ÿØÿà JFIF    ÿÛ „ ( %!1!%*+...983,7(-.- Zip/StringMember.pm000064400000003264152347641360010262 0ustar00package Archive::Zip::StringMember; use strict; use vars qw( $VERSION @ISA ); BEGIN { $VERSION = '1.60'; @ISA = qw( Archive::Zip::Member ); } use Archive::Zip qw( :CONSTANTS :ERROR_CODES ); # Create a new string member. Default is COMPRESSION_STORED. # Can take a ref to a string as well. sub _newFromString { my $class = shift; my $string = shift; my $name = shift; my $self = $class->new(@_); $self->contents($string); $self->fileName($name) if defined($name); # Set the file date to now $self->setLastModFileDateTimeFromUnix(time()); $self->unixFileAttributes($self->DEFAULT_FILE_PERMISSIONS); return $self; } sub _become { my $self = shift; my $newClass = shift; return $self if ref($self) eq $newClass; delete($self->{'contents'}); return $self->SUPER::_become($newClass); } # Get or set my contents. Note that we do not call the superclass # version of this, because it calls us. sub contents { my $self = shift; my $string = shift; if (defined($string)) { $self->{'contents'} = pack('C0a*', (ref($string) eq 'SCALAR') ? $$string : $string); $self->{'uncompressedSize'} = $self->{'compressedSize'} = length($self->{'contents'}); $self->{'compressionMethod'} = COMPRESSION_STORED; } return $self->{'contents'}; } # Return bytes read. Note that first parameter is a ref to a buffer. # my $data; # my ( $bytesRead, $status) = $self->readRawChunk( \$data, $chunkSize ); sub _readRawChunk { my ($self, $dataRef, $chunkSize) = @_; $$dataRef = substr($self->contents(), $self->_readOffset(), $chunkSize); return (length($$dataRef), AZ_OK); } 1; Zip/Archive.pm000064400000077125152347641370007255 0ustar00package Archive::Zip::Archive; # Represents a generic ZIP archive use strict; use File::Path; use File::Find (); use File::Spec (); use File::Copy (); use File::Basename; use Cwd; use Encode qw(encode_utf8 decode_utf8); use vars qw( $VERSION @ISA ); BEGIN { $VERSION = '1.60'; @ISA = qw( Archive::Zip ); } use Archive::Zip qw( :CONSTANTS :ERROR_CODES :PKZIP_CONSTANTS :UTILITY_METHODS ); our $UNICODE; # Note that this returns undef on read errors, else new zip object. sub new { my $class = shift; my $self = bless( { 'diskNumber' => 0, 'diskNumberWithStartOfCentralDirectory' => 0, 'numberOfCentralDirectoriesOnThisDisk' => 0, # should be # of members 'numberOfCentralDirectories' => 0, # should be # of members 'centralDirectorySize' => 0, # must re-compute on write 'centralDirectoryOffsetWRTStartingDiskNumber' => 0, # must re-compute 'writeEOCDOffset' => 0, 'writeCentralDirectoryOffset' => 0, 'zipfileComment' => '', 'eocdOffset' => 0, 'fileName' => '' }, $class ); $self->{'members'} = []; my $fileName = (ref($_[0]) eq 'HASH') ? shift->{filename} : shift; if ($fileName) { my $status = $self->read($fileName); return $status == AZ_OK ? $self : undef; } return $self; } sub storeSymbolicLink { my $self = shift; $self->{'storeSymbolicLink'} = shift; } sub members { @{shift->{'members'}}; } sub numberOfMembers { scalar(shift->members()); } sub memberNames { my $self = shift; return map { $_->fileName() } $self->members(); } # return ref to member with given name or undef sub memberNamed { my $self = shift; my $fileName = (ref($_[0]) eq 'HASH') ? shift->{zipName} : shift; foreach my $member ($self->members()) { return $member if $member->fileName() eq $fileName; } return undef; } sub membersMatching { my $self = shift; my $pattern = (ref($_[0]) eq 'HASH') ? shift->{regex} : shift; return grep { $_->fileName() =~ /$pattern/ } $self->members(); } sub diskNumber { shift->{'diskNumber'}; } sub diskNumberWithStartOfCentralDirectory { shift->{'diskNumberWithStartOfCentralDirectory'}; } sub numberOfCentralDirectoriesOnThisDisk { shift->{'numberOfCentralDirectoriesOnThisDisk'}; } sub numberOfCentralDirectories { shift->{'numberOfCentralDirectories'}; } sub centralDirectorySize { shift->{'centralDirectorySize'}; } sub centralDirectoryOffsetWRTStartingDiskNumber { shift->{'centralDirectoryOffsetWRTStartingDiskNumber'}; } sub zipfileComment { my $self = shift; my $comment = $self->{'zipfileComment'}; if (@_) { my $new_comment = (ref($_[0]) eq 'HASH') ? shift->{comment} : shift; $self->{'zipfileComment'} = pack('C0a*', $new_comment); # avoid Unicode } return $comment; } sub eocdOffset { shift->{'eocdOffset'}; } # Return the name of the file last read. sub fileName { shift->{'fileName'}; } sub removeMember { my $self = shift; my $member = (ref($_[0]) eq 'HASH') ? shift->{memberOrZipName} : shift; $member = $self->memberNamed($member) unless ref($member); return undef unless $member; my @newMembers = grep { $_ != $member } $self->members(); $self->{'members'} = \@newMembers; return $member; } sub replaceMember { my $self = shift; my ($oldMember, $newMember); if (ref($_[0]) eq 'HASH') { $oldMember = $_[0]->{memberOrZipName}; $newMember = $_[0]->{newMember}; } else { ($oldMember, $newMember) = @_; } $oldMember = $self->memberNamed($oldMember) unless ref($oldMember); return undef unless $oldMember; return undef unless $newMember; my @newMembers = map { ($_ == $oldMember) ? $newMember : $_ } $self->members(); $self->{'members'} = \@newMembers; return $oldMember; } sub extractMember { my $self = shift; my ($member, $name); if (ref($_[0]) eq 'HASH') { $member = $_[0]->{memberOrZipName}; $name = $_[0]->{name}; } else { ($member, $name) = @_; } $member = $self->memberNamed($member) unless ref($member); return _error('member not found') unless $member; my $originalSize = $member->compressedSize(); my ($volumeName, $dirName, $fileName); if (defined($name)) { ($volumeName, $dirName, $fileName) = File::Spec->splitpath($name); $dirName = File::Spec->catpath($volumeName, $dirName, ''); } else { $name = $member->fileName(); if ((my $ret = _extractionNameIsSafe($name)) != AZ_OK) { return $ret; } ($dirName = $name) =~ s{[^/]*$}{}; $dirName = Archive::Zip::_asLocalName($dirName); $name = Archive::Zip::_asLocalName($name); } if ($dirName && !-d $dirName) { mkpath($dirName); return _ioError("can't create dir $dirName") if (!-d $dirName); } my $rc = $member->extractToFileNamed($name, @_); # TODO refactor this fix into extractToFileNamed() $member->{'compressedSize'} = $originalSize; return $rc; } sub extractMemberWithoutPaths { my $self = shift; my ($member, $name); if (ref($_[0]) eq 'HASH') { $member = $_[0]->{memberOrZipName}; $name = $_[0]->{name}; } else { ($member, $name) = @_; } $member = $self->memberNamed($member) unless ref($member); return _error('member not found') unless $member; my $originalSize = $member->compressedSize(); return AZ_OK if $member->isDirectory(); unless ($name) { $name = $member->fileName(); $name =~ s{.*/}{}; # strip off directories, if any if ((my $ret = _extractionNameIsSafe($name)) != AZ_OK) { return $ret; } $name = Archive::Zip::_asLocalName($name); } my $rc = $member->extractToFileNamed($name, @_); $member->{'compressedSize'} = $originalSize; return $rc; } sub addMember { my $self = shift; my $newMember = (ref($_[0]) eq 'HASH') ? shift->{member} : shift; push(@{$self->{'members'}}, $newMember) if $newMember; if($newMember && ($newMember->{bitFlag} & 0x800) && !utf8::is_utf8($newMember->{fileName})){ $newMember->{fileName} = Encode::decode_utf8( $newMember->{fileName} ); } return $newMember; } sub addFile { my $self = shift; my ($fileName, $newName, $compressionLevel); if (ref($_[0]) eq 'HASH') { $fileName = $_[0]->{filename}; $newName = $_[0]->{zipName}; $compressionLevel = $_[0]->{compressionLevel}; } else { ($fileName, $newName, $compressionLevel) = @_; } if ($^O eq 'MSWin32' && $Archive::Zip::UNICODE) { $fileName = Win32::GetANSIPathName($fileName); } my $newMember = Archive::Zip::Member->newFromFile($fileName, $newName); $newMember->desiredCompressionLevel($compressionLevel); if ($self->{'storeSymbolicLink'} && -l $fileName) { my $newMember = Archive::Zip::Member->newFromString(readlink $fileName, $newName); # For symbolic links, External File Attribute is set to 0xA1FF0000 by Info-ZIP $newMember->{'externalFileAttributes'} = 0xA1FF0000; $self->addMember($newMember); } else { $self->addMember($newMember); } return $newMember; } sub addString { my $self = shift; my ($stringOrStringRef, $name, $compressionLevel); if (ref($_[0]) eq 'HASH') { $stringOrStringRef = $_[0]->{string}; $name = $_[0]->{zipName}; $compressionLevel = $_[0]->{compressionLevel}; } else { ($stringOrStringRef, $name, $compressionLevel) = @_; } my $newMember = Archive::Zip::Member->newFromString($stringOrStringRef, $name); $newMember->desiredCompressionLevel($compressionLevel); return $self->addMember($newMember); } sub addDirectory { my $self = shift; my ($name, $newName); if (ref($_[0]) eq 'HASH') { $name = $_[0]->{directoryName}; $newName = $_[0]->{zipName}; } else { ($name, $newName) = @_; } if ($^O eq 'MSWin32' && $Archive::Zip::UNICODE) { $name = Win32::GetANSIPathName($name); } my $newMember = Archive::Zip::Member->newDirectoryNamed($name, $newName); if ($self->{'storeSymbolicLink'} && -l $name) { my $link = readlink $name; ($newName =~ s{/$}{}) if $newName; # Strip trailing / my $newMember = Archive::Zip::Member->newFromString($link, $newName); # For symbolic links, External File Attribute is set to 0xA1FF0000 by Info-ZIP $newMember->{'externalFileAttributes'} = 0xA1FF0000; $self->addMember($newMember); } else { $self->addMember($newMember); } return $newMember; } # add either a file or a directory. sub addFileOrDirectory { my $self = shift; my ($name, $newName, $compressionLevel); if (ref($_[0]) eq 'HASH') { $name = $_[0]->{name}; $newName = $_[0]->{zipName}; $compressionLevel = $_[0]->{compressionLevel}; } else { ($name, $newName, $compressionLevel) = @_; } if ($^O eq 'MSWin32' && $Archive::Zip::UNICODE) { $name = Win32::GetANSIPathName($name); } $name =~ s{/$}{}; if ($newName) { $newName =~ s{/$}{}; } else { $newName = $name; } if (-f $name) { return $self->addFile($name, $newName, $compressionLevel); } elsif (-d $name) { return $self->addDirectory($name, $newName); } else { return _error("$name is neither a file nor a directory"); } } sub contents { my $self = shift; my ($member, $newContents); if (ref($_[0]) eq 'HASH') { $member = $_[0]->{memberOrZipName}; $newContents = $_[0]->{contents}; } else { ($member, $newContents) = @_; } return _error('No member name given') unless $member; $member = $self->memberNamed($member) unless ref($member); return undef unless $member; return $member->contents($newContents); } sub writeToFileNamed { my $self = shift; my $fileName = (ref($_[0]) eq 'HASH') ? shift->{filename} : shift; # local FS format foreach my $member ($self->members()) { if ($member->_usesFileNamed($fileName)) { return _error("$fileName is needed by member " . $member->fileName() . "; consider using overwrite() or overwriteAs() instead."); } } my ($status, $fh) = _newFileHandle($fileName, 'w'); return _ioError("Can't open $fileName for write") unless $status; my $retval = $self->writeToFileHandle($fh, 1); $fh->close(); $fh = undef; return $retval; } # It is possible to write data to the FH before calling this, # perhaps to make a self-extracting archive. sub writeToFileHandle { my $self = shift; my ($fh, $fhIsSeekable); if (ref($_[0]) eq 'HASH') { $fh = $_[0]->{fileHandle}; $fhIsSeekable = exists($_[0]->{seek}) ? $_[0]->{seek} : _isSeekable($fh); } else { $fh = shift; $fhIsSeekable = @_ ? shift : _isSeekable($fh); } return _error('No filehandle given') unless $fh; return _ioError('filehandle not open') unless $fh->opened(); _binmode($fh); # Find out where the current position is. my $offset = $fhIsSeekable ? $fh->tell() : 0; $offset = 0 if $offset < 0; foreach my $member ($self->members()) { my $retval = $member->_writeToFileHandle($fh, $fhIsSeekable, $offset); $member->endRead(); return $retval if $retval != AZ_OK; $offset += $member->_localHeaderSize() + $member->_writeOffset(); $offset += $member->hasDataDescriptor() ? DATA_DESCRIPTOR_LENGTH + SIGNATURE_LENGTH : 0; # changed this so it reflects the last successful position $self->{'writeCentralDirectoryOffset'} = $offset; } return $self->writeCentralDirectory($fh); } # Write zip back to the original file, # as safely as possible. # Returns AZ_OK if successful. sub overwrite { my $self = shift; return $self->overwriteAs($self->{'fileName'}); } # Write zip to the specified file, # as safely as possible. # Returns AZ_OK if successful. sub overwriteAs { my $self = shift; my $zipName = (ref($_[0]) eq 'HASH') ? $_[0]->{filename} : shift; return _error("no filename in overwriteAs()") unless defined($zipName); my ($fh, $tempName) = Archive::Zip::tempFile(); return _error("Can't open temp file", $!) unless $fh; (my $backupName = $zipName) =~ s{(\.[^.]*)?$}{.zbk}; my $status = $self->writeToFileHandle($fh); $fh->close(); $fh = undef; if ($status != AZ_OK) { unlink($tempName); _printError("Can't write to $tempName"); return $status; } my $err; # rename the zip if (-f $zipName && !rename($zipName, $backupName)) { $err = $!; unlink($tempName); return _error("Can't rename $zipName as $backupName", $err); } # move the temp to the original name (possibly copying) unless (File::Copy::move($tempName, $zipName) || File::Copy::copy($tempName, $zipName)) { $err = $!; rename($backupName, $zipName); unlink($tempName); return _error("Can't move $tempName to $zipName", $err); } # unlink the backup if (-f $backupName && !unlink($backupName)) { $err = $!; return _error("Can't unlink $backupName", $err); } return AZ_OK; } # Used only during writing sub _writeCentralDirectoryOffset { shift->{'writeCentralDirectoryOffset'}; } sub _writeEOCDOffset { shift->{'writeEOCDOffset'}; } # Expects to have _writeEOCDOffset() set sub _writeEndOfCentralDirectory { my ($self, $fh) = @_; $self->_print($fh, END_OF_CENTRAL_DIRECTORY_SIGNATURE_STRING) or return _ioError('writing EOCD Signature'); my $zipfileCommentLength = length($self->zipfileComment()); my $header = pack( END_OF_CENTRAL_DIRECTORY_FORMAT, 0, # {'diskNumber'}, 0, # {'diskNumberWithStartOfCentralDirectory'}, $self->numberOfMembers(), # {'numberOfCentralDirectoriesOnThisDisk'}, $self->numberOfMembers(), # {'numberOfCentralDirectories'}, $self->_writeEOCDOffset() - $self->_writeCentralDirectoryOffset(), $self->_writeCentralDirectoryOffset(), $zipfileCommentLength ); $self->_print($fh, $header) or return _ioError('writing EOCD header'); if ($zipfileCommentLength) { $self->_print($fh, $self->zipfileComment()) or return _ioError('writing zipfile comment'); } return AZ_OK; } # $offset can be specified to truncate a zip file. sub writeCentralDirectory { my $self = shift; my ($fh, $offset); if (ref($_[0]) eq 'HASH') { $fh = $_[0]->{fileHandle}; $offset = $_[0]->{offset}; } else { ($fh, $offset) = @_; } if (defined($offset)) { $self->{'writeCentralDirectoryOffset'} = $offset; $fh->seek($offset, IO::Seekable::SEEK_SET) or return _ioError('seeking to write central directory'); } else { $offset = $self->_writeCentralDirectoryOffset(); } foreach my $member ($self->members()) { my $status = $member->_writeCentralDirectoryFileHeader($fh); return $status if $status != AZ_OK; $offset += $member->_centralDirectoryHeaderSize(); $self->{'writeEOCDOffset'} = $offset; } return $self->_writeEndOfCentralDirectory($fh); } sub read { my $self = shift; my $fileName = (ref($_[0]) eq 'HASH') ? shift->{filename} : shift; return _error('No filename given') unless $fileName; my ($status, $fh) = _newFileHandle($fileName, 'r'); return _ioError("opening $fileName for read") unless $status; $status = $self->readFromFileHandle($fh, $fileName); return $status if $status != AZ_OK; $fh->close(); $self->{'fileName'} = $fileName; return AZ_OK; } sub readFromFileHandle { my $self = shift; my ($fh, $fileName); if (ref($_[0]) eq 'HASH') { $fh = $_[0]->{fileHandle}; $fileName = $_[0]->{filename}; } else { ($fh, $fileName) = @_; } $fileName = $fh unless defined($fileName); return _error('No filehandle given') unless $fh; return _ioError('filehandle not open') unless $fh->opened(); _binmode($fh); $self->{'fileName'} = "$fh"; # TODO: how to support non-seekable zips? return _error('file not seekable') unless _isSeekable($fh); $fh->seek(0, 0); # rewind the file my $status = $self->_findEndOfCentralDirectory($fh); return $status if $status != AZ_OK; my $eocdPosition = $fh->tell(); $status = $self->_readEndOfCentralDirectory($fh); return $status if $status != AZ_OK; $fh->seek($eocdPosition - $self->centralDirectorySize(), IO::Seekable::SEEK_SET) or return _ioError("Can't seek $fileName"); # Try to detect garbage at beginning of archives # This should be 0 $self->{'eocdOffset'} = $eocdPosition - $self->centralDirectorySize() # here - $self->centralDirectoryOffsetWRTStartingDiskNumber(); for (; ;) { my $newMember = Archive::Zip::Member->_newFromZipFile($fh, $fileName, $self->eocdOffset()); my $signature; ($status, $signature) = _readSignature($fh, $fileName); return $status if $status != AZ_OK; last if $signature == END_OF_CENTRAL_DIRECTORY_SIGNATURE; $status = $newMember->_readCentralDirectoryFileHeader(); return $status if $status != AZ_OK; $status = $newMember->endRead(); return $status if $status != AZ_OK; $newMember->_becomeDirectoryIfNecessary(); if(($newMember->{bitFlag} & 0x800) && !utf8::is_utf8($newMember->{fileName})){ $newMember->{fileName} = Encode::decode_utf8($newMember->{fileName}); } push(@{$self->{'members'}}, $newMember); } return AZ_OK; } # Read EOCD, starting from position before signature. # Return AZ_OK on success. sub _readEndOfCentralDirectory { my $self = shift; my $fh = shift; # Skip past signature $fh->seek(SIGNATURE_LENGTH, IO::Seekable::SEEK_CUR) or return _ioError("Can't seek past EOCD signature"); my $header = ''; my $bytesRead = $fh->read($header, END_OF_CENTRAL_DIRECTORY_LENGTH); if ($bytesRead != END_OF_CENTRAL_DIRECTORY_LENGTH) { return _ioError("reading end of central directory"); } my $zipfileCommentLength; ( $self->{'diskNumber'}, $self->{'diskNumberWithStartOfCentralDirectory'}, $self->{'numberOfCentralDirectoriesOnThisDisk'}, $self->{'numberOfCentralDirectories'}, $self->{'centralDirectorySize'}, $self->{'centralDirectoryOffsetWRTStartingDiskNumber'}, $zipfileCommentLength ) = unpack(END_OF_CENTRAL_DIRECTORY_FORMAT, $header); if ($self->{'diskNumber'} == 0xFFFF || $self->{'diskNumberWithStartOfCentralDirectory'} == 0xFFFF || $self->{'numberOfCentralDirectoriesOnThisDisk'} == 0xFFFF || $self->{'numberOfCentralDirectories'} == 0xFFFF || $self->{'centralDirectorySize'} == 0xFFFFFFFF || $self->{'centralDirectoryOffsetWRTStartingDiskNumber'} == 0xFFFFFFFF) { return _formatError("zip64 not supported" . Dumper($self)); } use Data::Dumper; if ($zipfileCommentLength) { my $zipfileComment = ''; $bytesRead = $fh->read($zipfileComment, $zipfileCommentLength); if ($bytesRead != $zipfileCommentLength) { return _ioError("reading zipfile comment"); } $self->{'zipfileComment'} = $zipfileComment; } return AZ_OK; } # Seek in my file to the end, then read backwards until we find the # signature of the central directory record. Leave the file positioned right # before the signature. Returns AZ_OK if success. sub _findEndOfCentralDirectory { my $self = shift; my $fh = shift; my $data = ''; $fh->seek(0, IO::Seekable::SEEK_END) or return _ioError("seeking to end"); my $fileLength = $fh->tell(); if ($fileLength < END_OF_CENTRAL_DIRECTORY_LENGTH + 4) { return _formatError("file is too short"); } my $seekOffset = 0; my $pos = -1; for (; ;) { $seekOffset += 512; $seekOffset = $fileLength if ($seekOffset > $fileLength); $fh->seek(-$seekOffset, IO::Seekable::SEEK_END) or return _ioError("seek failed"); my $bytesRead = $fh->read($data, $seekOffset); if ($bytesRead != $seekOffset) { return _ioError("read failed"); } $pos = rindex($data, END_OF_CENTRAL_DIRECTORY_SIGNATURE_STRING); last if ( $pos >= 0 or $seekOffset == $fileLength or $seekOffset >= $Archive::Zip::ChunkSize); } if ($pos >= 0) { $fh->seek($pos - $seekOffset, IO::Seekable::SEEK_CUR) or return _ioError("seeking to EOCD"); return AZ_OK; } else { return _formatError("can't find EOCD signature"); } } # Used to avoid taint problems when chdir'ing. # Not intended to increase security in any way; just intended to shut up the -T # complaints. If your Cwd module is giving you unreliable returns from cwd() # you have bigger problems than this. sub _untaintDir { my $dir = shift; $dir =~ m/\A(.+)\z/s; return $1; } sub addTree { my $self = shift; my ($root, $dest, $pred, $compressionLevel); if (ref($_[0]) eq 'HASH') { $root = $_[0]->{root}; $dest = $_[0]->{zipName}; $pred = $_[0]->{select}; $compressionLevel = $_[0]->{compressionLevel}; } else { ($root, $dest, $pred, $compressionLevel) = @_; } return _error("root arg missing in call to addTree()") unless defined($root); $dest = '' unless defined($dest); $pred = sub { -r } unless defined($pred); my @files; my $startDir = _untaintDir(cwd()); return _error('undef returned by _untaintDir on cwd ', cwd()) unless $startDir; # This avoids chdir'ing in Find, in a way compatible with older # versions of File::Find. my $wanted = sub { local $main::_ = $File::Find::name; my $dir = _untaintDir($File::Find::dir); chdir($startDir); if ($^O eq 'MSWin32' && $Archive::Zip::UNICODE) { push(@files, Win32::GetANSIPathName($File::Find::name)) if (&$pred); $dir = Win32::GetANSIPathName($dir); } else { push(@files, $File::Find::name) if (&$pred); } chdir($dir); }; if ($^O eq 'MSWin32' && $Archive::Zip::UNICODE) { $root = Win32::GetANSIPathName($root); } File::Find::find($wanted, $root); my $rootZipName = _asZipDirName($root, 1); # with trailing slash my $pattern = $rootZipName eq './' ? '^' : "^\Q$rootZipName\E"; $dest = _asZipDirName($dest, 1); # with trailing slash foreach my $fileName (@files) { my $isDir; if ($^O eq 'MSWin32' && $Archive::Zip::UNICODE) { $isDir = -d Win32::GetANSIPathName($fileName); } else { $isDir = -d $fileName; } # normalize, remove leading ./ my $archiveName = _asZipDirName($fileName, $isDir); if ($archiveName eq $rootZipName) { $archiveName = $dest } else { $archiveName =~ s{$pattern}{$dest} } next if $archiveName =~ m{^\.?/?$}; # skip current dir my $member = $isDir ? $self->addDirectory($fileName, $archiveName) : $self->addFile($fileName, $archiveName); $member->desiredCompressionLevel($compressionLevel); return _error("add $fileName failed in addTree()") if !$member; } return AZ_OK; } sub addTreeMatching { my $self = shift; my ($root, $dest, $pattern, $pred, $compressionLevel); if (ref($_[0]) eq 'HASH') { $root = $_[0]->{root}; $dest = $_[0]->{zipName}; $pattern = $_[0]->{pattern}; $pred = $_[0]->{select}; $compressionLevel = $_[0]->{compressionLevel}; } else { ($root, $dest, $pattern, $pred, $compressionLevel) = @_; } return _error("root arg missing in call to addTreeMatching()") unless defined($root); $dest = '' unless defined($dest); return _error("pattern missing in call to addTreeMatching()") unless defined($pattern); my $matcher = $pred ? sub { m{$pattern} && &$pred } : sub { m{$pattern} && -r }; return $self->addTree($root, $dest, $matcher, $compressionLevel); } # Check if one of the components of a path to the file or the file name # itself is an already existing symbolic link. If yes then return an # error. Continuing and writing to a file traversing a link posseses # a security threat, especially if the link was extracted from an # attacker-supplied archive. This would allow writing to an arbitrary # file. The same applies when using ".." to escape from a working # directory. sub _extractionNameIsSafe { my $name = shift; my ($volume, $directories) = File::Spec->splitpath($name, 1); my @directories = File::Spec->splitdir($directories); if (grep '..' eq $_, @directories) { return _error( "Could not extract $name safely: a parent directory is used"); } my @path; my $path; for my $directory (@directories) { push @path, $directory; $path = File::Spec->catpath($volume, File::Spec->catdir(@path), ''); if (-l $path) { return _error( "Could not extract $name safely: $path is an existing symbolic link"); } if (!-e $path) { last; } } return AZ_OK; } # $zip->extractTree( $root, $dest [, $volume] ); # # $root and $dest are Unix-style. # $volume is in local FS format. # sub extractTree { my $self = shift; my ($root, $dest, $volume); if (ref($_[0]) eq 'HASH') { $root = $_[0]->{root}; $dest = $_[0]->{zipName}; $volume = $_[0]->{volume}; } else { ($root, $dest, $volume) = @_; } $root = '' unless defined($root); if (defined $dest) { if ($dest !~ m{/$}) { $dest .= '/'; } } else { $dest = './'; } my $pattern = "^\Q$root"; my @members = $self->membersMatching($pattern); foreach my $member (@members) { my $fileName = $member->fileName(); # in Unix format $fileName =~ s{$pattern}{$dest}; # in Unix format # convert to platform format: $fileName = Archive::Zip::_asLocalName($fileName, $volume); if ((my $ret = _extractionNameIsSafe($fileName)) != AZ_OK) { return $ret; } my $status = $member->extractToFileNamed($fileName); return $status if $status != AZ_OK; } return AZ_OK; } # $zip->updateMember( $memberOrName, $fileName ); # Returns (possibly updated) member, if any; undef on errors. sub updateMember { my $self = shift; my ($oldMember, $fileName); if (ref($_[0]) eq 'HASH') { $oldMember = $_[0]->{memberOrZipName}; $fileName = $_[0]->{name}; } else { ($oldMember, $fileName) = @_; } if (!defined($fileName)) { _error("updateMember(): missing fileName argument"); return undef; } my @newStat = stat($fileName); if (!@newStat) { _ioError("Can't stat $fileName"); return undef; } my $isDir = -d _; my $memberName; if (ref($oldMember)) { $memberName = $oldMember->fileName(); } else { $oldMember = $self->memberNamed($memberName = $oldMember) || $self->memberNamed($memberName = _asZipDirName($oldMember, $isDir)); } unless (defined($oldMember) && $oldMember->lastModTime() == $newStat[9] && $oldMember->isDirectory() == $isDir && ($isDir || ($oldMember->uncompressedSize() == $newStat[7]))) { # create the new member my $newMember = $isDir ? Archive::Zip::Member->newDirectoryNamed($fileName, $memberName) : Archive::Zip::Member->newFromFile($fileName, $memberName); unless (defined($newMember)) { _error("creation of member $fileName failed in updateMember()"); return undef; } # replace old member or append new one if (defined($oldMember)) { $self->replaceMember($oldMember, $newMember); } else { $self->addMember($newMember); } return $newMember; } return $oldMember; } # $zip->updateTree( $root, [ $dest, [ $pred [, $mirror]]] ); # # This takes the same arguments as addTree, but first checks to see # whether the file or directory already exists in the zip file. # # If the fourth argument $mirror is true, then delete all my members # if corresponding files were not found. sub updateTree { my $self = shift; my ($root, $dest, $pred, $mirror, $compressionLevel); if (ref($_[0]) eq 'HASH') { $root = $_[0]->{root}; $dest = $_[0]->{zipName}; $pred = $_[0]->{select}; $mirror = $_[0]->{mirror}; $compressionLevel = $_[0]->{compressionLevel}; } else { ($root, $dest, $pred, $mirror, $compressionLevel) = @_; } return _error("root arg missing in call to updateTree()") unless defined($root); $dest = '' unless defined($dest); $pred = sub { -r } unless defined($pred); $dest = _asZipDirName($dest, 1); my $rootZipName = _asZipDirName($root, 1); # with trailing slash my $pattern = $rootZipName eq './' ? '^' : "^\Q$rootZipName\E"; my @files; my $startDir = _untaintDir(cwd()); return _error('undef returned by _untaintDir on cwd ', cwd()) unless $startDir; # This avoids chdir'ing in Find, in a way compatible with older # versions of File::Find. my $wanted = sub { local $main::_ = $File::Find::name; my $dir = _untaintDir($File::Find::dir); chdir($startDir); push(@files, $File::Find::name) if (&$pred); chdir($dir); }; File::Find::find($wanted, $root); # Now @files has all the files that I could potentially be adding to # the zip. Only add the ones that are necessary. # For each file (updated or not), add its member name to @done. my %done; foreach my $fileName (@files) { my @newStat = stat($fileName); my $isDir = -d _; # normalize, remove leading ./ my $memberName = _asZipDirName($fileName, $isDir); if ($memberName eq $rootZipName) { $memberName = $dest } else { $memberName =~ s{$pattern}{$dest} } next if $memberName =~ m{^\.?/?$}; # skip current dir $done{$memberName} = 1; my $changedMember = $self->updateMember($memberName, $fileName); $changedMember->desiredCompressionLevel($compressionLevel); return _error("updateTree failed to update $fileName") unless ref($changedMember); } # @done now has the archive names corresponding to all the found files. # If we're mirroring, delete all those members that aren't in @done. if ($mirror) { foreach my $member ($self->members()) { $self->removeMember($member) unless $done{$member->fileName()}; } } return AZ_OK; } 1; Zip/Tree.pm000064400000001460152347641370006560 0ustar00package Archive::Zip::Tree; use strict; use vars qw{$VERSION}; BEGIN { $VERSION = '1.60'; } use Archive::Zip; warn( "Archive::Zip::Tree is deprecated; its methods have been moved into Archive::Zip." ) if $^W; 1; __END__ =head1 NAME Archive::Zip::Tree - (DEPRECATED) methods for adding/extracting trees using Archive::Zip =head1 DESCRIPTION This module is deprecated, because all its methods were moved into the main Archive::Zip module. It is included in the distribution merely to avoid breaking old code. See L. =head1 AUTHOR Ned Konz, perl@bike-nomad.com =head1 COPYRIGHT Copyright (c) 2000-2002 Ned Konz. All rights reserved. This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself. =head1 SEE ALSO L =cut Zip/FileMember.pm000064400000002472152347641370007674 0ustar00package Archive::Zip::FileMember; use strict; use vars qw( $VERSION @ISA ); BEGIN { $VERSION = '1.60'; @ISA = qw ( Archive::Zip::Member ); } use Archive::Zip qw( :UTILITY_METHODS ); sub externalFileName { shift->{'externalFileName'}; } # Return true if I depend on the named file sub _usesFileNamed { my $self = shift; my $fileName = shift; my $xfn = $self->externalFileName(); return undef if ref($xfn); return $xfn eq $fileName; } sub fh { my $self = shift; $self->_openFile() if !defined($self->{'fh'}) || !$self->{'fh'}->opened(); return $self->{'fh'}; } # opens my file handle from my file name sub _openFile { my $self = shift; my ($status, $fh) = _newFileHandle($self->externalFileName(), 'r'); if (!$status) { _ioError("Can't open", $self->externalFileName()); return undef; } $self->{'fh'} = $fh; _binmode($fh); return $fh; } # Make sure I close my file handle sub endRead { my $self = shift; undef $self->{'fh'}; # _closeFile(); return $self->SUPER::endRead(@_); } sub _become { my $self = shift; my $newClass = shift; return $self if ref($self) eq $newClass; delete($self->{'externalFileName'}); delete($self->{'fh'}); return $self->SUPER::_become($newClass); } 1; Zip/Member.pm000064400000111437152347641370007076 0ustar00package Archive::Zip::Member; # A generic member of an archive use strict; use vars qw( $VERSION @ISA ); BEGIN { $VERSION = '1.60'; @ISA = qw( Archive::Zip ); if ($^O eq 'MSWin32') { require Win32; require Encode; Encode->import(qw{ decode_utf8 }); } } use Archive::Zip qw( :CONSTANTS :MISC_CONSTANTS :ERROR_CODES :PKZIP_CONSTANTS :UTILITY_METHODS ); use Time::Local (); use Compress::Raw::Zlib qw( Z_OK Z_STREAM_END MAX_WBITS ); use File::Path; use File::Basename; # Unix perms for default creation of files/dirs. use constant DEFAULT_DIRECTORY_PERMISSIONS => 040755; use constant DEFAULT_FILE_PERMISSIONS => 0100666; use constant DIRECTORY_ATTRIB => 040000; use constant FILE_ATTRIB => 0100000; # Returns self if successful, else undef # Assumes that fh is positioned at beginning of central directory file header. # Leaves fh positioned immediately after file header or EOCD signature. sub _newFromZipFile { my $class = shift; my $self = Archive::Zip::ZipFileMember->_newFromZipFile(@_); return $self; } sub newFromString { my $class = shift; my ($stringOrStringRef, $fileName); if (ref($_[0]) eq 'HASH') { $stringOrStringRef = $_[0]->{string}; $fileName = $_[0]->{zipName}; } else { ($stringOrStringRef, $fileName) = @_; } my $self = Archive::Zip::StringMember->_newFromString($stringOrStringRef, $fileName); return $self; } sub newFromFile { my $class = shift; my ($fileName, $zipName); if (ref($_[0]) eq 'HASH') { $fileName = $_[0]->{fileName}; $zipName = $_[0]->{zipName}; } else { ($fileName, $zipName) = @_; } my $self = Archive::Zip::NewFileMember->_newFromFileNamed($fileName, $zipName); return $self; } sub newDirectoryNamed { my $class = shift; my ($directoryName, $newName); if (ref($_[0]) eq 'HASH') { $directoryName = $_[0]->{directoryName}; $newName = $_[0]->{zipName}; } else { ($directoryName, $newName) = @_; } my $self = Archive::Zip::DirectoryMember->_newNamed($directoryName, $newName); return $self; } sub new { my $class = shift; my $self = { 'lastModFileDateTime' => 0, 'fileAttributeFormat' => FA_UNIX, 'versionMadeBy' => 20, 'versionNeededToExtract' => 20, 'bitFlag' => ($Archive::Zip::UNICODE ? 0x0800 : 0), 'compressionMethod' => COMPRESSION_STORED, 'desiredCompressionMethod' => COMPRESSION_STORED, 'desiredCompressionLevel' => COMPRESSION_LEVEL_NONE, 'internalFileAttributes' => 0, 'externalFileAttributes' => 0, # set later 'fileName' => '', 'cdExtraField' => '', 'localExtraField' => '', 'fileComment' => '', 'crc32' => 0, 'compressedSize' => 0, 'uncompressedSize' => 0, 'isSymbolicLink' => 0, 'password' => undef, # password for encrypted data 'crc32c' => -1, # crc for decrypted data @_ }; bless($self, $class); $self->unixFileAttributes($self->DEFAULT_FILE_PERMISSIONS); return $self; } sub _becomeDirectoryIfNecessary { my $self = shift; $self->_become('Archive::Zip::DirectoryMember') if $self->isDirectory(); return $self; } # Morph into given class (do whatever cleanup I need to do) sub _become { return bless($_[0], $_[1]); } sub versionMadeBy { shift->{'versionMadeBy'}; } sub fileAttributeFormat { my $self = shift; if (@_) { $self->{fileAttributeFormat} = (ref($_[0]) eq 'HASH') ? $_[0]->{format} : $_[0]; } else { return $self->{fileAttributeFormat}; } } sub versionNeededToExtract { shift->{'versionNeededToExtract'}; } sub bitFlag { my $self = shift; # Set General Purpose Bit Flags according to the desiredCompressionLevel setting if ( $self->desiredCompressionLevel == 1 || $self->desiredCompressionLevel == 2) { $self->{'bitFlag'} |= DEFLATING_COMPRESSION_FAST; } elsif ($self->desiredCompressionLevel == 3 || $self->desiredCompressionLevel == 4 || $self->desiredCompressionLevel == 5 || $self->desiredCompressionLevel == 6 || $self->desiredCompressionLevel == 7) { $self->{'bitFlag'} |= DEFLATING_COMPRESSION_NORMAL; } elsif ($self->desiredCompressionLevel == 8 || $self->desiredCompressionLevel == 9) { $self->{'bitFlag'} |= DEFLATING_COMPRESSION_MAXIMUM; } if ($Archive::Zip::UNICODE) { $self->{'bitFlag'} |= 0x0800; } $self->{'bitFlag'}; } sub password { my $self = shift; $self->{'password'} = shift if @_; $self->{'password'}; } sub compressionMethod { shift->{'compressionMethod'}; } sub desiredCompressionMethod { my $self = shift; my $newDesiredCompressionMethod = (ref($_[0]) eq 'HASH') ? shift->{compressionMethod} : shift; my $oldDesiredCompressionMethod = $self->{'desiredCompressionMethod'}; if (defined($newDesiredCompressionMethod)) { $self->{'desiredCompressionMethod'} = $newDesiredCompressionMethod; if ($newDesiredCompressionMethod == COMPRESSION_STORED) { $self->{'desiredCompressionLevel'} = 0; $self->{'bitFlag'} &= ~GPBF_HAS_DATA_DESCRIPTOR_MASK if $self->uncompressedSize() == 0; } elsif ($oldDesiredCompressionMethod == COMPRESSION_STORED) { $self->{'desiredCompressionLevel'} = COMPRESSION_LEVEL_DEFAULT; } } return $oldDesiredCompressionMethod; } sub desiredCompressionLevel { my $self = shift; my $newDesiredCompressionLevel = (ref($_[0]) eq 'HASH') ? shift->{compressionLevel} : shift; my $oldDesiredCompressionLevel = $self->{'desiredCompressionLevel'}; if (defined($newDesiredCompressionLevel)) { $self->{'desiredCompressionLevel'} = $newDesiredCompressionLevel; $self->{'desiredCompressionMethod'} = ( $newDesiredCompressionLevel ? COMPRESSION_DEFLATED : COMPRESSION_STORED ); } return $oldDesiredCompressionLevel; } sub fileName { my $self = shift; my $newName = shift; if (defined $newName) { $newName =~ s{[\\/]+}{/}g; # deal with dos/windoze problems $self->{'fileName'} = $newName; } return $self->{'fileName'}; } sub fileNameAsBytes { my $self = shift; my $bytes = $self->{'fileName'}; if($self->{'bitFlag'} & 0x800){ $bytes = Encode::encode_utf8($bytes); } return $bytes; } sub lastModFileDateTime { my $modTime = shift->{'lastModFileDateTime'}; $modTime =~ m/^(\d+)$/; # untaint return $1; } sub lastModTime { my $self = shift; return _dosToUnixTime($self->lastModFileDateTime()); } sub setLastModFileDateTimeFromUnix { my $self = shift; my $time_t = shift; $self->{'lastModFileDateTime'} = _unixToDosTime($time_t); } sub internalFileAttributes { shift->{'internalFileAttributes'}; } sub externalFileAttributes { shift->{'externalFileAttributes'}; } # Convert UNIX permissions into proper value for zip file # Usable as a function or a method sub _mapPermissionsFromUnix { my $self = shift; my $mode = shift; my $attribs = $mode << 16; # Microsoft Windows Explorer needs this bit set for directories if ($mode & DIRECTORY_ATTRIB) { $attribs |= 16; } return $attribs; # TODO: map more MS-DOS perms } # Convert ZIP permissions into Unix ones # # This was taken from Info-ZIP group's portable UnZip # zipfile-extraction program, version 5.50. # http://www.info-zip.org/pub/infozip/ # # See the mapattr() function in unix/unix.c # See the attribute format constants in unzpriv.h # # XXX Note that there's one situation that is not implemented # yet that depends on the "extra field." sub _mapPermissionsToUnix { my $self = shift; my $format = $self->{'fileAttributeFormat'}; my $attribs = $self->{'externalFileAttributes'}; my $mode = 0; if ($format == FA_AMIGA) { $attribs = $attribs >> 17 & 7; # Amiga RWE bits $mode = $attribs << 6 | $attribs << 3 | $attribs; return $mode; } if ($format == FA_THEOS) { $attribs &= 0xF1FFFFFF; if (($attribs & 0xF0000000) != 0x40000000) { $attribs &= 0x01FFFFFF; # not a dir, mask all ftype bits } else { $attribs &= 0x41FFFFFF; # leave directory bit as set } } if ( $format == FA_UNIX || $format == FA_VAX_VMS || $format == FA_ACORN || $format == FA_ATARI_ST || $format == FA_BEOS || $format == FA_QDOS || $format == FA_TANDEM) { $mode = $attribs >> 16; return $mode if $mode != 0 or not $self->localExtraField; # warn("local extra field is: ", $self->localExtraField, "\n"); # XXX This condition is not implemented # I'm just including the comments from the info-zip section for now. # Some (non-Info-ZIP) implementations of Zip for Unix and # VMS (and probably others ??) leave 0 in the upper 16-bit # part of the external_file_attributes field. Instead, they # store file permission attributes in some extra field. # As a work-around, we search for the presence of one of # these extra fields and fall back to the MSDOS compatible # part of external_file_attributes if one of the known # e.f. types has been detected. # Later, we might implement extraction of the permission # bits from the VMS extra field. But for now, the work-around # should be sufficient to provide "readable" extracted files. # (For ASI Unix e.f., an experimental remap from the e.f. # mode value IS already provided!) } # PKWARE's PKZip for Unix marks entries as FA_MSDOS, but stores the # Unix attributes in the upper 16 bits of the external attributes # field, just like Info-ZIP's Zip for Unix. We try to use that # value, after a check for consistency with the MSDOS attribute # bits (see below). if ($format == FA_MSDOS) { $mode = $attribs >> 16; } # FA_MSDOS, FA_OS2_HPFS, FA_WINDOWS_NTFS, FA_MACINTOSH, FA_TOPS20 $attribs = !($attribs & 1) << 1 | ($attribs & 0x10) >> 4; # keep previous $mode setting when its "owner" # part appears to be consistent with DOS attribute flags! return $mode if ($mode & 0700) == (0400 | $attribs << 6); $mode = 0444 | $attribs << 6 | $attribs << 3 | $attribs; return $mode; } sub unixFileAttributes { my $self = shift; my $oldPerms = $self->_mapPermissionsToUnix; my $perms; if (@_) { $perms = (ref($_[0]) eq 'HASH') ? $_[0]->{attributes} : $_[0]; if ($self->isDirectory) { $perms &= ~FILE_ATTRIB; $perms |= DIRECTORY_ATTRIB; } else { $perms &= ~DIRECTORY_ATTRIB; $perms |= FILE_ATTRIB; } $self->{externalFileAttributes} = $self->_mapPermissionsFromUnix($perms); } return $oldPerms; } sub localExtraField { my $self = shift; if (@_) { $self->{localExtraField} = (ref($_[0]) eq 'HASH') ? $_[0]->{field} : $_[0]; } else { return $self->{localExtraField}; } } sub cdExtraField { my $self = shift; if (@_) { $self->{cdExtraField} = (ref($_[0]) eq 'HASH') ? $_[0]->{field} : $_[0]; } else { return $self->{cdExtraField}; } } sub extraFields { my $self = shift; return $self->localExtraField() . $self->cdExtraField(); } sub fileComment { my $self = shift; if (@_) { $self->{fileComment} = (ref($_[0]) eq 'HASH') ? pack('C0a*', $_[0]->{comment}) : pack('C0a*', $_[0]); } else { return $self->{fileComment}; } } sub hasDataDescriptor { my $self = shift; if (@_) { my $shouldHave = shift; if ($shouldHave) { $self->{'bitFlag'} |= GPBF_HAS_DATA_DESCRIPTOR_MASK; } else { $self->{'bitFlag'} &= ~GPBF_HAS_DATA_DESCRIPTOR_MASK; } } return $self->{'bitFlag'} & GPBF_HAS_DATA_DESCRIPTOR_MASK; } sub crc32 { shift->{'crc32'}; } sub crc32String { sprintf("%08x", shift->{'crc32'}); } sub compressedSize { shift->{'compressedSize'}; } sub uncompressedSize { shift->{'uncompressedSize'}; } sub isEncrypted { shift->{'bitFlag'} & GPBF_ENCRYPTED_MASK; } sub isTextFile { my $self = shift; my $bit = $self->internalFileAttributes() & IFA_TEXT_FILE_MASK; if (@_) { my $flag = (ref($_[0]) eq 'HASH') ? shift->{flag} : shift; $self->{'internalFileAttributes'} &= ~IFA_TEXT_FILE_MASK; $self->{'internalFileAttributes'} |= ($flag ? IFA_TEXT_FILE : IFA_BINARY_FILE); } return $bit == IFA_TEXT_FILE; } sub isBinaryFile { my $self = shift; my $bit = $self->internalFileAttributes() & IFA_TEXT_FILE_MASK; if (@_) { my $flag = shift; $self->{'internalFileAttributes'} &= ~IFA_TEXT_FILE_MASK; $self->{'internalFileAttributes'} |= ($flag ? IFA_BINARY_FILE : IFA_TEXT_FILE); } return $bit == IFA_BINARY_FILE; } sub extractToFileNamed { my $self = shift; # local FS name my $name = (ref($_[0]) eq 'HASH') ? $_[0]->{name} : $_[0]; $self->{'isSymbolicLink'} = 0; # Check if the file / directory is a symbolic link or not if ($self->{'externalFileAttributes'} == 0xA1FF0000) { $self->{'isSymbolicLink'} = 1; $self->{'newName'} = $name; my ($status, $fh) = _newFileHandle($name, 'r'); my $retval = $self->extractToFileHandle($fh); $fh->close(); } else { #return _writeSymbolicLink($self, $name) if $self->isSymbolicLink(); my ($status, $fh); if ($^O eq 'MSWin32' && $Archive::Zip::UNICODE) { $name = decode_utf8(Win32::GetFullPathName($name)); mkpath_win32($name); Win32::CreateFile($name); ($status, $fh) = _newFileHandle(Win32::GetANSIPathName($name), 'w'); } else { mkpath(dirname($name)); # croaks on error ($status, $fh) = _newFileHandle($name, 'w'); } return _ioError("Can't open file $name for write") unless $status; my $retval = $self->extractToFileHandle($fh); $fh->close(); chmod($self->unixFileAttributes(), $name) or return _error("Can't chmod() ${name}: $!"); utime($self->lastModTime(), $self->lastModTime(), $name); return $retval; } } sub mkpath_win32 { my $path = shift; use File::Spec; my ($volume, @path) = File::Spec->splitdir($path); $path = File::Spec->catfile($volume, shift @path); pop @path; while (@path) { $path = File::Spec->catfile($path, shift @path); Win32::CreateDirectory($path); } } sub _writeSymbolicLink { my $self = shift; my $name = shift; my $chunkSize = $Archive::Zip::ChunkSize; #my ( $outRef, undef ) = $self->readChunk($chunkSize); my $fh; my $retval = $self->extractToFileHandle($fh); my ($outRef, undef) = $self->readChunk(100); } sub isSymbolicLink { my $self = shift; if ($self->{'externalFileAttributes'} == 0xA1FF0000) { $self->{'isSymbolicLink'} = 1; } else { return 0; } 1; } sub isDirectory { return 0; } sub externalFileName { return undef; } # The following are used when copying data sub _writeOffset { shift->{'writeOffset'}; } sub _readOffset { shift->{'readOffset'}; } sub writeLocalHeaderRelativeOffset { shift->{'writeLocalHeaderRelativeOffset'}; } sub wasWritten { shift->{'wasWritten'} } sub _dataEnded { shift->{'dataEnded'}; } sub _readDataRemaining { shift->{'readDataRemaining'}; } sub _inflater { shift->{'inflater'}; } sub _deflater { shift->{'deflater'}; } # Return the total size of my local header sub _localHeaderSize { my $self = shift; { use bytes; return SIGNATURE_LENGTH + LOCAL_FILE_HEADER_LENGTH + length($self->fileName()) + length($self->localExtraField()); } } # Return the total size of my CD header sub _centralDirectoryHeaderSize { my $self = shift; { use bytes; return SIGNATURE_LENGTH + CENTRAL_DIRECTORY_FILE_HEADER_LENGTH + length($self->fileName()) + length($self->cdExtraField()) + length($self->fileComment()); } } # DOS date/time format # 0-4 (5) Second divided by 2 # 5-10 (6) Minute (0-59) # 11-15 (5) Hour (0-23 on a 24-hour clock) # 16-20 (5) Day of the month (1-31) # 21-24 (4) Month (1 = January, 2 = February, etc.) # 25-31 (7) Year offset from 1980 (add 1980 to get actual year) # Convert DOS date/time format to unix time_t format # NOT AN OBJECT METHOD! sub _dosToUnixTime { my $dt = shift; return time() unless defined($dt); my $year = (($dt >> 25) & 0x7f) + 80; my $mon = (($dt >> 21) & 0x0f) - 1; my $mday = (($dt >> 16) & 0x1f); my $hour = (($dt >> 11) & 0x1f); my $min = (($dt >> 5) & 0x3f); my $sec = (($dt << 1) & 0x3e); # catch errors my $time_t = eval { Time::Local::timelocal($sec, $min, $hour, $mday, $mon, $year); }; return time() if ($@); return $time_t; } # Note, this is not exactly UTC 1980, it's 1980 + 12 hours and 1 # minute so that nothing timezoney can muck us up. my $safe_epoch = 31.606060; # convert a unix time to DOS date/time # NOT AN OBJECT METHOD! sub _unixToDosTime { my $time_t = shift; unless ($time_t) { _error("Tried to add member with zero or undef value for time"); $time_t = $safe_epoch; } if ($time_t < $safe_epoch) { _ioError("Unsupported date before 1980 encountered, moving to 1980"); $time_t = $safe_epoch; } my ($sec, $min, $hour, $mday, $mon, $year) = localtime($time_t); my $dt = 0; $dt += ($sec >> 1); $dt += ($min << 5); $dt += ($hour << 11); $dt += ($mday << 16); $dt += (($mon + 1) << 21); $dt += (($year - 80) << 25); return $dt; } sub head { my ($self, $mode) = (@_, 0); use bytes; return pack LOCAL_FILE_HEADER_FORMAT, $self->versionNeededToExtract(), $self->{'bitFlag'}, $self->desiredCompressionMethod(), $self->lastModFileDateTime(), $self->hasDataDescriptor() ? (0,0,0) # crc, compr & uncompr all zero if data descriptor present : ( $self->crc32(), $mode ? $self->_writeOffset() # compressed size : $self->compressedSize(), # may need to be re-written later $self->uncompressedSize(), ), length($self->fileNameAsBytes()), length($self->localExtraField()); } # Write my local header to a file handle. # Stores the offset to the start of the header in my # writeLocalHeaderRelativeOffset member. # Returns AZ_OK on success. sub _writeLocalFileHeader { my $self = shift; my $fh = shift; my $signatureData = pack(SIGNATURE_FORMAT, LOCAL_FILE_HEADER_SIGNATURE); $self->_print($fh, $signatureData) or return _ioError("writing local header signature"); my $header = $self->head(1); $self->_print($fh, $header) or return _ioError("writing local header"); # Check for a valid filename or a filename equal to a literal `0' if ($self->fileName() || $self->fileName eq '0') { $self->_print($fh, $self->fileNameAsBytes()) or return _ioError("writing local header filename"); } if ($self->localExtraField()) { $self->_print($fh, $self->localExtraField()) or return _ioError("writing local extra field"); } return AZ_OK; } sub _writeCentralDirectoryFileHeader { my $self = shift; my $fh = shift; my $sigData = pack(SIGNATURE_FORMAT, CENTRAL_DIRECTORY_FILE_HEADER_SIGNATURE); $self->_print($fh, $sigData) or return _ioError("writing central directory header signature"); my ($fileNameLength, $extraFieldLength, $fileCommentLength); { use bytes; $fileNameLength = length($self->fileNameAsBytes()); $extraFieldLength = length($self->cdExtraField()); $fileCommentLength = length($self->fileComment()); } my $header = pack( CENTRAL_DIRECTORY_FILE_HEADER_FORMAT, $self->versionMadeBy(), $self->fileAttributeFormat(), $self->versionNeededToExtract(), $self->bitFlag(), $self->desiredCompressionMethod(), $self->lastModFileDateTime(), $self->crc32(), # these three fields should have been updated $self->_writeOffset(), # by writing the data stream out $self->uncompressedSize(), # $fileNameLength, $extraFieldLength, $fileCommentLength, 0, # {'diskNumberStart'}, $self->internalFileAttributes(), $self->externalFileAttributes(), $self->writeLocalHeaderRelativeOffset()); $self->_print($fh, $header) or return _ioError("writing central directory header"); if ($fileNameLength) { $self->_print($fh, $self->fileNameAsBytes()) or return _ioError("writing central directory header signature"); } if ($extraFieldLength) { $self->_print($fh, $self->cdExtraField()) or return _ioError("writing central directory extra field"); } if ($fileCommentLength) { $self->_print($fh, $self->fileComment()) or return _ioError("writing central directory file comment"); } return AZ_OK; } # This writes a data descriptor to the given file handle. # Assumes that crc32, writeOffset, and uncompressedSize are # set correctly (they should be after a write). # Further, the local file header should have the # GPBF_HAS_DATA_DESCRIPTOR_MASK bit set. sub _writeDataDescriptor { my $self = shift; my $fh = shift; my $header = pack( SIGNATURE_FORMAT . DATA_DESCRIPTOR_FORMAT, DATA_DESCRIPTOR_SIGNATURE, $self->crc32(), $self->_writeOffset(), # compressed size $self->uncompressedSize()); $self->_print($fh, $header) or return _ioError("writing data descriptor"); return AZ_OK; } # Re-writes the local file header with new crc32 and compressedSize fields. # To be called after writing the data stream. # Assumes that filename and extraField sizes didn't change since last written. sub _refreshLocalFileHeader { my $self = shift; my $fh = shift; my $here = $fh->tell(); $fh->seek($self->writeLocalHeaderRelativeOffset() + SIGNATURE_LENGTH, IO::Seekable::SEEK_SET) or return _ioError("seeking to rewrite local header"); my $header = $self->head(1); $self->_print($fh, $header) or return _ioError("re-writing local header"); $fh->seek($here, IO::Seekable::SEEK_SET) or return _ioError("seeking after rewrite of local header"); return AZ_OK; } sub readChunk { my $self = shift; my $chunkSize = (ref($_[0]) eq 'HASH') ? $_[0]->{chunkSize} : $_[0]; if ($self->readIsDone()) { $self->endRead(); my $dummy = ''; return (\$dummy, AZ_STREAM_END); } $chunkSize = $Archive::Zip::ChunkSize if not defined($chunkSize); $chunkSize = $self->_readDataRemaining() if $chunkSize > $self->_readDataRemaining(); my $buffer = ''; my $outputRef; my ($bytesRead, $status) = $self->_readRawChunk(\$buffer, $chunkSize); return (\$buffer, $status) unless $status == AZ_OK; $buffer && $self->isEncrypted and $buffer = $self->_decode($buffer); $self->{'readDataRemaining'} -= $bytesRead; $self->{'readOffset'} += $bytesRead; if ($self->compressionMethod() == COMPRESSION_STORED) { $self->{'crc32'} = $self->computeCRC32($buffer, $self->{'crc32'}); } ($outputRef, $status) = &{$self->{'chunkHandler'}}($self, \$buffer); $self->{'writeOffset'} += length($$outputRef); $self->endRead() if $self->readIsDone(); return ($outputRef, $status); } # Read the next raw chunk of my data. Subclasses MUST implement. # my ( $bytesRead, $status) = $self->_readRawChunk( \$buffer, $chunkSize ); sub _readRawChunk { my $self = shift; return $self->_subclassResponsibility(); } # A place holder to catch rewindData errors if someone ignores # the error code. sub _noChunk { my $self = shift; return (\undef, _error("trying to copy chunk when init failed")); } # Basically a no-op so that I can have a consistent interface. # ( $outputRef, $status) = $self->_copyChunk( \$buffer ); sub _copyChunk { my ($self, $dataRef) = @_; return ($dataRef, AZ_OK); } # ( $outputRef, $status) = $self->_deflateChunk( \$buffer ); sub _deflateChunk { my ($self, $buffer) = @_; my ($status) = $self->_deflater()->deflate($buffer, my $out); if ($self->_readDataRemaining() == 0) { my $extraOutput; ($status) = $self->_deflater()->flush($extraOutput); $out .= $extraOutput; $self->endRead(); return (\$out, AZ_STREAM_END); } elsif ($status == Z_OK) { return (\$out, AZ_OK); } else { $self->endRead(); my $retval = _error('deflate error', $status); my $dummy = ''; return (\$dummy, $retval); } } # ( $outputRef, $status) = $self->_inflateChunk( \$buffer ); sub _inflateChunk { my ($self, $buffer) = @_; my ($status) = $self->_inflater()->inflate($buffer, my $out); my $retval; $self->endRead() unless $status == Z_OK; if ($status == Z_OK || $status == Z_STREAM_END) { $retval = ($status == Z_STREAM_END) ? AZ_STREAM_END : AZ_OK; return (\$out, $retval); } else { $retval = _error('inflate error', $status); my $dummy = ''; return (\$dummy, $retval); } } sub rewindData { my $self = shift; my $status; # set to trap init errors $self->{'chunkHandler'} = $self->can('_noChunk'); # Work around WinZip bug with 0-length DEFLATED files $self->desiredCompressionMethod(COMPRESSION_STORED) if $self->uncompressedSize() == 0; # assume that we're going to read the whole file, and compute the CRC anew. $self->{'crc32'} = 0 if ($self->compressionMethod() == COMPRESSION_STORED); # These are the only combinations of methods we deal with right now. if ( $self->compressionMethod() == COMPRESSION_STORED and $self->desiredCompressionMethod() == COMPRESSION_DEFLATED) { ($self->{'deflater'}, $status) = Compress::Raw::Zlib::Deflate->new( '-Level' => $self->desiredCompressionLevel(), '-WindowBits' => -MAX_WBITS(), # necessary magic '-Bufsize' => $Archive::Zip::ChunkSize, @_ ); # pass additional options return _error('deflateInit error:', $status) unless $status == Z_OK; $self->{'chunkHandler'} = $self->can('_deflateChunk'); } elsif ($self->compressionMethod() == COMPRESSION_DEFLATED and $self->desiredCompressionMethod() == COMPRESSION_STORED) { ($self->{'inflater'}, $status) = Compress::Raw::Zlib::Inflate->new( '-WindowBits' => -MAX_WBITS(), # necessary magic '-Bufsize' => $Archive::Zip::ChunkSize, @_ ); # pass additional options return _error('inflateInit error:', $status) unless $status == Z_OK; $self->{'chunkHandler'} = $self->can('_inflateChunk'); } elsif ($self->compressionMethod() == $self->desiredCompressionMethod()) { $self->{'chunkHandler'} = $self->can('_copyChunk'); } else { return _error( sprintf( "Unsupported compression combination: read %d, write %d", $self->compressionMethod(), $self->desiredCompressionMethod())); } $self->{'readDataRemaining'} = ($self->compressionMethod() == COMPRESSION_STORED) ? $self->uncompressedSize() : $self->compressedSize(); $self->{'dataEnded'} = 0; $self->{'readOffset'} = 0; return AZ_OK; } sub endRead { my $self = shift; delete $self->{'inflater'}; delete $self->{'deflater'}; $self->{'dataEnded'} = 1; $self->{'readDataRemaining'} = 0; return AZ_OK; } sub readIsDone { my $self = shift; return ($self->_dataEnded() or !$self->_readDataRemaining()); } sub contents { my $self = shift; my $newContents = shift; if (defined($newContents)) { # change our type and call the subclass contents method. $self->_become('Archive::Zip::StringMember'); return $self->contents(pack('C0a*', $newContents)); # in case of Unicode } else { my $oldCompression = $self->desiredCompressionMethod(COMPRESSION_STORED); my $status = $self->rewindData(@_); if ($status != AZ_OK) { $self->endRead(); return $status; } my $retval = ''; while ($status == AZ_OK) { my $ref; ($ref, $status) = $self->readChunk($self->_readDataRemaining()); # did we get it in one chunk? if (length($$ref) == $self->uncompressedSize()) { $retval = $$ref; } else { $retval .= $$ref } } $self->desiredCompressionMethod($oldCompression); $self->endRead(); $status = AZ_OK if $status == AZ_STREAM_END; $retval = undef unless $status == AZ_OK; return wantarray ? ($retval, $status) : $retval; } } sub extractToFileHandle { my $self = shift; my $fh = (ref($_[0]) eq 'HASH') ? shift->{fileHandle} : shift; _binmode($fh); my $oldCompression = $self->desiredCompressionMethod(COMPRESSION_STORED); my $status = $self->rewindData(@_); $status = $self->_writeData($fh) if $status == AZ_OK; $self->desiredCompressionMethod($oldCompression); $self->endRead(); return $status; } # write local header and data stream to file handle sub _writeToFileHandle { my $self = shift; my $fh = shift; my $fhIsSeekable = shift; my $offset = shift; return _error("no member name given for $self") if $self->fileName() eq ''; $self->{'writeLocalHeaderRelativeOffset'} = $offset; $self->{'wasWritten'} = 0; # Determine if I need to write a data descriptor # I need to do this if I can't refresh the header # and I don't know compressed size or crc32 fields. my $headerFieldsUnknown = ( ($self->uncompressedSize() > 0) and ($self->compressionMethod() == COMPRESSION_STORED or $self->desiredCompressionMethod() == COMPRESSION_DEFLATED)); my $shouldWriteDataDescriptor = ($headerFieldsUnknown and not $fhIsSeekable); $self->hasDataDescriptor(1) if ($shouldWriteDataDescriptor); $self->{'writeOffset'} = 0; my $status = $self->rewindData(); ($status = $self->_writeLocalFileHeader($fh)) if $status == AZ_OK; ($status = $self->_writeData($fh)) if $status == AZ_OK; if ($status == AZ_OK) { $self->{'wasWritten'} = 1; if ($self->hasDataDescriptor()) { $status = $self->_writeDataDescriptor($fh); } elsif ($headerFieldsUnknown) { $status = $self->_refreshLocalFileHeader($fh); } } return $status; } # Copy my (possibly compressed) data to given file handle. # Returns C on success sub _writeData { my $self = shift; my $writeFh = shift; # If symbolic link, just create one if the operating system is Linux, Unix, BSD or VMS # TODO: Add checks for other operating systems if ($self->{'isSymbolicLink'} == 1 && $^O eq 'linux') { my $chunkSize = $Archive::Zip::ChunkSize; my ($outRef, $status) = $self->readChunk($chunkSize); symlink $$outRef, $self->{'newName'}; } else { return AZ_OK if ($self->uncompressedSize() == 0); my $status; my $chunkSize = $Archive::Zip::ChunkSize; while ($self->_readDataRemaining() > 0) { my $outRef; ($outRef, $status) = $self->readChunk($chunkSize); return $status if ($status != AZ_OK and $status != AZ_STREAM_END); if (length($$outRef) > 0) { $self->_print($writeFh, $$outRef) or return _ioError("write error during copy"); } last if $status == AZ_STREAM_END; } } return AZ_OK; } # Return true if I depend on the named file sub _usesFileNamed { return 0; } # ############################################################################## # # Decrypt section # # H.Merijn Brand (Tux) 2011-06-28 # # ############################################################################## # This code is derived from the crypt source of unzip-6.0 dated 05 Jan 2007 # Its license states: # # --8<--- # Copyright (c) 1990-2007 Info-ZIP. All rights reserved. # See the accompanying file LICENSE, version 2005-Feb-10 or later # (the contents of which are also included in (un)zip.h) for terms of use. # If, for some reason, all these files are missing, the Info-ZIP license # also may be found at: ftp://ftp.info-zip.org/pub/infozip/license.html # # crypt.c (full version) by Info-ZIP. Last revised: [see crypt.h] # The main encryption/decryption source code for Info-Zip software was # originally written in Europe. To the best of our knowledge, it can # be freely distributed in both source and object forms from any country, # including the USA under License Exception TSU of the U.S. Export # Administration Regulations (section 740.13(e)) of 6 June 2002. # NOTE on copyright history: # Previous versions of this source package (up to version 2.8) were # not copyrighted and put in the public domain. If you cannot comply # with the Info-Zip LICENSE, you may want to look for one of those # public domain versions. # # This encryption code is a direct transcription of the algorithm from # Roger Schlafly, described by Phil Katz in the file appnote.txt. This # file (appnote.txt) is distributed with the PKZIP program (even in the # version without encryption capabilities). # -->8--- # As of January 2000, US export regulations were amended to allow export # of free encryption source code from the US. As of June 2002, these # regulations were further relaxed to allow export of encryption binaries # associated with free encryption source code. The Zip 2.31, UnZip 5.52 # and Wiz 5.02 archives now include full crypto source code. As of the # Zip 2.31 release, all official binaries include encryption support; the # former "zcr" archives ceased to exist. # (Note that restrictions may still exist in other countries, of course.) # For now, we just support the decrypt stuff # All below methods are supposed to be private # use Data::Peek; my @keys; my @crct = do { my $xor = 0xedb88320; my @crc = (0) x 1024; # generate a crc for every 8-bit value foreach my $n (0 .. 255) { my $c = $n; $c = $c & 1 ? $xor ^ ($c >> 1) : $c >> 1 for 1 .. 8; $crc[$n] = _revbe($c); } # generate crc for each value followed by one, two, and three zeros */ foreach my $n (0 .. 255) { my $c = ($crc[($crc[$n] >> 24) ^ 0] ^ ($crc[$n] << 8)) & 0xffffffff; $crc[$_ * 256 + $n] = $c for 1 .. 3; } map { _revbe($crc[$_]) } 0 .. 1023; }; sub _crc32 { my ($c, $b) = @_; return ($crct[($c ^ $b) & 0xff] ^ ($c >> 8)); } # _crc32 sub _revbe { my $w = shift; return (($w >> 24) + (($w >> 8) & 0xff00) + (($w & 0xff00) << 8) + (($w & 0xff) << 24)); } # _revbe sub _update_keys { use integer; my $c = shift; # signed int $keys[0] = _crc32($keys[0], $c); $keys[1] = (($keys[1] + ($keys[0] & 0xff)) * 0x08088405 + 1) & 0xffffffff; my $keyshift = $keys[1] >> 24; $keys[2] = _crc32($keys[2], $keyshift); } # _update_keys sub _zdecode ($) { my $c = shift; my $t = ($keys[2] & 0xffff) | 2; _update_keys($c ^= ((($t * ($t ^ 1)) >> 8) & 0xff)); return $c; } # _zdecode sub _decode { my $self = shift; my $buff = shift; $self->isEncrypted or return $buff; my $pass = $self->password; defined $pass or return ""; @keys = (0x12345678, 0x23456789, 0x34567890); _update_keys($_) for unpack "C*", $pass; # DDumper { uk => [ @keys ] }; my $head = substr $buff, 0, 12, ""; my @head = map { _zdecode($_) } unpack "C*", $head; my $x = $self->{externalFileAttributes} ? ($self->{lastModFileDateTime} >> 8) & 0xff : $self->{crc32} >> 24; $head[-1] == $x or return ""; # Password fail # Worth checking ... $self->{crc32c} = (unpack LOCAL_FILE_HEADER_FORMAT, pack "C*", @head)[3]; # DHexDump ($buff); $buff = pack "C*" => map { _zdecode($_) } unpack "C*" => $buff; # DHexDump ($buff); return $buff; } # _decode 1; Zip/BufferedFileHandle.pm000064400000005272152347641370011324 0ustar00package Archive::Zip::BufferedFileHandle; # File handle that uses a string internally and can seek # This is given as a demo for getting a zip file written # to a string. # I probably should just use IO::Scalar instead. # Ned Konz, March 2000 use strict; use IO::File; use Carp; use vars qw{$VERSION}; BEGIN { $VERSION = '1.60'; $VERSION = eval $VERSION; } sub new { my $class = shift || __PACKAGE__; $class = ref($class) || $class; my $self = bless( { content => '', position => 0, size => 0 }, $class ); return $self; } # Utility method to read entire file sub readFromFile { my $self = shift; my $fileName = shift; my $fh = IO::File->new($fileName, "r"); CORE::binmode($fh); if (!$fh) { Carp::carp("Can't open $fileName: $!\n"); return undef; } local $/ = undef; $self->{content} = <$fh>; $self->{size} = length($self->{content}); return $self; } sub contents { my $self = shift; if (@_) { $self->{content} = shift; $self->{size} = length($self->{content}); } return $self->{content}; } sub binmode { 1 } sub close { 1 } sub opened { 1 } sub eof { my $self = shift; return $self->{position} >= $self->{size}; } sub seek { my $self = shift; my $pos = shift; my $whence = shift; # SEEK_SET if ($whence == 0) { $self->{position} = $pos; } # SEEK_CUR elsif ($whence == 1) { $self->{position} += $pos; } # SEEK_END elsif ($whence == 2) { $self->{position} = $self->{size} + $pos; } else { return 0; } return 1; } sub tell { return shift->{position}; } # Copy my data to given buffer sub read { my $self = shift; my $buf = \($_[0]); shift; my $len = shift; my $offset = shift || 0; $$buf = '' if not defined($$buf); my $bytesRead = ($self->{position} + $len > $self->{size}) ? ($self->{size} - $self->{position}) : $len; substr($$buf, $offset, $bytesRead) = substr($self->{content}, $self->{position}, $bytesRead); $self->{position} += $bytesRead; return $bytesRead; } # Copy given buffer to me sub write { my $self = shift; my $buf = \($_[0]); shift; my $len = shift; my $offset = shift || 0; $$buf = '' if not defined($$buf); my $bufLen = length($$buf); my $bytesWritten = ($offset + $len > $bufLen) ? $bufLen - $offset : $len; substr($self->{content}, $self->{position}, $bytesWritten) = substr($$buf, $offset, $bytesWritten); $self->{size} = length($self->{content}); return $bytesWritten; } sub clearerr() { 1 } 1; Zip/DirectoryMember.pm000064400000003676152347641370010770 0ustar00package Archive::Zip::DirectoryMember; use strict; use File::Path; use vars qw( $VERSION @ISA ); BEGIN { $VERSION = '1.60'; @ISA = qw( Archive::Zip::Member ); } use Archive::Zip qw( :ERROR_CODES :UTILITY_METHODS ); sub _newNamed { my $class = shift; my $fileName = shift; # FS name my $newName = shift; # Zip name $newName = _asZipDirName($fileName) unless $newName; my $self = $class->new(@_); $self->{'externalFileName'} = $fileName; $self->fileName($newName); if (-e $fileName) { # -e does NOT do a full stat, so we need to do one now if (-d _ ) { my @stat = stat(_); $self->unixFileAttributes($stat[2]); my $mod_t = $stat[9]; if ($^O eq 'MSWin32' and !$mod_t) { $mod_t = time(); } $self->setLastModFileDateTimeFromUnix($mod_t); } else { # hmm.. trying to add a non-directory? _error($fileName, ' exists but is not a directory'); return undef; } } else { $self->unixFileAttributes($self->DEFAULT_DIRECTORY_PERMISSIONS); $self->setLastModFileDateTimeFromUnix(time()); } return $self; } sub externalFileName { shift->{'externalFileName'}; } sub isDirectory { return 1; } sub extractToFileNamed { my $self = shift; my $name = shift; # local FS name my $attribs = $self->unixFileAttributes() & 07777; mkpath($name, 0, $attribs); # croaks on error utime($self->lastModTime(), $self->lastModTime(), $name); return AZ_OK; } sub fileName { my $self = shift; my $newName = shift; $newName =~ s{/?$}{/} if defined($newName); return $self->SUPER::fileName($newName); } # So people don't get too confused. This way it looks like the problem # is in their code... sub contents { return wantarray ? (undef, AZ_OK) : undef; } 1; Zip/FAQ.pod000064400000030556152347641370006446 0ustar00=head1 NAME Archive::Zip::FAQ - Answers to a few frequently asked questions about Archive::Zip =head1 DESCRIPTION It seems that I keep answering the same questions over and over again. I assume that this is because my documentation is deficient, rather than that people don't read the documentation. So this FAQ is an attempt to cut down on the number of personal answers I have to give. At least I can now say "You I read the FAQ, right?". The questions are not in any particular order. The answers assume the current version of Archive::Zip; some of the answers depend on newly added/fixed functionality. =head1 Install problems on RedHat 8 or 9 with Perl 5.8.0 B Archive::Zip won't install on my RedHat 9 system! It's broke! B This has become something of a FAQ. Basically, RedHat broke some versions of Perl by setting LANG to UTF8. They apparently have a fixed version out as an update. You might try running CPAN or creating your Makefile after exporting the LANG environment variable as C L =head1 Why is my zip file so big? B My zip file is actually bigger than what I stored in it! Why? B Some things to make sure of: =over 4 =item Make sure that you are requesting COMPRESSION_DEFLATED if you are storing strings. $member->desiredCompressionMethod( COMPRESSION_DEFLATED ); =item Don't make lots of little files if you can help it. Since zip computes the compression tables for each member, small members without much entropy won't compress well. Instead, if you've got lots of repeated strings in your data, try to combine them into one big member. =item Make sure that you are requesting COMPRESSION_STORED if you are storing things that are already compressed. If you're storing a .zip, .jpg, .mp3, or other compressed file in a zip, then don't compress them again. They'll get bigger. =back =head1 Sample code? B Can you send me code to do (whatever)? B Have you looked in the C directory yet? It contains: =over 4 =item examples/calcSizes.pl -- How to find out how big a Zip file will be before writing it =item examples/copy.pl -- Copies one Zip file to another =item examples/extract.pl -- extract file(s) from a Zip =item examples/mailZip.pl -- make and mail a zip file =item examples/mfh.pl -- demo for use of MockFileHandle =item examples/readScalar.pl -- shows how to use IO::Scalar as the source of a Zip read =item examples/selfex.pl -- a brief example of a self-extracting Zip =item examples/unzipAll.pl -- uses Archive::Zip::Tree to unzip an entire Zip =item examples/updateZip.pl -- shows how to read/modify/write a Zip =item examples/updateTree.pl -- shows how to update a Zip in place =item examples/writeScalar.pl -- shows how to use IO::Scalar as the destination of a Zip write =item examples/writeScalar2.pl -- shows how to use IO::String as the destination of a Zip write =item examples/zip.pl -- Constructs a Zip file =item examples/zipcheck.pl -- One way to check a Zip file for validity =item examples/zipinfo.pl -- Prints out information about a Zip archive file =item examples/zipGrep.pl -- Searches for text in Zip files =item examples/ziptest.pl -- Lists a Zip file and checks member CRCs =item examples/ziprecent.pl -- Puts recent files into a zipfile =item examples/ziptest.pl -- Another way to check a Zip file for validity =back =head1 Can't Read/modify/write same Zip file B Why can't I open a Zip file, add a member, and write it back? I get an error message when I try. B Because Archive::Zip doesn't (and can't, generally) read file contents into memory, the original Zip file is required to stay around until the writing of the new file is completed. The best way to do this is to write the Zip to a temporary file and then rename the temporary file to have the old name (possibly after deleting the old one). Archive::Zip v1.02 added the archive methods C and C to do this simply and carefully. See C for an example of this technique. =head1 File creation time not set B Upon extracting files, I see that their modification (and access) times are set to the time in the Zip archive. However, their creation time is not set to the same time. Why? B Mostly because Perl doesn't give cross-platform access to I. Indeed, many systems (like Unix) don't support such a concept. However, if yours does, you can easily set it. Get the modification time from the member using C. =head1 Can't use Archive::Zip on gzip files B Can I use Archive::Zip to extract Unix gzip files? B No. There is a distinction between Unix gzip files, and Zip archives that also can use the gzip compression. Depending on the format of the gzip file, you can use L, or L to decompress it (and de-archive it in the case of Tar files). You can unzip PKZIP/WinZip/etc/ archives using Archive::Zip (that's what it's for) as long as any compressed members are compressed using Deflate compression. =head1 Add a directory/tree to a Zip B How can I add a directory (or tree) full of files to a Zip? B You can use the Archive::Zip::addTree*() methods: use Archive::Zip; my $zip = Archive::Zip->new(); # add all readable files and directories below . as xyz/* $zip->addTree( '.', 'xyz' ); # add all readable plain files below /abc as def/* $zip->addTree( '/abc', 'def', sub { -f && -r } ); # add all .c files below /tmp as stuff/* $zip->addTreeMatching( '/tmp', 'stuff', '\.c$' ); # add all .o files below /tmp as stuff/* if they aren't writable $zip->addTreeMatching( '/tmp', 'stuff', '\.o$', sub { ! -w } ); # add all .so files below /tmp that are smaller than 200 bytes as stuff/* $zip->addTreeMatching( '/tmp', 'stuff', '\.o$', sub { -s < 200 } ); # and write them into a file $zip->writeToFileNamed('xxx.zip'); =head1 Extract a directory/tree B How can I extract some (or all) files from a Zip into a different directory? B You can use the Archive::Zip::extractTree() method: ??? || # now extract the same files into /tmpx $zip->extractTree( 'stuff', '/tmpx' ); =head1 Update a directory/tree B How can I update a Zip from a directory tree, adding or replacing only the newer files? B You can use the Archive::Zip::updateTree() method that was added in version 1.09. =head1 Zip times might be off by 1 second B It bothers me greatly that my file times are wrong by one second about half the time. Why don't you do something about it? B Get over it. This is a result of the Zip format storing times in DOS format, which has a resolution of only two seconds. =head1 Zip times don't include time zone information B My file times don't respect time zones. What gives? B If this is important to you, please submit patches to read the various Extra Fields that encode times with time zones. I'm just using the DOS Date/Time, which doesn't have a time zone. =head1 How do I make a self-extracting Zip B I want to make a self-extracting Zip file. Can I do this? B Yes. You can write a self-extracting archive stub (that is, a version of unzip) to the output filehandle that you pass to writeToFileHandle(). See examples/selfex.pl for how to write a self-extracting archive. However, you should understand that this will only work on one kind of platform (the one for which the stub was compiled). =head1 How can I deal with Zips with prepended garbage (i.e. from Sircam) B How can I tell if a Zip has been damaged by adding garbage to the beginning or inside the file? B I added code for this for the Amavis virus scanner. You can query archives for their 'eocdOffset' property, which should be 0: if ($zip->eocdOffset > 0) { warn($zip->eocdOffset . " bytes of garbage at beginning or within Zip") } When members are extracted, this offset will be used to adjust the start of the member if necessary. =head1 Can't extract Shrunk files B I'm trying to extract a file out of a Zip produced by PKZIP, and keep getting this error message: error: Unsupported compression combination: read 6, write 0 B You can't uncompress this archive member. Archive::Zip only supports uncompressed members, and compressed members that are compressed using the compression supported by Compress::Raw::Zlib. That means only Deflated and Stored members. Your file is compressed using the Shrink format, which is not supported by Compress::Raw::Zlib. You could, perhaps, use a command-line UnZip program (like the Info-Zip one) to extract this. =head1 Can't do decryption B How do I decrypt encrypted Zip members? B With some other program or library. Archive::Zip doesn't support decryption, and probably never will (unless I write it). =head1 How to test file integrity? B How can Archive::Zip can test the validity of a Zip file? B If you try to decompress the file, the gzip streams will report errors if you have garbage. Most of the time. If you try to open the file and a central directory structure can't be found, an error will be reported. When a file is being read, if we can't find a proper PK.. signature in the right places we report a format error. If there is added garbage at the beginning of a Zip file (as inserted by some viruses), you can find out about it, but Archive::Zip will ignore it, and you can still use the archive. When it gets written back out the added stuff will be gone. There are two ready-to-use utilities in the examples directory that can be used to test file integrity, or that you can use as examples for your own code: =over 4 =item examples/zipcheck.pl shows how to use an attempted extraction to test a file. =item examples/ziptest.pl shows how to test CRCs in a file. =back =head1 Duplicate files in Zip? B Archive::Zip let me put the same file in my Zip twice! Why don't you prevent this? B As far as I can tell, this is not disallowed by the Zip spec. If you think it's a bad idea, check for it yourself: $zip->addFile($someFile, $someName) unless $zip->memberNamed($someName); I can even imagine cases where this might be useful (for instance, multiple versions of files). =head1 File ownership/permissions/ACLS/etc B Why doesn't Archive::Zip deal with file ownership, ACLs, etc.? B There is no standard way to represent these in the Zip file format. If you want to send me code to properly handle the various extra fields that have been used to represent these through the years, I'll look at it. =head1 I can't compile but ActiveState only has an old version of Archive::Zip B I've only installed modules using ActiveState's PPM program and repository. But they have a much older version of Archive::Zip than is in CPAN. Will you send me a newer PPM? B Probably not, unless I get lots of extra time. But there's no reason you can't install the version from CPAN. Archive::Zip is pure Perl, so all you need is NMAKE, which you can get for free from Microsoft (see the FAQ in the ActiveState documentation for details on how to install CPAN modules). =head1 My JPEGs (or MP3's) don't compress when I put them into Zips! B How come my JPEGs and MP3's don't compress much when I put them into Zips? B Because they're already compressed. =head1 Under Windows, things lock up/get damaged B I'm using Windows. When I try to use Archive::Zip, my machine locks up/makes funny sounds/displays a BSOD/corrupts data. How can I fix this? B First, try the newest version of Compress::Raw::Zlib. I know of Windows-related problems prior to v1.14 of that library. =head1 Zip contents in a scalar B I want to read a Zip file from (or write one to) a scalar variable instead of a file. How can I do this? B Use C and the C and C methods. See C and C. =head1 Reading from streams B How do I read from a stream (like for the Info-Zip C program)? B This is not currently supported, though writing to a stream is. Zip/ZipFileMember.pm000064400000032710152347641370010355 0ustar00package Archive::Zip::ZipFileMember; use strict; use vars qw( $VERSION @ISA ); BEGIN { $VERSION = '1.60'; @ISA = qw ( Archive::Zip::FileMember ); } use Archive::Zip qw( :CONSTANTS :ERROR_CODES :PKZIP_CONSTANTS :UTILITY_METHODS ); # Create a new Archive::Zip::ZipFileMember # given a filename and optional open file handle # sub _newFromZipFile { my $class = shift; my $fh = shift; my $externalFileName = shift; my $possibleEocdOffset = shift; # normally 0 my $self = $class->new( 'crc32' => 0, 'diskNumberStart' => 0, 'localHeaderRelativeOffset' => 0, 'dataOffset' => 0, # localHeaderRelativeOffset + header length @_ ); $self->{'externalFileName'} = $externalFileName; $self->{'fh'} = $fh; $self->{'possibleEocdOffset'} = $possibleEocdOffset; return $self; } sub isDirectory { my $self = shift; return (substr($self->fileName, -1, 1) eq '/' and $self->uncompressedSize == 0); } # Seek to the beginning of the local header, just past the signature. # Verify that the local header signature is in fact correct. # Update the localHeaderRelativeOffset if necessary by adding the possibleEocdOffset. # Returns status. sub _seekToLocalHeader { my $self = shift; my $where = shift; # optional my $previousWhere = shift; # optional $where = $self->localHeaderRelativeOffset() unless defined($where); # avoid loop on certain corrupt files (from Julian Field) return _formatError("corrupt zip file") if defined($previousWhere) && $where == $previousWhere; my $status; my $signature; $status = $self->fh()->seek($where, IO::Seekable::SEEK_SET); return _ioError("seeking to local header") unless $status; ($status, $signature) = _readSignature($self->fh(), $self->externalFileName(), LOCAL_FILE_HEADER_SIGNATURE); return $status if $status == AZ_IO_ERROR; # retry with EOCD offset if any was given. if ($status == AZ_FORMAT_ERROR && $self->{'possibleEocdOffset'}) { $status = $self->_seekToLocalHeader( $self->localHeaderRelativeOffset() + $self->{'possibleEocdOffset'}, $where ); if ($status == AZ_OK) { $self->{'localHeaderRelativeOffset'} += $self->{'possibleEocdOffset'}; $self->{'possibleEocdOffset'} = 0; } } return $status; } # Because I'm going to delete the file handle, read the local file # header if the file handle is seekable. If it is not, I assume that # I've already read the local header. # Return ( $status, $self ) sub _become { my $self = shift; my $newClass = shift; return $self if ref($self) eq $newClass; my $status = AZ_OK; if (_isSeekable($self->fh())) { my $here = $self->fh()->tell(); $status = $self->_seekToLocalHeader(); $status = $self->_readLocalFileHeader() if $status == AZ_OK; $self->fh()->seek($here, IO::Seekable::SEEK_SET); return $status unless $status == AZ_OK; } delete($self->{'eocdCrc32'}); delete($self->{'diskNumberStart'}); delete($self->{'localHeaderRelativeOffset'}); delete($self->{'dataOffset'}); return $self->SUPER::_become($newClass); } sub diskNumberStart { shift->{'diskNumberStart'}; } sub localHeaderRelativeOffset { shift->{'localHeaderRelativeOffset'}; } sub dataOffset { shift->{'dataOffset'}; } # Skip local file header, updating only extra field stuff. # Assumes that fh is positioned before signature. sub _skipLocalFileHeader { my $self = shift; my $header; my $bytesRead = $self->fh()->read($header, LOCAL_FILE_HEADER_LENGTH); if ($bytesRead != LOCAL_FILE_HEADER_LENGTH) { return _ioError("reading local file header"); } my $fileNameLength; my $extraFieldLength; my $bitFlag; ( undef, # $self->{'versionNeededToExtract'}, $bitFlag, undef, # $self->{'compressionMethod'}, undef, # $self->{'lastModFileDateTime'}, undef, # $crc32, undef, # $compressedSize, undef, # $uncompressedSize, $fileNameLength, $extraFieldLength ) = unpack(LOCAL_FILE_HEADER_FORMAT, $header); if ($fileNameLength) { $self->fh()->seek($fileNameLength, IO::Seekable::SEEK_CUR) or return _ioError("skipping local file name"); } if ($extraFieldLength) { $bytesRead = $self->fh()->read($self->{'localExtraField'}, $extraFieldLength); if ($bytesRead != $extraFieldLength) { return _ioError("reading local extra field"); } } $self->{'dataOffset'} = $self->fh()->tell(); if ($bitFlag & GPBF_HAS_DATA_DESCRIPTOR_MASK) { # Read the crc32, compressedSize, and uncompressedSize from the # extended data descriptor, which directly follows the compressed data. # # Skip over the compressed file data (assumes that EOCD compressedSize # was correct) $self->fh()->seek($self->{'compressedSize'}, IO::Seekable::SEEK_CUR) or return _ioError("seeking to extended local header"); # these values should be set correctly from before. my $oldCrc32 = $self->{'eocdCrc32'}; my $oldCompressedSize = $self->{'compressedSize'}; my $oldUncompressedSize = $self->{'uncompressedSize'}; my $status = $self->_readDataDescriptor(); return $status unless $status == AZ_OK; # The buffer withe encrypted data is prefixed with a new # encrypted 12 byte header. The size only changes when # the buffer is also compressed $self->isEncrypted && $oldUncompressedSize > $self->{uncompressedSize} and $oldUncompressedSize -= DATA_DESCRIPTOR_LENGTH; return _formatError( "CRC or size mismatch while skipping data descriptor") if ( $oldCrc32 != $self->{'crc32'} || $oldUncompressedSize != $self->{'uncompressedSize'}); $self->{'crc32'} = 0 if $self->compressionMethod() == COMPRESSION_STORED ; } return AZ_OK; } # Read from a local file header into myself. Returns AZ_OK if successful. # Assumes that fh is positioned after signature. # Note that crc32, compressedSize, and uncompressedSize will be 0 if # GPBF_HAS_DATA_DESCRIPTOR_MASK is set in the bitFlag. sub _readLocalFileHeader { my $self = shift; my $header; my $bytesRead = $self->fh()->read($header, LOCAL_FILE_HEADER_LENGTH); if ($bytesRead != LOCAL_FILE_HEADER_LENGTH) { return _ioError("reading local file header"); } my $fileNameLength; my $crc32; my $compressedSize; my $uncompressedSize; my $extraFieldLength; ( $self->{'versionNeededToExtract'}, $self->{'bitFlag'}, $self->{'compressionMethod'}, $self->{'lastModFileDateTime'}, $crc32, $compressedSize, $uncompressedSize, $fileNameLength, $extraFieldLength ) = unpack(LOCAL_FILE_HEADER_FORMAT, $header); if ($fileNameLength) { my $fileName; $bytesRead = $self->fh()->read($fileName, $fileNameLength); if ($bytesRead != $fileNameLength) { return _ioError("reading local file name"); } $self->fileName($fileName); } if ($extraFieldLength) { $bytesRead = $self->fh()->read($self->{'localExtraField'}, $extraFieldLength); if ($bytesRead != $extraFieldLength) { return _ioError("reading local extra field"); } } $self->{'dataOffset'} = $self->fh()->tell(); if ($self->hasDataDescriptor()) { # Read the crc32, compressedSize, and uncompressedSize from the # extended data descriptor. # Skip over the compressed file data (assumes that EOCD compressedSize # was correct) $self->fh()->seek($self->{'compressedSize'}, IO::Seekable::SEEK_CUR) or return _ioError("seeking to extended local header"); my $status = $self->_readDataDescriptor(); return $status unless $status == AZ_OK; } else { return _formatError( "CRC or size mismatch after reading data descriptor") if ( $self->{'crc32'} != $crc32 || $self->{'uncompressedSize'} != $uncompressedSize); } return AZ_OK; } # This will read the data descriptor, which is after the end of compressed file # data in members that have GPBF_HAS_DATA_DESCRIPTOR_MASK set in their bitFlag. # The only reliable way to find these is to rely on the EOCD compressedSize. # Assumes that file is positioned immediately after the compressed data. # Returns status; sets crc32, compressedSize, and uncompressedSize. sub _readDataDescriptor { my $self = shift; my $signatureData; my $header; my $crc32; my $compressedSize; my $uncompressedSize; my $bytesRead = $self->fh()->read($signatureData, SIGNATURE_LENGTH); return _ioError("reading header signature") if $bytesRead != SIGNATURE_LENGTH; my $signature = unpack(SIGNATURE_FORMAT, $signatureData); # unfortunately, the signature appears to be optional. if ($signature == DATA_DESCRIPTOR_SIGNATURE && ($signature != $self->{'crc32'})) { $bytesRead = $self->fh()->read($header, DATA_DESCRIPTOR_LENGTH); return _ioError("reading data descriptor") if $bytesRead != DATA_DESCRIPTOR_LENGTH; ($crc32, $compressedSize, $uncompressedSize) = unpack(DATA_DESCRIPTOR_FORMAT, $header); } else { $bytesRead = $self->fh()->read($header, DATA_DESCRIPTOR_LENGTH_NO_SIG); return _ioError("reading data descriptor") if $bytesRead != DATA_DESCRIPTOR_LENGTH_NO_SIG; $crc32 = $signature; ($compressedSize, $uncompressedSize) = unpack(DATA_DESCRIPTOR_FORMAT_NO_SIG, $header); } $self->{'eocdCrc32'} = $self->{'crc32'} unless defined($self->{'eocdCrc32'}); $self->{'crc32'} = $crc32; $self->{'compressedSize'} = $compressedSize; $self->{'uncompressedSize'} = $uncompressedSize; return AZ_OK; } # Read a Central Directory header. Return AZ_OK on success. # Assumes that fh is positioned right after the signature. sub _readCentralDirectoryFileHeader { my $self = shift; my $fh = $self->fh(); my $header = ''; my $bytesRead = $fh->read($header, CENTRAL_DIRECTORY_FILE_HEADER_LENGTH); if ($bytesRead != CENTRAL_DIRECTORY_FILE_HEADER_LENGTH) { return _ioError("reading central dir header"); } my ($fileNameLength, $extraFieldLength, $fileCommentLength); ( $self->{'versionMadeBy'}, $self->{'fileAttributeFormat'}, $self->{'versionNeededToExtract'}, $self->{'bitFlag'}, $self->{'compressionMethod'}, $self->{'lastModFileDateTime'}, $self->{'crc32'}, $self->{'compressedSize'}, $self->{'uncompressedSize'}, $fileNameLength, $extraFieldLength, $fileCommentLength, $self->{'diskNumberStart'}, $self->{'internalFileAttributes'}, $self->{'externalFileAttributes'}, $self->{'localHeaderRelativeOffset'} ) = unpack(CENTRAL_DIRECTORY_FILE_HEADER_FORMAT, $header); $self->{'eocdCrc32'} = $self->{'crc32'}; if ($fileNameLength) { $bytesRead = $fh->read($self->{'fileName'}, $fileNameLength); if ($bytesRead != $fileNameLength) { _ioError("reading central dir filename"); } } if ($extraFieldLength) { $bytesRead = $fh->read($self->{'cdExtraField'}, $extraFieldLength); if ($bytesRead != $extraFieldLength) { return _ioError("reading central dir extra field"); } } if ($fileCommentLength) { $bytesRead = $fh->read($self->{'fileComment'}, $fileCommentLength); if ($bytesRead != $fileCommentLength) { return _ioError("reading central dir file comment"); } } # NK 10/21/04: added to avoid problems with manipulated headers if ( $self->{'uncompressedSize'} != $self->{'compressedSize'} and $self->{'compressionMethod'} == COMPRESSION_STORED) { $self->{'uncompressedSize'} = $self->{'compressedSize'}; } $self->desiredCompressionMethod($self->compressionMethod()); return AZ_OK; } sub rewindData { my $self = shift; my $status = $self->SUPER::rewindData(@_); return $status unless $status == AZ_OK; return AZ_IO_ERROR unless $self->fh(); $self->fh()->clearerr(); # Seek to local file header. # The only reason that I'm doing this this way is that the extraField # length seems to be different between the CD header and the LF header. $status = $self->_seekToLocalHeader(); return $status unless $status == AZ_OK; # skip local file header $status = $self->_skipLocalFileHeader(); return $status unless $status == AZ_OK; # Seek to beginning of file data $self->fh()->seek($self->dataOffset(), IO::Seekable::SEEK_SET) or return _ioError("seeking to beginning of file data"); return AZ_OK; } # Return bytes read. Note that first parameter is a ref to a buffer. # my $data; # my ( $bytesRead, $status) = $self->readRawChunk( \$data, $chunkSize ); sub _readRawChunk { my ($self, $dataRef, $chunkSize) = @_; return (0, AZ_OK) unless $chunkSize; my $bytesRead = $self->fh()->read($$dataRef, $chunkSize) or return (0, _ioError("reading data")); return ($bytesRead, AZ_OK); } 1; Zip/MockFileHandle.pm000064400000002465152347641370010474 0ustar00package Archive::Zip::MockFileHandle; # Output file handle that calls a custom write routine # Ned Konz, March 2000 # This is provided to help with writing zip files # when you have to process them a chunk at a time. use strict; use vars qw{$VERSION}; BEGIN { $VERSION = '1.60'; $VERSION = eval $VERSION; } sub new { my $class = shift || __PACKAGE__; $class = ref($class) || $class; my $self = bless( { 'position' => 0, 'size' => 0 }, $class ); return $self; } sub eof { my $self = shift; return $self->{'position'} >= $self->{'size'}; } # Copy given buffer to me sub print { my $self = shift; my $bytes = join('', @_); my $bytesWritten = $self->writeHook($bytes); if ($self->{'position'} + $bytesWritten > $self->{'size'}) { $self->{'size'} = $self->{'position'} + $bytesWritten; } $self->{'position'} += $bytesWritten; return $bytesWritten; } # Called on each write. # Override in subclasses. # Return number of bytes written (0 on error). sub writeHook { my $self = shift; my $bytes = shift; return length($bytes); } sub binmode { 1 } sub close { 1 } sub clearerr { 1 } # I'm write-only! sub read { 0 } sub tell { return shift->{'position'} } sub opened { 1 } 1; Zip/NewFileMember.pm000064400000004212152347641370010340 0ustar00package Archive::Zip::NewFileMember; use strict; use vars qw( $VERSION @ISA ); BEGIN { $VERSION = '1.60'; @ISA = qw ( Archive::Zip::FileMember ); } use Archive::Zip qw( :CONSTANTS :ERROR_CODES :UTILITY_METHODS ); # Given a file name, set up for eventual writing. sub _newFromFileNamed { my $class = shift; my $fileName = shift; # local FS format my $newName = shift; $newName = _asZipDirName($fileName) unless defined($newName); return undef unless (stat($fileName) && -r _ && !-d _ ); my $self = $class->new(@_); $self->{'fileName'} = $newName; $self->{'externalFileName'} = $fileName; $self->{'compressionMethod'} = COMPRESSION_STORED; my @stat = stat(_); $self->{'compressedSize'} = $self->{'uncompressedSize'} = $stat[7]; $self->desiredCompressionMethod( ($self->compressedSize() > 0) ? COMPRESSION_DEFLATED : COMPRESSION_STORED ); $self->unixFileAttributes($stat[2]); $self->setLastModFileDateTimeFromUnix($stat[9]); $self->isTextFile(-T _ ); return $self; } sub rewindData { my $self = shift; my $status = $self->SUPER::rewindData(@_); return $status unless $status == AZ_OK; return AZ_IO_ERROR unless $self->fh(); $self->fh()->clearerr(); $self->fh()->seek(0, IO::Seekable::SEEK_SET) or return _ioError("rewinding", $self->externalFileName()); return AZ_OK; } # Return bytes read. Note that first parameter is a ref to a buffer. # my $data; # my ( $bytesRead, $status) = $self->readRawChunk( \$data, $chunkSize ); sub _readRawChunk { my ($self, $dataRef, $chunkSize) = @_; return (0, AZ_OK) unless $chunkSize; my $bytesRead = $self->fh()->read($$dataRef, $chunkSize) or return (0, _ioError("reading data")); return ($bytesRead, AZ_OK); } # If I already exist, extraction is a no-op. sub extractToFileNamed { my $self = shift; my $name = shift; # local FS name if (File::Spec->rel2abs($name) eq File::Spec->rel2abs($self->externalFileName()) and -r $name) { return AZ_OK; } else { return $self->SUPER::extractToFileNamed($name, @_); } } 1; Zip/MemberRead.pm000064400000017213152347641370007667 0ustar00package Archive::Zip::MemberRead; =head1 NAME Archive::Zip::MemberRead - A wrapper that lets you read Zip archive members as if they were files. =cut =head1 SYNOPSIS use Archive::Zip; use Archive::Zip::MemberRead; $zip = Archive::Zip->new("file.zip"); $fh = Archive::Zip::MemberRead->new($zip, "subdir/abc.txt"); while (defined($line = $fh->getline())) { print $fh->input_line_number . "#: $line\n"; } $read = $fh->read($buffer, 32*1024); print "Read $read bytes as :$buffer:\n"; =head1 DESCRIPTION The Archive::Zip::MemberRead module lets you read Zip archive member data just like you read data from files. =head1 METHODS =over 4 =cut use strict; use Archive::Zip qw( :ERROR_CODES :CONSTANTS ); use vars qw{$VERSION}; my $nl; BEGIN { $VERSION = '1.60'; $VERSION = eval $VERSION; # Requirement for newline conversion. Should check for e.g., DOS and OS/2 as well, but am too lazy. $nl = $^O eq 'MSWin32' ? "\r\n" : "\n"; } =item Archive::Zip::Member::readFileHandle() You can get a C from an archive member by calling C: my $member = $zip->memberNamed('abc/def.c'); my $fh = $member->readFileHandle(); while (defined($line = $fh->getline())) { # ... } $fh->close(); =cut sub Archive::Zip::Member::readFileHandle { return Archive::Zip::MemberRead->new(shift()); } =item Archive::Zip::MemberRead->new($zip, $fileName) =item Archive::Zip::MemberRead->new($zip, $member) =item Archive::Zip::MemberRead->new($member) Construct a new Archive::Zip::MemberRead on the specified member. my $fh = Archive::Zip::MemberRead->new($zip, 'fred.c') =cut sub new { my ($class, $zip, $file) = @_; my ($self, $member); if ($zip && $file) # zip and filename, or zip and member { $member = ref($file) ? $file : $zip->memberNamed($file); } elsif ($zip && !$file && ref($zip)) # just member { $member = $zip; } else { die( 'Archive::Zip::MemberRead::new needs a zip and filename, zip and member, or member' ); } $self = {}; bless($self, $class); $self->set_member($member); return $self; } sub set_member { my ($self, $member) = @_; $self->{member} = $member; $self->set_compression(COMPRESSION_STORED); $self->rewind(); } sub set_compression { my ($self, $compression) = @_; $self->{member}->desiredCompressionMethod($compression) if $self->{member}; } =item setLineEnd(expr) Set the line end character to use. This is set to \n by default except on Windows systems where it is set to \r\n. You will only need to set this on systems which are not Windows or Unix based and require a line end different from \n. This is a class method so call as C->C =cut sub setLineEnd { shift; $nl = shift; } =item rewind() Rewinds an C so that you can read from it again starting at the beginning. =cut sub rewind { my $self = shift; $self->_reset_vars(); $self->{member}->rewindData() if $self->{member}; } sub _reset_vars { my $self = shift; $self->{line_no} = 0; $self->{at_end} = 0; delete $self->{buffer}; } =item input_record_separator(expr) If the argument is given, input_record_separator for this instance is set to it. The current setting (which may be the global $/) is always returned. =cut sub input_record_separator { my $self = shift; if (@_) { $self->{sep} = shift; $self->{sep_re} = _sep_as_re($self->{sep}); # Cache the RE as an optimization } return exists $self->{sep} ? $self->{sep} : $/; } # Return the input_record_separator in use as an RE fragment # Note that if we have a per-instance input_record_separator # we can just return the already converted value. Otherwise, # the conversion must be done on $/ every time since we cannot # know whether it has changed or not. sub _sep_re { my $self = shift; # Important to phrase this way: sep's value may be undef. return exists $self->{sep} ? $self->{sep_re} : _sep_as_re($/); } # Convert the input record separator into an RE and return it. sub _sep_as_re { my $sep = shift; if (defined $sep) { if ($sep eq '') { return "(?:$nl){2,}"; } else { $sep =~ s/\n/$nl/og; return quotemeta $sep; } } else { return undef; } } =item input_line_number() Returns the current line number, but only if you're using C. Using C will not update the line number. =cut sub input_line_number { my $self = shift; return $self->{line_no}; } =item close() Closes the given file handle. =cut sub close { my $self = shift; $self->_reset_vars(); $self->{member}->endRead(); } =item buffer_size([ $size ]) Gets or sets the buffer size used for reads. Default is the chunk size used by Archive::Zip. =cut sub buffer_size { my ($self, $size) = @_; if (!$size) { return $self->{chunkSize} || Archive::Zip::chunkSize(); } else { $self->{chunkSize} = $size; } } =item getline() Returns the next line from the currently open member. Makes sense only for text files. A read error is considered fatal enough to die. Returns undef on eof. All subsequent calls would return undef, unless a rewind() is called. Note: The line returned has the input_record_separator (default: newline) removed. =item getline( { preserve_line_ending => 1 } ) Returns the next line including the line ending. =cut sub getline { my ($self, $argref) = @_; my $size = $self->buffer_size(); my $sep = $self->_sep_re(); my $preserve_line_ending; if (ref $argref eq 'HASH') { $preserve_line_ending = $argref->{'preserve_line_ending'}; $sep =~ s/\\([^A-Za-z_0-9])+/$1/g; } for (; ;) { if ( $sep && defined($self->{buffer}) && $self->{buffer} =~ s/^(.*?)$sep//s) { my $line = $1; $self->{line_no}++; if ($preserve_line_ending) { return $line . $sep; } else { return $line; } } elsif ($self->{at_end}) { $self->{line_no}++ if $self->{buffer}; return delete $self->{buffer}; } my ($temp, $status) = $self->{member}->readChunk($size); if ($status != AZ_OK && $status != AZ_STREAM_END) { die "ERROR: Error reading chunk from archive - $status"; } $self->{at_end} = $status == AZ_STREAM_END; $self->{buffer} .= $$temp; } } =item read($buffer, $num_bytes_to_read) Simulates a normal C system call. Returns the no. of bytes read. C on error, 0 on eof, I: $fh = Archive::Zip::MemberRead->new($zip, "sreeji/secrets.bin"); while (1) { $read = $fh->read($buffer, 1024); die "FATAL ERROR reading my secrets !\n" if (!defined($read)); last if (!$read); # Do processing. .... } =cut # # All these $_ are required to emulate read(). # sub read { my $self = $_[0]; my $size = $_[2]; my ($temp, $status, $ret); ($temp, $status) = $self->{member}->readChunk($size); if ($status != AZ_OK && $status != AZ_STREAM_END) { $_[1] = undef; $ret = undef; } else { $_[1] = $$temp; $ret = length($$temp); } return $ret; } 1; =back =head1 AUTHOR Sreeji K. Das Esreeji_k@yahoo.comE See L by Ned Konz without which this module does not make any sense! Minor mods by Ned Konz. =head1 COPYRIGHT Copyright 2002 Sreeji K. Das. This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself. =cut Zip.pm000064400000202144152347641370005663 0ustar00package Archive::Zip; use 5.006; use strict; use Carp (); use Cwd (); use IO::File (); use IO::Seekable (); use Compress::Raw::Zlib (); use File::Spec (); use File::Temp (); use FileHandle (); use vars qw( $VERSION @ISA ); BEGIN { $VERSION = '1.60'; require Exporter; @ISA = qw( Exporter ); } use vars qw( $ChunkSize $ErrorHandler ); BEGIN { # This is the size we'll try to read, write, and (de)compress. # You could set it to something different if you had lots of memory # and needed more speed. $ChunkSize ||= 32768; $ErrorHandler = \&Carp::carp; } # BEGIN block is necessary here so that other modules can use the constants. use vars qw( @EXPORT_OK %EXPORT_TAGS ); BEGIN { @EXPORT_OK = ('computeCRC32'); %EXPORT_TAGS = ( CONSTANTS => [ qw( FA_MSDOS FA_UNIX GPBF_ENCRYPTED_MASK GPBF_DEFLATING_COMPRESSION_MASK GPBF_HAS_DATA_DESCRIPTOR_MASK COMPRESSION_STORED COMPRESSION_DEFLATED COMPRESSION_LEVEL_NONE COMPRESSION_LEVEL_DEFAULT COMPRESSION_LEVEL_FASTEST COMPRESSION_LEVEL_BEST_COMPRESSION IFA_TEXT_FILE_MASK IFA_TEXT_FILE IFA_BINARY_FILE ) ], MISC_CONSTANTS => [ qw( FA_AMIGA FA_VAX_VMS FA_VM_CMS FA_ATARI_ST FA_OS2_HPFS FA_MACINTOSH FA_Z_SYSTEM FA_CPM FA_TOPS20 FA_WINDOWS_NTFS FA_QDOS FA_ACORN FA_VFAT FA_MVS FA_BEOS FA_TANDEM FA_THEOS GPBF_IMPLODING_8K_SLIDING_DICTIONARY_MASK GPBF_IMPLODING_3_SHANNON_FANO_TREES_MASK GPBF_IS_COMPRESSED_PATCHED_DATA_MASK COMPRESSION_SHRUNK DEFLATING_COMPRESSION_NORMAL DEFLATING_COMPRESSION_MAXIMUM DEFLATING_COMPRESSION_FAST DEFLATING_COMPRESSION_SUPER_FAST COMPRESSION_REDUCED_1 COMPRESSION_REDUCED_2 COMPRESSION_REDUCED_3 COMPRESSION_REDUCED_4 COMPRESSION_IMPLODED COMPRESSION_TOKENIZED COMPRESSION_DEFLATED_ENHANCED COMPRESSION_PKWARE_DATA_COMPRESSION_LIBRARY_IMPLODED ) ], ERROR_CODES => [ qw( AZ_OK AZ_STREAM_END AZ_ERROR AZ_FORMAT_ERROR AZ_IO_ERROR ) ], # For Internal Use Only PKZIP_CONSTANTS => [ qw( SIGNATURE_FORMAT SIGNATURE_LENGTH LOCAL_FILE_HEADER_SIGNATURE LOCAL_FILE_HEADER_FORMAT LOCAL_FILE_HEADER_LENGTH DATA_DESCRIPTOR_SIGNATURE DATA_DESCRIPTOR_FORMAT DATA_DESCRIPTOR_LENGTH DATA_DESCRIPTOR_FORMAT_NO_SIG DATA_DESCRIPTOR_LENGTH_NO_SIG CENTRAL_DIRECTORY_FILE_HEADER_SIGNATURE CENTRAL_DIRECTORY_FILE_HEADER_FORMAT CENTRAL_DIRECTORY_FILE_HEADER_LENGTH ZIP64_END_OF_CENTRAL_DIRECTORY_RECORD_SIGNATURE ZIP64_END_OF_CENTRAL_DIRECTORY_RECORD_FORMAT ZIP64_END_OF_CENTRAL_DIRECTORY_RECORD_LENGTH ZIP64_END_OF_CENTRAL_DIRECTORY_LOCATOR_SIGNATURE ZIP64_END_OF_CENTRAL_DIRECTORY_LOCATOR_FORMAT ZIP64_END_OF_CENTRAL_DIRECTORY_LOCATOR_LENGTH END_OF_CENTRAL_DIRECTORY_SIGNATURE END_OF_CENTRAL_DIRECTORY_FORMAT END_OF_CENTRAL_DIRECTORY_LENGTH END_OF_CENTRAL_DIRECTORY_SIGNATURE_STRING ) ], # For Internal Use Only UTILITY_METHODS => [ qw( _error _printError _ioError _formatError _subclassResponsibility _binmode _isSeekable _newFileHandle _readSignature _asZipDirName ) ], ); # Add all the constant names and error code names to @EXPORT_OK Exporter::export_ok_tags( qw( CONSTANTS ERROR_CODES PKZIP_CONSTANTS UTILITY_METHODS MISC_CONSTANTS )); } # Error codes use constant AZ_OK => 0; use constant AZ_STREAM_END => 1; use constant AZ_ERROR => 2; use constant AZ_FORMAT_ERROR => 3; use constant AZ_IO_ERROR => 4; # File types # Values of Archive::Zip::Member->fileAttributeFormat() use constant FA_MSDOS => 0; use constant FA_AMIGA => 1; use constant FA_VAX_VMS => 2; use constant FA_UNIX => 3; use constant FA_VM_CMS => 4; use constant FA_ATARI_ST => 5; use constant FA_OS2_HPFS => 6; use constant FA_MACINTOSH => 7; use constant FA_Z_SYSTEM => 8; use constant FA_CPM => 9; use constant FA_TOPS20 => 10; use constant FA_WINDOWS_NTFS => 11; use constant FA_QDOS => 12; use constant FA_ACORN => 13; use constant FA_VFAT => 14; use constant FA_MVS => 15; use constant FA_BEOS => 16; use constant FA_TANDEM => 17; use constant FA_THEOS => 18; # general-purpose bit flag masks # Found in Archive::Zip::Member->bitFlag() use constant GPBF_ENCRYPTED_MASK => 1 << 0; use constant GPBF_DEFLATING_COMPRESSION_MASK => 3 << 1; use constant GPBF_HAS_DATA_DESCRIPTOR_MASK => 1 << 3; # deflating compression types, if compressionMethod == COMPRESSION_DEFLATED # ( Archive::Zip::Member->bitFlag() & GPBF_DEFLATING_COMPRESSION_MASK ) use constant DEFLATING_COMPRESSION_NORMAL => 0 << 1; use constant DEFLATING_COMPRESSION_MAXIMUM => 1 << 1; use constant DEFLATING_COMPRESSION_FAST => 2 << 1; use constant DEFLATING_COMPRESSION_SUPER_FAST => 3 << 1; # compression method # these two are the only ones supported in this module use constant COMPRESSION_STORED => 0; # file is stored (no compression) use constant COMPRESSION_DEFLATED => 8; # file is Deflated use constant COMPRESSION_LEVEL_NONE => 0; use constant COMPRESSION_LEVEL_DEFAULT => -1; use constant COMPRESSION_LEVEL_FASTEST => 1; use constant COMPRESSION_LEVEL_BEST_COMPRESSION => 9; # internal file attribute bits # Found in Archive::Zip::Member::internalFileAttributes() use constant IFA_TEXT_FILE_MASK => 1; use constant IFA_TEXT_FILE => 1; use constant IFA_BINARY_FILE => 0; # PKZIP file format miscellaneous constants (for internal use only) use constant SIGNATURE_FORMAT => "V"; use constant SIGNATURE_LENGTH => 4; # these lengths are without the signature. use constant LOCAL_FILE_HEADER_SIGNATURE => 0x04034b50; use constant LOCAL_FILE_HEADER_FORMAT => "v3 V4 v2"; use constant LOCAL_FILE_HEADER_LENGTH => 26; # PKZIP docs don't mention the signature, but Info-Zip writes it. use constant DATA_DESCRIPTOR_SIGNATURE => 0x08074b50; use constant DATA_DESCRIPTOR_FORMAT => "V3"; use constant DATA_DESCRIPTOR_LENGTH => 12; # but the signature is apparently optional. use constant DATA_DESCRIPTOR_FORMAT_NO_SIG => "V2"; use constant DATA_DESCRIPTOR_LENGTH_NO_SIG => 8; use constant CENTRAL_DIRECTORY_FILE_HEADER_SIGNATURE => 0x02014b50; use constant CENTRAL_DIRECTORY_FILE_HEADER_FORMAT => "C2 v3 V4 v5 V2"; use constant CENTRAL_DIRECTORY_FILE_HEADER_LENGTH => 42; # zip64 support use constant ZIP64_END_OF_CENTRAL_DIRECTORY_RECORD_SIGNATURE => 0x06064b50; use constant ZIP64_END_OF_CENTRAL_DIRECTORY_RECORD_FORMAT => 0; use constant ZIP64_END_OF_CENTRAL_DIRECTORY_RECORD_LENGTH => 0; use constant ZIP64_END_OF_CENTRAL_DIRECTORY_LOCATOR_SIGNATURE => 0x07064b50; use constant ZIP64_END_OF_CENTRAL_DIRECTORY_LOCATOR_FORMAT => 0; use constant ZIP64_END_OF_CENTRAL_DIRECTORY_LOCATOR_LENGTH => 0; use constant END_OF_CENTRAL_DIRECTORY_SIGNATURE => 0x06054b50; use constant END_OF_CENTRAL_DIRECTORY_SIGNATURE_STRING => pack("V", END_OF_CENTRAL_DIRECTORY_SIGNATURE); use constant END_OF_CENTRAL_DIRECTORY_FORMAT => "v4 V2 v"; use constant END_OF_CENTRAL_DIRECTORY_LENGTH => 18; use constant GPBF_IMPLODING_8K_SLIDING_DICTIONARY_MASK => 1 << 1; use constant GPBF_IMPLODING_3_SHANNON_FANO_TREES_MASK => 1 << 2; use constant GPBF_IS_COMPRESSED_PATCHED_DATA_MASK => 1 << 5; # the rest of these are not supported in this module use constant COMPRESSION_SHRUNK => 1; # file is Shrunk use constant COMPRESSION_REDUCED_1 => 2; # file is Reduced CF=1 use constant COMPRESSION_REDUCED_2 => 3; # file is Reduced CF=2 use constant COMPRESSION_REDUCED_3 => 4; # file is Reduced CF=3 use constant COMPRESSION_REDUCED_4 => 5; # file is Reduced CF=4 use constant COMPRESSION_IMPLODED => 6; # file is Imploded use constant COMPRESSION_TOKENIZED => 7; # reserved for Tokenizing compr. use constant COMPRESSION_DEFLATED_ENHANCED => 9; # reserved for enh. Deflating use constant COMPRESSION_PKWARE_DATA_COMPRESSION_LIBRARY_IMPLODED => 10; # Load the various required classes require Archive::Zip::Archive; require Archive::Zip::Member; require Archive::Zip::FileMember; require Archive::Zip::DirectoryMember; require Archive::Zip::ZipFileMember; require Archive::Zip::NewFileMember; require Archive::Zip::StringMember; # Convenience functions sub _ISA ($$) { # Can't rely on Scalar::Util, so use the next best way local $@; !!eval { ref $_[0] and $_[0]->isa($_[1]) }; } sub _CAN ($$) { local $@; !!eval { ref $_[0] and $_[0]->can($_[1]) }; } ##################################################################### # Methods sub new { my $class = shift; return Archive::Zip::Archive->new(@_); } sub computeCRC32 { my ($data, $crc); if (ref($_[0]) eq 'HASH') { $data = $_[0]->{string}; $crc = $_[0]->{checksum}; } else { $data = shift; $data = shift if ref($data); $crc = shift; } return Compress::Raw::Zlib::crc32($data, $crc); } # Report or change chunk size used for reading and writing. # Also sets Zlib's default buffer size (eventually). sub setChunkSize { shift if ref($_[0]) eq 'Archive::Zip::Archive'; my $chunkSize = (ref($_[0]) eq 'HASH') ? shift->{chunkSize} : shift; my $oldChunkSize = $Archive::Zip::ChunkSize; $Archive::Zip::ChunkSize = $chunkSize if ($chunkSize); return $oldChunkSize; } sub chunkSize { return $Archive::Zip::ChunkSize; } sub setErrorHandler { my $errorHandler = (ref($_[0]) eq 'HASH') ? shift->{subroutine} : shift; $errorHandler = \&Carp::carp unless defined($errorHandler); my $oldErrorHandler = $Archive::Zip::ErrorHandler; $Archive::Zip::ErrorHandler = $errorHandler; return $oldErrorHandler; } ###################################################################### # Private utility functions (not methods). sub _printError { my $string = join(' ', @_, "\n"); my $oldCarpLevel = $Carp::CarpLevel; $Carp::CarpLevel += 2; &{$ErrorHandler}($string); $Carp::CarpLevel = $oldCarpLevel; } # This is called on format errors. sub _formatError { shift if ref($_[0]); _printError('format error:', @_); return AZ_FORMAT_ERROR; } # This is called on IO errors. sub _ioError { shift if ref($_[0]); _printError('IO error:', @_, ':', $!); return AZ_IO_ERROR; } # This is called on generic errors. sub _error { shift if ref($_[0]); _printError('error:', @_); return AZ_ERROR; } # Called when a subclass should have implemented # something but didn't sub _subclassResponsibility { Carp::croak("subclass Responsibility\n"); } # Try to set the given file handle or object into binary mode. sub _binmode { my $fh = shift; return _CAN($fh, 'binmode') ? $fh->binmode() : binmode($fh); } # Attempt to guess whether file handle is seekable. # Because of problems with Windows, this only returns true when # the file handle is a real file. sub _isSeekable { my $fh = shift; return 0 unless ref $fh; _ISA($fh, "IO::Scalar") # IO::Scalar objects are brokenly-seekable and return 0; _ISA($fh, "IO::String") and return 1; if (_ISA($fh, "IO::Seekable")) { # Unfortunately, some things like FileHandle objects # return true for Seekable, but AREN'T!!!!! _ISA($fh, "FileHandle") and return 0; return 1; } # open my $fh, "+<", \$data; ref $fh eq "GLOB" && eval { seek $fh, 0, 1 } and return 1; _CAN($fh, "stat") and return -f $fh; return (_CAN($fh, "seek") and _CAN($fh, "tell")) ? 1 : 0; } # Print to the filehandle, while making sure the pesky Perl special global # variables don't interfere. sub _print { my ($self, $fh, @data) = @_; local $\; return $fh->print(@data); } # Return an opened IO::Handle # my ( $status, fh ) = _newFileHandle( 'fileName', 'w' ); # Can take a filename, file handle, or ref to GLOB # Or, if given something that is a ref but not an IO::Handle, # passes back the same thing. sub _newFileHandle { my $fd = shift; my $status = 1; my $handle; if (ref($fd)) { if (_ISA($fd, 'IO::Scalar') or _ISA($fd, 'IO::String')) { $handle = $fd; } elsif (_ISA($fd, 'IO::Handle') or ref($fd) eq 'GLOB') { $handle = IO::File->new; $status = $handle->fdopen($fd, @_); } else { $handle = $fd; } } else { $handle = IO::File->new; $status = $handle->open($fd, @_); } return ($status, $handle); } # Returns next signature from given file handle, leaves # file handle positioned afterwards. # In list context, returns ($status, $signature) # ( $status, $signature) = _readSignature( $fh, $fileName ); sub _readSignature { my $fh = shift; my $fileName = shift; my $expectedSignature = shift; # optional my $signatureData; my $bytesRead = $fh->read($signatureData, SIGNATURE_LENGTH); if ($bytesRead != SIGNATURE_LENGTH) { return _ioError("reading header signature"); } my $signature = unpack(SIGNATURE_FORMAT, $signatureData); my $status = AZ_OK; # compare with expected signature, if any, or any known signature. if ( (defined($expectedSignature) && $signature != $expectedSignature) || ( !defined($expectedSignature) && $signature != CENTRAL_DIRECTORY_FILE_HEADER_SIGNATURE && $signature != LOCAL_FILE_HEADER_SIGNATURE && $signature != END_OF_CENTRAL_DIRECTORY_SIGNATURE && $signature != DATA_DESCRIPTOR_SIGNATURE && $signature != ZIP64_END_OF_CENTRAL_DIRECTORY_RECORD_SIGNATURE && $signature != ZIP64_END_OF_CENTRAL_DIRECTORY_LOCATOR_SIGNATURE ) ) { my $errmsg = sprintf("bad signature: 0x%08x", $signature); if (_isSeekable($fh)) { $errmsg .= sprintf(" at offset %d", $fh->tell() - SIGNATURE_LENGTH); } $status = _formatError("$errmsg in file $fileName"); } return ($status, $signature); } # Utility method to make and open a temp file. # Will create $temp_dir if it does not exist. # Returns file handle and name: # # my ($fh, $name) = Archive::Zip::tempFile(); # my ($fh, $name) = Archive::Zip::tempFile('mytempdir'); # sub tempFile { my $dir = (ref($_[0]) eq 'HASH') ? shift->{tempDir} : shift; my ($fh, $filename) = File::Temp::tempfile( SUFFIX => '.zip', UNLINK => 1, $dir ? (DIR => $dir) : ()); return (undef, undef) unless $fh; my ($status, $newfh) = _newFileHandle($fh, 'w+'); $fh->close(); return ($newfh, $filename); } # Return the normalized directory name as used in a zip file (path # separators become slashes, etc.). # Will translate internal slashes in path components (i.e. on Macs) to # underscores. Discards volume names. # When $forceDir is set, returns paths with trailing slashes (or arrays # with trailing blank members). # # If third argument is a reference, returns volume information there. # # input output # . ('.') '.' # ./a ('a') a # ./a/b ('a','b') a/b # ./a/b/ ('a','b') a/b # a/b/ ('a','b') a/b # /a/b/ ('','a','b') a/b # c:\a\b\c.doc ('','a','b','c.doc') a/b/c.doc # on Windows # "i/o maps:whatever" ('i_o maps', 'whatever') "i_o maps/whatever" # on Macs sub _asZipDirName { my $name = shift; my $forceDir = shift; my $volReturn = shift; my ($volume, $directories, $file) = File::Spec->splitpath(File::Spec->canonpath($name), $forceDir); $$volReturn = $volume if (ref($volReturn)); my @dirs = map { $_ =~ s{/}{_}g; $_ } File::Spec->splitdir($directories); if (@dirs > 0) { pop(@dirs) unless $dirs[-1] } # remove empty component push(@dirs, defined($file) ? $file : ''); #return wantarray ? @dirs : join ( '/', @dirs ); my $normalised_path = join '/', @dirs; # Leading directory separators should not be stored in zip archives. # Example: # C:\a\b\c\ a/b/c # C:\a\b\c.txt a/b/c.txt # /a/b/c/ a/b/c # /a/b/c.txt a/b/c.txt $normalised_path =~ s{^/}{}; # remove leading separator return $normalised_path; } # Return an absolute local name for a zip name. # Assume a directory if zip name has trailing slash. # Takes an optional volume name in FS format (like 'a:'). # sub _asLocalName { my $name = shift; # zip format my $volume = shift; $volume = '' unless defined($volume); # local FS format my @paths = split(/\//, $name); my $filename = pop(@paths); $filename = '' unless defined($filename); my $localDirs = @paths ? File::Spec->catdir(@paths) : ''; my $localName = File::Spec->catpath($volume, $localDirs, $filename); unless ($volume) { $localName = File::Spec->rel2abs($localName, Cwd::getcwd()); } return $localName; } 1; __END__ =pod =encoding utf8 =head1 NAME Archive::Zip - Provide an interface to ZIP archive files. =head1 SYNOPSIS # Create a Zip file use Archive::Zip qw( :ERROR_CODES :CONSTANTS ); my $zip = Archive::Zip->new(); # Add a directory my $dir_member = $zip->addDirectory( 'dirname/' ); # Add a file from a string with compression my $string_member = $zip->addString( 'This is a test', 'stringMember.txt' ); $string_member->desiredCompressionMethod( COMPRESSION_DEFLATED ); # Add a file from disk my $file_member = $zip->addFile( 'xyz.pl', 'AnotherName.pl' ); # Save the Zip file unless ( $zip->writeToFileNamed('someZip.zip') == AZ_OK ) { die 'write error'; } # Read a Zip file my $somezip = Archive::Zip->new(); unless ( $somezip->read( 'someZip.zip' ) == AZ_OK ) { die 'read error'; } # Change the compression type for a file in the Zip my $member = $somezip->memberNamed( 'stringMember.txt' ); $member->desiredCompressionMethod( COMPRESSION_STORED ); unless ( $zip->writeToFileNamed( 'someOtherZip.zip' ) == AZ_OK ) { die 'write error'; } =head1 DESCRIPTION The Archive::Zip module allows a Perl program to create, manipulate, read, and write Zip archive files. Zip archives can be created, or you can read from existing zip files. Once created, they can be written to files, streams, or strings. Members can be added, removed, extracted, replaced, rearranged, and enumerated. They can also be renamed or have their dates, comments, or other attributes queried or modified. Their data can be compressed or uncompressed as needed. Members can be created from members in existing Zip files, or from existing directories, files, or strings. This module uses the L library to read and write the compressed streams inside the files. One can use L to read the zip file archive members as if they were files. =head2 File Naming Regardless of what your local file system uses for file naming, names in a Zip file are in Unix format (I slashes (/) separating directory names, etc.). C tries to be consistent with file naming conventions, and will translate back and forth between native and Zip file names. However, it can't guess which format names are in. So two rules control what kind of file name you must pass various routines: =over 4 =item Names of files are in local format. C and C are used for various file operations. When you're referring to a file on your system, use its file naming conventions. =item Names of archive members are in Unix format. This applies to every method that refers to an archive member, or provides a name for new archive members. The C methods that can take one or two names will convert from local to zip names if you call them with a single name. =back =head2 Archive::Zip Object Model =head3 Overview Archive::Zip::Archive objects are what you ordinarily deal with. These maintain the structure of a zip file, without necessarily holding data. When a zip is read from a disk file, the (possibly compressed) data still lives in the file, not in memory. Archive members hold information about the individual members, but not (usually) the actual member data. When the zip is written to a (different) file, the member data is compressed or copied as needed. It is possible to make archive members whose data is held in a string in memory, but this is not done when a zip file is read. Directory members don't have any data. =head2 Inheritance Exporter Archive::Zip Common base class, has defs. Archive::Zip::Archive A Zip archive. Archive::Zip::Member Abstract superclass for all members. Archive::Zip::StringMember Member made from a string Archive::Zip::FileMember Member made from an external file Archive::Zip::ZipFileMember Member that lives in a zip file Archive::Zip::NewFileMember Member whose data is in a file Archive::Zip::DirectoryMember Member that is a directory =head1 EXPORTS =over 4 =item :CONSTANTS Exports the following constants: FA_MSDOS FA_UNIX GPBF_ENCRYPTED_MASK GPBF_DEFLATING_COMPRESSION_MASK GPBF_HAS_DATA_DESCRIPTOR_MASK COMPRESSION_STORED COMPRESSION_DEFLATED IFA_TEXT_FILE_MASK IFA_TEXT_FILE IFA_BINARY_FILE COMPRESSION_LEVEL_NONE COMPRESSION_LEVEL_DEFAULT COMPRESSION_LEVEL_FASTEST COMPRESSION_LEVEL_BEST_COMPRESSION =item :MISC_CONSTANTS Exports the following constants (only necessary for extending the module): FA_AMIGA FA_VAX_VMS FA_VM_CMS FA_ATARI_ST FA_OS2_HPFS FA_MACINTOSH FA_Z_SYSTEM FA_CPM FA_WINDOWS_NTFS GPBF_IMPLODING_8K_SLIDING_DICTIONARY_MASK GPBF_IMPLODING_3_SHANNON_FANO_TREES_MASK GPBF_IS_COMPRESSED_PATCHED_DATA_MASK COMPRESSION_SHRUNK DEFLATING_COMPRESSION_NORMAL DEFLATING_COMPRESSION_MAXIMUM DEFLATING_COMPRESSION_FAST DEFLATING_COMPRESSION_SUPER_FAST COMPRESSION_REDUCED_1 COMPRESSION_REDUCED_2 COMPRESSION_REDUCED_3 COMPRESSION_REDUCED_4 COMPRESSION_IMPLODED COMPRESSION_TOKENIZED COMPRESSION_DEFLATED_ENHANCED COMPRESSION_PKWARE_DATA_COMPRESSION_LIBRARY_IMPLODED =item :ERROR_CODES Explained below. Returned from most methods. AZ_OK AZ_STREAM_END AZ_ERROR AZ_FORMAT_ERROR AZ_IO_ERROR =back =head1 ERROR CODES Many of the methods in Archive::Zip return error codes. These are implemented as inline subroutines, using the C pragma. They can be imported into your namespace using the C<:ERROR_CODES> tag: use Archive::Zip qw( :ERROR_CODES ); ... unless ( $zip->read( 'myfile.zip' ) == AZ_OK ) { die "whoops!"; } =over 4 =item AZ_OK (0) Everything is fine. =item AZ_STREAM_END (1) The read stream (or central directory) ended normally. =item AZ_ERROR (2) There was some generic kind of error. =item AZ_FORMAT_ERROR (3) There is a format error in a ZIP file being read. =item AZ_IO_ERROR (4) There was an IO error. =back =head2 Compression Archive::Zip allows each member of a ZIP file to be compressed (using the Deflate algorithm) or uncompressed. Other compression algorithms that some versions of ZIP have been able to produce are not supported. Each member has two compression methods: the one it's stored as (this is always COMPRESSION_STORED for string and external file members), and the one you desire for the member in the zip file. These can be different, of course, so you can make a zip member that is not compressed out of one that is, and vice versa. You can inquire about the current compression and set the desired compression method: my $member = $zip->memberNamed( 'xyz.txt' ); $member->compressionMethod(); # return current compression # set to read uncompressed $member->desiredCompressionMethod( COMPRESSION_STORED ); # set to read compressed $member->desiredCompressionMethod( COMPRESSION_DEFLATED ); There are two different compression methods: =over 4 =item COMPRESSION_STORED File is stored (no compression) =item COMPRESSION_DEFLATED File is Deflated =back =head2 Compression Levels If a member's desiredCompressionMethod is COMPRESSION_DEFLATED, you can choose different compression levels. This choice may affect the speed of compression and decompression, as well as the size of the compressed member data. $member->desiredCompressionLevel( 9 ); The levels given can be: =over 4 =item * 0 or COMPRESSION_LEVEL_NONE This is the same as saying $member->desiredCompressionMethod( COMPRESSION_STORED ); =item * 1 .. 9 1 gives the best speed and worst compression, and 9 gives the best compression and worst speed. =item * COMPRESSION_LEVEL_FASTEST This is a synonym for level 1. =item * COMPRESSION_LEVEL_BEST_COMPRESSION This is a synonym for level 9. =item * COMPRESSION_LEVEL_DEFAULT This gives a good compromise between speed and compression, and is currently equivalent to 6 (this is in the zlib code). This is the level that will be used if not specified. =back =head1 Archive::Zip Methods The Archive::Zip class (and its invisible subclass Archive::Zip::Archive) implement generic zip file functionality. Creating a new Archive::Zip object actually makes an Archive::Zip::Archive object, but you don't have to worry about this unless you're subclassing. =head2 Constructor =over 4 =item new( [$fileName] ) =item new( { filename => $fileName } ) Make a new, empty zip archive. my $zip = Archive::Zip->new(); If an additional argument is passed, new() will call read() to read the contents of an archive: my $zip = Archive::Zip->new( 'xyz.zip' ); If a filename argument is passed and the read fails for any reason, new will return undef. For this reason, it may be better to call read separately. =back =head2 Zip Archive Utility Methods These Archive::Zip methods may be called as functions or as object methods. Do not call them as class methods: $zip = Archive::Zip->new(); $crc = Archive::Zip::computeCRC32( 'ghijkl' ); # OK $crc = $zip->computeCRC32( 'ghijkl' ); # also OK $crc = Archive::Zip->computeCRC32( 'ghijkl' ); # NOT OK =over 4 =item Archive::Zip::computeCRC32( $string [, $crc] ) =item Archive::Zip::computeCRC32( { string => $string [, checksum => $crc ] } ) This is a utility function that uses the Compress::Raw::Zlib CRC routine to compute a CRC-32. You can get the CRC of a string: $crc = Archive::Zip::computeCRC32( $string ); Or you can compute the running CRC: $crc = 0; $crc = Archive::Zip::computeCRC32( 'abcdef', $crc ); $crc = Archive::Zip::computeCRC32( 'ghijkl', $crc ); =item Archive::Zip::setChunkSize( $number ) =item Archive::Zip::setChunkSize( { chunkSize => $number } ) Report or change chunk size used for reading and writing. This can make big differences in dealing with large files. Currently, this defaults to 32K. This also changes the chunk size used for Compress::Raw::Zlib. You must call setChunkSize() before reading or writing. This is not exportable, so you must call it like: Archive::Zip::setChunkSize( 4096 ); or as a method on a zip (though this is a global setting). Returns old chunk size. =item Archive::Zip::chunkSize() Returns the current chunk size: my $chunkSize = Archive::Zip::chunkSize(); =item Archive::Zip::setErrorHandler( \&subroutine ) =item Archive::Zip::setErrorHandler( { subroutine => \&subroutine } ) Change the subroutine called with error strings. This defaults to \&Carp::carp, but you may want to change it to get the error strings. This is not exportable, so you must call it like: Archive::Zip::setErrorHandler( \&myErrorHandler ); If myErrorHandler is undef, resets handler to default. Returns old error handler. Note that if you call Carp::carp or a similar routine or if you're chaining to the default error handler from your error handler, you may want to increment the number of caller levels that are skipped (do not just set it to a number): $Carp::CarpLevel++; =item Archive::Zip::tempFile( [ $tmpdir ] ) =item Archive::Zip::tempFile( { tempDir => $tmpdir } ) Create a uniquely named temp file. It will be returned open for read/write. If C<$tmpdir> is given, it is used as the name of a directory to create the file in. If not given, creates the file using C. Generally, you can override this choice using the $ENV{TMPDIR} environment variable. But see the L documentation for your system. Note that on many systems, if you're running in taint mode, then you must make sure that C<$ENV{TMPDIR}> is untainted for it to be used. Will I create C<$tmpdir> if it does not exist (this is a change from prior versions!). Returns file handle and name: my ($fh, $name) = Archive::Zip::tempFile(); my ($fh, $name) = Archive::Zip::tempFile('myTempDir'); my $fh = Archive::Zip::tempFile(); # if you don't need the name =back =head2 Zip Archive Accessors =over 4 =item members() Return a copy of the members array my @members = $zip->members(); =item numberOfMembers() Return the number of members I have =item memberNames() Return a list of the (internal) file names of the zip members =item memberNamed( $string ) =item memberNamed( { zipName => $string } ) Return ref to member whose filename equals given filename or undef. C<$string> must be in Zip (Unix) filename format. =item membersMatching( $regex ) =item membersMatching( { regex => $regex } ) Return array of members whose filenames match given regular expression in list context. Returns number of matching members in scalar context. my @textFileMembers = $zip->membersMatching( '.*\.txt' ); # or my $numberOfTextFiles = $zip->membersMatching( '.*\.txt' ); =item diskNumber() Return the disk that I start on. Not used for writing zips, but might be interesting if you read a zip in. This should be 0, as Archive::Zip does not handle multi-volume archives. =item diskNumberWithStartOfCentralDirectory() Return the disk number that holds the beginning of the central directory. Not used for writing zips, but might be interesting if you read a zip in. This should be 0, as Archive::Zip does not handle multi-volume archives. =item numberOfCentralDirectoriesOnThisDisk() Return the number of CD structures in the zipfile last read in. Not used for writing zips, but might be interesting if you read a zip in. =item numberOfCentralDirectories() Return the number of CD structures in the zipfile last read in. Not used for writing zips, but might be interesting if you read a zip in. =item centralDirectorySize() Returns central directory size, as read from an external zip file. Not used for writing zips, but might be interesting if you read a zip in. =item centralDirectoryOffsetWRTStartingDiskNumber() Returns the offset into the zip file where the CD begins. Not used for writing zips, but might be interesting if you read a zip in. =item zipfileComment( [ $string ] ) =item zipfileComment( [ { comment => $string } ] ) Get or set the zipfile comment. Returns the old comment. print $zip->zipfileComment(); $zip->zipfileComment( 'New Comment' ); =item eocdOffset() Returns the (unexpected) number of bytes between where the EOCD was found and where it expected to be. This is normally 0, but would be positive if something (a virus, perhaps) had added bytes somewhere before the EOCD. Not used for writing zips, but might be interesting if you read a zip in. Here is an example of how you can diagnose this: my $zip = Archive::Zip->new('somefile.zip'); if ($zip->eocdOffset()) { warn "A virus has added ", $zip->eocdOffset, " bytes of garbage\n"; } The C is used to adjust the starting position of member headers, if necessary. =item fileName() Returns the name of the file last read from. If nothing has been read yet, returns an empty string; if read from a file handle, returns the handle in string form. =back =head2 Zip Archive Member Operations Various operations on a zip file modify members. When a member is passed as an argument, you can either use a reference to the member itself, or the name of a member. Of course, using the name requires that names be unique within a zip (this is not enforced). =over 4 =item removeMember( $memberOrName ) =item removeMember( { memberOrZipName => $memberOrName } ) Remove and return the given member, or match its name and remove it. Returns undef if member or name does not exist in this Zip. No-op if member does not belong to this zip. =item replaceMember( $memberOrName, $newMember ) =item replaceMember( { memberOrZipName => $memberOrName, newMember => $newMember } ) Remove and return the given member, or match its name and remove it. Replace with new member. Returns undef if member or name does not exist in this Zip, or if C<$newMember> is undefined. It is an (undiagnosed) error to provide a C<$newMember> that is a member of the zip being modified. my $member1 = $zip->removeMember( 'xyz' ); my $member2 = $zip->replaceMember( 'abc', $member1 ); # now, $member2 (named 'abc') is not in $zip, # and $member1 (named 'xyz') is, having taken $member2's place. =item extractMember( $memberOrName [, $extractedName ] ) =item extractMember( { memberOrZipName => $memberOrName [, name => $extractedName ] } ) Extract the given member, or match its name and extract it. Returns undef if member does not exist in this Zip. If optional second arg is given, use it as the name of the extracted member. Otherwise, the internal filename of the member is used as the name of the extracted file or directory. If you pass C<$extractedName>, it should be in the local file system's format. If you do not pass C<$extractedName> and the internal filename traverses a parent directory or a symbolic link, the extraction will be aborted with C for security reason. All necessary directories will be created. Returns C on success. =item extractMemberWithoutPaths( $memberOrName [, $extractedName ] ) =item extractMemberWithoutPaths( { memberOrZipName => $memberOrName [, name => $extractedName ] } ) Extract the given member, or match its name and extract it. Does not use path information (extracts into the current directory). Returns undef if member does not exist in this Zip. If optional second arg is given, use it as the name of the extracted member (its paths will be deleted too). Otherwise, the internal filename of the member (minus paths) is used as the name of the extracted file or directory. Returns C on success. If you do not pass C<$extractedName> and the internal filename is equalled to a local symbolic link, the extraction will be aborted with C for security reason. =item addMember( $member ) =item addMember( { member => $member } ) Append a member (possibly from another zip file) to the zip file. Returns the new member. Generally, you will use addFile(), addDirectory(), addFileOrDirectory(), addString(), or read() to add members. # Move member named 'abc' to end of zip: my $member = $zip->removeMember( 'abc' ); $zip->addMember( $member ); =item updateMember( $memberOrName, $fileName ) =item updateMember( { memberOrZipName => $memberOrName, name => $fileName } ) Update a single member from the file or directory named C<$fileName>. Returns the (possibly added or updated) member, if any; C on errors. The comparison is based on C and (in the case of a non-directory) the size of the file. =item addFile( $fileName [, $newName, $compressionLevel ] ) =item addFile( { filename => $fileName [, zipName => $newName, compressionLevel => $compressionLevel } ] ) Append a member whose data comes from an external file, returning the member or undef. The member will have its file name set to the name of the external file, and its desiredCompressionMethod set to COMPRESSION_DEFLATED. The file attributes and last modification time will be set from the file. If the name given does not represent a readable plain file or symbolic link, undef will be returned. C<$fileName> must be in the format required for the local file system. The optional C<$newName> argument sets the internal file name to something different than the given $fileName. C<$newName>, if given, must be in Zip name format (i.e. Unix). The text mode bit will be set if the contents appears to be text (as returned by the C<-T> perl operator). I that you should not (generally) use absolute path names in zip member names, as this will cause problems with some zip tools as well as introduce a security hole and make the zip harder to use. =item addDirectory( $directoryName [, $fileName ] ) =item addDirectory( { directoryName => $directoryName [, zipName => $fileName ] } ) Append a member created from the given directory name. The directory name does not have to name an existing directory. If the named directory exists, the file modification time and permissions are set from the existing directory, otherwise they are set to now and permissive default permissions. C<$directoryName> must be in local file system format. The optional second argument sets the name of the archive member (which defaults to C<$directoryName>). If given, it must be in Zip (Unix) format. Returns the new member. =item addFileOrDirectory( $name [, $newName, $compressionLevel ] ) =item addFileOrDirectory( { name => $name [, zipName => $newName, compressionLevel => $compressionLevel ] } ) Append a member from the file or directory named $name. If $newName is given, use it for the name of the new member. Will add or remove trailing slashes from $newName as needed. C<$name> must be in local file system format. The optional second argument sets the name of the archive member (which defaults to C<$name>). If given, it must be in Zip (Unix) format. =item addString( $stringOrStringRef, $name, [$compressionLevel] ) =item addString( { string => $stringOrStringRef [, zipName => $name, compressionLevel => $compressionLevel ] } ) Append a member created from the given string or string reference. The name is given by the second argument. Returns the new member. The last modification time will be set to now, and the file attributes will be set to permissive defaults. my $member = $zip->addString( 'This is a test', 'test.txt' ); =item contents( $memberOrMemberName [, $newContents ] ) =item contents( { memberOrZipName => $memberOrMemberName [, contents => $newContents ] } ) Returns the uncompressed data for a particular member, or undef. print "xyz.txt contains " . $zip->contents( 'xyz.txt' ); Also can change the contents of a member: $zip->contents( 'xyz.txt', 'This is the new contents' ); If called expecting an array as the return value, it will include the status as the second value in the array. ($content, $status) = $zip->contents( 'xyz.txt'); =back =head2 Zip Archive I/O operations A Zip archive can be written to a file or file handle, or read from one. =over 4 =item writeToFileNamed( $fileName ) =item writeToFileNamed( { fileName => $fileName } ) Write a zip archive to named file. Returns C on success. my $status = $zip->writeToFileNamed( 'xx.zip' ); die "error somewhere" if $status != AZ_OK; Note that if you use the same name as an existing zip file that you read in, you will clobber ZipFileMembers. So instead, write to a different file name, then delete the original. If you use the C or C methods, you can re-write the original zip in this way. C<$fileName> should be a valid file name on your system. =item writeToFileHandle( $fileHandle [, $seekable] ) Write a zip archive to a file handle. Return AZ_OK on success. The optional second arg tells whether or not to try to seek backwards to re-write headers. If not provided, it is set if the Perl C<-f> test returns true. This could fail on some operating systems, though. my $fh = IO::File->new( 'someFile.zip', 'w' ); unless ( $zip->writeToFileHandle( $fh ) == AZ_OK ) { # error handling } If you pass a file handle that is not seekable (like if you're writing to a pipe or a socket), pass a false second argument: my $fh = IO::File->new( '| cat > somefile.zip', 'w' ); $zip->writeToFileHandle( $fh, 0 ); # fh is not seekable If this method fails during the write of a member, that member and all following it will return false from C. See writeCentralDirectory() for a way to deal with this. If you want, you can write data to the file handle before passing it to writeToFileHandle(); this could be used (for instance) for making self-extracting archives. However, this only works reliably when writing to a real file (as opposed to STDOUT or some other possible non-file). See examples/selfex.pl for how to write a self-extracting archive. =item writeCentralDirectory( $fileHandle [, $offset ] ) =item writeCentralDirectory( { fileHandle => $fileHandle [, offset => $offset ] } ) Writes the central directory structure to the given file handle. Returns AZ_OK on success. If given an $offset, will seek to that point before writing. This can be used for recovery in cases where writeToFileHandle or writeToFileNamed returns an IO error because of running out of space on the destination file. You can truncate the zip by seeking backwards and then writing the directory: my $fh = IO::File->new( 'someFile.zip', 'w' ); my $retval = $zip->writeToFileHandle( $fh ); if ( $retval == AZ_IO_ERROR ) { my @unwritten = grep { not $_->wasWritten() } $zip->members(); if (@unwritten) { $zip->removeMember( $member ) foreach my $member ( @unwritten ); $zip->writeCentralDirectory( $fh, $unwritten[0]->writeLocalHeaderRelativeOffset()); } } =item overwriteAs( $newName ) =item overwriteAs( { filename => $newName } ) Write the zip to the specified file, as safely as possible. This is done by first writing to a temp file, then renaming the original if it exists, then renaming the temp file, then deleting the renamed original if it exists. Returns AZ_OK if successful. =item overwrite() Write back to the original zip file. See overwriteAs() above. If the zip was not ever read from a file, this generates an error. =item read( $fileName ) =item read( { filename => $fileName } ) Read zipfile headers from a zip file, appending new members. Returns C or error code. my $zipFile = Archive::Zip->new(); my $status = $zipFile->read( '/some/FileName.zip' ); =item readFromFileHandle( $fileHandle, $filename ) =item readFromFileHandle( { fileHandle => $fileHandle, filename => $filename } ) Read zipfile headers from an already-opened file handle, appending new members. Does not close the file handle. Returns C or error code. Note that this requires a seekable file handle; reading from a stream is not yet supported, but using in-memory data is. my $fh = IO::File->new( '/some/FileName.zip', 'r' ); my $zip1 = Archive::Zip->new(); my $status = $zip1->readFromFileHandle( $fh ); my $zip2 = Archive::Zip->new(); $status = $zip2->readFromFileHandle( $fh ); Read zip using in-memory data (recursable): open my $fh, "<", "archive.zip" or die $!; my $zip_data = do { local $.; <$fh> }; my $zip = Archive::Zip->new; open my $dh, "+<", \$zip_data; $zip->readFromFileHandle ($dh); =back =head2 Zip Archive Tree operations These used to be in Archive::Zip::Tree but got moved into Archive::Zip. They enable operation on an entire tree of members or files. A usage example: use Archive::Zip; my $zip = Archive::Zip->new(); # add all readable files and directories below . as xyz/* $zip->addTree( '.', 'xyz' ); # add all readable plain files below /abc as def/* $zip->addTree( '/abc', 'def', sub { -f && -r } ); # add all .c files below /tmp as stuff/* $zip->addTreeMatching( '/tmp', 'stuff', '\.c$' ); # add all .o files below /tmp as stuff/* if they aren't writable $zip->addTreeMatching( '/tmp', 'stuff', '\.o$', sub { ! -w } ); # add all .so files below /tmp that are smaller than 200 bytes as stuff/* $zip->addTreeMatching( '/tmp', 'stuff', '\.o$', sub { -s < 200 } ); # and write them into a file $zip->writeToFileNamed('xxx.zip'); # now extract the same files into /tmpx $zip->extractTree( 'stuff', '/tmpx' ); =over 4 =item $zip->addTree( $root, $dest [, $pred, $compressionLevel ] ) -- Add tree of files to a zip =item $zip->addTree( { root => $root, zipName => $dest [, select => $pred, compressionLevel => $compressionLevel ] ) C<$root> is the root of the tree of files and directories to be added. It is a valid directory name on your system. C<$dest> is the name for the root in the zip file (undef or blank means to use relative pathnames). It is a valid ZIP directory name (that is, it uses forward slashes (/) for separating directory components). C<$pred> is an optional subroutine reference to select files: it is passed the name of the prospective file or directory using C<$_>, and if it returns true, the file or directory will be included. The default is to add all readable files and directories. For instance, using my $pred = sub { /\.txt/ }; $zip->addTree( '.', '', $pred ); will add all the .txt files in and below the current directory, using relative names, and making the names identical in the zipfile: original name zip member name ./xyz xyz ./a/ a/ ./a/b a/b To translate absolute to relative pathnames, just pass them in: $zip->addTree( '/c/d', 'a' ); original name zip member name /c/d/xyz a/xyz /c/d/a/ a/a/ /c/d/a/b a/a/b Returns AZ_OK on success. Note that this will not follow symbolic links to directories. Note also that this does not check for the validity of filenames. Note that you generally I want to make zip archive member names absolute. =item $zip->addTreeMatching( $root, $dest, $pattern [, $pred, $compressionLevel ] ) =item $zip->addTreeMatching( { root => $root, zipName => $dest, pattern => $pattern [, select => $pred, compressionLevel => $compressionLevel ] } ) $root is the root of the tree of files and directories to be added $dest is the name for the root in the zip file (undef means to use relative pathnames) $pattern is a (non-anchored) regular expression for filenames to match $pred is an optional subroutine reference to select files: it is passed the name of the prospective file or directory in C<$_>, and if it returns true, the file or directory will be included. The default is to add all readable files and directories. To add all files in and below the current directory whose names end in C<.pl>, and make them extract into a subdirectory named C, do this: $zip->addTreeMatching( '.', 'xyz', '\.pl$' ) To add all I files in and below the directory named C whose names end in C<.pl>, and make them extract into a subdirectory named C, do this: $zip->addTreeMatching( '/abc', 'xyz', '\.pl$', sub { -w } ) Returns AZ_OK on success. Note that this will not follow symbolic links to directories. =item $zip->updateTree( $root [, $dest , $pred , $mirror, $compressionLevel ] ); =item $zip->updateTree( { root => $root [, zipName => $dest, select => $pred, mirror => $mirror, compressionLevel => $compressionLevel ] } ); Update a zip file from a directory tree. C takes the same arguments as C, but first checks to see whether the file or directory already exists in the zip file, and whether it has been changed. If the fourth argument C<$mirror> is true, then delete all my members if corresponding files were not found. Returns an error code or AZ_OK if all is well. =item $zip->extractTree( [ $root, $dest, $volume } ] ) =item $zip->extractTree( [ { root => $root, zipName => $dest, volume => $volume } ] ) If you don't give any arguments at all, will extract all the files in the zip with their original names. If you supply one argument for C<$root>, C will extract all the members whose names start with C<$root> into the current directory, stripping off C<$root> first. C<$root> is in Zip (Unix) format. For instance, $zip->extractTree( 'a' ); when applied to a zip containing the files: a/x a/b/c ax/d/e d/e will extract: a/x as ./x a/b/c as ./b/c If you give two arguments, C extracts all the members whose names start with C<$root>. It will translate C<$root> into C<$dest> to construct the destination file name. C<$root> and C<$dest> are in Zip (Unix) format. For instance, $zip->extractTree( 'a', 'd/e' ); when applied to a zip containing the files: a/x a/b/c ax/d/e d/e will extract: a/x to d/e/x a/b/c to d/e/b/c and ignore ax/d/e and d/e If you give three arguments, C extracts all the members whose names start with C<$root>. It will translate C<$root> into C<$dest> to construct the destination file name, and then it will convert to local file system format, using C<$volume> as the name of the destination volume. C<$root> and C<$dest> are in Zip (Unix) format. C<$volume> is in local file system format. For instance, under Windows, $zip->extractTree( 'a', 'd/e', 'f:' ); when applied to a zip containing the files: a/x a/b/c ax/d/e d/e will extract: a/x to f:d/e/x a/b/c to f:d/e/b/c and ignore ax/d/e and d/e If you want absolute paths (the prior example used paths relative to the current directory on the destination volume, you can specify these in C<$dest>: $zip->extractTree( 'a', '/d/e', 'f:' ); when applied to a zip containing the files: a/x a/b/c ax/d/e d/e will extract: a/x to f:\d\e\x a/b/c to f:\d\e\b\c and ignore ax/d/e and d/e If the path to the extracted file traverses a parent directory or a symbolic link, the extraction will be aborted with C for security reason. Returns an error code or AZ_OK if everything worked OK. =back =head1 Archive::Zip Global Variables =over 4 =item $Archive::Zip::UNICODE This variable governs how Unicode file and directory names are added to or extracted from an archive. If set, file and directory names are considered to be UTF-8 encoded. This is I. Please report problems. { local $Archive::Zip::UNICODE = 1; $zip->addFile('Déjà vu.txt'); } =back =head1 MEMBER OPERATIONS =head2 Member Class Methods Several constructors allow you to construct members without adding them to a zip archive. These work the same as the addFile(), addDirectory(), and addString() zip instance methods described above, but they don't add the new members to a zip. =over 4 =item Archive::Zip::Member->newFromString( $stringOrStringRef [, $fileName ] ) =item Archive::Zip::Member->newFromString( { string => $stringOrStringRef [, zipName => $fileName ] ) Construct a new member from the given string. Returns undef on error. my $member = Archive::Zip::Member->newFromString( 'This is a test', =item newFromFile( $fileName [, $zipName ] ) =item newFromFile( { filename => $fileName [, zipName => $zipName ] } ) Construct a new member from the given file. Returns undef on error. my $member = Archive::Zip::Member->newFromFile( 'xyz.txt' ); =item newDirectoryNamed( $directoryName [, $zipname ] ) =item newDirectoryNamed( { directoryName => $directoryName [, zipName => $zipname ] } ) Construct a new member from the given directory. C<$directoryName> must be a valid name on your file system; it does not have to exist. If given, C<$zipname> will be the name of the zip member; it must be a valid Zip (Unix) name. If not given, it will be converted from C<$directoryName>. Returns undef on error. my $member = Archive::Zip::Member->newDirectoryNamed( 'CVS/' ); =back =head2 Member Simple accessors These methods get (and/or set) member attribute values. =over 4 =item versionMadeBy() Gets the field from the member header. =item fileAttributeFormat( [ $format ] ) =item fileAttributeFormat( [ { format => $format ] } ) Gets or sets the field from the member header. These are C values. =item versionNeededToExtract() Gets the field from the member header. =item bitFlag() Gets the general purpose bit field from the member header. This is where the C bits live. =item compressionMethod() Returns the member compression method. This is the method that is currently being used to compress the member data. This will be COMPRESSION_STORED for added string or file members, or any of the C values for members from a zip file. However, this module can only handle members whose data is in COMPRESSION_STORED or COMPRESSION_DEFLATED format. =item desiredCompressionMethod( [ $method ] ) =item desiredCompressionMethod( [ { compressionMethod => $method } ] ) Get or set the member's C. This is the compression method that will be used when the member is written. Returns prior desiredCompressionMethod. Only COMPRESSION_DEFLATED or COMPRESSION_STORED are valid arguments. Changing to COMPRESSION_STORED will change the member desiredCompressionLevel to 0; changing to COMPRESSION_DEFLATED will change the member desiredCompressionLevel to COMPRESSION_LEVEL_DEFAULT. =item desiredCompressionLevel( [ $level ] ) =item desiredCompressionLevel( [ { compressionLevel => $level } ] ) Get or set the member's desiredCompressionLevel This is the method that will be used to write. Returns prior desiredCompressionLevel. Valid arguments are 0 through 9, COMPRESSION_LEVEL_NONE, COMPRESSION_LEVEL_DEFAULT, COMPRESSION_LEVEL_BEST_COMPRESSION, and COMPRESSION_LEVEL_FASTEST. 0 or COMPRESSION_LEVEL_NONE will change the desiredCompressionMethod to COMPRESSION_STORED. All other arguments will change the desiredCompressionMethod to COMPRESSION_DEFLATED. =item externalFileName() Return the member's external file name, if any, or undef. =item fileName() Get or set the member's internal filename. Returns the (possibly new) filename. Names will have backslashes converted to forward slashes, and will have multiple consecutive slashes converted to single ones. =item lastModFileDateTime() Return the member's last modification date/time stamp in MS-DOS format. =item lastModTime() Return the member's last modification date/time stamp, converted to unix localtime format. print "Mod Time: " . scalar( localtime( $member->lastModTime() ) ); =item setLastModFileDateTimeFromUnix() Set the member's lastModFileDateTime from the given unix time. $member->setLastModFileDateTimeFromUnix( time() ); =item internalFileAttributes() Return the internal file attributes field from the zip header. This is only set for members read from a zip file. =item externalFileAttributes() Return member attributes as read from the ZIP file. Note that these are NOT UNIX! =item unixFileAttributes( [ $newAttributes ] ) =item unixFileAttributes( [ { attributes => $newAttributes } ] ) Get or set the member's file attributes using UNIX file attributes. Returns old attributes. my $oldAttribs = $member->unixFileAttributes( 0666 ); Note that the return value has more than just the file permissions, so you will have to mask off the lowest bits for comparisons. =item localExtraField( [ $newField ] ) =item localExtraField( [ { field => $newField } ] ) Gets or sets the extra field that was read from the local header. This is not set for a member from a zip file until after the member has been written out. The extra field must be in the proper format. =item cdExtraField( [ $newField ] ) =item cdExtraField( [ { field => $newField } ] ) Gets or sets the extra field that was read from the central directory header. The extra field must be in the proper format. =item extraFields() Return both local and CD extra fields, concatenated. =item fileComment( [ $newComment ] ) =item fileComment( [ { comment => $newComment } ] ) Get or set the member's file comment. =item hasDataDescriptor() Get or set the data descriptor flag. If this is set, the local header will not necessarily have the correct data sizes. Instead, a small structure will be stored at the end of the member data with these values. This should be transparent in normal operation. =item crc32() Return the CRC-32 value for this member. This will not be set for members that were constructed from strings or external files until after the member has been written. =item crc32String() Return the CRC-32 value for this member as an 8 character printable hex string. This will not be set for members that were constructed from strings or external files until after the member has been written. =item compressedSize() Return the compressed size for this member. This will not be set for members that were constructed from strings or external files until after the member has been written. =item uncompressedSize() Return the uncompressed size for this member. =item password( [ $password ] ) Returns the password for this member to be used on decryption. If $password is given, it will set the password for the decryption. =item isEncrypted() Return true if this member is encrypted. The Archive::Zip module does not currently support creation of encrypted members. Decryption works more or less like this: my $zip = Archive::Zip->new; $zip->read ("encrypted.zip"); for my $m (map { $zip->memberNamed ($_) } $zip->memberNames) { $m->password ("secret"); $m->contents; # is "" when password was wrong That shows that the password has to be set per member, and not per archive. This might change in the future. =item isTextFile( [ $flag ] ) =item isTextFile( [ { flag => $flag } ] ) Returns true if I am a text file. Also can set the status if given an argument (then returns old state). Note that this module does not currently do anything with this flag upon extraction or storage. That is, bytes are stored in native format whether or not they came from a text file. =item isBinaryFile() Returns true if I am a binary file. Also can set the status if given an argument (then returns old state). Note that this module does not currently do anything with this flag upon extraction or storage. That is, bytes are stored in native format whether or not they came from a text file. =item extractToFileNamed( $fileName ) =item extractToFileNamed( { name => $fileName } ) Extract me to a file with the given name. The file will be created with default modes. Directories will be created as needed. The C<$fileName> argument should be a valid file name on your file system. Returns AZ_OK on success. =item isDirectory() Returns true if I am a directory. =item writeLocalHeaderRelativeOffset() Returns the file offset in bytes the last time I was written. =item wasWritten() Returns true if I was successfully written. Reset at the beginning of a write attempt. =back =head2 Low-level member data reading It is possible to use lower-level routines to access member data streams, rather than the extract* methods and contents(). For instance, here is how to print the uncompressed contents of a member in chunks using these methods: my ( $member, $status, $bufferRef ); $member = $zip->memberNamed( 'xyz.txt' ); $member->desiredCompressionMethod( COMPRESSION_STORED ); $status = $member->rewindData(); die "error $status" unless $status == AZ_OK; while ( ! $member->readIsDone() ) { ( $bufferRef, $status ) = $member->readChunk(); die "error $status" if $status != AZ_OK && $status != AZ_STREAM_END; # do something with $bufferRef: print $$bufferRef; } $member->endRead(); =over 4 =item readChunk( [ $chunkSize ] ) =item readChunk( [ { chunkSize => $chunkSize } ] ) This reads the next chunk of given size from the member's data stream and compresses or uncompresses it as necessary, returning a reference to the bytes read and a status. If size argument is not given, defaults to global set by Archive::Zip::setChunkSize. Status is AZ_OK on success until the last chunk, where it returns AZ_STREAM_END. Returns C<( \$bytes, $status)>. my ( $outRef, $status ) = $self->readChunk(); print $$outRef if $status != AZ_OK && $status != AZ_STREAM_END; =item rewindData() Rewind data and set up for reading data streams or writing zip files. Can take options for C or C, but this is not likely to be necessary. Subclass overrides should call this method. Returns C on success. =item endRead() Reset the read variables and free the inflater or deflater. Must be called to close files, etc. Returns AZ_OK on success. =item readIsDone() Return true if the read has run out of data or encountered an error. =item contents() Return the entire uncompressed member data or undef in scalar context. When called in array context, returns C<( $string, $status )>; status will be AZ_OK on success: my $string = $member->contents(); # or my ( $string, $status ) = $member->contents(); die "error $status" unless $status == AZ_OK; Can also be used to set the contents of a member (this may change the class of the member): $member->contents( "this is my new contents" ); =item extractToFileHandle( $fh ) =item extractToFileHandle( { fileHandle => $fh } ) Extract (and uncompress, if necessary) the member's contents to the given file handle. Return AZ_OK on success. =back =head1 Archive::Zip::FileMember methods The Archive::Zip::FileMember class extends Archive::Zip::Member. It is the base class for both ZipFileMember and NewFileMember classes. This class adds an C and an C member to keep track of the external file. =over 4 =item externalFileName() Return the member's external filename. =item fh() Return the member's read file handle. Automatically opens file if necessary. =back =head1 Archive::Zip::ZipFileMember methods The Archive::Zip::ZipFileMember class represents members that have been read from external zip files. =over 4 =item diskNumberStart() Returns the disk number that the member's local header resides in. Should be 0. =item localHeaderRelativeOffset() Returns the offset into the zip file where the member's local header is. =item dataOffset() Returns the offset from the beginning of the zip file to the member's data. =back =head1 REQUIRED MODULES L requires several other modules: L L L L L L L L L L L =head1 BUGS AND CAVEATS =head2 When not to use Archive::Zip If you are just going to be extracting zips (and/or other archives) you are recommended to look at using L instead, as it is much easier to use and factors out archive-specific functionality. =head2 Try to avoid IO::Scalar One of the most common ways to use Archive::Zip is to generate Zip files in-memory. Most people use L for this purpose. Unfortunately, as of 1.11 this module no longer works with L as it incorrectly implements seeking. Anybody using L should consider porting to L, which is smaller, lighter, and is implemented to be perfectly compatible with regular seekable filehandles. Support for L most likely will B be restored in the future, as L itself cannot change the way it is implemented due to back-compatibility issues. =head2 Wrong password for encrypted members When an encrypted member is read using the wrong password, you currently have to re-read the entire archive to try again with the correct password. =head1 TO DO * auto-choosing storing vs compression * extra field hooks (see notes.txt) * check for duplicates on addition/renaming? * Text file extraction (line end translation) * Reading zip files from non-seekable inputs (Perhaps by proxying through IO::String?) * separate unused constants into separate module * cookbook style docs * Handle tainted paths correctly * Work on better compatibility with other IO:: modules * Support encryption * More user-friendly decryption =head1 SUPPORT Bugs should be reported via the CPAN bug tracker L For other issues contact the maintainer =head1 AUTHOR Currently maintained by Fred Moyer Previously maintained by Adam Kennedy Previously maintained by Steve Peters Esteve@fisharerojo.orgE. File attributes code by Maurice Aubrey Emaurice@lovelyfilth.comE. Originally by Ned Konz Enedkonz@cpan.orgE. =head1 COPYRIGHT Some parts copyright 2006 - 2012 Adam Kennedy. Some parts copyright 2005 Steve Peters. Original work copyright 2000 - 2004 Ned Konz. This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself. =head1 SEE ALSO Look at L which is a wrapper that allows one to read Zip archive members as if they were files. L, L, L =cut Tar.pm000064400000230107152347641370005647 0ustar00### the gnu tar specification: ### http://www.gnu.org/software/tar/manual/tar.html ### ### and the pax format spec, which tar derives from: ### http://www.opengroup.org/onlinepubs/007904975/utilities/pax.html package Archive::Tar; require 5.005_03; use Cwd; use IO::Zlib; use IO::File; use Carp qw(carp croak); use File::Spec (); use File::Spec::Unix (); use File::Path (); use Archive::Tar::File; use Archive::Tar::Constant; require Exporter; use strict; use vars qw[$DEBUG $error $VERSION $WARN $FOLLOW_SYMLINK $CHOWN $CHMOD $DO_NOT_USE_PREFIX $HAS_PERLIO $HAS_IO_STRING $SAME_PERMISSIONS $INSECURE_EXTRACT_MODE $ZERO_PAD_NUMBERS @ISA @EXPORT $RESOLVE_SYMLINK $MAX_FILE_SIZE ]; @ISA = qw[Exporter]; @EXPORT = qw[ COMPRESS_GZIP COMPRESS_BZIP ]; $DEBUG = 0; $WARN = 1; $FOLLOW_SYMLINK = 0; $VERSION = "2.30"; $CHOWN = 1; $CHMOD = 1; $SAME_PERMISSIONS = $> == 0 ? 1 : 0; $DO_NOT_USE_PREFIX = 0; $INSECURE_EXTRACT_MODE = 0; $ZERO_PAD_NUMBERS = 0; $RESOLVE_SYMLINK = $ENV{'PERL5_AT_RESOLVE_SYMLINK'} || 'speed'; $MAX_FILE_SIZE = 1024 * 1024 * 1024; BEGIN { use Config; $HAS_PERLIO = $Config::Config{useperlio}; ### try and load IO::String anyway, so you can dynamically ### switch between perlio and IO::String $HAS_IO_STRING = eval { require IO::String; import IO::String; 1; } || 0; } =head1 NAME Archive::Tar - module for manipulations of tar archives =head1 SYNOPSIS use Archive::Tar; my $tar = Archive::Tar->new; $tar->read('origin.tgz'); $tar->extract(); $tar->add_files('file/foo.pl', 'docs/README'); $tar->add_data('file/baz.txt', 'This is the contents now'); $tar->rename('oldname', 'new/file/name'); $tar->chown('/', 'root'); $tar->chown('/', 'root:root'); $tar->chmod('/tmp', '1777'); $tar->write('files.tar'); # plain tar $tar->write('files.tgz', COMPRESS_GZIP); # gzip compressed $tar->write('files.tbz', COMPRESS_BZIP); # bzip2 compressed =head1 DESCRIPTION Archive::Tar provides an object oriented mechanism for handling tar files. It provides class methods for quick and easy files handling while also allowing for the creation of tar file objects for custom manipulation. If you have the IO::Zlib module installed, Archive::Tar will also support compressed or gzipped tar files. An object of class Archive::Tar represents a .tar(.gz) archive full of files and things. =head1 Object Methods =head2 Archive::Tar->new( [$file, $compressed] ) Returns a new Tar object. If given any arguments, C calls the C method automatically, passing on the arguments provided to the C method. If C is invoked with arguments and the C method fails for any reason, C returns undef. =cut my $tmpl = { _data => [ ], _file => 'Unknown', }; ### install get/set accessors for this object. for my $key ( keys %$tmpl ) { no strict 'refs'; *{__PACKAGE__."::$key"} = sub { my $self = shift; $self->{$key} = $_[0] if @_; return $self->{$key}; } } sub new { my $class = shift; $class = ref $class if ref $class; ### copying $tmpl here since a shallow copy makes it use the ### same aref, causing for files to remain in memory always. my $obj = bless { _data => [ ], _file => 'Unknown', _error => '' }, $class; if (@_) { unless ( $obj->read( @_ ) ) { $obj->_error(qq[No data could be read from file]); return; } } return $obj; } =head2 $tar->read ( $filename|$handle, [$compressed, {opt => 'val'}] ) Read the given tar file into memory. The first argument can either be the name of a file or a reference to an already open filehandle (or an IO::Zlib object if it's compressed) The C will I any previous content in C<$tar>! The second argument may be considered optional, but remains for backwards compatibility. Archive::Tar now looks at the file magic to determine what class should be used to open the file and will transparently Do The Right Thing. Archive::Tar will warn if you try to pass a bzip2 compressed file and the IO::Zlib / IO::Uncompress::Bunzip2 modules are not available and simply return. Note that you can currently B pass a C compressed filehandle, which is not opened with C, a C compressed filehandle, which is not opened with C, nor a string containing the full archive information (either compressed or uncompressed). These are worth while features, but not currently implemented. See the C section. The third argument can be a hash reference with options. Note that all options are case-sensitive. =over 4 =item limit Do not read more than C files. This is useful if you have very big archives, and are only interested in the first few files. =item filter Can be set to a regular expression. Only files with names that match the expression will be read. =item md5 Set to 1 and the md5sum of files will be returned (instead of file data) my $iter = Archive::Tar->iter( $file, 1, {md5 => 1} ); while( my $f = $iter->() ) { print $f->data . "\t" . $f->full_path . $/; } =item extract If set to true, immediately extract entries when reading them. This gives you the same memory break as the C function. Note however that entries will not be read into memory, but written straight to disk. This means no C objects are created for you to inspect. =back All files are stored internally as C objects. Please consult the L documentation for details. Returns the number of files read in scalar context, and a list of C objects in list context. =cut sub read { my $self = shift; my $file = shift; my $gzip = shift || 0; my $opts = shift || {}; unless( defined $file ) { $self->_error( qq[No file to read from!] ); return; } else { $self->_file( $file ); } my $handle = $self->_get_handle($file, $gzip, READ_ONLY->( ZLIB ) ) or return; my $data = $self->_read_tar( $handle, $opts ) or return; $self->_data( $data ); return wantarray ? @$data : scalar @$data; } sub _get_handle { my $self = shift; my $file = shift; return unless defined $file; my $compress = shift || 0; my $mode = shift || READ_ONLY->( ZLIB ); # default to read only ### Check if file is a file handle or IO glob if ( ref $file ) { return $file if eval{ *$file{IO} }; return $file if eval{ $file->isa(q{IO::Handle}) }; $file = q{}.$file; } ### get a FH opened to the right class, so we can use it transparently ### throughout the program my $fh; { ### reading magic only makes sense if we're opening a file for ### reading. otherwise, just use what the user requested. my $magic = ''; if( MODE_READ->($mode) ) { open my $tmp, $file or do { $self->_error( qq[Could not open '$file' for reading: $!] ); return; }; ### read the first 4 bites of the file to figure out which class to ### use to open the file. sysread( $tmp, $magic, 4 ); close $tmp; } ### is it bzip? ### if you asked specifically for bzip compression, or if we're in ### read mode and the magic numbers add up, use bzip if( BZIP and ( ($compress eq COMPRESS_BZIP) or ( MODE_READ->($mode) and $magic =~ BZIP_MAGIC_NUM ) ) ) { ### different reader/writer modules, different error vars... sigh if( MODE_READ->($mode) ) { $fh = IO::Uncompress::Bunzip2->new( $file, MultiStream => 1 ) or do { $self->_error( qq[Could not read '$file': ] . $IO::Uncompress::Bunzip2::Bunzip2Error ); return; }; } else { $fh = IO::Compress::Bzip2->new( $file ) or do { $self->_error( qq[Could not write to '$file': ] . $IO::Compress::Bzip2::Bzip2Error ); return; }; } ### is it gzip? ### if you asked for compression, if you wanted to read or the gzip ### magic number is present (redundant with read) } elsif( ZLIB and ( $compress or MODE_READ->($mode) or $magic =~ GZIP_MAGIC_NUM ) ) { $fh = IO::Zlib->new; unless( $fh->open( $file, $mode ) ) { $self->_error(qq[Could not create filehandle for '$file': $!]); return; } ### is it plain tar? } else { $fh = IO::File->new; unless( $fh->open( $file, $mode ) ) { $self->_error(qq[Could not create filehandle for '$file': $!]); return; } ### enable bin mode on tar archives binmode $fh; } } return $fh; } sub _read_tar { my $self = shift; my $handle = shift or return; my $opts = shift || {}; my $count = $opts->{limit} || 0; my $filter = $opts->{filter}; my $md5 = $opts->{md5} || 0; # cdrake my $filter_cb = $opts->{filter_cb}; my $extract = $opts->{extract} || 0; ### set a cap on the amount of files to extract ### my $limit = 0; $limit = 1 if $count > 0; my $tarfile = [ ]; my $chunk; my $read = 0; my $real_name; # to set the name of a file when # we're encountering @longlink my $data; LOOP: while( $handle->read( $chunk, HEAD ) ) { ### IO::Zlib doesn't support this yet my $offset; if ( ref($handle) ne 'IO::Zlib' ) { local $@; $offset = eval { tell $handle } || 'unknown'; $@ = ''; } else { $offset = 'unknown'; } unless( $read++ ) { my $gzip = GZIP_MAGIC_NUM; if( $chunk =~ /$gzip/ ) { $self->_error( qq[Cannot read compressed format in tar-mode] ); return; } ### size is < HEAD, which means a corrupted file, as the minimum ### length is _at least_ HEAD if (length $chunk != HEAD) { $self->_error( qq[Cannot read enough bytes from the tarfile] ); return; } } ### if we can't read in all bytes... ### last if length $chunk != HEAD; ### 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. next if $chunk eq TAR_END; ### according to the posix spec, the last 12 bytes of the header are ### null bytes, to pad it to a 512 byte block. That means if these ### bytes are NOT null bytes, it's a corrupt header. See: ### www.koders.com/c/fidCE473AD3D9F835D690259D60AD5654591D91D5BA.aspx ### line 111 { my $nulls = join '', "\0" x 12; unless( $nulls eq substr( $chunk, 500, 12 ) ) { $self->_error( qq[Invalid header block at offset $offset] ); next LOOP; } } ### pass the realname, so we can set it 'proper' right away ### some of the heuristics are done on the name, so important ### to set it ASAP my $entry; { my %extra_args = (); $extra_args{'name'} = $$real_name if defined $real_name; unless( $entry = Archive::Tar::File->new( chunk => $chunk, %extra_args ) ) { $self->_error( qq[Couldn't read chunk at offset $offset] ); next LOOP; } } ### ignore labels: ### http://www.gnu.org/software/tar/manual/html_chapter/Media.html#SEC159 next if $entry->is_label; if( length $entry->type and ($entry->is_file || $entry->is_longlink) ) { if ( $entry->is_file && !$entry->validate ) { ### sometimes the chunk is rather fux0r3d and a whole 512 ### bytes ends up in the ->name area. ### clean it up, if need be my $name = $entry->name; $name = substr($name, 0, 100) if length $name > 100; $name =~ s/\n/ /g; $self->_error( $name . qq[: checksum error] ); next LOOP; } my $block = BLOCK_SIZE->( $entry->size ); if ( $MAX_FILE_SIZE && $entry->size > $MAX_FILE_SIZE ) { $self->_error( qq[Entry '] . $entry->full_path . qq[' declared size ] . $entry->size . qq[ bytes exceeds \$Archive::Tar::MAX_FILE_SIZE ] . qq[($MAX_FILE_SIZE); refusing to allocate] ); next LOOP; } $data = $entry->get_content_by_ref; my $skip = 0; my $ctx; # cdrake ### skip this entry if we're filtering if($md5) { # cdrake $ctx = Digest::MD5->new; # cdrake $skip=5; # cdrake } elsif ($filter && $entry->name !~ $filter) { $skip = 1; } elsif ($filter_cb && ! $filter_cb->($entry)) { $skip = 2; ### skip this entry if it's a pax header. This is a special file added ### by, among others, git-generated tarballs. It holds comments and is ### not meant for extracting. See #38932: pax_global_header extracted } elsif ( $entry->name eq PAX_HEADER or $entry->type =~ /^(x|g)$/ ) { $skip = 3; } if ($skip) { # # Since we're skipping, do not allocate memory for the # whole file. Read it 64 BLOCKS at a time. Do not # complete the skip yet because maybe what we read is a # longlink and it won't get skipped after all # my $amt = $block; my $fsz=$entry->size; # cdrake while ($amt > 0) { $$data = ''; my $this = 64 * BLOCK; $this = $amt if $this > $amt; if( $handle->read( $$data, $this ) < $this ) { $self->_error( qq[Read error on tarfile (missing data) ']. $entry->full_path ."' at offset $offset" ); next LOOP; } $amt -= $this; $fsz -= $this; # cdrake substr ($$data, $fsz) = "" if ($fsz<0); # remove external junk prior to md5 # cdrake $ctx->add($$data) if($skip==5); # cdrake } $$data = $ctx->hexdigest if($skip==5 && !$entry->is_longlink && !$entry->is_unknown && !$entry->is_label ) ; # cdrake } else { ### just read everything into memory ### can't do lazy loading since IO::Zlib doesn't support 'seek' ### this is because Compress::Zlib doesn't support it =/ ### this reads in the whole data in one read() call. if ( $handle->read( $$data, $block ) < $block ) { $self->_error( qq[Read error on tarfile (missing data) ']. $entry->full_path ."' at offset $offset" ); next LOOP; } ### throw away trailing garbage ### substr ($$data, $entry->size) = "" if defined $$data; } ### part II of the @LongLink munging -- need to do /after/ ### the checksum check. if( $entry->is_longlink ) { ### weird thing in tarfiles -- if the file is actually a ### @LongLink, the data part seems to have a trailing ^@ ### (unprintable) char. to display, pipe output through less. ### but that doesn't *always* happen.. so check if the last ### character is a control character, and if so remove it ### at any rate, we better remove that character here, or tests ### like 'eq' and hash lookups based on names will SO not work ### remove it by calculating the proper size, and then ### tossing out everything that's longer than that size. ### count number of nulls my $nulls = $$data =~ tr/\0/\0/; ### cut data + size by that many bytes $entry->size( $entry->size - $nulls ); substr ($$data, $entry->size) = ""; } } ### clean up of the entries.. posix tar /apparently/ has some ### weird 'feature' that allows for filenames > 255 characters ### they'll put a header in with as name '././@LongLink' and the ### contents will be the name of the /next/ file in the archive ### pretty crappy and kludgy if you ask me ### set the name for the next entry if this is a @LongLink; ### this is one ugly hack =/ but needed for direct extraction if( $entry->is_longlink ) { $real_name = $data; next LOOP; } elsif ( defined $real_name ) { $entry->name( $$real_name ); $entry->prefix(''); undef $real_name; } if ($filter && $entry->name !~ $filter) { next LOOP; } elsif ($filter_cb && ! $filter_cb->($entry)) { next LOOP; ### skip this entry if it's a pax header. This is a special file added ### by, among others, git-generated tarballs. It holds comments and is ### not meant for extracting. See #38932: pax_global_header extracted } elsif ( $entry->name eq PAX_HEADER or $entry->type =~ /^(x|g)$/ ) { next LOOP; } if ( $extract && !$entry->is_longlink && !$entry->is_unknown && !$entry->is_label ) { $self->_extract_file( $entry ) or return; } ### Guard against tarfiles with garbage at the end last LOOP if $entry->name eq ''; ### push only the name on the rv if we're extracting ### -- for extract_archive push @$tarfile, ($extract ? $entry->name : $entry); if( $limit ) { $count-- unless $entry->is_longlink || $entry->is_dir; last LOOP unless $count; } } continue { undef $data; } return $tarfile; } =head2 $tar->contains_file( $filename ) Check if the archive contains a certain file. It will return true if the file is in the archive, false otherwise. Note however, that this function does an exact match using C on the full path. So it cannot compensate for case-insensitive file- systems or compare 2 paths to see if they would point to the same underlying file. =cut sub contains_file { my $self = shift; my $full = shift; return unless defined $full; ### don't warn if the entry isn't there.. that's what this function ### is for after all. local $WARN = 0; return 1 if $self->_find_entry($full); return; } =head2 $tar->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 is called without a list of file names, the entire contents of the archive are extracted. Returns a list of filenames extracted. =cut sub extract { my $self = shift; my @args = @_; my @files; my $hashmap; # use the speed optimization for all extracted files local($self->{cwd}) = cwd() unless $self->{cwd}; ### you requested the extraction of only certain files if( @args ) { for my $file ( @args ) { ### it's already an object? if( UNIVERSAL::isa( $file, 'Archive::Tar::File' ) ) { push @files, $file; next; ### go find it then } else { # create hash-map once to speed up lookup $hashmap = $hashmap || { map { $_->full_path, $_ } @{$self->_data} }; if (exists $hashmap->{$file}) { ### we found the file you're looking for push @files, $hashmap->{$file}; } else { return $self->_error( qq[Could not find '$file' in archive] ); } } } ### just grab all the file items } else { @files = $self->get_files; } ### nothing found? that's an error unless( scalar @files ) { $self->_error( qq[No files found for ] . $self->_file ); return; } ### now extract them for my $entry ( @files ) { unless( $self->_extract_file( $entry ) ) { $self->_error(q[Could not extract ']. $entry->full_path .q['] ); return; } } return @files; } =head2 $tar->extract_file( $file, [$extract_path] ) Write an entry, whose name is equivalent to the file name provided to disk. Optionally takes a second parameter, which is the full native path (including filename) the entry will be written to. For example: $tar->extract_file( 'name/in/archive', 'name/i/want/to/give/it' ); $tar->extract_file( $at_file_object, 'name/i/want/to/give/it' ); Returns true on success, false on failure. =cut sub extract_file { my $self = shift; my $file = shift; return unless defined $file; my $alt = shift; my $entry = $self->_find_entry( $file ) or $self->_error( qq[Could not find an entry for '$file'] ), return; return $self->_extract_file( $entry, $alt ); } sub _extract_file { my $self = shift; my $entry = shift or return; my $alt = shift; ### you wanted an alternate extraction location ### my $name = defined $alt ? $alt : $entry->full_path; ### splitpath takes a bool at the end to indicate ### that it's splitting a dir my ($vol,$dirs,$file); if ( defined $alt ) { # It's a local-OS path ($vol,$dirs,$file) = File::Spec->splitpath( $alt, $entry->is_dir ); } else { ($vol,$dirs,$file) = File::Spec::Unix->splitpath( $name, $entry->is_dir ); } my $dir; ### is $name an absolute path? ### if( $vol || File::Spec->file_name_is_absolute( $dirs ) ) { ### absolute names are not allowed to be in tarballs under ### strict mode, so only allow it if a user tells us to do it if( not defined $alt and not $INSECURE_EXTRACT_MODE ) { $self->_error( q[Entry ']. $entry->full_path .q[' is an absolute path. ]. q[Not extracting absolute paths under SECURE EXTRACT MODE] ); return; } ### user asked us to, it's fine. $dir = File::Spec->catpath( $vol, $dirs, "" ); ### it's a relative path ### } else { my $cwd = (ref $self and defined $self->{cwd}) ? $self->{cwd} : cwd(); my @dirs = defined $alt ? File::Spec->splitdir( $dirs ) # It's a local-OS path : File::Spec::Unix->splitdir( $dirs ); # it's UNIX-style, likely # straight from the tarball if( not defined $alt and not $INSECURE_EXTRACT_MODE ) { ### paths that leave the current directory are not allowed under ### strict mode, so only allow it if a user tells us to do this. if( grep { $_ eq '..' } @dirs ) { $self->_error( q[Entry ']. $entry->full_path .q[' is attempting to leave ]. q[the current working directory. Not extracting under ]. q[SECURE EXTRACT MODE] ); return; } ### the archive may be asking us to extract into a symlink. This ### is not sane and a possible security issue, as outlined here: ### https://rt.cpan.org/Ticket/Display.html?id=30380 ### https://bugzilla.redhat.com/show_bug.cgi?id=295021 ### https://issues.rpath.com/browse/RPL-1716 my $full_path = $cwd; for my $d ( @dirs ) { $full_path = File::Spec->catdir( $full_path, $d ); ### we've already checked this one, and it's safe. Move on. next if ref $self and $self->{_link_cache}->{$full_path}; if( -l $full_path ) { my $to = readlink $full_path; my $diag = "symlinked directory ($full_path => $to)"; $self->_error( q[Entry ']. $entry->full_path .q[' is attempting to ]. qq[extract to a $diag. This is considered a security ]. q[vulnerability and not allowed under SECURE EXTRACT ]. q[MODE] ); return; } ### XXX keep a cache if possible, so the stats become cheaper: $self->{_link_cache}->{$full_path} = 1 if ref $self; } } ### '.' is the directory delimiter on VMS, which has to be escaped ### or changed to '_' on vms. vmsify is used, because older versions ### of vmspath do not handle this properly. ### Must not add a '/' to an empty directory though. map { length() ? VMS::Filespec::vmsify($_.'/') : $_ } @dirs if ON_VMS; my ($cwd_vol,$cwd_dir,$cwd_file) = File::Spec->splitpath( $cwd ); my @cwd = File::Spec->splitdir( $cwd_dir ); push @cwd, $cwd_file if length $cwd_file; ### We need to pass '' as the last element to catpath. Craig Berry ### explains why (msgid ): ### The root problem is that splitpath on UNIX always returns the ### final path element as a file even if it is a directory, and of ### course there is no way it can know the difference without checking ### against the filesystem, which it is documented as not doing. When ### you turn around and call catpath, on VMS you have to know which bits ### are directory bits and which bits are file bits. In this case we ### know the result should be a directory. I had thought you could omit ### the file argument to catpath in such a case, but apparently on UNIX ### you can't. $dir = File::Spec->catpath( $cwd_vol, File::Spec->catdir( @cwd, @dirs ), '' ); ### catdir() returns undef if the path is longer than 255 chars on ### older VMS systems. unless ( defined $dir ) { $^W && $self->_error( qq[Could not compose a path for '$dirs'\n] ); return; } } if( -e $dir && !-d _ ) { $^W && $self->_error( qq['$dir' exists, but it's not a directory!\n] ); return; } unless ( -d _ ) { eval { File::Path::mkpath( $dir, 0, 0777 ) }; if( $@ ) { my $fp = $entry->full_path; $self->_error(qq[Could not create directory '$dir' for '$fp': $@]); return; } ### XXX chown here? that might not be the same as in the archive ### as we're only chown'ing to the owner of the file we're extracting ### not to the owner of the directory itself, which may or may not ### be another entry in the archive ### Answer: no, gnu tar doesn't do it either, it'd be the wrong ### way to go. #if( $CHOWN && CAN_CHOWN ) { # chown $entry->uid, $entry->gid, $dir or # $self->_error( qq[Could not set uid/gid on '$dir'] ); #} } ### we're done if we just needed to create a dir ### return 1 if $entry->is_dir; my $full = File::Spec->catfile( $dir, $file ); if( $entry->is_unknown ) { $self->_error( qq[Unknown file type for file '$full'] ); return; } ### If a file system already contains a block device with the same name as ### the being extracted regular file, we would write the file's content ### to the block device. So remove the existing file (block device) now. ### If an archive contains multiple same-named entries, the last one ### should replace the previous ones. So remove the old file now. ### If the old entry is a symlink to a file outside of the CWD, the new ### entry would create a file there. This is CVE-2018-12015 ### . if (-l $full || -e _) { if (!unlink $full) { $self->_error( qq[Could not remove old file '$full': $!] ); return; } } if( length $entry->type && $entry->is_file ) { my $fh = IO::File->new; $fh->open( $full, '>' ) or ( $self->_error( qq[Could not open file '$full': $!] ), return ); if( $entry->size ) { binmode $fh; syswrite $fh, $entry->data or ( $self->_error( qq[Could not write data to '$full'] ), return ); } close $fh or ( $self->_error( qq[Could not close file '$full'] ), return ); } else { $self->_make_special_file( $entry, $full ) or return; } ### only update the timestamp if it's not a symlink; that will change the ### timestamp of the original. This addresses bug #33669: Could not update ### timestamp warning on symlinks if( not -l $full ) { utime time, $entry->mtime - TIME_OFFSET, $full or $self->_error( qq[Could not update timestamp] ); } if( $CHOWN && CAN_CHOWN->() and not -l $full ) { chown $entry->uid, $entry->gid, $full or $self->_error( qq[Could not set uid/gid on '$full'] ); } ### only chmod if we're allowed to, but never chmod symlinks, since they'll ### change the perms on the file they're linking too... if( $CHMOD and not -l $full ) { my $mode = $entry->mode; unless ($SAME_PERMISSIONS) { $mode &= ~(oct(7000) | umask); } chmod $mode, $full or $self->_error( qq[Could not chown '$full' to ] . $entry->mode ); } return 1; } sub _make_special_file { my $self = shift; my $entry = shift or return; my $file = shift; return unless defined $file; my $err; if( $entry->is_symlink ) { if( !$INSECURE_EXTRACT_MODE ) { my $linkname = $entry->linkname; if( File::Spec->file_name_is_absolute($linkname) ) { $self->_error( qq[Symlink '] . $entry->full_path . qq[' has absolute target. Not extracting under SECURE EXTRACT MODE] ); return; } if( !defined _symlinks_resolver( $entry->full_path, $linkname, 1 ) ) { $self->_error( qq[Symlink '] . $entry->full_path . qq[' target attempts traversal. Not extracting under SECURE EXTRACT MODE] ); return; } } my $fail; if( ON_UNIX ) { symlink( $entry->linkname, $file ) or $fail++; } else { $self->_extract_special_file_as_plain_file( $entry, $file ) or $fail++; } $err = qq[Making symbolic link '$file' to '] . $entry->linkname .q[' failed] if $fail; } elsif ( $entry->is_hardlink ) { if( !$INSECURE_EXTRACT_MODE ) { my $linkname = $entry->linkname; if( File::Spec->file_name_is_absolute($linkname) ) { $self->_error( qq[Hardlink '] . $entry->full_path . qq[' has absolute target '$linkname'. Not extracting ] . qq[under SECURE EXTRACT MODE: extraction itself chmods ] . qq[the shared inode.] ); return; } if( !defined _symlinks_resolver( $entry->full_path, $linkname, 1 ) ) { $self->_error( qq[Hardlink '] . $entry->full_path . qq[' target '$linkname' attempts traversal. Not ] . qq[extracting under SECURE EXTRACT MODE: extraction ] . qq[itself chmods the shared inode.] ); return; } } my $fail; if( ON_UNIX ) { link( $entry->linkname, $file ) or $fail++; } else { $self->_extract_special_file_as_plain_file( $entry, $file ) or $fail++; } $err = qq[Making hard link from '] . $entry->linkname . qq[' to '$file' failed] if $fail; } elsif ( $entry->is_fifo ) { ON_UNIX && !system('mknod', $file, 'p') or $err = qq[Making fifo ']. $entry->name .qq[' failed]; } elsif ( $entry->is_blockdev or $entry->is_chardev ) { my $mode = $entry->is_blockdev ? 'b' : 'c'; ON_UNIX && !system('mknod', $file, $mode, $entry->devmajor, $entry->devminor) or $err = qq[Making block device ']. $entry->name .qq[' (maj=] . $entry->devmajor . qq[ min=] . $entry->devminor . qq[) failed.]; } elsif ( $entry->is_socket ) { ### the original doesn't do anything special for sockets.... ### 1; } return $err ? $self->_error( $err ) : 1; } ### don't know how to make symlinks, let's just extract the file as ### a plain file sub _extract_special_file_as_plain_file { my $self = shift; my $entry = shift or return; my $file = shift; return unless defined $file; my $err; TRY: { my $orig = $self->_find_entry( $entry->linkname, $entry ); unless( $orig ) { $err = qq[Could not find file '] . $entry->linkname . qq[' in memory.]; last TRY; } ### clone the entry, make it appear as a normal file ### my $clone = $orig->clone; $clone->_downgrade_to_plainfile; $self->_extract_file( $clone, $file ) or last TRY; return 1; } return $self->_error($err); } =head2 $tar->list_files( [\@properties] ) Returns a list of the names of all the files in the archive. If C 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, making it equivalent to calling C without arguments. =cut sub list_files { my $self = shift; my $aref = shift || [ ]; unless( $self->_data ) { $self->read() or return; } if( @$aref == 0 or ( @$aref == 1 and $aref->[0] eq 'name' ) ) { return map { $_->full_path } @{$self->_data}; } else { #my @rv; #for my $obj ( @{$self->_data} ) { # push @rv, { map { $_ => $obj->$_() } @$aref }; #} #return @rv; ### this does the same as the above.. just needs a +{ } ### to make sure perl doesn't confuse it for a block return map { my $o=$_; +{ map { $_ => $o->$_() } @$aref } } @{$self->_data}; } } sub _find_entry { my $self = shift; my $file = shift; unless( defined $file ) { $self->_error( qq[No file specified] ); return; } ### it's an object already return $file if UNIVERSAL::isa( $file, 'Archive::Tar::File' ); seach_entry: if($self->_data){ for my $entry ( @{$self->_data} ) { my $path = $entry->full_path; return $entry if $path eq $file; } } if($Archive::Tar::RESOLVE_SYMLINK!~/none/){ if(my $link_entry = shift()){#fallback mode when symlinks are using relative notations ( ../a/./b/text.bin ) $file = _symlinks_resolver( $link_entry->name, $file ); goto seach_entry if $self->_data; #this will be slower than never, but won't failed! my $iterargs = $link_entry->{'_archive'}; if($Archive::Tar::RESOLVE_SYMLINK=~/speed/ && @$iterargs==3){ #faster but whole archive will be read in memory #read whole archive and share data my $archive = Archive::Tar->new; $archive->read( @$iterargs ); push @$iterargs, $archive; #take a trace for destruction if($archive->_data){ $self->_data( $archive->_data ); goto seach_entry; } }#faster {#slower but lower memory usage # $iterargs = [$filename, $compressed, $opts]; my $next = Archive::Tar->iter( @$iterargs ); while(my $e = $next->()){ if($e->full_path eq $file){ undef $next; return $e; } } }#slower } } $self->_error( qq[No such file in archive: '$file'] ); return; } =head2 $tar->get_files( [@filenames] ) Returns the C objects matching the filenames provided. If no filename list was passed, all C objects in the current Tar object are returned. Please refer to the C documentation on how to handle these objects. =cut sub get_files { my $self = shift; return @{ $self->_data } unless @_; my @list; for my $file ( @_ ) { push @list, grep { defined } $self->_find_entry( $file ); } return @list; } =head2 $tar->get_content( $file ) Return the content of the named file. =cut sub get_content { my $self = shift; my $entry = $self->_find_entry( shift ) or return; return $entry->data; } =head2 $tar->replace_content( $file, $content ) Make the string $content be the content for the file named $file. =cut sub replace_content { my $self = shift; my $entry = $self->_find_entry( shift ) or return; return $entry->replace_content( shift ); } =head2 $tar->rename( $file, $new_name ) Rename the file of the in-memory archive to $new_name. Note that you must specify a Unix path for $new_name, since per tar standard, all files in the archive must be Unix paths. Returns true on success and false on failure. =cut sub rename { my $self = shift; my $file = shift; return unless defined $file; my $new = shift; return unless defined $new; my $entry = $self->_find_entry( $file ) or return; return $entry->rename( $new ); } =head2 $tar->chmod( $file, $mode ) Change mode of $file to $mode. Returns true on success and false on failure. =cut sub chmod { my $self = shift; my $file = shift; return unless defined $file; my $mode = shift; return unless defined $mode && $mode =~ /^[0-7]{1,4}$/; my @args = ("$mode"); my $entry = $self->_find_entry( $file ) or return; my $x = $entry->chmod( @args ); return $x; } =head2 $tar->chown( $file, $uname [, $gname] ) Change owner $file to $uname and $gname. Returns true on success and false on failure. =cut sub chown { my $self = shift; my $file = shift; return unless defined $file; my $uname = shift; return unless defined $uname; my @args = ($uname); push(@args, shift); my $entry = $self->_find_entry( $file ) or return; my $x = $entry->chown( @args ); return $x; } =head2 $tar->remove (@filenamelist) Removes any entries with names matching any of the given filenames from the in-memory archive. Returns a list of C objects that remain. =cut sub remove { my $self = shift; my @list = @_; my %seen = map { $_->full_path => $_ } @{$self->_data}; delete $seen{ $_ } for @list; $self->_data( [values %seen] ); return values %seen; } =head2 $tar->clear C clears the current in-memory archive. This effectively gives you a 'blank' object, ready to be filled again. Note that C only has effect on the object, not the underlying tarfile. =cut sub clear { my $self = shift or return; $self->_data( [] ); $self->_file( '' ); return 1; } =head2 $tar->write ( [$file, $compressed, $prefix] ) 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 filehandle (a GLOB reference). The second argument is used to indicate compression. You can either compress using C or C. If you pass a digit, it's assumed to be the C compression level (between 1 and 9), but the use of constants is preferred: # write a gzip compressed file $tar->write( 'out.tgz', COMPRESS_GZIP ); # write a bzip compressed file $tar->write( 'out.tbz', COMPRESS_BZIP ); Note that when you pass in a filehandle, the compression argument is ignored, as all files are printed verbatim to your filehandle. If you wish to enable compression with filehandles, use an C or C filehandle instead. The third argument is an optional prefix. All files will be tucked away in the directory you specify as prefix. So if you have files 'a' and 'b' in your archive, and you specify 'foo' as prefix, they will be written to the archive as 'foo/a' and 'foo/b'. If no arguments are given, C 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. =cut sub write { my $self = shift; my $file = shift; $file = '' unless defined $file; my $gzip = shift || 0; my $ext_prefix = shift; $ext_prefix = '' unless defined $ext_prefix; my $dummy = ''; ### only need a handle if we have a file to print to ### my $handle = length($file) ? ( $self->_get_handle($file, $gzip, WRITE_ONLY->($gzip) ) or return ) : $HAS_PERLIO ? do { open my $h, '>', \$dummy; $h } : $HAS_IO_STRING ? IO::String->new : __PACKAGE__->no_string_support(); ### Addresses: #41798: Nonempty $\ when writing a TAR file produces a ### corrupt TAR file. Must clear out $\ to make sure no garbage is ### printed to the archive local $\; for my $entry ( @{$self->_data} ) { ### entries to be written to the tarfile ### my @write_me; ### only now will we change the object to reflect the current state ### of the name and prefix fields -- this needs to be limited to ### write() only! my $clone = $entry->clone; ### so, if you don't want use to use the prefix, we'll stuff ### everything in the name field instead if( $DO_NOT_USE_PREFIX ) { ### you might have an extended prefix, if so, set it in the clone ### XXX is ::Unix right? $clone->name( length $ext_prefix ? File::Spec::Unix->catdir( $ext_prefix, $clone->full_path) : $clone->full_path ); $clone->prefix( '' ); ### otherwise, we'll have to set it properly -- prefix part in the ### prefix and name part in the name field. } else { ### split them here, not before! my ($prefix,$name) = $clone->_prefix_and_file( $clone->full_path ); ### you might have an extended prefix, if so, set it in the clone ### XXX is ::Unix right? $prefix = File::Spec::Unix->catdir( $ext_prefix, $prefix ) if length $ext_prefix; $clone->prefix( $prefix ); $clone->name( $name ); } ### names are too long, and will get truncated if we don't add a ### '@LongLink' file... my $make_longlink = ( length($clone->name) > NAME_LENGTH or length($clone->prefix) > PREFIX_LENGTH ) || 0; ### perhaps we need to make a longlink file? if( $make_longlink ) { my $longlink = Archive::Tar::File->new( data => LONGLINK_NAME, $clone->full_path, { type => LONGLINK } ); unless( $longlink ) { $self->_error( qq[Could not create 'LongLink' entry for ] . qq[oversize file '] . $clone->full_path ."'" ); return; }; push @write_me, $longlink; } push @write_me, $clone; ### write the one, optionally 2 a::t::file objects to the handle for my $clone (@write_me) { ### if the file is a symlink, there are 2 options: ### either we leave the symlink intact, but then we don't write any ### data OR we follow the symlink, which means we actually make a ### copy. if we do the latter, we have to change the TYPE of the ### clone to 'FILE' my $link_ok = $clone->is_symlink && $Archive::Tar::FOLLOW_SYMLINK; my $data_ok = !$clone->is_symlink && $clone->has_content; ### downgrade to a 'normal' file if it's a symlink we're going to ### treat as a regular file $clone->_downgrade_to_plainfile if $link_ok; ### get the header for this block my $header = $self->_format_tar_entry( $clone ); unless( $header ) { $self->_error(q[Could not format header for: ] . $clone->full_path ); return; } unless( print $handle $header ) { $self->_error(q[Could not write header for: ] . $clone->full_path); return; } if( $link_ok or $data_ok ) { unless( print $handle $clone->data ) { $self->_error(q[Could not write data for: ] . $clone->full_path); return; } ### pad the end of the clone if required ### print $handle TAR_PAD->( $clone->size ) if $clone->size % BLOCK } } ### done writing these entries } ### write the end markers ### print $handle TAR_END x 2 or return $self->_error( qq[Could not write tar end markers] ); ### did you want it written to a file, or returned as a string? ### my $rv = length($file) ? 1 : $HAS_PERLIO ? $dummy : do { seek $handle, 0, 0; local $/; <$handle> }; ### make sure to close the handle if we created it if ( $file ne $handle ) { unless( close $handle ) { $self->_error( qq[Could not write tar] ); return; } } return $rv; } sub _format_tar_entry { my $self = shift; my $entry = shift or return; my $ext_prefix = shift; $ext_prefix = '' unless defined $ext_prefix; my $no_prefix = shift || 0; my $file = $entry->name; my $prefix = $entry->prefix; $prefix = '' unless defined $prefix; ### remove the prefix from the file name ### not sure if this is still needed --kane ### no it's not -- Archive::Tar::File->_new_from_file will take care of ### this for us. Even worse, this would break if we tried to add a file ### like x/x. #if( length $prefix ) { # $file =~ s/^$match//; #} $prefix = File::Spec::Unix->catdir($ext_prefix, $prefix) if length $ext_prefix; ### not sure why this is... ### my $l = PREFIX_LENGTH; # is ambiguous otherwise... substr ($prefix, 0, -$l) = "" if length $prefix >= PREFIX_LENGTH; my $f1 = "%06o"; my $f2 = $ZERO_PAD_NUMBERS ? "%011o" : "%11o"; ### this might be optimizable with a 'changed' flag in the file objects ### my $tar = pack ( PACK, $file, (map { sprintf( $f1, $entry->$_() ) } qw[mode uid gid]), (map { sprintf( $f2, $entry->$_() ) } qw[size mtime]), "", # checksum field - space padded a bit down (map { $entry->$_() } qw[type linkname magic]), $entry->version || TAR_VERSION, (map { $entry->$_() } qw[uname gname]), (map { sprintf( $f1, $entry->$_() ) } qw[devmajor devminor]), ($no_prefix ? '' : $prefix) ); ### add the checksum ### my $checksum_fmt = $ZERO_PAD_NUMBERS ? "%06o\0" : "%06o\0"; substr($tar,148,7) = sprintf("%6o\0", unpack("%16C*",$tar)); return $tar; } =head2 $tar->add_files( @filenamelist ) Takes a list of filenames and adds them to the in-memory archive. The path to the file is automatically converted to a Unix like equivalent for use in the archive, and, if on MacOS, the file's modification time is converted from the MacOS epoch to the Unix epoch. So tar archives created on MacOS with B can be read both with I on Unix and applications like I or I 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. Instead of a filename, you can also pass it an existing C object from, for example, another archive. The object will be clone, and effectively be a copy of the original, not an alias. Returns a list of C objects that were just added. =cut sub add_files { my $self = shift; my @files = @_ or return; my @rv; for my $file ( @files ) { ### you passed an Archive::Tar::File object ### clone it so we don't accidentally have a reference to ### an object from another archive if( UNIVERSAL::isa( $file,'Archive::Tar::File' ) ) { push @rv, $file->clone; next; } eval { if( utf8::is_utf8( $file )) { utf8::encode( $file ); } }; unless( -e $file || -l $file ) { $self->_error( qq[No such file: '$file'] ); next; } my $obj = Archive::Tar::File->new( file => $file ); unless( $obj ) { $self->_error( qq[Unable to add file: '$file'] ); next; } push @rv, $obj; } push @{$self->{_data}}, @rv; return @rv; } =head2 $tar->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, type. (On MacOS, the file's path and modification times are converted to Unix equivalents.) Valid values for the file type are the following constants defined by Archive::Tar::Constant: =over 4 =item FILE Regular file. =item HARDLINK =item SYMLINK Hard and symbolic ("soft") links; linkname should specify target. =item CHARDEV =item BLOCKDEV Character and block devices. devmajor and devminor should specify the major and minor device numbers. =item DIR Directory. =item FIFO FIFO (named pipe). =item SOCKET Socket. =back Returns the C object that was just added, or C on failure. =cut sub add_data { my $self = shift; my ($file, $data, $opt) = @_; my $obj = Archive::Tar::File->new( data => $file, $data, $opt ); unless( $obj ) { $self->_error( qq[Unable to add file: '$file'] ); return; } push @{$self->{_data}}, $obj; return $obj; } =head2 $tar->error( [$BOOL] ) Returns the current error string (usually, the last error reported). If a true value was specified, it will give the C equivalent of the error, in effect giving you a stacktrace. For backwards compatibility, this error is also available as C<$Archive::Tar::error> although it is much recommended you use the method call instead. =cut { $error = ''; my $longmess; sub _error { my $self = shift; my $msg = $error = shift; $longmess = Carp::longmess($error); if (ref $self) { $self->{_error} = $error; $self->{_longmess} = $longmess; } ### set Archive::Tar::WARN to 0 to disable printing ### of errors if( $WARN ) { carp $DEBUG ? $longmess : $msg; } return; } sub error { my $self = shift; if (ref $self) { return shift() ? $self->{_longmess} : $self->{_error}; } else { return shift() ? $longmess : $error; } } } =head2 $tar->setcwd( $cwd ); C needs to know the current directory, and it will run C I time it extracts a I entry from the tarfile and saves it in the file system. (As of version 1.30, however, C will use the speed optimization described below automatically, so it's only relevant if you're using C). Since C doesn't change the current directory internally while it is extracting the items in a tarball, all calls to C can be avoided if we can guarantee that the current directory doesn't get changed externally. To use this performance boost, set the current directory via use Cwd; $tar->setcwd( cwd() ); once before calling a function like C and C will use the current directory setting from then on and won't call C internally. To switch back to the default behaviour, use $tar->setcwd( undef ); and C will call C internally again. If you're using C's C method, C will be called for you. =cut sub setcwd { my $self = shift; my $cwd = shift; $self->{cwd} = $cwd; } =head1 Class Methods =head2 Archive::Tar->create_archive($file, $compressed, @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 is used to indicate compression. You can either compress using C or C. If you pass a digit, it's assumed to be the C compression level (between 1 and 9), but the use of constants is preferred: # write a gzip compressed file Archive::Tar->create_archive( 'out.tgz', COMPRESS_GZIP, @filelist ); # write a bzip compressed file Archive::Tar->create_archive( 'out.tbz', COMPRESS_BZIP, @filelist ); Note that when you pass in a filehandle, the compression argument is ignored, as all files are printed verbatim to your filehandle. If you wish to enable compression with filehandles, use an C or C filehandle instead. 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 will return false. Please use the C method to find the cause of the failure. Note that this method does not write C as it were; it still reads all the files into memory before writing out the archive. Consult the FAQ below if this is a problem. =cut sub create_archive { my $class = shift; my $file = shift; return unless defined $file; my $gzip = shift || 0; my @files = @_; unless( @files ) { return $class->_error( qq[Cowardly refusing to create empty archive!] ); } my $tar = $class->new; $tar->add_files( @files ); return $tar->write( $file, $gzip ); } =head2 Archive::Tar->iter( $filename, [ $compressed, {opt => $val} ] ) Returns an iterator function that reads the tar file without loading it all in memory. Each time the function is called it will return the next file in the tarball. The files are returned as C objects. The iterator function returns the empty list once it has exhausted the files contained. The second argument can be a hash reference with options, which are identical to the arguments passed to C. Example usage: my $next = Archive::Tar->iter( "example.tar.gz", 1, {filter => qr/\.pm$/} ); while( my $f = $next->() ) { print $f->name, "\n"; $f->extract or warn "Extraction failed"; # .... } =cut sub iter { my $class = shift; my $filename = shift; return unless defined $filename; my $compressed = shift || 0; my $opts = shift || {}; ### get a handle to read from. my $handle = $class->_get_handle( $filename, $compressed, READ_ONLY->( ZLIB ) ) or return; my @data; my $CONSTRUCT_ARGS = [ $filename, $compressed, $opts ]; return sub { return shift(@data) if @data; # more than one file returned? return unless $handle; # handle exhausted? ### read data, should only return file my $tarfile = $class->_read_tar($handle, { %$opts, limit => 1 }); @data = @$tarfile if ref $tarfile && ref $tarfile eq 'ARRAY'; if($Archive::Tar::RESOLVE_SYMLINK!~/none/){ foreach(@data){ #may refine this heuristic for ON_UNIX? if($_->linkname){ #is there a better slot to store/share it ? $_->{'_archive'} = $CONSTRUCT_ARGS; } } } ### return one piece of data return shift(@data) if @data; ### data is exhausted, free the filehandle undef $handle; if(@$CONSTRUCT_ARGS == 4){ #free archive in memory undef $CONSTRUCT_ARGS->[-1]; } return; }; } =head2 Archive::Tar->list_archive($file, $compressed, [\@properties]) 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 list or a reference to an open file handle (e.g. a GLOB reference). If C is passed an array reference as its third argument it returns a list of hash references containing the requested properties of each file. The following list of properties is supported: full_path, name, size, mtime (last modified date), mode, uid, gid, linkname, uname, gname, devmajor, devminor, prefix, type. See C for details about supported properties. 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. =cut sub list_archive { my $class = shift; my $file = shift; return unless defined $file; my $gzip = shift || 0; my $tar = $class->new($file, $gzip); return unless $tar; return $tar->list_files( @_ ); } =head2 Archive::Tar->extract_archive($file, $compressed) 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. C will return a list of files it extracted. If the archive extraction fails for any reason, C will return false. Please use the C method to find the cause of the failure. =cut sub extract_archive { my $class = shift; my $file = shift; return unless defined $file; my $gzip = shift || 0; my $tar = $class->new( ) or return; return $tar->read( $file, $gzip, { extract => 1 } ); } =head2 $bool = Archive::Tar->has_io_string Returns true if we currently have C support loaded. Either C or C support is needed to support writing stringified archives. Currently, C is the preferred method, if available. See the C section to see how to change this preference. =cut sub has_io_string { return $HAS_IO_STRING; } =head2 $bool = Archive::Tar->has_perlio Returns true if we currently have C support loaded. This requires C or higher, compiled with C Either C or C support is needed to support writing stringified archives. Currently, C is the preferred method, if available. See the C section to see how to change this preference. =cut sub has_perlio { return $HAS_PERLIO; } =head2 $bool = Archive::Tar->has_zlib_support Returns true if C can extract C compressed archives =cut sub has_zlib_support { return ZLIB } =head2 $bool = Archive::Tar->has_bzip2_support Returns true if C can extract C compressed archives =cut sub has_bzip2_support { return BZIP } =head2 Archive::Tar->can_handle_compressed_files A simple checking routine, which will return true if C is able to uncompress compressed archives on the fly with C and C or false if not both are installed. You can use this as a shortcut to determine whether C will do what you think before passing compressed archives to its C method. =cut sub can_handle_compressed_files { return ZLIB && BZIP ? 1 : 0 } sub no_string_support { croak("You have to install IO::String to support writing archives to strings"); } sub _symlinks_resolver{ my ($src, $trg, $strict) = @_; my @src = split /[\/\\]/, $src; my @trg = split /[\/\\]/, $trg; pop @src; #strip out current object name if(@trg and $trg[0] eq ''){ shift @trg; #restart path from scratch @src = ( ); } foreach my $part ( @trg ){ next if $part eq '.'; #ignore current if($part eq '..'){ #got to parent return if $strict && !@src; pop @src; } else{ #append it push @src, $part; } } my $path = join('/', @src); warn "_symlinks_resolver('$src','$trg') = $path" if $DEBUG; return $path; } 1; __END__ =head1 GLOBAL VARIABLES =head2 $Archive::Tar::FOLLOW_SYMLINK Set this variable to C<1> to make C effectively make a copy of the file when extracting. Default is C<0>, which means the symlink stays intact. Of course, you will have to pack the file linked to as well. This option is checked when you write out the tarfile using C or C. This works just like C's C<-h> option. =head2 $Archive::Tar::CHOWN By default, C will try to C your files if it is able to. In some cases, this may not be desired. In that case, set this variable to C<0> to disable C-ing, even if it were possible. The default is C<1>. =head2 $Archive::Tar::CHMOD By default, C will try to C your files to whatever mode was specified for the particular file in the archive. In some cases, this may not be desired. In that case, set this variable to C<0> to disable C-ing. The default is C<1>. =head2 $Archive::Tar::SAME_PERMISSIONS When, C<$Archive::Tar::CHMOD> is enabled, this setting controls whether the permissions on files from the archive are used without modification of if they are filtered by removing any setid bits and applying the current umask. The default is C<1> for the root user and C<0> for normal users. =head2 $Archive::Tar::DO_NOT_USE_PREFIX By default, C will try to put paths that are over 100 characters in the C field of your tar header, as defined per POSIX-standard. However, some (older) tar programs do not implement this spec. To retain compatibility with these older or non-POSIX compliant versions, you can set the C<$DO_NOT_USE_PREFIX> variable to a true value, and C will use an alternate way of dealing with paths over 100 characters by using the C feature. Note that clients who do not support the C feature will not be able to read these archives. Such clients include tars on C, C and C. The default is C<0>. =head2 $Archive::Tar::DEBUG Set this variable to C<1> to always get the C output of the warnings, instead of the regular C. This is the same message you would get by doing: $tar->error(1); Defaults to C<0>. =head2 $Archive::Tar::WARN Set this variable to C<0> if you do not want any warnings printed. Personally I recommend against doing this, but people asked for the option. Also, be advised that this is of course not threadsafe. Defaults to C<1>. =head2 $Archive::Tar::error Holds the last reported error. Kept for historical reasons, but its use is very much discouraged. Use the C method instead: warn $tar->error unless $tar->extract; Note that in older versions of this module, the C method would return an effectively global value even when called an instance method as above. This has since been fixed, and multiple instances of C now have separate error strings. =head2 $Archive::Tar::INSECURE_EXTRACT_MODE This variable indicates whether C should allow files to be extracted outside their current working directory. Allowing this could have security implications, as a malicious tar archive could alter or replace any file the extracting user has permissions to. Therefor, the default is to not allow insecure extractions. If you trust the archive, or have other reasons to allow the archive to write files outside your current working directory, set this variable to C. Note that this is a backwards incompatible change from version C<1.36> and before. =head2 $Archive::Tar::HAS_PERLIO This variable holds a boolean indicating if we currently have C support loaded. This will be enabled for any perl greater than C<5.8> compiled with C. If you feel strongly about disabling it, set this variable to C. Note that you will then need C installed to support writing stringified archives. Don't change this variable unless you B know what you're doing. =head2 $Archive::Tar::HAS_IO_STRING This variable holds a boolean indicating if we currently have C support loaded. This will be enabled for any perl that has a loadable C module. If you feel strongly about disabling it, set this variable to C. Note that you will then need C support from your perl to be able to write stringified archives. Don't change this variable unless you B know what you're doing. =head2 $Archive::Tar::ZERO_PAD_NUMBERS This variable holds a boolean indicating if we will create zero padded numbers for C, C and C. The default is C<0>, indicating that we will create space padded numbers. Added for compatibility with C implementations. =head2 Tuning the way RESOLVE_SYMLINK will works You can tune the behaviour by setting the $Archive::Tar::RESOLVE_SYMLINK variable, or $ENV{PERL5_AT_RESOLVE_SYMLINK} before loading the module Archive::Tar. Values can be one of the following: none Disable this mechanism and failed as it was in previous version (<1.88) speed (default) If you prefer speed this will read again the whole archive using read() so all entries will be available memory If you prefer memory Limitation It won't work for terminal, pipe or sockets or every non seekable source. =head2 $Archive::Tar::MAX_FILE_SIZE This variable holds an upper bound on the per-entry declared size that C will accept when reading an archive. Entries whose header claims a larger size are refused with an error before any read allocation. Defaults to 1 GiB. Set to 0 to disable the cap. =cut =head1 FAQ =over 4 =item What's the minimum perl version required to run Archive::Tar? You will need perl version 5.005_03 or newer. =item Isn't Archive::Tar slow? Yes it is. It's pure perl, so it's a lot slower then your C However, it's very portable. If speed is an issue, consider using C instead. =item Isn't Archive::Tar heavier on memory than /bin/tar? Yes it is, see previous answer. Since C and therefore C doesn't support C on their filehandles, there is little choice but to read the archive into memory. This is ok if you want to do in-memory manipulation of the archive. If you just want to extract, use the C class method instead. It will optimize and write to disk immediately. Another option is to use the C class method to iterate over the files in the tarball without reading them all in memory at once. =item Can you lazy-load data instead? In some cases, yes. You can use the C class method to iterate over the files in the tarball without reading them all in memory at once. =item How much memory will an X kb tar file need? Probably more than X kb, since it will all be read into memory. If this is a problem, and you don't need to do in memory manipulation of the archive, consider using the C class method, or C instead. =item What do you do with unsupported filetypes in an archive? C has a few filetypes that aren't supported on other platforms, like C. If we encounter a C or C we'll just try to make a copy of the original file, rather than throwing an error. This does require you to read the entire archive in to memory first, since otherwise we wouldn't know what data to fill the copy with. (This means that you cannot use the class methods, including C on archives that have incompatible filetypes and still expect things to work). For other filetypes, like C and C we'll warn that the extraction of this particular item didn't work. =item I'm using WinZip, or some other non-POSIX client, and files are not being extracted properly! By default, C is in a completely POSIX-compatible mode, which uses the POSIX-specification of C to store files. For paths greater than 100 characters, this is done using the C. Non-POSIX-compatible clients may not support this part of the specification, and may only support the C functionality. To facilitate those clients, you can set the C<$Archive::Tar::DO_NOT_USE_PREFIX> variable to C. See the C section for details on this variable. Note that GNU tar earlier than version 1.14 does not cope well with the C. If you use such a version, consider setting the C<$Archive::Tar::DO_NOT_USE_PREFIX> variable to C. =item How do I extract only files that have property X from an archive? Sometimes, you might not wish to extract a complete archive, just the files that are relevant to you, based on some criteria. You can do this by filtering a list of C objects based on your criteria. For example, to extract only files that have the string C in their title, you would use: $tar->extract( grep { $_->full_path =~ /foo/ } $tar->get_files ); This way, you can filter on any attribute of the files in the archive. Consult the C documentation on how to use these objects. =item How do I access .tar.Z files? The C module can optionally use C (via the C module) to access tar files that have been compressed with C. Unfortunately tar files compressed with the Unix C utility cannot be read by C and so cannot be directly accesses by C. If the C or C programs are available, you can use one of these workarounds to read C<.tar.Z> files from C Firstly with C use Archive::Tar; open F, "uncompress -c $filename |"; my $tar = Archive::Tar->new(*F); ... and this with C use Archive::Tar; open F, "gunzip -c $filename |"; my $tar = Archive::Tar->new(*F); ... Similarly, if the C program is available, you can use this to write a C<.tar.Z> file use Archive::Tar; use IO::File; my $fh = new IO::File "| compress -c >$filename"; my $tar = Archive::Tar->new(); ... $tar->write($fh); $fh->close ; =item How do I handle Unicode strings? C uses byte semantics for any files it reads from or writes to disk. This is not a problem if you only deal with files and never look at their content or work solely with byte strings. But if you use Unicode strings with character semantics, some additional steps need to be taken. For example, if you add a Unicode string like # Problem $tar->add_data('file.txt', "Euro: \x{20AC}"); then there will be a problem later when the tarfile gets written out to disk via C<< $tar->write() >>: Wide character in print at .../Archive/Tar.pm line 1014. The data was added as a Unicode string and when writing it out to disk, the C<:utf8> line discipline wasn't set by C, so Perl tried to convert the string to ISO-8859 and failed. The written file now contains garbage. For this reason, Unicode strings need to be converted to UTF-8-encoded bytestrings before they are handed off to C: use Encode; my $data = "Accented character: \x{20AC}"; $data = encode('utf8', $data); $tar->add_data('file.txt', $data); A opposite problem occurs if you extract a UTF8-encoded file from a tarball. Using C on the C object will return its content as a bytestring, not as a Unicode string. If you want it to be a Unicode string (because you want character semantics with operations like regular expression matching), you need to decode the UTF8-encoded content and have Perl convert it into a Unicode string: use Encode; my $data = $tar->get_content(); # Make it a Unicode string $data = decode('utf8', $data); There is no easy way to provide this functionality in C, because a tarball can contain many files, and each of which could be encoded in a different way. =back =head1 CAVEATS The AIX tar does not fill all unused space in the tar archive with 0x00. This sometimes leads to warning messages from C. Invalid header block at offset nnn A fix for that problem is scheduled to be released in the following levels of AIX, all of which should be coming out in the 4th quarter of 2009: AIX 5.3 TL7 SP10 AIX 5.3 TL8 SP8 AIX 5.3 TL9 SP5 AIX 5.3 TL10 SP2 AIX 6.1 TL0 SP11 AIX 6.1 TL1 SP7 AIX 6.1 TL2 SP6 AIX 6.1 TL3 SP3 The IBM APAR number for this problem is IZ50240 (Reported component ID: 5765G0300 / AIX 5.3). It is possible to get an ifix for that problem. If you need an ifix please contact your local IBM AIX support. =head1 TODO =over 4 =item Check if passed in handles are open for read/write Currently I don't know of any portable pure perl way to do this. Suggestions welcome. =item Allow archives to be passed in as string Currently, we only allow opened filehandles or filenames, but not strings. The internals would need some reworking to facilitate stringified archives. =item Facilitate processing an opened filehandle of a compressed archive Currently, we only support this if the filehandle is an IO::Zlib object. Environments, like apache, will present you with an opened filehandle to an uploaded file, which might be a compressed archive. =back =head1 SEE ALSO =over 4 =item The GNU tar specification C =item The PAX format specification The specification which tar derives from; C< http://www.opengroup.org/onlinepubs/007904975/utilities/pax.html> =item A comparison of GNU and POSIX tar standards; C =item GNU tar intends to switch to POSIX compatibility GNU Tar authors have expressed their intention to become completely POSIX-compatible; C =item A Comparison between various tar implementations Lists known issues and incompatibilities; C =back =head1 AUTHOR This module by Jos Boumans Ekane@cpan.orgE. Please reports bugs to Ebug-archive-tar@rt.cpan.orgE. =head1 ACKNOWLEDGEMENTS Thanks to Sean Burke, Chris Nandor, Chip Salzenberg, Tim Heaney, Gisle Aas, Rainer Tammer and especially Andrew Savige for their help and suggestions. =head1 COPYRIGHT This module is copyright (c) 2002 - 2009 Jos Boumans Ekane@cpan.orgE. All rights reserved. This library is free software; you may redistribute and/or modify it under the same terms as Perl itself. =cut Tar/Constant.pm000064400000010153152347641370007435 0ustar00package Archive::Tar::Constant; BEGIN { require Exporter; $VERSION = '2.30'; @ISA = qw[Exporter]; require Time::Local if $^O eq "MacOS"; } @EXPORT = Archive::Tar::Constant->_list_consts( __PACKAGE__ ); use strict; use warnings; use constant FILE => 0; use constant HARDLINK => 1; use constant SYMLINK => 2; use constant CHARDEV => 3; use constant BLOCKDEV => 4; use constant DIR => 5; use constant FIFO => 6; use constant SOCKET => 8; use constant UNKNOWN => 9; use constant LONGLINK => 'L'; use constant LABEL => 'V'; use constant BUFFER => 4096; use constant HEAD => 512; use constant BLOCK => 512; use constant COMPRESS_GZIP => 9; use constant COMPRESS_BZIP => 'bzip2'; use constant BLOCK_SIZE => sub { my $n = int($_[0]/BLOCK); $n++ if $_[0] % BLOCK; $n * BLOCK }; use constant TAR_PAD => sub { my $x = shift || return; return "\0" x (BLOCK - ($x % BLOCK) ) }; use constant TAR_END => "\0" x BLOCK; use constant READ_ONLY => sub { shift() ? 'rb' : 'r' }; use constant WRITE_ONLY => sub { $_[0] ? 'wb' . shift : 'w' }; use constant MODE_READ => sub { $_[0] =~ /^r/ ? 1 : 0 }; # Pointless assignment to make -w shut up my $getpwuid; $getpwuid = 'unknown' unless eval { my $f = getpwuid (0); }; my $getgrgid; $getgrgid = 'unknown' unless eval { my $f = getgrgid (0); }; use constant UNAME => sub { $getpwuid || scalar getpwuid( shift() ) || '' }; use constant GNAME => sub { $getgrgid || scalar getgrgid( shift() ) || '' }; use constant UID => $>; use constant GID => (split ' ', $) )[0]; use constant MODE => do { 0666 & (0777 & ~umask) }; use constant STRIP_MODE => sub { shift() & 0777 }; use constant CHECK_SUM => " "; use constant UNPACK => 'a100 a8 a8 a8 a12 a12 a8 a1 a100 A6 a2 a32 a32 a8 a8 a155 x12'; # cdrake - size must be a12 - not A12 - or else screws up huge file sizes (>8gb) use constant PACK => 'a100 a8 a8 a8 a12 a12 A8 a1 a100 a6 a2 a32 a32 a8 a8 a155 x12'; use constant NAME_LENGTH => 100; use constant PREFIX_LENGTH => 155; use constant TIME_OFFSET => ($^O eq "MacOS") ? Time::Local::timelocal(0,0,0,1,0,70) : 0; use constant MAGIC => "ustar"; use constant TAR_VERSION => "00"; use constant LONGLINK_NAME => '././@LongLink'; use constant PAX_HEADER => 'pax_global_header'; ### allow ZLIB to be turned off using ENV: DEBUG only use constant ZLIB => do { !$ENV{'PERL5_AT_NO_ZLIB'} and eval { require IO::Zlib }; $ENV{'PERL5_AT_NO_ZLIB'} || $@ ? 0 : 1 }; ### allow BZIP to be turned off using ENV: DEBUG only use constant BZIP => do { !$ENV{'PERL5_AT_NO_BZIP'} and eval { require IO::Uncompress::Bunzip2; require IO::Compress::Bzip2; }; $ENV{'PERL5_AT_NO_BZIP'} || $@ ? 0 : 1 }; use constant GZIP_MAGIC_NUM => qr/^(?:\037\213|\037\235)/; use constant BZIP_MAGIC_NUM => qr/^BZh\d/; use constant CAN_CHOWN => sub { ($> == 0 and $^O ne "MacOS" and $^O ne "MSWin32") }; use constant CAN_READLINK => ($^O ne 'MSWin32' and $^O !~ /RISC(?:[ _])?OS/i and $^O ne 'VMS'); use constant ON_UNIX => ($^O ne 'MSWin32' and $^O ne 'MacOS' and $^O ne 'VMS'); use constant ON_VMS => $^O eq 'VMS'; sub _list_consts { my $class = shift; my $pkg = shift; return unless defined $pkg; # some joker might use '0' as a pkg... my @rv; { no strict 'refs'; my $stash = $pkg . '::'; for my $name (sort keys %$stash ) { ### is it a subentry? my $sub = $pkg->can( $name ); next unless defined $sub; next unless defined prototype($sub) and not length prototype($sub); push @rv, $name; } } return sort @rv; } 1; Tar/File.pm000064400000044041152347641370006526 0ustar00package Archive::Tar::File; use strict; use Carp (); use IO::File; use File::Spec::Unix (); use File::Spec (); use File::Basename (); ### avoid circular use, so only require; require Archive::Tar; use Archive::Tar::Constant; use vars qw[@ISA $VERSION]; #@ISA = qw[Archive::Tar]; $VERSION = '2.30'; ### set value to 1 to oct() it during the unpack ### my $tmpl = [ name => 0, # string A100 mode => 1, # octal A8 uid => 1, # octal A8 gid => 1, # octal A8 size => 0, # octal # cdrake - not *always* octal.. A12 mtime => 1, # octal A12 chksum => 1, # octal A8 type => 0, # character A1 linkname => 0, # string A100 magic => 0, # string A6 version => 0, # 2 bytes A2 uname => 0, # string A32 gname => 0, # string A32 devmajor => 1, # octal A8 devminor => 1, # octal A8 prefix => 0, # A155 x 12 ### end UNPACK items ### raw => 0, # the raw data chunk data => 0, # the data associated with the file -- # This might be very memory intensive ]; ### install get/set accessors for this object. for ( my $i=0; $i[$i]; no strict 'refs'; *{__PACKAGE__."::$key"} = sub { my $self = shift; $self->{$key} = $_[0] if @_; ### just in case the key is not there or undef or something ### { local $^W = 0; return $self->{$key}; } } } =head1 NAME Archive::Tar::File - a subclass for in-memory extracted file from Archive::Tar =head1 SYNOPSIS my @items = $tar->get_files; print $_->name, ' ', $_->size, "\n" for @items; print $object->get_content; $object->replace_content('new content'); $object->rename( 'new/full/path/to/file.c' ); =head1 DESCRIPTION Archive::Tar::Files provides a neat little object layer for in-memory extracted files. It's mostly used internally in Archive::Tar to tidy up the code, but there's no reason users shouldn't use this API as well. =head2 Accessors A lot of the methods in this package are accessors to the various fields in the tar header: =over 4 =item name The file's name =item mode The file's mode =item uid The user id owning the file =item gid The group id owning the file =item size File size in bytes =item mtime Modification time. Adjusted to mac-time on MacOS if required =item chksum Checksum field for the tar header =item type File type -- numeric, but comparable to exported constants -- see Archive::Tar's documentation =item linkname If the file is a symlink, the file it's pointing to =item magic Tar magic string -- not useful for most users =item version Tar version string -- not useful for most users =item uname The user name that owns the file =item gname The group name that owns the file =item devmajor Device major number in case of a special file =item devminor Device minor number in case of a special file =item prefix Any directory to prefix to the extraction path, if any =item raw Raw tar header -- not useful for most users =back =head1 Methods =head2 Archive::Tar::File->new( file => $path ) Returns a new Archive::Tar::File object from an existing file. Returns undef on failure. =head2 Archive::Tar::File->new( data => $path, $data, $opt ) Returns a new Archive::Tar::File object from data. C<$path> defines the file name (which need not exist), C<$data> the file contents, and C<$opt> is a reference to a hash of attributes which may be used to override the default attributes (fields in the tar header), which are described above in the Accessors section. Returns undef on failure. =head2 Archive::Tar::File->new( chunk => $chunk ) Returns a new Archive::Tar::File object from a raw 512-byte tar archive chunk. Returns undef on failure. =cut sub new { my $class = shift; my $what = shift; my $obj = ($what eq 'chunk') ? __PACKAGE__->_new_from_chunk( @_ ) : ($what eq 'file' ) ? __PACKAGE__->_new_from_file( @_ ) : ($what eq 'data' ) ? __PACKAGE__->_new_from_data( @_ ) : undef; return $obj; } ### copies the data, creates a clone ### sub clone { my $self = shift; return bless { %$self }, ref $self; } sub _new_from_chunk { my $class = shift; my $chunk = shift or return; # 512 bytes of tar header my %hash = @_; ### filter any arguments on defined-ness of values. ### this allows overriding from what the tar-header is saying ### about this tar-entry. Particularly useful for @LongLink files my %args = map { $_ => $hash{$_} } grep { defined $hash{$_} } keys %hash; ### makes it start at 0 actually... :) ### my $i = -1; my %entry = map { my ($s,$v)=($tmpl->[++$i],$tmpl->[++$i]); # cdrake ($_)=($_=~/^([^\0]*)/) unless($s eq 'size'); # cdrake $s=> $v ? oct $_ : $_ # cdrake # $tmpl->[++$i] => $tmpl->[++$i] ? oct $_ : $_ # removed by cdrake - mucks up binary sizes >8gb } unpack( UNPACK, $chunk ); # cdrake # } map { /^([^\0]*)/ } unpack( UNPACK, $chunk ); # old - replaced now by cdrake if(substr($entry{'size'}, 0, 1) eq "\x80") { # binary size extension for files >8gigs (> octal 77777777777777) # cdrake my @sz=unpack("aCSNN",$entry{'size'}); $entry{'size'}=$sz[4]+(2**32)*$sz[3]+$sz[2]*(2**64); # Use the low 80 bits (should use the upper 15 as well, but as at year 2011, that seems unlikely to ever be needed - the numbers are just too big...) # cdrake } else { # cdrake ($entry{'size'})=($entry{'size'}=~/^([^\0]*)/); $entry{'size'}=oct $entry{'size'}; # cdrake } # cdrake my $obj = bless { %entry, %args }, $class; ### magic is a filetype string.. it should have something like 'ustar' or ### something similar... if the chunk is garbage, skip it return unless $obj->magic !~ /\W/; ### store the original chunk ### $obj->raw( $chunk ); $obj->type(FILE) if ( (!length $obj->type) or ($obj->type =~ /\W/) ); $obj->type(DIR) if ( ($obj->is_file) && ($obj->name =~ m|/$|) ); return $obj; } sub _new_from_file { my $class = shift; my $path = shift; ### path has to at least exist return unless defined $path; my $type = __PACKAGE__->_filetype($path); my $data = ''; READ: { unless ($type == DIR ) { my $fh = IO::File->new; unless( $fh->open($path) ) { ### dangling symlinks are fine, stop reading but continue ### creating the object last READ if $type == SYMLINK; ### otherwise, return from this function -- ### anything that's *not* a symlink should be ### resolvable return; } ### binmode needed to read files properly on win32 ### binmode $fh; $data = do { local $/; <$fh> }; close $fh; } } my @items = qw[mode uid gid size mtime]; my %hash = map { shift(@items), $_ } (lstat $path)[2,4,5,7,9]; if (ON_VMS) { ### VMS has two UID modes, traditional and POSIX. Normally POSIX is ### not used. We currently do not have an easy way to see if we are in ### POSIX mode. In traditional mode, the UID is actually the VMS UIC. ### The VMS UIC has the upper 16 bits is the GID, which in many cases ### the VMS UIC will be larger than 209715, the largest that TAR can ### handle. So for now, assume it is traditional if the UID is larger ### than 0x10000. if ($hash{uid} > 0x10000) { $hash{uid} = $hash{uid} & 0xFFFF; } ### The file length from stat() is the physical length of the file ### However the amount of data read in may be more for some file types. ### Fixed length files are read past the logical EOF to end of the block ### containing. Other file types get expanded on read because record ### delimiters are added. my $data_len = length $data; $hash{size} = $data_len if $hash{size} < $data_len; } ### you *must* set size == 0 on symlinks, or the next entry will be ### though of as the contents of the symlink, which is wrong. ### this fixes bug #7937 $hash{size} = 0 if ($type == DIR or $type == SYMLINK); $hash{mtime} -= TIME_OFFSET; ### strip the high bits off the mode, which we don't need to store $hash{mode} = STRIP_MODE->( $hash{mode} ); ### probably requires some file path munging here ... ### ### name and prefix are set later my $obj = { %hash, name => '', chksum => CHECK_SUM, type => $type, linkname => ($type == SYMLINK and CAN_READLINK) ? readlink $path : '', magic => MAGIC, version => TAR_VERSION, uname => UNAME->( $hash{uid} ), gname => GNAME->( $hash{gid} ), devmajor => 0, # not handled devminor => 0, # not handled prefix => '', data => $data, }; bless $obj, $class; ### fix up the prefix and file from the path my($prefix,$file) = $obj->_prefix_and_file( $path ); $obj->prefix( $prefix ); $obj->name( $file ); return $obj; } sub _new_from_data { my $class = shift; my $path = shift; return unless defined $path; my $data = shift; return unless defined $data; my $opt = shift; my $obj = { data => $data, name => '', mode => MODE, uid => UID, gid => GID, size => length $data, mtime => time - TIME_OFFSET, chksum => CHECK_SUM, type => FILE, linkname => '', magic => MAGIC, version => TAR_VERSION, uname => UNAME->( UID ), gname => GNAME->( GID ), devminor => 0, devmajor => 0, prefix => '', }; ### overwrite with user options, if provided ### if( $opt and ref $opt eq 'HASH' ) { for my $key ( keys %$opt ) { ### don't write bogus options ### next unless exists $obj->{$key}; $obj->{$key} = $opt->{$key}; } } bless $obj, $class; ### fix up the prefix and file from the path my($prefix,$file) = $obj->_prefix_and_file( $path ); $obj->prefix( $prefix ); $obj->name( $file ); return $obj; } sub _prefix_and_file { my $self = shift; my $path = shift; my ($vol, $dirs, $file) = File::Spec->splitpath( $path, $self->is_dir ); my @dirs = File::Spec->splitdir( File::Spec->canonpath($dirs) ); ### if it's a directory, then $file might be empty $file = pop @dirs if $self->is_dir and not length $file; ### splitting ../ gives you the relative path in native syntax map { $_ = '..' if $_ eq '-' } @dirs if ON_VMS; my $prefix = File::Spec::Unix->catdir(@dirs); return( $prefix, $file ); } sub _filetype { my $self = shift; my $file = shift; return unless defined $file; 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 ### shouldn't happen, this is when making archives, not reading ### return LONGLINK if ( $file eq LONGLINK_NAME ); return UNKNOWN; # Something else (like what?) } ### this method 'downgrades' a file to plain file -- this is used for ### symlinks when FOLLOW_SYMLINKS is true. sub _downgrade_to_plainfile { my $entry = shift; $entry->type( FILE ); $entry->mode( MODE ); $entry->linkname(''); return 1; } =head2 $bool = $file->extract( [ $alternative_name ] ) Extract this object, optionally to an alternative name. See C<< Archive::Tar->extract_file >> for details. Returns true on success and false on failure. =cut sub extract { my $self = shift; local $Carp::CarpLevel += 1; return Archive::Tar->_extract_file( $self, @_ ); } =head2 $path = $file->full_path Returns the full path from the tar header; this is basically a concatenation of the C and C fields. =cut sub full_path { my $self = shift; ### if prefix field is empty return $self->name unless defined $self->prefix and length $self->prefix; ### or otherwise, catfile'd return File::Spec::Unix->catfile( $self->prefix, $self->name ); } =head2 $bool = $file->validate Done by Archive::Tar internally when reading the tar file: validate the header against the checksum to ensure integer tar file. Returns true on success, false on failure =cut sub validate { my $self = shift; my $raw = $self->raw; ### don't know why this one is different from the one we /write/ ### substr ($raw, 148, 8) = " "; ### bug #43513: [PATCH] Accept wrong checksums from SunOS and HP-UX tar ### like GNU tar does. See here for details: ### http://www.gnu.org/software/tar/manual/tar.html#SEC139 ### so we do both a signed AND unsigned validate. if one succeeds, that's ### good enough return ( (unpack ("%16C*", $raw) == $self->chksum) or (unpack ("%16c*", $raw) == $self->chksum)) ? 1 : 0; } =head2 $bool = $file->has_content Returns a boolean to indicate whether the current object has content. Some special files like directories and so on never will have any content. This method is mainly to make sure you don't get warnings for using uninitialized values when looking at an object's content. =cut sub has_content { my $self = shift; return defined $self->data() && length $self->data() ? 1 : 0; } =head2 $content = $file->get_content Returns the current content for the in-memory file =cut sub get_content { my $self = shift; $self->data( ); } =head2 $cref = $file->get_content_by_ref Returns the current content for the in-memory file as a scalar reference. Normal users won't need this, but it will save memory if you are dealing with very large data files in your tar archive, since it will pass the contents by reference, rather than make a copy of it first. =cut sub get_content_by_ref { my $self = shift; return \$self->{data}; } =head2 $bool = $file->replace_content( $content ) Replace the current content of the file with the new content. This only affects the in-memory archive, not the on-disk version until you write it. Returns true on success, false on failure. =cut sub replace_content { my $self = shift; my $data = shift || ''; $self->data( $data ); $self->size( length $data ); return 1; } =head2 $bool = $file->rename( $new_name ) Rename the current file to $new_name. Note that you must specify a Unix path for $new_name, since per tar standard, all files in the archive must be Unix paths. Returns true on success and false on failure. =cut sub rename { my $self = shift; my $path = shift; return unless defined $path; my ($prefix,$file) = $self->_prefix_and_file( $path ); $self->name( $file ); $self->prefix( $prefix ); return 1; } =head2 $bool = $file->chmod $mode) Change mode of $file to $mode. The mode can be a string or a number which is interpreted as octal whether or not a leading 0 is given. Returns true on success and false on failure. =cut sub chmod { my $self = shift; my $mode = shift; return unless defined $mode && $mode =~ /^[0-7]{1,4}$/; $self->{mode} = oct($mode); return 1; } =head2 $bool = $file->chown( $user [, $group]) Change owner of $file to $user. If a $group is given that is changed as well. You can also pass a single parameter with a colon separating the use and group as in 'root:wheel'. Returns true on success and false on failure. =cut sub chown { my $self = shift; my $uname = shift; return unless defined $uname; my $gname; if (-1 != index($uname, ':')) { ($uname, $gname) = split(/:/, $uname); } else { $gname = shift if @_ > 0; } $self->uname( $uname ); $self->gname( $gname ) if $gname; return 1; } =head1 Convenience methods To quickly check the type of a C object, you can use the following methods: =over 4 =item $file->is_file Returns true if the file is of type C =item $file->is_dir Returns true if the file is of type C =item $file->is_hardlink Returns true if the file is of type C =item $file->is_symlink Returns true if the file is of type C =item $file->is_chardev Returns true if the file is of type C =item $file->is_blockdev Returns true if the file is of type C =item $file->is_fifo Returns true if the file is of type C =item $file->is_socket Returns true if the file is of type C =item $file->is_longlink Returns true if the file is of type C. Should not happen after a successful C. =item $file->is_label Returns true if the file is of type C