7 # Multi-level database module for storing hash trees, arrays and simple
8 # key/value pairs into FTP-able, cross-platform binary database files.
10 # Type `perldoc DBM::Deep` for complete documentation.
14 # tie %db, 'DBM::Deep', 'my_database.db'; # standard tie() method
16 # my $db = new DBM::Deep( 'my_database.db' ); # preferred OO method
18 # $db->{my_scalar} = 'hello world';
19 # $db->{my_hash} = { larry => 'genius', hashes => 'fast' };
20 # $db->{my_array} = [ 1, 2, 3, time() ];
21 # $db->{my_complex} = [ 'hello', { perl => 'rules' }, 42, 99 ];
22 # push @{$db->{my_array}}, 'another value';
23 # my @key_list = keys %{$db->{my_hash}};
24 # print "This module " . $db->{my_complex}->[1]->{perl} . "!\n";
27 # (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
28 # This program is free software; you can redistribute it and/or
29 # modify it under the same terms as Perl itself.
37 our $VERSION = q(0.99_03);
39 use Fcntl qw( :DEFAULT :flock :seek );
41 use Clone::Any '_clone_data';
43 use FileHandle::Fmode ();
46 use DBM::Deep::Engine2;
50 # Setup constants for users to pass to new()
52 sub TYPE_HASH () { DBM::Deep::Engine2->SIG_HASH }
53 sub TYPE_ARRAY () { DBM::Deep::Engine2->SIG_ARRAY }
61 $proto->_throw_error( "Odd number of parameters to " . (caller(1))[2] );
66 unless ( eval { local $SIG{'__DIE__'}; %{$_[0]} || 1 } ) {
67 $proto->_throw_error( "Not a hashref in args to " . (caller(1))[2] );
72 $args = { file => shift };
80 # Class constructor method for Perl OO interface.
81 # Calls tie() and returns blessed reference to tied hash or array,
82 # providing a hybrid OO/tie interface.
85 my $args = $class->_get_args( @_ );
88 # Check if we want a tied hash or array.
91 if (defined($args->{type}) && $args->{type} eq TYPE_ARRAY) {
92 $class = 'DBM::Deep::Array';
93 require DBM::Deep::Array;
94 tie @$self, $class, %$args;
97 $class = 'DBM::Deep::Hash';
98 require DBM::Deep::Hash;
99 tie %$self, $class, %$args;
102 return bless $self, $class;
105 # This initializer is called from the various TIE* methods. new() calls tie(),
106 # which allows for a single point of entry.
111 $args->{storage} = DBM::Deep::File->new( $args )
112 unless exists $args->{storage};
114 # locking implicitly enables autoflush
115 if ($args->{locking}) { $args->{autoflush} = 1; }
117 # These are the defaults to be optionally overridden below
120 base_offset => undef,
127 $self->{engine} = DBM::Deep::Engine2->new( { %{$args}, obj => $self } );
129 # Grab the parameters we want to use
130 foreach my $param ( keys %$self ) {
131 next unless exists $args->{$param};
132 $self->{$param} = $args->{$param};
135 $self->_engine->setup_fh( $self );
137 $self->_storage->set_db( $self );
144 require DBM::Deep::Hash;
145 return DBM::Deep::Hash->TIEHASH( @_ );
150 require DBM::Deep::Array;
151 return DBM::Deep::Array->TIEARRAY( @_ );
155 my $self = shift->_get_self;
156 return $self->_storage->lock( $self, @_ );
160 my $self = shift->_get_self;
161 return $self->_storage->unlock( $self, @_ );
165 my $self = shift->_get_self;
166 my ($spot, $value) = @_;
171 elsif ( eval { local $SIG{__DIE__}; $value->isa( 'DBM::Deep' ) } ) {
172 ${$spot} = $value->_repr;
173 $value->_copy_node( ${$spot} );
176 my $r = Scalar::Util::reftype( $value );
177 my $c = Scalar::Util::blessed( $value );
178 if ( $r eq 'ARRAY' ) {
179 ${$spot} = [ @{$value} ];
182 ${$spot} = { %{$value} };
184 ${$spot} = bless ${$spot}, $c
192 die "Must be implemented in a child class\n";
196 die "Must be implemented in a child class\n";
201 # Recursively export into standard Perl hashes and arrays.
203 my $self = shift->_get_self;
205 my $temp = $self->_repr;
208 $self->_copy_node( $temp );
211 # This will always work because $self, after _get_self() is a HASH
212 if ( $self->{parent} ) {
213 my $c = Scalar::Util::blessed(
214 $self->{parent}->get($self->{parent_key})
216 if ( $c && !$c->isa( 'DBM::Deep' ) ) {
226 # Recursively import Perl hash/array structure
228 if (!ref($_[0])) { return; } # Perl calls import() on use -- ignore
230 my $self = shift->_get_self;
233 # struct is not a reference, so just import based on our type
235 $struct = $self->_repr( @_ );
238 #XXX This isn't the best solution. Better would be to use Data::Walker,
239 #XXX but that's a lot more thinking than I want to do right now.
242 $self->_import( _clone_data( $struct ) );
252 #XXX Need to keep track of who has a fh to this file in order to
253 #XXX close them all prior to optimize on Win32/cygwin
256 # Rebuild entire database into new file, then move
257 # it back on top of original.
259 my $self = shift->_get_self;
261 #XXX Need to create a new test for this
262 # if ($self->_storage->{links} > 1) {
263 # $self->_throw_error("Cannot optimize: reference count is greater than 1");
266 #XXX Do we have to lock the tempfile?
268 my $db_temp = DBM::Deep->new(
269 file => $self->_storage->{file} . '.tmp',
274 $self->_copy_node( $db_temp );
278 # Attempt to copy user, group and permissions over to new file
280 my @stats = stat($self->_fh);
281 my $perms = $stats[2] & 07777;
284 chown( $uid, $gid, $self->_storage->{file} . '.tmp' );
285 chmod( $perms, $self->_storage->{file} . '.tmp' );
287 # q.v. perlport for more information on this variable
288 if ( $^O eq 'MSWin32' || $^O eq 'cygwin' ) {
290 # Potential race condition when optmizing on Win32 with locking.
291 # The Windows filesystem requires that the filehandle be closed
292 # before it is overwritten with rename(). This could be redone
296 $self->_storage->close;
299 if (!rename $self->_storage->{file} . '.tmp', $self->_storage->{file}) {
300 unlink $self->_storage->{file} . '.tmp';
302 $self->_throw_error("Optimize failed: Cannot copy temp file over original: $!");
306 $self->_storage->close;
307 $self->_storage->open;
308 $self->_engine->setup_fh( $self );
315 # Make copy of object and return
317 my $self = shift->_get_self;
319 return DBM::Deep->new(
320 type => $self->_type,
321 base_offset => $self->_base_offset,
322 storage => $self->_storage,
323 parent => $self->{parent},
324 parent_key => $self->{parent_key},
329 my %is_legal_filter = map {
332 store_key store_value
333 fetch_key fetch_value
338 # Setup filter function for storing or fetching the key or value
340 my $self = shift->_get_self;
344 if ( $is_legal_filter{$type} ) {
345 $self->_storage->{"filter_$type"} = $func;
354 my $self = shift->_get_self;
355 return $self->_storage->begin_transaction;
359 my $self = shift->_get_self;
360 return $self->_storage->end_transaction;
364 my $self = shift->_get_self;
365 return $self->_storage->commit_transaction;
373 my $self = $_[0]->_get_self;
374 return $self->{engine};
378 my $self = $_[0]->_get_self;
379 return $self->{storage};
383 my $self = $_[0]->_get_self;
384 return $self->{type};
388 my $self = $_[0]->_get_self;
389 return $self->{base_offset};
393 my $self = $_[0]->_get_self;
394 return $self->_storage->{fh};
402 die "DBM::Deep: $_[1]\n";
409 #XXX This if() is redundant
410 if ( my $parent = $self->{parent} ) {
412 while ( $parent->{parent} ) {
414 $parent->_type eq TYPE_HASH
415 ? "\{q{$child->{parent_key}}\}"
416 : "\[$child->{parent_key}\]"
420 $parent = $parent->{parent};
424 $base = "\$db->get( q{$child->{parent_key}} )->" . $base;
427 $base = "\$db->get( q{$child->{parent_key}} )";
435 # Store single hash key/value or array element in database.
437 my $self = shift->_get_self;
438 my ($key, $value, $orig_key) = @_;
439 $orig_key = $key unless defined $orig_key;
441 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
442 $self->_throw_error( 'Cannot write to a readonly filehandle' );
445 #XXX The second condition needs to disappear
446 if ( !( $self->_type eq TYPE_ARRAY && $orig_key eq 'length') ) {
449 my $r = Scalar::Util::reftype( $value ) || '';
450 if ( $r eq 'HASH' ) {
453 elsif ( $r eq 'ARRAY' ) {
456 elsif ( defined $value ) {
463 if ( my $c = Scalar::Util::blessed( $value ) ) {
464 $rhs = "bless $rhs, '$c'";
467 my $lhs = $self->_find_parent;
469 if ( $self->_type eq TYPE_HASH ) {
470 $lhs .= "->\{q{$orig_key}\}";
473 $lhs .= "->\[$orig_key\]";
479 $lhs = "\$db->put(q{$orig_key},$rhs);";
482 $self->_storage->audit($lhs);
486 # Request exclusive lock for writing
488 $self->lock( LOCK_EX );
490 # User may be storing a complex value, in which case we do not want it run
491 # through the filtering system.
492 if ( !ref($value) && $self->_storage->{filter_store_value} ) {
493 $value = $self->_storage->{filter_store_value}->( $value );
496 $self->_engine->write_value( $self->_storage->transaction_id, $self->_base_offset, $key, $value, $orig_key );
505 # Fetch single value or element given plain key or array index
507 my $self = shift->_get_self;
508 my ($key, $orig_key) = @_;
509 $orig_key = $key unless defined $orig_key;
512 # Request shared lock for reading
514 $self->lock( LOCK_SH );
516 my $result = $self->_engine->read_value( $self->_storage->transaction_id, $self->_base_offset, $key, $orig_key );
520 # Filters only apply to scalar values, so the ref check is making
521 # sure the fetched bucket is a scalar, not a child hash or array.
522 return ($result && !ref($result) && $self->_storage->{filter_fetch_value})
523 ? $self->_storage->{filter_fetch_value}->($result)
529 # Delete single key/value pair or element given plain key or array index
531 my $self = shift->_get_self;
532 my ($key, $orig_key) = @_;
533 $orig_key = $key unless defined $orig_key;
535 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
536 $self->_throw_error( 'Cannot write to a readonly filehandle' );
539 if ( defined $orig_key ) {
540 my $lhs = $self->_find_parent;
542 $self->_storage->audit( "delete $lhs;" );
545 $self->_storage->audit( "\$db->delete('$orig_key');" );
550 # Request exclusive lock for writing
552 $self->lock( LOCK_EX );
557 my $value = $self->_engine->delete_key( $self->_storage->transaction_id, $self->_base_offset, $key, $orig_key );
559 if (defined $value && !ref($value) && $self->_storage->{filter_fetch_value}) {
560 $value = $self->_storage->{filter_fetch_value}->($value);
570 # Check if a single key or element exists given plain key or array index
572 my $self = shift->_get_self;
576 # Request shared lock for reading
578 $self->lock( LOCK_SH );
580 my $result = $self->_engine->key_exists( $self->_storage->transaction_id, $self->_base_offset, $key );
589 # Clear all keys from hash, or all elements from array.
591 my $self = shift->_get_self;
593 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
594 $self->_throw_error( 'Cannot write to a readonly filehandle' );
598 my $lhs = $self->_find_parent;
600 if ( $self->_type eq TYPE_HASH ) {
601 $lhs = '%{' . $lhs . '}';
604 $lhs = '@{' . $lhs . '}';
607 $self->_storage->audit( "$lhs = ();" );
611 # Request exclusive lock for writing
613 $self->lock( LOCK_EX );
615 if ( $self->_type eq TYPE_HASH ) {
616 my $key = $self->first_key;
618 # Retrieve the key before deleting because we depend on next_key
619 my $next_key = $self->next_key( $key );
620 $self->_engine->delete_key( $self->_storage->transaction_id, $self->_base_offset, $key, $key );
625 my $size = $self->FETCHSIZE;
626 for my $key ( 0 .. $size - 1 ) {
627 $self->_engine->delete_key( $self->_storage->transaction_id, $self->_base_offset, $key, $key );
629 $self->STORESIZE( 0 );
631 #XXX This needs updating to use _release_space
632 # $self->_engine->write_tag(
633 # $self->_base_offset, $self->_type,
634 # chr(0)x$self->_engine->{index_size},
643 # Public method aliases
645 sub put { (shift)->STORE( @_ ) }
646 sub store { (shift)->STORE( @_ ) }
647 sub get { (shift)->FETCH( @_ ) }
648 sub fetch { (shift)->FETCH( @_ ) }
649 sub delete { (shift)->DELETE( @_ ) }
650 sub exists { (shift)->EXISTS( @_ ) }
651 sub clear { (shift)->CLEAR( @_ ) }
658 DBM::Deep - A pure perl multi-level hash/array DBM
663 my $db = DBM::Deep->new( "foo.db" );
665 $db->{key} = 'value';
668 $db->put('key' => 'value');
669 print $db->get('key');
671 # true multi-level support
672 $db->{my_complex} = [
673 'hello', { perl => 'rules' },
677 tie my %db, 'DBM::Deep', 'foo.db';
681 tied(%db)->put('key' => 'value');
682 print tied(%db)->get('key');
686 A unique flat-file database module, written in pure perl. True multi-level
687 hash/array support (unlike MLDBM, which is faked), hybrid OO / tie()
688 interface, cross-platform FTPable files, ACID transactions, and is quite fast.
689 Can handle millions of keys and unlimited levels without significant
690 slow-down. Written from the ground-up in pure perl -- this is NOT a wrapper
691 around a C-based DBM. Out-of-the-box compatibility with Unix, Mac OS X and
694 =head1 VERSION DIFFERENCES
696 B<NOTE>: 0.99_01 and above have significant file format differences from 0.983 and
697 before. There will be a backwards-compatibility layer in 1.00, but that is
698 slated for a later 0.99_x release. This version is B<NOT> backwards compatible
699 with 0.983 and before.
703 Construction can be done OO-style (which is the recommended way), or using
704 Perl's tie() function. Both are examined here.
706 =head2 OO CONSTRUCTION
708 The recommended way to construct a DBM::Deep object is to use the new()
709 method, which gets you a blessed I<and> tied hash (or array) reference.
711 my $db = DBM::Deep->new( "foo.db" );
713 This opens a new database handle, mapped to the file "foo.db". If this
714 file does not exist, it will automatically be created. DB files are
715 opened in "r+" (read/write) mode, and the type of object returned is a
716 hash, unless otherwise specified (see L<OPTIONS> below).
718 You can pass a number of options to the constructor to specify things like
719 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
721 my $db = DBM::Deep->new(
727 Notice that the filename is now specified I<inside> the hash with
728 the "file" parameter, as opposed to being the sole argument to the
729 constructor. This is required if any options are specified.
730 See L<OPTIONS> below for the complete list.
732 You can also start with an array instead of a hash. For this, you must
733 specify the C<type> parameter:
735 my $db = DBM::Deep->new(
737 type => DBM::Deep->TYPE_ARRAY
740 B<Note:> Specifing the C<type> parameter only takes effect when beginning
741 a new DB file. If you create a DBM::Deep object with an existing file, the
742 C<type> will be loaded from the file header, and an error will be thrown if
743 the wrong type is passed in.
745 =head2 TIE CONSTRUCTION
747 Alternately, you can create a DBM::Deep handle by using Perl's built-in
748 tie() function. The object returned from tie() can be used to call methods,
749 such as lock() and unlock(). (That object can be retrieved from the tied
750 variable at any time using tied() - please see L<perltie/> for more info.
753 my $db = tie %hash, "DBM::Deep", "foo.db";
756 my $db = tie @array, "DBM::Deep", "bar.db";
758 As with the OO constructor, you can replace the DB filename parameter with
759 a hash containing one or more options (see L<OPTIONS> just below for the
762 tie %hash, "DBM::Deep", {
770 There are a number of options that can be passed in when constructing your
771 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
777 Filename of the DB file to link the handle to. You can pass a full absolute
778 filesystem path, partial path, or a plain filename if the file is in the
779 current working directory. This is a required parameter (though q.v. fh).
783 If you want, you can pass in the fh instead of the file. This is most useful for doing
786 my $db = DBM::Deep->new( { fh => \*DATA } );
788 You are responsible for making sure that the fh has been opened appropriately for your
789 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
790 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
791 needs to read from the fh.
793 =item * audit_file / audit_fh
795 These are just like file/fh, except for auditing. Please see L</AUDITING> for
800 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
801 not need to set this. However, it's there if you want it.
803 If you pass in fh and do not set this, it will be set appropriately.
807 This parameter specifies what type of object to create, a hash or array. Use
808 one of these two constants:
812 =item * C<DBM::Deep-E<gt>TYPE_HASH>
814 =item * C<DBM::Deep-E<gt>TYPE_ARRAY>.
818 This only takes effect when beginning a new file. This is an optional
819 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
823 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
824 function to lock the database in exclusive mode for writes, and shared mode
825 for reads. Pass any true value to enable. This affects the base DB handle
826 I<and any child hashes or arrays> that use the same DB file. This is an
827 optional parameter, and defaults to 0 (disabled). See L<LOCKING> below for
832 Specifies whether autoflush is to be enabled on the underlying filehandle.
833 This obviously slows down write operations, but is required if you may have
834 multiple processes accessing the same DB file (also consider enable I<locking>).
835 Pass any true value to enable. This is an optional parameter, and defaults to 0
840 If I<autobless> mode is enabled, DBM::Deep will preserve the class something
841 is blessed into, and restores it when fetched. This is an optional parameter, and defaults to 1 (enabled).
843 B<Note:> If you use the OO-interface, you will not be able to call any methods
844 of DBM::Deep on the blessed item. This is considered to be a feature.
848 See L</FILTERS> below.
854 With DBM::Deep you can access your databases using Perl's standard hash/array
855 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
856 treat them as such. DBM::Deep will intercept all reads/writes and direct them
857 to the right place -- the DB file. This has nothing to do with the
858 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
859 using regular hashes and arrays, rather than calling functions like C<get()>
860 and C<put()> (although those work too). It is entirely up to you how to want
861 to access your databases.
865 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
866 or even nested hashes (or arrays) using standard Perl syntax:
868 my $db = DBM::Deep->new( "foo.db" );
870 $db->{mykey} = "myvalue";
872 $db->{myhash}->{subkey} = "subvalue";
874 print $db->{myhash}->{subkey} . "\n";
876 You can even step through hash keys using the normal Perl C<keys()> function:
878 foreach my $key (keys %$db) {
879 print "$key: " . $db->{$key} . "\n";
882 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
883 pushes them onto an array, all before the loop even begins. If you have an
884 extremely large hash, this may exhaust Perl's memory. Instead, consider using
885 Perl's C<each()> function, which pulls keys/values one at a time, using very
888 while (my ($key, $value) = each %$db) {
889 print "$key: $value\n";
892 Please note that when using C<each()>, you should always pass a direct
893 hash reference, not a lookup. Meaning, you should B<never> do this:
896 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
898 This causes an infinite loop, because for each iteration, Perl is calling
899 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
900 it effectively keeps returning the first key over and over again. Instead,
901 assign a temporary variable to C<$db->{foo}>, then pass that to each().
905 As with hashes, you can treat any DBM::Deep object like a normal Perl array
906 reference. This includes inserting, removing and manipulating elements,
907 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
908 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
909 or simply be a nested array reference inside a hash. Example:
911 my $db = DBM::Deep->new(
912 file => "foo-array.db",
913 type => DBM::Deep->TYPE_ARRAY
917 push @$db, "bar", "baz";
920 my $last_elem = pop @$db; # baz
921 my $first_elem = shift @$db; # bah
922 my $second_elem = $db->[1]; # bar
924 my $num_elements = scalar @$db;
928 In addition to the I<tie()> interface, you can also use a standard OO interface
929 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
930 array) has its own methods, but both types share the following common methods:
931 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
932 C<store(> are aliases to C<put()> and C<get()>, respectively.
936 =item * new() / clone()
938 These are the constructor and copy-functions.
940 =item * put() / store()
942 Stores a new hash key/value pair, or sets an array element value. Takes two
943 arguments, the hash key or array index, and the new value. The value can be
944 a scalar, hash ref or array ref. Returns true on success, false on failure.
946 $db->put("foo", "bar"); # for hashes
947 $db->put(1, "bar"); # for arrays
949 =item * get() / fetch()
951 Fetches the value of a hash key or array element. Takes one argument: the hash
952 key or array index. Returns a scalar, hash ref or array ref, depending on the
955 my $value = $db->get("foo"); # for hashes
956 my $value = $db->get(1); # for arrays
960 Checks if a hash key or array index exists. Takes one argument: the hash key
961 or array index. Returns true if it exists, false if not.
963 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
964 if ($db->exists(1)) { print "yay!\n"; } # for arrays
968 Deletes one hash key/value pair or array element. Takes one argument: the hash
969 key or array index. Returns true on success, false if not found. For arrays,
970 the remaining elements located after the deleted element are NOT moved over.
971 The deleted element is essentially just undefined, which is exactly how Perl's
972 internal arrays work. Please note that the space occupied by the deleted
973 key/value or element is B<not> reused again -- see L<UNUSED SPACE RECOVERY>
974 below for details and workarounds.
976 $db->delete("foo"); # for hashes
977 $db->delete(1); # for arrays
981 Deletes B<all> hash keys or array elements. Takes no arguments. No return
982 value. Please note that the space occupied by the deleted keys/values or
983 elements is B<not> reused again -- see L<UNUSED SPACE RECOVERY> below for
984 details and workarounds.
986 $db->clear(); # hashes or arrays
988 =item * lock() / unlock()
994 Recover lost disk space. This is important to do, especially if you use
997 =item * import() / export()
999 Data going in and out.
1005 For hashes, DBM::Deep supports all the common methods described above, and the
1006 following additional methods: C<first_key()> and C<next_key()>.
1012 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
1013 fetched in an undefined order (which appears random). Takes no arguments,
1014 returns the key as a scalar value.
1016 my $key = $db->first_key();
1020 Returns the "next" key in the hash, given the previous one as the sole argument.
1021 Returns undef if there are no more keys to be fetched.
1023 $key = $db->next_key($key);
1027 Here are some examples of using hashes:
1029 my $db = DBM::Deep->new( "foo.db" );
1031 $db->put("foo", "bar");
1032 print "foo: " . $db->get("foo") . "\n";
1034 $db->put("baz", {}); # new child hash ref
1035 $db->get("baz")->put("buz", "biz");
1036 print "buz: " . $db->get("baz")->get("buz") . "\n";
1038 my $key = $db->first_key();
1040 print "$key: " . $db->get($key) . "\n";
1041 $key = $db->next_key($key);
1044 if ($db->exists("foo")) { $db->delete("foo"); }
1048 For arrays, DBM::Deep supports all the common methods described above, and the
1049 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
1050 C<unshift()> and C<splice()>.
1056 Returns the number of elements in the array. Takes no arguments.
1058 my $len = $db->length();
1062 Adds one or more elements onto the end of the array. Accepts scalars, hash
1063 refs or array refs. No return value.
1065 $db->push("foo", "bar", {});
1069 Fetches the last element in the array, and deletes it. Takes no arguments.
1070 Returns undef if array is empty. Returns the element value.
1072 my $elem = $db->pop();
1076 Fetches the first element in the array, deletes it, then shifts all the
1077 remaining elements over to take up the space. Returns the element value. This
1078 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
1081 my $elem = $db->shift();
1085 Inserts one or more elements onto the beginning of the array, shifting all
1086 existing elements over to make room. Accepts scalars, hash refs or array refs.
1087 No return value. This method is not recommended with large arrays -- see
1088 <LARGE ARRAYS> below for details.
1090 $db->unshift("foo", "bar", {});
1094 Performs exactly like Perl's built-in function of the same name. See L<perldoc
1095 -f splice> for usage -- it is too complicated to document here. This method is
1096 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
1100 Here are some examples of using arrays:
1102 my $db = DBM::Deep->new(
1104 type => DBM::Deep->TYPE_ARRAY
1107 $db->push("bar", "baz");
1108 $db->unshift("foo");
1111 my $len = $db->length();
1112 print "length: $len\n"; # 4
1114 for (my $k=0; $k<$len; $k++) {
1115 print "$k: " . $db->get($k) . "\n";
1118 $db->splice(1, 2, "biz", "baf");
1120 while (my $elem = shift @$db) {
1121 print "shifted: $elem\n";
1126 Enable automatic file locking by passing a true value to the C<locking>
1127 parameter when constructing your DBM::Deep object (see L<SETUP> above).
1129 my $db = DBM::Deep->new(
1134 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
1135 mode for writes, and shared mode for reads. This is required if you have
1136 multiple processes accessing the same database file, to avoid file corruption.
1137 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
1138 NFS> below for more.
1140 =head2 EXPLICIT LOCKING
1142 You can explicitly lock a database, so it remains locked for multiple
1143 transactions. This is done by calling the C<lock()> method, and passing an
1144 optional lock mode argument (defaults to exclusive mode). This is particularly
1145 useful for things like counters, where the current value needs to be fetched,
1146 then incremented, then stored again.
1149 my $counter = $db->get("counter");
1151 $db->put("counter", $counter);
1160 You can pass C<lock()> an optional argument, which specifies which mode to use
1161 (exclusive or shared). Use one of these two constants:
1162 C<DBM::Deep-E<gt>LOCK_EX> or C<DBM::Deep-E<gt>LOCK_SH>. These are passed
1163 directly to C<flock()>, and are the same as the constants defined in Perl's
1166 $db->lock( $db->LOCK_SH );
1170 =head1 IMPORTING/EXPORTING
1172 You can import existing complex structures by calling the C<import()> method,
1173 and export an entire database into an in-memory structure using the C<export()>
1174 method. Both are examined here.
1178 Say you have an existing hash with nested hashes/arrays inside it. Instead of
1179 walking the structure and adding keys/elements to the database as you go,
1180 simply pass a reference to the C<import()> method. This recursively adds
1181 everything to an existing DBM::Deep object for you. Here is an example:
1186 array1 => [ "elem0", "elem1", "elem2" ],
1188 subkey1 => "subvalue1",
1189 subkey2 => "subvalue2"
1193 my $db = DBM::Deep->new( "foo.db" );
1194 $db->import( $struct );
1196 print $db->{key1} . "\n"; # prints "value1"
1198 This recursively imports the entire C<$struct> object into C<$db>, including
1199 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
1200 keys are merged with the existing ones, replacing if they already exist.
1201 The C<import()> method can be called on any database level (not just the base
1202 level), and works with both hash and array DB types.
1204 B<Note:> Make sure your existing structure has no circular references in it.
1205 These will cause an infinite loop when importing. There are plans to fix this
1210 Calling the C<export()> method on an existing DBM::Deep object will return
1211 a reference to a new in-memory copy of the database. The export is done
1212 recursively, so all nested hashes/arrays are all exported to standard Perl
1213 objects. Here is an example:
1215 my $db = DBM::Deep->new( "foo.db" );
1217 $db->{key1} = "value1";
1218 $db->{key2} = "value2";
1220 $db->{hash1}->{subkey1} = "subvalue1";
1221 $db->{hash1}->{subkey2} = "subvalue2";
1223 my $struct = $db->export();
1225 print $struct->{key1} . "\n"; # prints "value1"
1227 This makes a complete copy of the database in memory, and returns a reference
1228 to it. The C<export()> method can be called on any database level (not just
1229 the base level), and works with both hash and array DB types. Be careful of
1230 large databases -- you can store a lot more data in a DBM::Deep object than an
1231 in-memory Perl structure.
1233 B<Note:> Make sure your database has no circular references in it.
1234 These will cause an infinite loop when exporting. There are plans to fix this
1239 DBM::Deep has a number of hooks where you can specify your own Perl function
1240 to perform filtering on incoming or outgoing data. This is a perfect
1241 way to extend the engine, and implement things like real-time compression or
1242 encryption. Filtering applies to the base DB level, and all child hashes /
1243 arrays. Filter hooks can be specified when your DBM::Deep object is first
1244 constructed, or by calling the C<set_filter()> method at any time. There are
1245 four available filter hooks, described below:
1249 =item * filter_store_key
1251 This filter is called whenever a hash key is stored. It
1252 is passed the incoming key, and expected to return a transformed key.
1254 =item * filter_store_value
1256 This filter is called whenever a hash key or array element is stored. It
1257 is passed the incoming value, and expected to return a transformed value.
1259 =item * filter_fetch_key
1261 This filter is called whenever a hash key is fetched (i.e. via
1262 C<first_key()> or C<next_key()>). It is passed the transformed key,
1263 and expected to return the plain key.
1265 =item * filter_fetch_value
1267 This filter is called whenever a hash key or array element is fetched.
1268 It is passed the transformed value, and expected to return the plain value.
1272 Here are the two ways to setup a filter hook:
1274 my $db = DBM::Deep->new(
1276 filter_store_value => \&my_filter_store,
1277 filter_fetch_value => \&my_filter_fetch
1282 $db->set_filter( "filter_store_value", \&my_filter_store );
1283 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
1285 Your filter function will be called only when dealing with SCALAR keys or
1286 values. When nested hashes and arrays are being stored/fetched, filtering
1287 is bypassed. Filters are called as static functions, passed a single SCALAR
1288 argument, and expected to return a single SCALAR value. If you want to
1289 remove a filter, set the function reference to C<undef>:
1291 $db->set_filter( "filter_store_value", undef );
1293 =head2 REAL-TIME ENCRYPTION EXAMPLE
1295 Here is a working example that uses the I<Crypt::Blowfish> module to
1296 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
1297 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
1298 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
1301 use Crypt::Blowfish;
1304 my $cipher = Crypt::CBC->new({
1305 'key' => 'my secret key',
1306 'cipher' => 'Blowfish',
1308 'regenerate_key' => 0,
1309 'padding' => 'space',
1313 my $db = DBM::Deep->new(
1314 file => "foo-encrypt.db",
1315 filter_store_key => \&my_encrypt,
1316 filter_store_value => \&my_encrypt,
1317 filter_fetch_key => \&my_decrypt,
1318 filter_fetch_value => \&my_decrypt,
1321 $db->{key1} = "value1";
1322 $db->{key2} = "value2";
1323 print "key1: " . $db->{key1} . "\n";
1324 print "key2: " . $db->{key2} . "\n";
1330 return $cipher->encrypt( $_[0] );
1333 return $cipher->decrypt( $_[0] );
1336 =head2 REAL-TIME COMPRESSION EXAMPLE
1338 Here is a working example that uses the I<Compress::Zlib> module to do real-time
1339 compression / decompression of keys & values with DBM::Deep Filters.
1340 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
1341 more on I<Compress::Zlib>.
1346 my $db = DBM::Deep->new(
1347 file => "foo-compress.db",
1348 filter_store_key => \&my_compress,
1349 filter_store_value => \&my_compress,
1350 filter_fetch_key => \&my_decompress,
1351 filter_fetch_value => \&my_decompress,
1354 $db->{key1} = "value1";
1355 $db->{key2} = "value2";
1356 print "key1: " . $db->{key1} . "\n";
1357 print "key2: " . $db->{key2} . "\n";
1363 return Compress::Zlib::memGzip( $_[0] ) ;
1366 return Compress::Zlib::memGunzip( $_[0] ) ;
1369 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
1370 actually numerical index numbers, and are not filtered.
1372 =head1 ERROR HANDLING
1374 Most DBM::Deep methods return a true value for success, and call die() on
1375 failure. You can wrap calls in an eval block to catch the die.
1377 my $db = DBM::Deep->new( "foo.db" ); # create hash
1378 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
1380 print $@; # prints error message
1382 =head1 LARGEFILE SUPPORT
1384 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
1385 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
1386 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
1387 by specifying the 'pack_size' parameter when constructing the file.
1390 filename => $filename,
1391 pack_size => 'large',
1394 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
1395 instead of 32-bit longs. After setting these values your DB files have a
1396 theoretical maximum size of 16 XB (exabytes).
1398 You can also use C<pack_size =E<gt> 'small'> in order to use 16-bit file
1401 B<Note:> Changing these values will B<NOT> work for existing database files.
1402 Only change this for new files. Once the value has been set, it is stored in
1403 the file's header and cannot be changed for the life of the file. These
1404 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
1407 B<Note:> We have not personally tested files larger than 2 GB -- all my
1408 systems have only a 32-bit Perl. However, I have received user reports that
1409 this does indeed work!
1411 =head1 LOW-LEVEL ACCESS
1413 If you require low-level access to the underlying filehandle that DBM::Deep uses,
1414 you can call the C<_fh()> method, which returns the handle:
1416 my $fh = $db->_fh();
1418 This method can be called on the root level of the datbase, or any child
1419 hashes or arrays. All levels share a I<root> structure, which contains things
1420 like the filehandle, a reference counter, and all the options specified
1421 when you created the object. You can get access to this file object by
1422 calling the C<_storage()> method.
1424 my $file_obj = $db->_storage();
1426 This is useful for changing options after the object has already been created,
1427 such as enabling/disabling locking. You can also store your own temporary user
1428 data in this structure (be wary of name collision), which is then accessible from
1429 any child hash or array.
1431 =head1 CUSTOM DIGEST ALGORITHM
1433 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
1434 keys. However you can override this, and use another algorithm (such as SHA-256)
1435 or even write your own. But please note that DBM::Deep currently expects zero
1436 collisions, so your algorithm has to be I<perfect>, so to speak. Collision
1437 detection may be introduced in a later version.
1439 You can specify a custom digest algorithm by passing it into the parameter
1440 list for new(), passing a reference to a subroutine as the 'digest' parameter,
1441 and the length of the algorithm's hashes (in bytes) as the 'hash_size'
1442 parameter. Here is a working example that uses a 256-bit hash from the
1443 I<Digest::SHA256> module. Please see
1444 L<http://search.cpan.org/search?module=Digest::SHA256> for more information.
1449 my $context = Digest::SHA256::new(256);
1451 my $db = DBM::Deep->new(
1452 filename => "foo-sha.db",
1453 digest => \&my_digest,
1457 $db->{key1} = "value1";
1458 $db->{key2} = "value2";
1459 print "key1: " . $db->{key1} . "\n";
1460 print "key2: " . $db->{key2} . "\n";
1466 return substr( $context->hash($_[0]), 0, 32 );
1469 B<Note:> Your returned digest strings must be B<EXACTLY> the number
1470 of bytes you specify in the hash_size parameter (in this case 32).
1472 B<Note:> If you do choose to use a custom digest algorithm, you must set it
1473 every time you access this file. Otherwise, the default (MD5) will be used.
1475 =head1 CIRCULAR REFERENCES
1477 DBM::Deep has B<experimental> support for circular references. Meaning you
1478 can have a nested hash key or array element that points to a parent object.
1479 This relationship is stored in the DB file, and is preserved between sessions.
1482 my $db = DBM::Deep->new( "foo.db" );
1485 $db->{circle} = $db; # ref to self
1487 print $db->{foo} . "\n"; # prints "bar"
1488 print $db->{circle}->{foo} . "\n"; # prints "bar" again
1490 B<Note>: Passing the object to a function that recursively walks the
1491 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
1492 C<export()> methods) will result in an infinite loop. This will be fixed in
1497 New in 0.99_01 is the ability to audit your databases actions. By passing in
1498 audit_file (or audit_fh) to the constructor, all actions will be logged to
1499 that file. The format is one that is suitable for eval'ing against the
1500 database to replay the actions. Please see t/33_audit_trail.t for an example
1505 New in 0.99_01 is ACID transactions. Every DBM::Deep object is completely
1506 transaction-ready - it is not an option you have to turn on. Three new methods
1507 have been added to support them. They are:
1511 =item * begin_work()
1513 This starts a transaction.
1517 This applies the changes done within the transaction to the mainline and ends
1522 This discards the changes done within the transaction to the mainline and ends
1527 Transactions in DBM::Deep are done using the MVCC method, the same method used
1528 by the InnoDB MySQL table type.
1530 =head1 CAVEATS / ISSUES / BUGS
1532 This section describes all the known issues with DBM::Deep. It you have found
1533 something that is not listed here, please send e-mail to L<jhuckaby@cpan.org>.
1535 =head2 UNUSED SPACE RECOVERY
1537 One major caveat with DBM::Deep is that space occupied by existing keys and
1538 values is not recovered when they are deleted. Meaning if you keep deleting
1539 and adding new keys, your file will continuously grow. I am working on this,
1540 but in the meantime you can call the built-in C<optimize()> method from time to
1541 time (perhaps in a crontab or something) to recover all your unused space.
1543 $db->optimize(); # returns true on success
1545 This rebuilds the ENTIRE database into a new file, then moves it on top of
1546 the original. The new file will have no unused space, thus it will take up as
1547 little disk space as possible. Please note that this operation can take
1548 a long time for large files, and you need enough disk space to temporarily hold
1549 2 copies of your DB file. The temporary file is created in the same directory
1550 as the original, named with a ".tmp" extension, and is deleted when the
1551 operation completes. Oh, and if locking is enabled, the DB is automatically
1552 locked for the entire duration of the copy.
1554 B<WARNING:> Only call optimize() on the top-level node of the database, and
1555 make sure there are no child references lying around. DBM::Deep keeps a reference
1556 counter, and if it is greater than 1, optimize() will abort and return undef.
1560 (The reasons given assume a high level of Perl understanding, specifically of
1561 references. You can safely skip this section.)
1563 Currently, the only references supported are HASH and ARRAY. The other reference
1564 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
1570 These are things like filehandles and other sockets. They can't be supported
1571 because it's completely unclear how DBM::Deep should serialize them.
1573 =item * SCALAR / REF
1575 The discussion here refers to the following type of example:
1582 # In some other process ...
1584 my $val = ${ $db->{key1} };
1586 is( $val, 50, "What actually gets stored in the DB file?" );
1588 The problem is one of synchronization. When the variable being referred to
1589 changes value, the reference isn't notified. This means that the new value won't
1590 be stored in the datafile for other processes to read. There is no TIEREF.
1592 It is theoretically possible to store references to values already within a
1593 DBM::Deep object because everything already is synchronized, but the change to
1594 the internals would be quite large. Specifically, DBM::Deep would have to tie
1595 every single value that is stored. This would bloat the RAM footprint of
1596 DBM::Deep at least twofold (if not more) and be a significant performance drain,
1597 all to support a feature that has never been requested.
1601 L<Data::Dump::Streamer/> provides a mechanism for serializing coderefs,
1602 including saving off all closure state. However, just as for SCALAR and REF,
1603 that closure state may change without notifying the DBM::Deep object storing
1608 =head2 FILE CORRUPTION
1610 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
1611 for a 32-bit signature when opened, but other corruption in files can cause
1612 segmentation faults. DBM::Deep may try to seek() past the end of a file, or get
1613 stuck in an infinite loop depending on the level of corruption. File write
1614 operations are not checked for failure (for speed), so if you happen to run
1615 out of disk space, DBM::Deep will probably fail in a bad way. These things will
1616 be addressed in a later version of DBM::Deep.
1620 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
1621 well on local filesystems, but will NOT protect you from file corruption over
1622 NFS. I've heard about setting up your NFS server with a locking daemon, then
1623 using lockf() to lock your files, but your mileage may vary there as well.
1624 From what I understand, there is no real way to do it. However, if you need
1625 access to the underlying filehandle in DBM::Deep for using some other kind of
1626 locking scheme like lockf(), see the L<LOW-LEVEL ACCESS> section above.
1628 =head2 COPYING OBJECTS
1630 Beware of copying tied objects in Perl. Very strange things can happen.
1631 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
1632 returns a new, blessed, tied hash or array to the same level in the DB.
1634 my $copy = $db->clone();
1636 B<Note>: Since clone() here is cloning the object, not the database location, any
1637 modifications to either $db or $copy will be visible to both.
1641 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1642 These functions cause every element in the array to move, which can be murder
1643 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1644 a different location. This will be addressed in the forthcoming version 1.00.
1646 =head2 WRITEONLY FILES
1648 If you pass in a filehandle to new(), you may have opened it in either a readonly or
1649 writeonly mode. STORE will verify that the filehandle is writable. However, there
1650 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
1651 filehandle isn't readable, it's not clear what will happen. So, don't do that.
1653 =head1 CODE COVERAGE
1655 B<Devel::Cover> is used to test the code coverage of the tests. Below is the
1656 B<Devel::Cover> report on this distribution's test suite.
1658 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1659 File stmt bran cond sub pod time total
1660 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1661 blib/lib/DBM/Deep.pm 96.2 89.0 75.0 95.8 89.5 36.0 92.9
1662 blib/lib/DBM/Deep/Array.pm 96.1 88.3 100.0 96.4 100.0 15.9 94.7
1663 blib/lib/DBM/Deep/Engine.pm 96.6 86.6 89.5 100.0 0.0 20.0 91.0
1664 blib/lib/DBM/Deep/File.pm 99.4 88.3 55.6 100.0 0.0 19.6 89.5
1665 blib/lib/DBM/Deep/Hash.pm 98.5 83.3 100.0 100.0 100.0 8.5 96.3
1666 Total 96.9 87.4 81.2 98.0 38.5 100.0 92.1
1667 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1669 =head1 MORE INFORMATION
1671 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1672 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1675 The source code repository is at L<http://svn.perl.org/modules/DBM-Deep>
1679 Rob Kinyon, L<rkinyon@cpan.org>
1681 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1683 Special thanks to Adam Sah and Rich Gaushell! You know why :-)
1687 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1688 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1692 Copyright (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
1693 This is free software, you may use it and distribute it under the
1694 same terms as Perl itself.