IO::Compress::Bzip2::Bzip2Error
265 ### if you asked for compression, if you wanted to read or the gzip
266 ### magic number is present (redundant with read)
268 $compress or MODE_READ->($mode) or $magic =~ GZIP_MAGIC_NUM
273 unless( $fh->open( $file, $mode ) ) {
274 $self->_error(qq[Could not create filehandle for '$file': $!]);
282 unless( $fh->open( $file, $mode ) ) {
283 $self->_error(qq[Could not create filehandle for '$file': $!]);
287 ### enable bin mode on tar archives
298 my $handle = shift or return;
299 my $opts = shift || {};
301 my $count = $opts->{limit} || 0;
302 my $filter = $opts->{filter};
303 my $extract = $opts->{extract} || 0;
305 ### set a cap on the amount of files to extract ###
307 $limit = 1 if $count > 0;
312 my $real_name; # to set the name of a file when
313 # we're encountering @longlink
317 while( $handle->read( $chunk, HEAD ) ) {
318 ### IO::Zlib doesn't support this yet
319 my $offset = eval { tell $handle } || 'unknown';
322 my $gzip = GZIP_MAGIC_NUM;
323 if( $chunk =~ /$gzip/ ) {
324 $self->_error( qq[Cannot read compressed format in tar-mode] );
328 ### size is < HEAD, which means a corrupted file, as the minimum
329 ### length is _at least_ HEAD
330 if (length $chunk != HEAD) {
331 $self->_error( qq[Cannot read enough bytes from the tarfile] );
336 ### if we can't read in all bytes... ###
337 last if length $chunk != HEAD;
339 ### Apparently this should really be two blocks of 512 zeroes,
340 ### but GNU tar sometimes gets it wrong. See comment in the
341 ### source code (tar.c) to GNU cpio.
342 next if $chunk eq TAR_END;
344 ### according to the posix spec, the last 12 bytes of the header are
345 ### null bytes, to pad it to a 512 byte block. That means if these
346 ### bytes are NOT null bytes, it's a corrrupt header. See:
347 ### www.koders.com/c/fidCE473AD3D9F835D690259D60AD5654591D91D5BA.aspx
349 { my $nulls = join '', "\0" x 12;
350 unless( $nulls eq substr( $chunk, 500, 12 ) ) {
351 $self->_error( qq[Invalid header block at offset $offset] );
356 ### pass the realname, so we can set it 'proper' right away
357 ### some of the heuristics are done on the name, so important
360 { my %extra_args = ();
361 $extra_args{'name'} = $$real_name if defined $real_name;
363 unless( $entry = Archive::Tar::File->new( chunk => $chunk,
366 $self->_error( qq[Couldn't read chunk at offset $offset] );
372 ### http://www.gnu.org/manual/tar/html_node/tar_139.html
373 next if $entry->is_label;
375 if( length $entry->type and ($entry->is_file || $entry->is_longlink) ) {
377 if ( $entry->is_file && !$entry->validate ) {
378 ### sometimes the chunk is rather fux0r3d and a whole 512
379 ### bytes ends up in the ->name area.
380 ### clean it up, if need be
381 my $name = $entry->name;
382 $name = substr($name, 0, 100) if length $name > 100;
385 $self->_error( $name . qq[: checksum error] );
389 my $block = BLOCK_SIZE->( $entry->size );
391 $data = $entry->get_content_by_ref;
393 ### just read everything into memory
394 ### can't do lazy loading since IO::Zlib doesn't support 'seek'
395 ### this is because Compress::Zlib doesn't support it =/
396 ### this reads in the whole data in one read() call.
397 if( $handle->read( $$data, $block ) < $block ) {
398 $self->_error( qq[Read error on tarfile (missing data) '].
399 $entry->full_path ."' at offset $offset" );
403 ### throw away trailing garbage ###
404 substr ($$data, $entry->size) = "" if defined $$data;
406 ### part II of the @LongLink munging -- need to do /after/
407 ### the checksum check.
408 if( $entry->is_longlink ) {
409 ### weird thing in tarfiles -- if the file is actually a
410 ### @LongLink, the data part seems to have a trailing ^@
411 ### (unprintable) char. to display, pipe output through less.
412 ### but that doesn't *always* happen.. so check if the last
413 ### character is a control character, and if so remove it
414 ### at any rate, we better remove that character here, or tests
415 ### like 'eq' and hashlook ups based on names will SO not work
416 ### remove it by calculating the proper size, and then
417 ### tossing out everything that's longer than that size.
419 ### count number of nulls
420 my $nulls = $$data =~ tr/\0/\0/;
422 ### cut data + size by that many bytes
423 $entry->size( $entry->size - $nulls );
424 substr ($$data, $entry->size) = "";
428 ### clean up of the entries.. posix tar /apparently/ has some
429 ### weird 'feature' that allows for filenames > 255 characters
430 ### they'll put a header in with as name '././@LongLink' and the
431 ### contents will be the name of the /next/ file in the archive
432 ### pretty crappy and kludgy if you ask me
434 ### set the name for the next entry if this is a @LongLink;
435 ### this is one ugly hack =/ but needed for direct extraction
436 if( $entry->is_longlink ) {
439 } elsif ( defined $real_name ) {
440 $entry->name( $$real_name );
445 ### skip this entry if we're filtering
446 if ($filter && $entry->name !~ $filter) {
449 ### skip this entry if it's a pax header. This is a special file added
450 ### by, among others, git-generated tarballs. It holds comments and is
451 ### not meant for extracting. See #38932: pax_global_header extracted
452 } elsif ( $entry->name eq PAX_HEADER ) {
456 $self->_extract_file( $entry ) if $extract
457 && !$entry->is_longlink
458 && !$entry->is_unknown
459 && !$entry->is_label;
461 ### Guard against tarfiles with garbage at the end
462 last LOOP if $entry->name eq '';
464 ### push only the name on the rv if we're extracting
465 ### -- for extract_archive
466 push @$tarfile, ($extract ? $entry->name : $entry);
469 $count-- unless $entry->is_longlink || $entry->is_dir;
470 last LOOP unless $count;
479 =head2 $tar->contains_file( $filename )
481 Check if the archive contains a certain file.
482 It will return true if the file is in the archive, false otherwise.
484 Note however, that this function does an exact match using C<eq>
485 on the full path. So it cannot compensate for case-insensitive file-
486 systems or compare 2 paths to see if they would point to the same
495 return unless defined $full;
497 ### don't warn if the entry isn't there.. that's what this function
498 ### is for after all.
500 return 1 if $self->_find_entry($full);
504 =head2 $tar->extract( [@filenames] )
506 Write files whose names are equivalent to any of the names in
507 C<@filenames> to disk, creating subdirectories as necessary. This
508 might not work too well under VMS.
509 Under MacPerl, the file's modification time will be converted to the
510 MacOS zero of time, and appropriate conversions will be done to the
511 path. However, the length of each element of the path is not
512 inspected to see whether it's longer than MacOS currently allows (32
515 If C<extract> is called without a list of file names, the entire
516 contents of the archive are extracted.
518 Returns a list of filenames extracted.
527 # use the speed optimization for all extracted files
528 local($self->{cwd}) = cwd() unless $self->{cwd};
530 ### you requested the extraction of only certian files
532 for my $file ( @args ) {
534 ### it's already an object?
535 if( UNIVERSAL::isa( $file, 'Archive::Tar::File' ) ) {
543 for my $entry ( @{$self->_data} ) {
544 next unless $file eq $entry->full_path;
546 ### we found the file you're looking for
552 return $self->_error(
553 qq[Could not find '$file' in archive] );
558 ### just grab all the file items
560 @files = $self->get_files;
563 ### nothing found? that's an error
564 unless( scalar @files ) {
565 $self->_error( qq[No files found for ] . $self->_file );
570 for my $entry ( @files ) {
571 unless( $self->_extract_file( $entry ) ) {
572 $self->_error(q[Could not extract ']. $entry->full_path .q['] );
580 =head2 $tar->extract_file( $file, [$extract_path] )
582 Write an entry, whose name is equivalent to the file name provided to
583 disk. Optionally takes a second parameter, which is the full native
584 path (including filename) the entry will be written to.
588 $tar->extract_file( 'name/in/archive', 'name/i/want/to/give/it' );
590 $tar->extract_file( $at_file_object, 'name/i/want/to/give/it' );
592 Returns true on success, false on failure.
598 my $file = shift; return unless defined $file;
601 my $entry = $self->_find_entry( $file )
602 or $self->_error( qq[Could not find an entry for '$file'] ), return;
604 return $self->_extract_file( $entry, $alt );
609 my $entry = shift or return;
612 ### you wanted an alternate extraction location ###
613 my $name = defined $alt ? $alt : $entry->full_path;
615 ### splitpath takes a bool at the end to indicate
616 ### that it's splitting a dir
617 my ($vol,$dirs,$file);
618 if ( defined $alt ) { # It's a local-OS path
619 ($vol,$dirs,$file) = File::Spec->splitpath( $alt,
622 ($vol,$dirs,$file) = File::Spec::Unix->splitpath( $name,
627 ### is $name an absolute path? ###
628 if( $vol || File::Spec->file_name_is_absolute( $dirs ) ) {
630 ### absolute names are not allowed to be in tarballs under
631 ### strict mode, so only allow it if a user tells us to do it
632 if( not defined $alt and not $INSECURE_EXTRACT_MODE ) {
634 q[Entry ']. $entry->full_path .q[' is an absolute path. ].
635 q[Not extracting absolute paths under SECURE EXTRACT MODE]
640 ### user asked us to, it's fine.
641 $dir = File::Spec->catpath( $vol, $dirs, "" );
643 ### it's a relative path ###
645 my $cwd = (ref $self and defined $self->{cwd})
649 my @dirs = defined $alt
650 ? File::Spec->splitdir( $dirs ) # It's a local-OS path
651 : File::Spec::Unix->splitdir( $dirs ); # it's UNIX-style, likely
652 # straight from the tarball
654 if( not defined $alt and
655 not $INSECURE_EXTRACT_MODE
658 ### paths that leave the current directory are not allowed under
659 ### strict mode, so only allow it if a user tells us to do this.
660 if( grep { $_ eq '..' } @dirs ) {
663 q[Entry ']. $entry->full_path .q[' is attempting to leave ].
664 q[the current working directory. Not extracting under ].
665 q[SECURE EXTRACT MODE]
670 ### the archive may be asking us to extract into a symlink. This
671 ### is not sane and a possible security issue, as outlined here:
672 ### https://rt.cpan.org/Ticket/Display.html?id=30380
673 ### https://bugzilla.redhat.com/show_bug.cgi?id=295021
674 ### https://issues.rpath.com/browse/RPL-1716
675 my $full_path = $cwd;
676 for my $d ( @dirs ) {
677 $full_path = File::Spec->catdir( $full_path, $d );
679 ### we've already checked this one, and it's safe. Move on.
680 next if ref $self and $self->{_link_cache}->{$full_path};
682 if( -l $full_path ) {
683 my $to = readlink $full_path;
684 my $diag = "symlinked directory ($full_path => $to)";
687 q[Entry ']. $entry->full_path .q[' is attempting to ].
688 qq[extract to a $diag. This is considered a security ].
689 q[vulnerability and not allowed under SECURE EXTRACT ].
695 ### XXX keep a cache if possible, so the stats become cheaper:
696 $self->{_link_cache}->{$full_path} = 1 if ref $self;
700 ### '.' is the directory delimiter on VMS, which has to be escaped
701 ### or changed to '_' on vms. vmsify is used, because older versions
702 ### of vmspath do not handle this properly.
703 ### Must not add a '/' to an empty directory though.
704 map { length() ? VMS::Filespec::vmsify($_.'/') : $_ } @dirs if ON_VMS;
706 my ($cwd_vol,$cwd_dir,$cwd_file)
707 = File::Spec->splitpath( $cwd );
708 my @cwd = File::Spec->splitdir( $cwd_dir );
709 push @cwd, $cwd_file if length $cwd_file;
711 ### We need to pass '' as the last elemant to catpath. Craig Berry
712 ### explains why (msgid <p0624083dc311ae541393@[172.16.52.1]>):
713 ### The root problem is that splitpath on UNIX always returns the
714 ### final path element as a file even if it is a directory, and of
715 ### course there is no way it can know the difference without checking
716 ### against the filesystem, which it is documented as not doing. When
717 ### you turn around and call catpath, on VMS you have to know which bits
718 ### are directory bits and which bits are file bits. In this case we
719 ### know the result should be a directory. I had thought you could omit
720 ### the file argument to catpath in such a case, but apparently on UNIX
722 $dir = File::Spec->catpath(
723 $cwd_vol, File::Spec->catdir( @cwd, @dirs ), ''
726 ### catdir() returns undef if the path is longer than 255 chars on
727 ### older VMS systems.
728 unless ( defined $dir ) {
729 $^W && $self->_error( qq[Could not compose a path for '$dirs'\n] );
735 if( -e $dir && !-d _ ) {
736 $^W && $self->_error( qq['$dir' exists, but it's not a directory!\n] );
741 eval { File::Path::mkpath( $dir, 0, 0777 ) };
743 my $fp = $entry->full_path;
744 $self->_error(qq[Could not create directory '$dir' for '$fp': $@]);
748 ### XXX chown here? that might not be the same as in the archive
749 ### as we're only chown'ing to the owner of the file we're extracting
750 ### not to the owner of the directory itself, which may or may not
751 ### be another entry in the archive
752 ### Answer: no, gnu tar doesn't do it either, it'd be the wrong
754 #if( $CHOWN && CAN_CHOWN ) {
755 # chown $entry->uid, $entry->gid, $dir or
756 # $self->_error( qq[Could not set uid/gid on '$dir'] );
760 ### we're done if we just needed to create a dir ###
761 return 1 if $entry->is_dir;
763 my $full = File::Spec->catfile( $dir, $file );
765 if( $entry->is_unknown ) {
766 $self->_error( qq[Unknown file type for file '$full'] );
770 if( length $entry->type && $entry->is_file ) {
771 my $fh = IO::File->new;
772 $fh->open( '>' . $full ) or (
773 $self->_error( qq[Could not open file '$full': $!] ),
779 syswrite $fh, $entry->data or (
780 $self->_error( qq[Could not write data to '$full'] ),
786 $self->_error( qq[Could not close file '$full'] ),
791 $self->_make_special_file( $entry, $full ) or return;
794 ### only update the timestamp if it's not a symlink; that will change the
795 ### timestamp of the original. This addresses bug #33669: Could not update
796 ### timestamp warning on symlinks
798 utime time, $entry->mtime - TIME_OFFSET, $full or
799 $self->_error( qq[Could not update timestamp] );
802 if( $CHOWN && CAN_CHOWN->() ) {
803 chown $entry->uid, $entry->gid, $full or
804 $self->_error( qq[Could not set uid/gid on '$full'] );
807 ### only chmod if we're allowed to, but never chmod symlinks, since they'll
808 ### change the perms on the file they're linking too...
809 if( $CHMOD and not -l $full ) {
810 my $mode = $entry->mode;
811 unless ($SAME_PERMISSIONS) {
812 $mode &= ~(oct(7000) | umask);
814 chmod $mode, $full or
815 $self->_error( qq[Could not chown '$full' to ] . $entry->mode );
821 sub _make_special_file {
823 my $entry = shift or return;
824 my $file = shift; return unless defined $file;
828 if( $entry->is_symlink ) {
831 symlink( $entry->linkname, $file ) or $fail++;
834 $self->_extract_special_file_as_plain_file( $entry, $file )
838 $err = qq[Making symbolic link '$file' to '] .
839 $entry->linkname .q[' failed] if $fail;
841 } elsif ( $entry->is_hardlink ) {
844 link( $entry->linkname, $file ) or $fail++;
847 $self->_extract_special_file_as_plain_file( $entry, $file )
851 $err = qq[Making hard link from '] . $entry->linkname .
852 qq[' to '$file' failed] if $fail;
854 } elsif ( $entry->is_fifo ) {
855 ON_UNIX && !system('mknod', $file, 'p') or
856 $err = qq[Making fifo ']. $entry->name .qq[' failed];
858 } elsif ( $entry->is_blockdev or $entry->is_chardev ) {
859 my $mode = $entry->is_blockdev ? 'b' : 'c';
861 ON_UNIX && !system('mknod', $file, $mode,
862 $entry->devmajor, $entry->devminor) or
863 $err = qq[Making block device ']. $entry->name .qq[' (maj=] .
864 $entry->devmajor . qq[ min=] . $entry->devminor .
867 } elsif ( $entry->is_socket ) {
868 ### the original doesn't do anything special for sockets.... ###
872 return $err ? $self->_error( $err ) : 1;
875 ### don't know how to make symlinks, let's just extract the file as
877 sub _extract_special_file_as_plain_file {
879 my $entry = shift or return;
880 my $file = shift; return unless defined $file;
884 my $orig = $self->_find_entry( $entry->linkname );
887 $err = qq[Could not find file '] . $entry->linkname .
892 ### clone the entry, make it appear as a normal file ###
893 my $clone = $entry->clone;
894 $clone->_downgrade_to_plainfile;
895 $self->_extract_file( $clone, $file ) or last TRY;
900 return $self->_error($err);
903 =head2 $tar->list_files( [\@properties] )
905 Returns a list of the names of all the files in the archive.
907 If C<list_files()> is passed an array reference as its first argument
908 it returns a list of hash references containing the requested
909 properties of each file. The following list of properties is
910 supported: name, size, mtime (last modified date), mode, uid, gid,
911 linkname, uname, gname, devmajor, devminor, prefix.
913 Passing an array reference containing only one element, 'name', is
914 special cased to return a list of names rather than a list of hash
915 references, making it equivalent to calling C<list_files> without
922 my $aref = shift || [ ];
924 unless( $self->_data ) {
925 $self->read() or return;
928 if( @$aref == 0 or ( @$aref == 1 and $aref->[0] eq 'name' ) ) {
929 return map { $_->full_path } @{$self->_data};
933 #for my $obj ( @{$self->_data} ) {
934 # push @rv, { map { $_ => $obj->$_() } @$aref };
938 ### this does the same as the above.. just needs a +{ }
939 ### to make sure perl doesn't confuse it for a block
940 return map { my $o=$_;
941 +{ map { $_ => $o->$_() } @$aref }
950 unless( defined $file ) {
951 $self->_error( qq[No file specified] );
955 ### it's an object already
956 return $file if UNIVERSAL::isa( $file, 'Archive::Tar::File' );
958 for my $entry ( @{$self->_data} ) {
959 my $path = $entry->full_path;
960 return $entry if $path eq $file;
963 $self->_error( qq[No such file in archive: '$file'] );
967 =head2 $tar->get_files( [@filenames] )
969 Returns the C<Archive::Tar::File> objects matching the filenames
970 provided. If no filename list was passed, all C<Archive::Tar::File>
971 objects in the current Tar object are returned.
973 Please refer to the C<Archive::Tar::File> documentation on how to
974 handle these objects.
981 return @{ $self->_data } unless @_;
984 for my $file ( @_ ) {
985 push @list, grep { defined } $self->_find_entry( $file );
991 =head2 $tar->get_content( $file )
993 Return the content of the named file.
999 my $entry = $self->_find_entry( shift ) or return;
1001 return $entry->data;
1004 =head2 $tar->replace_content( $file, $content )
1006 Make the string $content be the content for the file named $file.
1010 sub replace_content {
1012 my $entry = $self->_find_entry( shift ) or return;
1014 return $entry->replace_content( shift );
1017 =head2 $tar->rename( $file, $new_name )
1019 Rename the file of the in-memory archive to $new_name.
1021 Note that you must specify a Unix path for $new_name, since per tar
1022 standard, all files in the archive must be Unix paths.
1024 Returns true on success and false on failure.
1030 my $file = shift; return unless defined $file;
1031 my $new = shift; return unless defined $new;
1033 my $entry = $self->_find_entry( $file ) or return;
1035 return $entry->rename( $new );
1038 =head2 $tar->remove (@filenamelist)
1040 Removes any entries with names matching any of the given filenames
1041 from the in-memory archive. Returns a list of C<Archive::Tar::File>
1042 objects that remain.
1050 my %seen = map { $_->full_path => $_ } @{$self->_data};
1051 delete $seen{ $_ } for @list;
1053 $self->_data( [values %seen] );
1055 return values %seen;
1060 C<clear> clears the current in-memory archive. This effectively gives
1061 you a 'blank' object, ready to be filled again. Note that C<clear>
1062 only has effect on the object, not the underlying tarfile.
1067 my $self = shift or return;
1076 =head2 $tar->write ( [$file, $compressed, $prefix] )
1078 Write the in-memory archive to disk. The first argument can either
1079 be the name of a file or a reference to an already open filehandle (a
1082 The second argument is used to indicate compression. You can either
1083 compress using C<gzip> or C<bzip2>. If you pass a digit, it's assumed
1084 to be the C<gzip> compression level (between 1 and 9), but the use of
1085 constants is prefered:
1087 # write a gzip compressed file
1088 $tar->write( 'out.tgz', COMPRESS_GZIP );
1090 # write a bzip compressed file
1091 $tar->write( 'out.tbz', COMPRESS_BZIP );
1093 Note that when you pass in a filehandle, the compression argument
1094 is ignored, as all files are printed verbatim to your filehandle.
1095 If you wish to enable compression with filehandles, use an
1096 C<IO::Zlib> or C<IO::Compress::Bzip2> filehandle instead.
1098 The third argument is an optional prefix. All files will be tucked
1099 away in the directory you specify as prefix. So if you have files
1100 'a' and 'b' in your archive, and you specify 'foo' as prefix, they
1101 will be written to the archive as 'foo/a' and 'foo/b'.
1103 If no arguments are given, C<write> returns the entire formatted
1104 archive as a string, which could be useful if you'd like to stuff the
1105 archive into a socket or a pipe to gzip or something.
1112 my $file = shift; $file = '' unless defined $file;
1113 my $gzip = shift || 0;
1114 my $ext_prefix = shift; $ext_prefix = '' unless defined $ext_prefix;
1117 ### only need a handle if we have a file to print to ###
1118 my $handle = length($file)
1119 ? ( $self->_get_handle($file, $gzip, WRITE_ONLY->($gzip) )
1121 : $HAS_PERLIO ? do { open my $h, '>', \$dummy; $h }
1122 : $HAS_IO_STRING ? IO::String->new
1123 : __PACKAGE__->no_string_support();
1125 ### Addresses: #41798: Nonempty $\ when writing a TAR file produces a
1126 ### corrupt TAR file. Must clear out $\ to make sure no garbage is
1127 ### printed to the archive
1130 for my $entry ( @{$self->_data} ) {
1131 ### entries to be written to the tarfile ###
1134 ### only now will we change the object to reflect the current state
1135 ### of the name and prefix fields -- this needs to be limited to
1137 my $clone = $entry->clone;
1140 ### so, if you don't want use to use the prefix, we'll stuff
1141 ### everything in the name field instead
1142 if( $DO_NOT_USE_PREFIX ) {
1144 ### you might have an extended prefix, if so, set it in the clone
1145 ### XXX is ::Unix right?
1146 $clone->name( length $ext_prefix
1147 ? File::Spec::Unix->catdir( $ext_prefix,
1149 : $clone->full_path );
1150 $clone->prefix( '' );
1152 ### otherwise, we'll have to set it properly -- prefix part in the
1153 ### prefix and name part in the name field.
1156 ### split them here, not before!
1157 my ($prefix,$name) = $clone->_prefix_and_file( $clone->full_path );
1159 ### you might have an extended prefix, if so, set it in the clone
1160 ### XXX is ::Unix right?
1161 $prefix = File::Spec::Unix->catdir( $ext_prefix, $prefix )
1162 if length $ext_prefix;
1164 $clone->prefix( $prefix );
1165 $clone->name( $name );
1168 ### names are too long, and will get truncated if we don't add a
1169 ### '@LongLink' file...
1170 my $make_longlink = ( length($clone->name) > NAME_LENGTH or
1171 length($clone->prefix) > PREFIX_LENGTH
1174 ### perhaps we need to make a longlink file?
1175 if( $make_longlink ) {
1176 my $longlink = Archive::Tar::File->new(
1177 data => LONGLINK_NAME,
1179 { type => LONGLINK }
1182 unless( $longlink ) {
1183 $self->_error( qq[Could not create 'LongLink' entry for ] .
1184 qq[oversize file '] . $clone->full_path ."'" );
1188 push @write_me, $longlink;
1191 push @write_me, $clone;
1193 ### write the one, optionally 2 a::t::file objects to the handle
1194 for my $clone (@write_me) {
1196 ### if the file is a symlink, there are 2 options:
1197 ### either we leave the symlink intact, but then we don't write any
1198 ### data OR we follow the symlink, which means we actually make a
1199 ### copy. if we do the latter, we have to change the TYPE of the
1201 my $link_ok = $clone->is_symlink && $Archive::Tar::FOLLOW_SYMLINK;
1202 my $data_ok = !$clone->is_symlink && $clone->has_content;
1204 ### downgrade to a 'normal' file if it's a symlink we're going to
1205 ### treat as a regular file
1206 $clone->_downgrade_to_plainfile if $link_ok;
1208 ### get the header for this block
1209 my $header = $self->_format_tar_entry( $clone );
1211 $self->_error(q[Could not format header for: ] .
1212 $clone->full_path );
1216 unless( print $handle $header ) {
1217 $self->_error(q[Could not write header for: ] .
1222 if( $link_ok or $data_ok ) {
1223 unless( print $handle $clone->data ) {
1224 $self->_error(q[Could not write data for: ] .
1229 ### pad the end of the clone if required ###
1230 print $handle TAR_PAD->( $clone->size ) if $clone->size % BLOCK
1233 } ### done writing these entries
1236 ### write the end markers ###
1237 print $handle TAR_END x 2 or
1238 return $self->_error( qq[Could not write tar end markers] );
1240 ### did you want it written to a file, or returned as a string? ###
1241 my $rv = length($file) ? 1
1242 : $HAS_PERLIO ? $dummy
1243 : do { seek $handle, 0, 0; local $/; <$handle> };
1245 ### make sure to close the handle;
1251 sub _format_tar_entry {
1253 my $entry = shift or return;
1254 my $ext_prefix = shift; $ext_prefix = '' unless defined $ext_prefix;
1255 my $no_prefix = shift || 0;
1257 my $file = $entry->name;
1258 my $prefix = $entry->prefix; $prefix = '' unless defined $prefix;
1260 ### remove the prefix from the file name
1261 ### not sure if this is still neeeded --kane
1262 ### no it's not -- Archive::Tar::File->_new_from_file will take care of
1263 ### this for us. Even worse, this would break if we tried to add a file
1265 #if( length $prefix ) {
1266 # $file =~ s/^$match//;
1269 $prefix = File::Spec::Unix->catdir($ext_prefix, $prefix)
1270 if length $ext_prefix;
1272 ### not sure why this is... ###
1273 my $l = PREFIX_LENGTH; # is ambiguous otherwise...
1274 substr ($prefix, 0, -$l) = "" if length $prefix >= PREFIX_LENGTH;
1276 my $f1 = "%06o"; my $f2 = "%11o";
1278 ### this might be optimizable with a 'changed' flag in the file objects ###
1283 (map { sprintf( $f1, $entry->$_() ) } qw[mode uid gid]),
1284 (map { sprintf( $f2, $entry->$_() ) } qw[size mtime]),
1286 "", # checksum field - space padded a bit down
1288 (map { $entry->$_() } qw[type linkname magic]),
1290 $entry->version || TAR_VERSION,
1292 (map { $entry->$_() } qw[uname gname]),
1293 (map { sprintf( $f1, $entry->$_() ) } qw[devmajor devminor]),
1295 ($no_prefix ? '' : $prefix)
1298 ### add the checksum ###
1299 substr($tar,148,7) = sprintf("%6o\0", unpack("%16C*",$tar));
1304 =head2 $tar->add_files( @filenamelist )
1306 Takes a list of filenames and adds them to the in-memory archive.
1308 The path to the file is automatically converted to a Unix like
1309 equivalent for use in the archive, and, if on MacOS, the file's
1310 modification time is converted from the MacOS epoch to the Unix epoch.
1311 So tar archives created on MacOS with B<Archive::Tar> can be read
1312 both with I<tar> on Unix and applications like I<suntar> or
1313 I<Stuffit Expander> on MacOS.
1315 Be aware that the file's type/creator and resource fork will be lost,
1316 which is usually what you want in cross-platform archives.
1318 Instead of a filename, you can also pass it an existing C<Archive::Tar::File>
1319 object from, for example, another archive. The object will be clone, and
1320 effectively be a copy of the original, not an alias.
1322 Returns a list of C<Archive::Tar::File> objects that were just added.
1328 my @files = @_ or return;
1331 for my $file ( @files ) {
1333 ### you passed an Archive::Tar::File object
1334 ### clone it so we don't accidentally have a reference to
1335 ### an object from another archive
1336 if( UNIVERSAL::isa( $file,'Archive::Tar::File' ) ) {
1337 push @rv, $file->clone;
1341 unless( -e $file || -l $file ) {
1342 $self->_error( qq[No such file: '$file'] );
1346 my $obj = Archive::Tar::File->new( file => $file );
1348 $self->_error( qq[Unable to add file: '$file'] );
1355 push @{$self->{_data}}, @rv;
1360 =head2 $tar->add_data ( $filename, $data, [$opthashref] )
1362 Takes a filename, a scalar full of data and optionally a reference to
1363 a hash with specific options.
1365 Will add a file to the in-memory archive, with name C<$filename> and
1366 content C<$data>. Specific properties can be set using C<$opthashref>.
1367 The following list of properties is supported: name, size, mtime
1368 (last modified date), mode, uid, gid, linkname, uname, gname,
1369 devmajor, devminor, prefix, type. (On MacOS, the file's path and
1370 modification times are converted to Unix equivalents.)
1372 Valid values for the file type are the following constants defined in
1373 Archive::Tar::Constants:
1385 Hard and symbolic ("soft") links; linkname should specify target.
1391 Character and block devices. devmajor and devminor should specify the major
1392 and minor device numbers.
1408 Returns the C<Archive::Tar::File> object that was just added, or
1409 C<undef> on failure.
1415 my ($file, $data, $opt) = @_;
1417 my $obj = Archive::Tar::File->new( data => $file, $data, $opt );
1419 $self->_error( qq[Unable to add file: '$file'] );
1423 push @{$self->{_data}}, $obj;
1428 =head2 $tar->error( [$BOOL] )
1430 Returns the current errorstring (usually, the last error reported).
1431 If a true value was specified, it will give the C<Carp::longmess>
1432 equivalent of the error, in effect giving you a stacktrace.
1434 For backwards compatibility, this error is also available as
1435 C<$Archive::Tar::error> although it is much recommended you use the
1436 method call instead.
1446 my $msg = $error = shift;
1447 $longmess = Carp::longmess($error);
1449 ### set Archive::Tar::WARN to 0 to disable printing
1452 carp $DEBUG ? $longmess : $msg;
1460 return shift() ? $longmess : $error;
1464 =head2 $tar->setcwd( $cwd );
1466 C<Archive::Tar> needs to know the current directory, and it will run
1467 C<Cwd::cwd()> I<every> time it extracts a I<relative> entry from the
1468 tarfile and saves it in the file system. (As of version 1.30, however,
1469 C<Archive::Tar> will use the speed optimization described below
1470 automatically, so it's only relevant if you're using C<extract_file()>).
1472 Since C<Archive::Tar> doesn't change the current directory internally
1473 while it is extracting the items in a tarball, all calls to C<Cwd::cwd()>
1474 can be avoided if we can guarantee that the current directory doesn't
1475 get changed externally.
1477 To use this performance boost, set the current directory via
1480 $tar->setcwd( cwd() );
1482 once before calling a function like C<extract_file> and
1483 C<Archive::Tar> will use the current directory setting from then on
1484 and won't call C<Cwd::cwd()> internally.
1486 To switch back to the default behaviour, use
1488 $tar->setcwd( undef );
1490 and C<Archive::Tar> will call C<Cwd::cwd()> internally again.
1492 If you're using C<Archive::Tar>'s C<exract()> method, C<setcwd()> will
1501 $self->{cwd} = $cwd;
1504 =head1 Class Methods
1506 =head2 Archive::Tar->create_archive($file, $compressed, @filelist)
1508 Creates a tar file from the list of files provided. The first
1509 argument can either be the name of the tar file to create or a
1510 reference to an open file handle (e.g. a GLOB reference).
1512 The second argument is used to indicate compression. You can either
1513 compress using C<gzip> or C<bzip2>. If you pass a digit, it's assumed
1514 to be the C<gzip> compression level (between 1 and 9), but the use of
1515 constants is prefered:
1517 # write a gzip compressed file
1518 Archive::Tar->create_archive( 'out.tgz', COMPRESS_GZIP, @filelist );
1520 # write a bzip compressed file
1521 Archive::Tar->create_archive( 'out.tbz', COMPRESS_BZIP, @filelist );
1523 Note that when you pass in a filehandle, the compression argument
1524 is ignored, as all files are printed verbatim to your filehandle.
1525 If you wish to enable compression with filehandles, use an
1526 C<IO::Zlib> or C<IO::Compress::Bzip2> filehandle instead.
1528 The remaining arguments list the files to be included in the tar file.
1529 These files must all exist. Any files which don't exist or can't be
1530 read are silently ignored.
1532 If the archive creation fails for any reason, C<create_archive> will
1533 return false. Please use the C<error> method to find the cause of the
1536 Note that this method does not write C<on the fly> as it were; it
1537 still reads all the files into memory before writing out the archive.
1538 Consult the FAQ below if this is a problem.
1542 sub create_archive {
1545 my $file = shift; return unless defined $file;
1546 my $gzip = shift || 0;
1550 return $class->_error( qq[Cowardly refusing to create empty archive!] );
1553 my $tar = $class->new;
1554 $tar->add_files( @files );
1555 return $tar->write( $file, $gzip );
1558 =head2 Archive::Tar->iter( $filename, [ $compressed, {opt => $val} ] )
1560 Returns an iterator function that reads the tar file without loading
1561 it all in memory. Each time the function is called it will return the
1562 next file in the tarball. The files are returned as
1563 C<Archive::Tar::File> objects. The iterator function returns the
1564 empty list once it has exhausted the the files contained.
1566 The second argument can be a hash reference with options, which are
1567 identical to the arguments passed to C<read()>.
1571 my $next = Archive::Tar->iter( "example.tar.gz", 1, {filter => qr/\.pm$/} );
1573 while( my $f = $next->() ) {
1574 print $f->name, "\n";
1576 $f->extract or warn "Extraction failed";
1586 my $filename = shift or return;
1587 my $compressed = shift or 0;
1588 my $opts = shift || {};
1590 ### get a handle to read from.
1591 my $handle = $class->_get_handle(
1599 return shift(@data) if @data; # more than one file returned?
1600 return unless $handle; # handle exhausted?
1602 ### read data, should only return file
1603 @data = @{ $class->_read_tar($handle, { %$opts, limit => 1 }) };
1605 ### return one piece of data
1606 return shift(@data) if @data;
1608 ### data is exhausted, free the filehandle
1614 =head2 Archive::Tar->list_archive($file, $compressed, [\@properties])
1616 Returns a list of the names of all the files in the archive. The
1617 first argument can either be the name of the tar file to list or a
1618 reference to an open file handle (e.g. a GLOB reference).
1620 If C<list_archive()> is passed an array reference as its third
1621 argument it returns a list of hash references containing the requested
1622 properties of each file. The following list of properties is
1623 supported: full_path, name, size, mtime (last modified date), mode,
1624 uid, gid, linkname, uname, gname, devmajor, devminor, prefix.
1626 See C<Archive::Tar::File> for details about supported properties.
1628 Passing an array reference containing only one element, 'name', is
1629 special cased to return a list of names rather than a list of hash
1636 my $file = shift; return unless defined $file;
1637 my $gzip = shift || 0;
1639 my $tar = $class->new($file, $gzip);
1642 return $tar->list_files( @_ );
1645 =head2 Archive::Tar->extract_archive($file, $compressed)
1647 Extracts the contents of the tar file. The first argument can either
1648 be the name of the tar file to create or a reference to an open file
1649 handle (e.g. a GLOB reference). All relative paths in the tar file will
1650 be created underneath the current working directory.
1652 C<extract_archive> will return a list of files it extracted.
1653 If the archive extraction fails for any reason, C<extract_archive>
1654 will return false. Please use the C<error> method to find the cause
1659 sub extract_archive {
1661 my $file = shift; return unless defined $file;
1662 my $gzip = shift || 0;
1664 my $tar = $class->new( ) or return;
1666 return $tar->read( $file, $gzip, { extract => 1 } );
1669 =head2 $bool = Archive::Tar->has_io_string
1671 Returns true if we currently have C<IO::String> support loaded.
1673 Either C<IO::String> or C<perlio> support is needed to support writing
1674 stringified archives. Currently, C<perlio> is the preferred method, if
1677 See the C<GLOBAL VARIABLES> section to see how to change this preference.
1681 sub has_io_string { return $HAS_IO_STRING; }
1683 =head2 $bool = Archive::Tar->has_perlio
1685 Returns true if we currently have C<perlio> support loaded.
1687 This requires C<perl-5.8> or higher, compiled with C<perlio>
1689 Either C<IO::String> or C<perlio> support is needed to support writing
1690 stringified archives. Currently, C<perlio> is the preferred method, if
1693 See the C<GLOBAL VARIABLES> section to see how to change this preference.
1697 sub has_perlio { return $HAS_PERLIO; }
1699 =head2 $bool = Archive::Tar->has_zlib_support
1701 Returns true if C<Archive::Tar> can extract C<zlib> compressed archives
1705 sub has_zlib_support { return ZLIB }
1707 =head2 $bool = Archive::Tar->has_bzip2_support
1709 Returns true if C<Archive::Tar> can extract C<bzip2> compressed archives
1713 sub has_bzip2_support { return BZIP }
1715 =head2 Archive::Tar->can_handle_compressed_files
1717 A simple checking routine, which will return true if C<Archive::Tar>
1718 is able to uncompress compressed archives on the fly with C<IO::Zlib>
1719 and C<IO::Compress::Bzip2> or false if not both are installed.
1721 You can use this as a shortcut to determine whether C<Archive::Tar>
1722 will do what you think before passing compressed archives to its
1727 sub can_handle_compressed_files { return ZLIB && BZIP ? 1 : 0 }
1729 sub no_string_support {
1730 croak("You have to install IO::String to support writing archives to strings");
1737 =head1 GLOBAL VARIABLES
1739 =head2 $Archive::Tar::FOLLOW_SYMLINK
1741 Set this variable to C<1> to make C<Archive::Tar> effectively make a
1742 copy of the file when extracting. Default is C<0>, which
1743 means the symlink stays intact. Of course, you will have to pack the
1744 file linked to as well.
1746 This option is checked when you write out the tarfile using C<write>
1747 or C<create_archive>.
1749 This works just like C</bin/tar>'s C<-h> option.
1751 =head2 $Archive::Tar::CHOWN
1753 By default, C<Archive::Tar> will try to C<chown> your files if it is
1754 able to. In some cases, this may not be desired. In that case, set
1755 this variable to C<0> to disable C<chown>-ing, even if it were
1758 The default is C<1>.
1760 =head2 $Archive::Tar::CHMOD
1762 By default, C<Archive::Tar> will try to C<chmod> your files to
1763 whatever mode was specified for the particular file in the archive.
1764 In some cases, this may not be desired. In that case, set this
1765 variable to C<0> to disable C<chmod>-ing.
1767 The default is C<1>.
1769 =head2 $Archive::Tar::SAME_PERMISSIONS
1771 When, C<$Archive::Tar::CHMOD> is enabled, this setting controls whether
1772 the permissions on files from the archive are used without modification
1773 of if they are filtered by removing any setid bits and applying the
1776 The default is C<1> for the root user and C<0> for normal users.
1778 =head2 $Archive::Tar::DO_NOT_USE_PREFIX
1780 By default, C<Archive::Tar> will try to put paths that are over
1781 100 characters in the C<prefix> field of your tar header, as
1782 defined per POSIX-standard. However, some (older) tar programs
1783 do not implement this spec. To retain compatibility with these older
1784 or non-POSIX compliant versions, you can set the C<$DO_NOT_USE_PREFIX>
1785 variable to a true value, and C<Archive::Tar> will use an alternate
1786 way of dealing with paths over 100 characters by using the
1787 C<GNU Extended Header> feature.
1789 Note that clients who do not support the C<GNU Extended Header>
1790 feature will not be able to read these archives. Such clients include
1791 tars on C<Solaris>, C<Irix> and C<AIX>.
1793 The default is C<0>.
1795 =head2 $Archive::Tar::DEBUG
1797 Set this variable to C<1> to always get the C<Carp::longmess> output
1798 of the warnings, instead of the regular C<carp>. This is the same
1799 message you would get by doing:
1805 =head2 $Archive::Tar::WARN
1807 Set this variable to C<0> if you do not want any warnings printed.
1808 Personally I recommend against doing this, but people asked for the
1809 option. Also, be advised that this is of course not threadsafe.
1813 =head2 $Archive::Tar::error
1815 Holds the last reported error. Kept for historical reasons, but its
1816 use is very much discouraged. Use the C<error()> method instead:
1818 warn $tar->error unless $tar->extract;
1820 =head2 $Archive::Tar::INSECURE_EXTRACT_MODE
1822 This variable indicates whether C<Archive::Tar> should allow
1823 files to be extracted outside their current working directory.
1825 Allowing this could have security implications, as a malicious
1826 tar archive could alter or replace any file the extracting user
1827 has permissions to. Therefor, the default is to not allow
1828 insecure extractions.
1830 If you trust the archive, or have other reasons to allow the
1831 archive to write files outside your current working directory,
1832 set this variable to C<true>.
1834 Note that this is a backwards incompatible change from version
1837 =head2 $Archive::Tar::HAS_PERLIO
1839 This variable holds a boolean indicating if we currently have
1840 C<perlio> support loaded. This will be enabled for any perl
1841 greater than C<5.8> compiled with C<perlio>.
1843 If you feel strongly about disabling it, set this variable to
1844 C<false>. Note that you will then need C<IO::String> installed
1845 to support writing stringified archives.
1847 Don't change this variable unless you B<really> know what you're
1850 =head2 $Archive::Tar::HAS_IO_STRING
1852 This variable holds a boolean indicating if we currently have
1853 C<IO::String> support loaded. This will be enabled for any perl
1854 that has a loadable C<IO::String> module.
1856 If you feel strongly about disabling it, set this variable to
1857 C<false>. Note that you will then need C<perlio> support from
1858 your perl to be able to write stringified archives.
1860 Don't change this variable unless you B<really> know what you're
1867 =item What's the minimum perl version required to run Archive::Tar?
1869 You will need perl version 5.005_03 or newer.
1871 =item Isn't Archive::Tar slow?
1873 Yes it is. It's pure perl, so it's a lot slower then your C</bin/tar>
1874 However, it's very portable. If speed is an issue, consider using
1875 C</bin/tar> instead.
1877 =item Isn't Archive::Tar heavier on memory than /bin/tar?
1879 Yes it is, see previous answer. Since C<Compress::Zlib> and therefore
1880 C<IO::Zlib> doesn't support C<seek> on their filehandles, there is little
1881 choice but to read the archive into memory.
1882 This is ok if you want to do in-memory manipulation of the archive.
1884 If you just want to extract, use the C<extract_archive> class method
1885 instead. It will optimize and write to disk immediately.
1887 Another option is to use the C<iter> class method to iterate over
1888 the files in the tarball without reading them all in memory at once.
1890 =item Can you lazy-load data instead?
1892 In some cases, yes. You can use the C<iter> class method to iterate
1893 over the files in the tarball without reading them all in memory at once.
1895 =item How much memory will an X kb tar file need?
1897 Probably more than X kb, since it will all be read into memory. If
1898 this is a problem, and you don't need to do in memory manipulation
1899 of the archive, consider using the C<iter> class method, or C</bin/tar>
1902 =item What do you do with unsupported filetypes in an archive?
1904 C<Unix> has a few filetypes that aren't supported on other platforms,
1905 like C<Win32>. If we encounter a C<hardlink> or C<symlink> we'll just
1906 try to make a copy of the original file, rather than throwing an error.
1908 This does require you to read the entire archive in to memory first,
1909 since otherwise we wouldn't know what data to fill the copy with.
1910 (This means that you cannot use the class methods, including C<iter>
1911 on archives that have incompatible filetypes and still expect things
1914 For other filetypes, like C<chardevs> and C<blockdevs> we'll warn that
1915 the extraction of this particular item didn't work.
1917 =item I'm using WinZip, or some other non-POSIX client, and files are not being extracted properly!
1919 By default, C<Archive::Tar> is in a completely POSIX-compatible
1920 mode, which uses the POSIX-specification of C<tar> to store files.
1921 For paths greather than 100 characters, this is done using the
1922 C<POSIX header prefix>. Non-POSIX-compatible clients may not support
1923 this part of the specification, and may only support the C<GNU Extended
1924 Header> functionality. To facilitate those clients, you can set the
1925 C<$Archive::Tar::DO_NOT_USE_PREFIX> variable to C<true>. See the
1926 C<GLOBAL VARIABLES> section for details on this variable.
1928 Note that GNU tar earlier than version 1.14 does not cope well with
1929 the C<POSIX header prefix>. If you use such a version, consider setting
1930 the C<$Archive::Tar::DO_NOT_USE_PREFIX> variable to C<true>.
1932 =item How do I extract only files that have property X from an archive?
1934 Sometimes, you might not wish to extract a complete archive, just
1935 the files that are relevant to you, based on some criteria.
1937 You can do this by filtering a list of C<Archive::Tar::File> objects
1938 based on your criteria. For example, to extract only files that have
1939 the string C<foo> in their title, you would use:
1942 grep { $_->full_path =~ /foo/ } $tar->get_files
1945 This way, you can filter on any attribute of the files in the archive.
1946 Consult the C<Archive::Tar::File> documentation on how to use these
1949 =item How do I access .tar.Z files?
1951 The C<Archive::Tar> module can optionally use C<Compress::Zlib> (via
1952 the C<IO::Zlib> module) to access tar files that have been compressed
1953 with C<gzip>. Unfortunately tar files compressed with the Unix C<compress>
1954 utility cannot be read by C<Compress::Zlib> and so cannot be directly
1955 accesses by C<Archive::Tar>.
1957 If the C<uncompress> or C<gunzip> programs are available, you can use
1958 one of these workarounds to read C<.tar.Z> files from C<Archive::Tar>
1960 Firstly with C<uncompress>
1964 open F, "uncompress -c $filename |";
1965 my $tar = Archive::Tar->new(*F);
1968 and this with C<gunzip>
1972 open F, "gunzip -c $filename |";
1973 my $tar = Archive::Tar->new(*F);
1976 Similarly, if the C<compress> program is available, you can use this to
1977 write a C<.tar.Z> file
1982 my $fh = new IO::File "| compress -c >$filename";
1983 my $tar = Archive::Tar->new();
1988 =item How do I handle Unicode strings?
1990 C<Archive::Tar> uses byte semantics for any files it reads from or writes
1991 to disk. This is not a problem if you only deal with files and never
1992 look at their content or work solely with byte strings. But if you use
1993 Unicode strings with character semantics, some additional steps need
1996 For example, if you add a Unicode string like
1999 $tar->add_data('file.txt', "Euro: \x{20AC}");
2001 then there will be a problem later when the tarfile gets written out
2002 to disk via C<$tar->write()>:
2004 Wide character in print at .../Archive/Tar.pm line 1014.
2006 The data was added as a Unicode string and when writing it out to disk,
2007 the C<:utf8> line discipline wasn't set by C<Archive::Tar>, so Perl
2008 tried to convert the string to ISO-8859 and failed. The written file
2009 now contains garbage.
2011 For this reason, Unicode strings need to be converted to UTF-8-encoded
2012 bytestrings before they are handed off to C<add_data()>:
2015 my $data = "Accented character: \x{20AC}";
2016 $data = encode('utf8', $data);
2018 $tar->add_data('file.txt', $data);
2020 A opposite problem occurs if you extract a UTF8-encoded file from a
2021 tarball. Using C<get_content()> on the C<Archive::Tar::File> object
2022 will return its content as a bytestring, not as a Unicode string.
2024 If you want it to be a Unicode string (because you want character
2025 semantics with operations like regular expression matching), you need
2026 to decode the UTF8-encoded content and have Perl convert it into
2030 my $data = $tar->get_content();
2032 # Make it a Unicode string
2033 $data = decode('utf8', $data);
2035 There is no easy way to provide this functionality in C<Archive::Tar>,
2036 because a tarball can contain many files, and each of which could be
2037 encoded in a different way.
2043 The AIX tar does not fill all unused space in the tar archive with 0x00.
2044 This sometimes leads to warning messages from C<Archive::Tar>.
2046 Invalid header block at offset nnn
2048 A fix for that problem is scheduled to be released in the following levels
2049 of AIX, all of which should be coming out in the 4th quarter of 2009:
2061 The IBM APAR number for this problem is IZ50240 (Reported component ID:
2062 5765G0300 / AIX 5.3). It is possible to get an ifix for that problem.
2063 If you need an ifix please contact your local IBM AIX support.
2069 =item Check if passed in handles are open for read/write
2071 Currently I don't know of any portable pure perl way to do this.
2072 Suggestions welcome.
2074 =item Allow archives to be passed in as string
2076 Currently, we only allow opened filehandles or filenames, but
2077 not strings. The internals would need some reworking to facilitate
2078 stringified archives.
2080 =item Facilitate processing an opened filehandle of a compressed archive
2082 Currently, we only support this if the filehandle is an IO::Zlib object.
2083 Environments, like apache, will present you with an opened filehandle
2084 to an uploaded file, which might be a compressed archive.
2092 =item The GNU tar specification
2094 C<http://www.gnu.org/software/tar/manual/tar.html>
2096 =item The PAX format specication
2098 The specifcation which tar derives from; C< http://www.opengroup.org/onlinepubs/007904975/utilities/pax.html>
2100 =item A comparison of GNU and POSIX tar standards; C<http://www.delorie.com/gnu/docs/tar/tar_114.html>
2102 =item GNU tar intends to switch to POSIX compatibility
2104 GNU Tar authors have expressed their intention to become completely
2105 POSIX-compatible; C<http://www.gnu.org/software/tar/manual/html_node/Formats.html>
2107 =item A Comparison between various tar implementations
2109 Lists known issues and incompatibilities; C<http://gd.tuwien.ac.at/utils/archivers/star/README.otherbugs>
2115 This module by Jos Boumans E<lt>kane@cpan.orgE<gt>.
2117 Please reports bugs to E<lt>bug-archive-tar@rt.cpan.orgE<gt>.
2119 =head1 ACKNOWLEDGEMENTS
2121 Thanks to Sean Burke, Chris Nandor, Chip Salzenberg, Tim Heaney, Gisle Aas,
2122 Rainer Tammer and especially Andrew Savige for their help and suggestions.
2126 This module is copyright (c) 2002 - 2009 Jos Boumans
2127 E<lt>kane@cpan.orgE<gt>. All rights reserved.
2129 This library is free software; you may redistribute and/or modify
2130 it under the same terms as Perl itself.