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::Engine;
50 # Setup constants for users to pass to new()
52 sub TYPE_HASH () { DBM::Deep::Engine->SIG_HASH }
53 sub TYPE_ARRAY () { DBM::Deep::Engine->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 $self->{engine} = DBM::Deep::Engine->new( { %{$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};
423 $base = "\$db->get( q{$child->{parent_key}} )->" . $base;
426 $base = "\$db->get( q{$child->{parent_key}} )";
434 # Store single hash key/value or array element in database.
436 my $self = shift->_get_self;
437 my ($key, $value, $orig_key) = @_;
438 $orig_key = $key unless defined $orig_key;
440 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
441 $self->_throw_error( 'Cannot write to a readonly filehandle' );
444 #XXX The second condition needs to disappear
445 if ( !( $self->_type eq TYPE_ARRAY && $orig_key eq 'length') ) {
448 my $r = Scalar::Util::reftype( $value ) || '';
449 if ( $r eq 'HASH' ) {
452 elsif ( $r eq 'ARRAY' ) {
455 elsif ( defined $value ) {
462 if ( my $c = Scalar::Util::blessed( $value ) ) {
463 $rhs = "bless $rhs, '$c'";
466 my $lhs = $self->_find_parent;
468 if ( $self->_type eq TYPE_HASH ) {
469 $lhs .= "->\{q{$orig_key}\}";
472 $lhs .= "->\[$orig_key\]";
478 $lhs = "\$db->put(q{$orig_key},$rhs);";
481 $self->_storage->audit($lhs);
485 # Request exclusive lock for writing
487 $self->lock( LOCK_EX );
489 # User may be storing a complex value, in which case we do not want it run
490 # through the filtering system.
491 if ( !ref($value) && $self->_storage->{filter_store_value} ) {
492 $value = $self->_storage->{filter_store_value}->( $value );
495 $self->_engine->write_value( $self->_base_offset, $key, $value, $orig_key );
504 # Fetch single value or element given plain key or array index
506 my $self = shift->_get_self;
507 my ($key, $orig_key) = @_;
508 $orig_key = $key unless defined $orig_key;
511 # Request shared lock for reading
513 $self->lock( LOCK_SH );
515 my $result = $self->_engine->read_value( $self->_base_offset, $key, $orig_key );
519 # Filters only apply to scalar values, so the ref check is making
520 # sure the fetched bucket is a scalar, not a child hash or array.
521 return ($result && !ref($result) && $self->_storage->{filter_fetch_value})
522 ? $self->_storage->{filter_fetch_value}->($result)
528 # Delete single key/value pair or element given plain key or array index
530 my $self = shift->_get_self;
531 my ($key, $orig_key) = @_;
532 $orig_key = $key unless defined $orig_key;
534 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
535 $self->_throw_error( 'Cannot write to a readonly filehandle' );
538 if ( defined $orig_key ) {
539 my $lhs = $self->_find_parent;
541 $self->_storage->audit( "delete $lhs;" );
544 $self->_storage->audit( "\$db->delete('$orig_key');" );
549 # Request exclusive lock for writing
551 $self->lock( LOCK_EX );
556 my $value = $self->_engine->delete_key( $self->_base_offset, $key, $orig_key );
558 if (defined $value && !ref($value) && $self->_storage->{filter_fetch_value}) {
559 $value = $self->_storage->{filter_fetch_value}->($value);
569 # Check if a single key or element exists given plain key or array index
571 my $self = shift->_get_self;
575 # Request shared lock for reading
577 $self->lock( LOCK_SH );
579 my $result = $self->_engine->key_exists( $self->_base_offset, $key );
588 # Clear all keys from hash, or all elements from array.
590 my $self = shift->_get_self;
592 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
593 $self->_throw_error( 'Cannot write to a readonly filehandle' );
597 my $lhs = $self->_find_parent;
599 if ( $self->_type eq TYPE_HASH ) {
600 $lhs = '%{' . $lhs . '}';
603 $lhs = '@{' . $lhs . '}';
606 $self->_storage->audit( "$lhs = ();" );
610 # Request exclusive lock for writing
612 $self->lock( LOCK_EX );
614 if ( $self->_type eq TYPE_HASH ) {
615 my $key = $self->first_key;
617 # Retrieve the key before deleting because we depend on next_key
618 my $next_key = $self->next_key( $key );
619 $self->_engine->delete_key( $self->_base_offset, $key, $key );
624 my $size = $self->FETCHSIZE;
625 for my $key ( 0 .. $size - 1 ) {
626 $self->_engine->delete_key( $self->_base_offset, $key, $key );
628 $self->STORESIZE( 0 );
630 #XXX This needs updating to use _release_space
631 # $self->_engine->write_tag(
632 # $self->_base_offset, $self->_type,
633 # chr(0)x$self->_engine->{index_size},
642 # Public method aliases
644 sub put { (shift)->STORE( @_ ) }
645 sub store { (shift)->STORE( @_ ) }
646 sub get { (shift)->FETCH( @_ ) }
647 sub fetch { (shift)->FETCH( @_ ) }
648 sub delete { (shift)->DELETE( @_ ) }
649 sub exists { (shift)->EXISTS( @_ ) }
650 sub clear { (shift)->CLEAR( @_ ) }
657 DBM::Deep - A pure perl multi-level hash/array DBM
662 my $db = DBM::Deep->new( "foo.db" );
664 $db->{key} = 'value';
667 $db->put('key' => 'value');
668 print $db->get('key');
670 # true multi-level support
671 $db->{my_complex} = [
672 'hello', { perl => 'rules' },
676 tie my %db, 'DBM::Deep', 'foo.db';
680 tied(%db)->put('key' => 'value');
681 print tied(%db)->get('key');
685 A unique flat-file database module, written in pure perl. True multi-level
686 hash/array support (unlike MLDBM, which is faked), hybrid OO / tie()
687 interface, cross-platform FTPable files, ACID transactions, and is quite fast.
688 Can handle millions of keys and unlimited levels without significant
689 slow-down. Written from the ground-up in pure perl -- this is NOT a wrapper
690 around a C-based DBM. Out-of-the-box compatibility with Unix, Mac OS X and
693 =head1 VERSION DIFFERENCES
695 B<NOTE>: 0.99_01 and above have significant file format differences from 0.983 and
696 before. There will be a backwards-compatibility layer in 1.00, but that is
697 slated for a later 0.99_x release. This version is B<NOT> backwards compatible
698 with 0.983 and before.
702 Construction can be done OO-style (which is the recommended way), or using
703 Perl's tie() function. Both are examined here.
705 =head2 OO CONSTRUCTION
707 The recommended way to construct a DBM::Deep object is to use the new()
708 method, which gets you a blessed I<and> tied hash (or array) reference.
710 my $db = DBM::Deep->new( "foo.db" );
712 This opens a new database handle, mapped to the file "foo.db". If this
713 file does not exist, it will automatically be created. DB files are
714 opened in "r+" (read/write) mode, and the type of object returned is a
715 hash, unless otherwise specified (see L<OPTIONS> below).
717 You can pass a number of options to the constructor to specify things like
718 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
720 my $db = DBM::Deep->new(
726 Notice that the filename is now specified I<inside> the hash with
727 the "file" parameter, as opposed to being the sole argument to the
728 constructor. This is required if any options are specified.
729 See L<OPTIONS> below for the complete list.
731 You can also start with an array instead of a hash. For this, you must
732 specify the C<type> parameter:
734 my $db = DBM::Deep->new(
736 type => DBM::Deep->TYPE_ARRAY
739 B<Note:> Specifing the C<type> parameter only takes effect when beginning
740 a new DB file. If you create a DBM::Deep object with an existing file, the
741 C<type> will be loaded from the file header, and an error will be thrown if
742 the wrong type is passed in.
744 =head2 TIE CONSTRUCTION
746 Alternately, you can create a DBM::Deep handle by using Perl's built-in
747 tie() function. The object returned from tie() can be used to call methods,
748 such as lock() and unlock(). (That object can be retrieved from the tied
749 variable at any time using tied() - please see L<perltie/> for more info.
752 my $db = tie %hash, "DBM::Deep", "foo.db";
755 my $db = tie @array, "DBM::Deep", "bar.db";
757 As with the OO constructor, you can replace the DB filename parameter with
758 a hash containing one or more options (see L<OPTIONS> just below for the
761 tie %hash, "DBM::Deep", {
769 There are a number of options that can be passed in when constructing your
770 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
776 Filename of the DB file to link the handle to. You can pass a full absolute
777 filesystem path, partial path, or a plain filename if the file is in the
778 current working directory. This is a required parameter (though q.v. fh).
782 If you want, you can pass in the fh instead of the file. This is most useful for doing
785 my $db = DBM::Deep->new( { fh => \*DATA } );
787 You are responsible for making sure that the fh has been opened appropriately for your
788 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
789 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
790 needs to read from the fh.
792 =item * audit_file / audit_fh
794 These are just like file/fh, except for auditing. Please see L</AUDITING> for
799 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
800 not need to set this. However, it's there if you want it.
802 If you pass in fh and do not set this, it will be set appropriately.
806 This parameter specifies what type of object to create, a hash or array. Use
807 one of these two constants:
811 =item * C<DBM::Deep-E<gt>TYPE_HASH>
813 =item * C<DBM::Deep-E<gt>TYPE_ARRAY>.
817 This only takes effect when beginning a new file. This is an optional
818 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
822 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
823 function to lock the database in exclusive mode for writes, and shared mode
824 for reads. Pass any true value to enable. This affects the base DB handle
825 I<and any child hashes or arrays> that use the same DB file. This is an
826 optional parameter, and defaults to 0 (disabled). See L<LOCKING> below for
831 Specifies whether autoflush is to be enabled on the underlying filehandle.
832 This obviously slows down write operations, but is required if you may have
833 multiple processes accessing the same DB file (also consider enable I<locking>).
834 Pass any true value to enable. This is an optional parameter, and defaults to 0
839 If I<autobless> mode is enabled, DBM::Deep will preserve the class something
840 is blessed into, and restores it when fetched. This is an optional parameter, and defaults to 1 (enabled).
842 B<Note:> If you use the OO-interface, you will not be able to call any methods
843 of DBM::Deep on the blessed item. This is considered to be a feature.
847 See L</FILTERS> below.
853 With DBM::Deep you can access your databases using Perl's standard hash/array
854 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
855 treat them as such. DBM::Deep will intercept all reads/writes and direct them
856 to the right place -- the DB file. This has nothing to do with the
857 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
858 using regular hashes and arrays, rather than calling functions like C<get()>
859 and C<put()> (although those work too). It is entirely up to you how to want
860 to access your databases.
864 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
865 or even nested hashes (or arrays) using standard Perl syntax:
867 my $db = DBM::Deep->new( "foo.db" );
869 $db->{mykey} = "myvalue";
871 $db->{myhash}->{subkey} = "subvalue";
873 print $db->{myhash}->{subkey} . "\n";
875 You can even step through hash keys using the normal Perl C<keys()> function:
877 foreach my $key (keys %$db) {
878 print "$key: " . $db->{$key} . "\n";
881 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
882 pushes them onto an array, all before the loop even begins. If you have an
883 extremely large hash, this may exhaust Perl's memory. Instead, consider using
884 Perl's C<each()> function, which pulls keys/values one at a time, using very
887 while (my ($key, $value) = each %$db) {
888 print "$key: $value\n";
891 Please note that when using C<each()>, you should always pass a direct
892 hash reference, not a lookup. Meaning, you should B<never> do this:
895 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
897 This causes an infinite loop, because for each iteration, Perl is calling
898 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
899 it effectively keeps returning the first key over and over again. Instead,
900 assign a temporary variable to C<$db->{foo}>, then pass that to each().
904 As with hashes, you can treat any DBM::Deep object like a normal Perl array
905 reference. This includes inserting, removing and manipulating elements,
906 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
907 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
908 or simply be a nested array reference inside a hash. Example:
910 my $db = DBM::Deep->new(
911 file => "foo-array.db",
912 type => DBM::Deep->TYPE_ARRAY
916 push @$db, "bar", "baz";
919 my $last_elem = pop @$db; # baz
920 my $first_elem = shift @$db; # bah
921 my $second_elem = $db->[1]; # bar
923 my $num_elements = scalar @$db;
927 In addition to the I<tie()> interface, you can also use a standard OO interface
928 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
929 array) has its own methods, but both types share the following common methods:
930 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
931 C<store(> are aliases to C<put()> and C<get()>, respectively.
935 =item * new() / clone()
937 These are the constructor and copy-functions.
939 =item * put() / store()
941 Stores a new hash key/value pair, or sets an array element value. Takes two
942 arguments, the hash key or array index, and the new value. The value can be
943 a scalar, hash ref or array ref. Returns true on success, false on failure.
945 $db->put("foo", "bar"); # for hashes
946 $db->put(1, "bar"); # for arrays
948 =item * get() / fetch()
950 Fetches the value of a hash key or array element. Takes one argument: the hash
951 key or array index. Returns a scalar, hash ref or array ref, depending on the
954 my $value = $db->get("foo"); # for hashes
955 my $value = $db->get(1); # for arrays
959 Checks if a hash key or array index exists. Takes one argument: the hash key
960 or array index. Returns true if it exists, false if not.
962 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
963 if ($db->exists(1)) { print "yay!\n"; } # for arrays
967 Deletes one hash key/value pair or array element. Takes one argument: the hash
968 key or array index. Returns true on success, false if not found. For arrays,
969 the remaining elements located after the deleted element are NOT moved over.
970 The deleted element is essentially just undefined, which is exactly how Perl's
971 internal arrays work. Please note that the space occupied by the deleted
972 key/value or element is B<not> reused again -- see L<UNUSED SPACE RECOVERY>
973 below for details and workarounds.
975 $db->delete("foo"); # for hashes
976 $db->delete(1); # for arrays
980 Deletes B<all> hash keys or array elements. Takes no arguments. No return
981 value. Please note that the space occupied by the deleted keys/values or
982 elements is B<not> reused again -- see L<UNUSED SPACE RECOVERY> below for
983 details and workarounds.
985 $db->clear(); # hashes or arrays
987 =item * lock() / unlock()
993 Recover lost disk space. This is important to do, especially if you use
996 =item * import() / export()
998 Data going in and out.
1004 For hashes, DBM::Deep supports all the common methods described above, and the
1005 following additional methods: C<first_key()> and C<next_key()>.
1011 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
1012 fetched in an undefined order (which appears random). Takes no arguments,
1013 returns the key as a scalar value.
1015 my $key = $db->first_key();
1019 Returns the "next" key in the hash, given the previous one as the sole argument.
1020 Returns undef if there are no more keys to be fetched.
1022 $key = $db->next_key($key);
1026 Here are some examples of using hashes:
1028 my $db = DBM::Deep->new( "foo.db" );
1030 $db->put("foo", "bar");
1031 print "foo: " . $db->get("foo") . "\n";
1033 $db->put("baz", {}); # new child hash ref
1034 $db->get("baz")->put("buz", "biz");
1035 print "buz: " . $db->get("baz")->get("buz") . "\n";
1037 my $key = $db->first_key();
1039 print "$key: " . $db->get($key) . "\n";
1040 $key = $db->next_key($key);
1043 if ($db->exists("foo")) { $db->delete("foo"); }
1047 For arrays, DBM::Deep supports all the common methods described above, and the
1048 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
1049 C<unshift()> and C<splice()>.
1055 Returns the number of elements in the array. Takes no arguments.
1057 my $len = $db->length();
1061 Adds one or more elements onto the end of the array. Accepts scalars, hash
1062 refs or array refs. No return value.
1064 $db->push("foo", "bar", {});
1068 Fetches the last element in the array, and deletes it. Takes no arguments.
1069 Returns undef if array is empty. Returns the element value.
1071 my $elem = $db->pop();
1075 Fetches the first element in the array, deletes it, then shifts all the
1076 remaining elements over to take up the space. Returns the element value. This
1077 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
1080 my $elem = $db->shift();
1084 Inserts one or more elements onto the beginning of the array, shifting all
1085 existing elements over to make room. Accepts scalars, hash refs or array refs.
1086 No return value. This method is not recommended with large arrays -- see
1087 <LARGE ARRAYS> below for details.
1089 $db->unshift("foo", "bar", {});
1093 Performs exactly like Perl's built-in function of the same name. See L<perldoc
1094 -f splice> for usage -- it is too complicated to document here. This method is
1095 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
1099 Here are some examples of using arrays:
1101 my $db = DBM::Deep->new(
1103 type => DBM::Deep->TYPE_ARRAY
1106 $db->push("bar", "baz");
1107 $db->unshift("foo");
1110 my $len = $db->length();
1111 print "length: $len\n"; # 4
1113 for (my $k=0; $k<$len; $k++) {
1114 print "$k: " . $db->get($k) . "\n";
1117 $db->splice(1, 2, "biz", "baf");
1119 while (my $elem = shift @$db) {
1120 print "shifted: $elem\n";
1125 Enable automatic file locking by passing a true value to the C<locking>
1126 parameter when constructing your DBM::Deep object (see L<SETUP> above).
1128 my $db = DBM::Deep->new(
1133 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
1134 mode for writes, and shared mode for reads. This is required if you have
1135 multiple processes accessing the same database file, to avoid file corruption.
1136 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
1137 NFS> below for more.
1139 =head2 EXPLICIT LOCKING
1141 You can explicitly lock a database, so it remains locked for multiple
1142 transactions. This is done by calling the C<lock()> method, and passing an
1143 optional lock mode argument (defaults to exclusive mode). This is particularly
1144 useful for things like counters, where the current value needs to be fetched,
1145 then incremented, then stored again.
1148 my $counter = $db->get("counter");
1150 $db->put("counter", $counter);
1159 You can pass C<lock()> an optional argument, which specifies which mode to use
1160 (exclusive or shared). Use one of these two constants:
1161 C<DBM::Deep-E<gt>LOCK_EX> or C<DBM::Deep-E<gt>LOCK_SH>. These are passed
1162 directly to C<flock()>, and are the same as the constants defined in Perl's
1165 $db->lock( $db->LOCK_SH );
1169 =head1 IMPORTING/EXPORTING
1171 You can import existing complex structures by calling the C<import()> method,
1172 and export an entire database into an in-memory structure using the C<export()>
1173 method. Both are examined here.
1177 Say you have an existing hash with nested hashes/arrays inside it. Instead of
1178 walking the structure and adding keys/elements to the database as you go,
1179 simply pass a reference to the C<import()> method. This recursively adds
1180 everything to an existing DBM::Deep object for you. Here is an example:
1185 array1 => [ "elem0", "elem1", "elem2" ],
1187 subkey1 => "subvalue1",
1188 subkey2 => "subvalue2"
1192 my $db = DBM::Deep->new( "foo.db" );
1193 $db->import( $struct );
1195 print $db->{key1} . "\n"; # prints "value1"
1197 This recursively imports the entire C<$struct> object into C<$db>, including
1198 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
1199 keys are merged with the existing ones, replacing if they already exist.
1200 The C<import()> method can be called on any database level (not just the base
1201 level), and works with both hash and array DB types.
1203 B<Note:> Make sure your existing structure has no circular references in it.
1204 These will cause an infinite loop when importing. There are plans to fix this
1209 Calling the C<export()> method on an existing DBM::Deep object will return
1210 a reference to a new in-memory copy of the database. The export is done
1211 recursively, so all nested hashes/arrays are all exported to standard Perl
1212 objects. Here is an example:
1214 my $db = DBM::Deep->new( "foo.db" );
1216 $db->{key1} = "value1";
1217 $db->{key2} = "value2";
1219 $db->{hash1}->{subkey1} = "subvalue1";
1220 $db->{hash1}->{subkey2} = "subvalue2";
1222 my $struct = $db->export();
1224 print $struct->{key1} . "\n"; # prints "value1"
1226 This makes a complete copy of the database in memory, and returns a reference
1227 to it. The C<export()> method can be called on any database level (not just
1228 the base level), and works with both hash and array DB types. Be careful of
1229 large databases -- you can store a lot more data in a DBM::Deep object than an
1230 in-memory Perl structure.
1232 B<Note:> Make sure your database has no circular references in it.
1233 These will cause an infinite loop when exporting. There are plans to fix this
1238 DBM::Deep has a number of hooks where you can specify your own Perl function
1239 to perform filtering on incoming or outgoing data. This is a perfect
1240 way to extend the engine, and implement things like real-time compression or
1241 encryption. Filtering applies to the base DB level, and all child hashes /
1242 arrays. Filter hooks can be specified when your DBM::Deep object is first
1243 constructed, or by calling the C<set_filter()> method at any time. There are
1244 four available filter hooks, described below:
1248 =item * filter_store_key
1250 This filter is called whenever a hash key is stored. It
1251 is passed the incoming key, and expected to return a transformed key.
1253 =item * filter_store_value
1255 This filter is called whenever a hash key or array element is stored. It
1256 is passed the incoming value, and expected to return a transformed value.
1258 =item * filter_fetch_key
1260 This filter is called whenever a hash key is fetched (i.e. via
1261 C<first_key()> or C<next_key()>). It is passed the transformed key,
1262 and expected to return the plain key.
1264 =item * filter_fetch_value
1266 This filter is called whenever a hash key or array element is fetched.
1267 It is passed the transformed value, and expected to return the plain value.
1271 Here are the two ways to setup a filter hook:
1273 my $db = DBM::Deep->new(
1275 filter_store_value => \&my_filter_store,
1276 filter_fetch_value => \&my_filter_fetch
1281 $db->set_filter( "filter_store_value", \&my_filter_store );
1282 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
1284 Your filter function will be called only when dealing with SCALAR keys or
1285 values. When nested hashes and arrays are being stored/fetched, filtering
1286 is bypassed. Filters are called as static functions, passed a single SCALAR
1287 argument, and expected to return a single SCALAR value. If you want to
1288 remove a filter, set the function reference to C<undef>:
1290 $db->set_filter( "filter_store_value", undef );
1292 =head2 REAL-TIME ENCRYPTION EXAMPLE
1294 Here is a working example that uses the I<Crypt::Blowfish> module to
1295 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
1296 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
1297 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
1300 use Crypt::Blowfish;
1303 my $cipher = Crypt::CBC->new({
1304 'key' => 'my secret key',
1305 'cipher' => 'Blowfish',
1307 'regenerate_key' => 0,
1308 'padding' => 'space',
1312 my $db = DBM::Deep->new(
1313 file => "foo-encrypt.db",
1314 filter_store_key => \&my_encrypt,
1315 filter_store_value => \&my_encrypt,
1316 filter_fetch_key => \&my_decrypt,
1317 filter_fetch_value => \&my_decrypt,
1320 $db->{key1} = "value1";
1321 $db->{key2} = "value2";
1322 print "key1: " . $db->{key1} . "\n";
1323 print "key2: " . $db->{key2} . "\n";
1329 return $cipher->encrypt( $_[0] );
1332 return $cipher->decrypt( $_[0] );
1335 =head2 REAL-TIME COMPRESSION EXAMPLE
1337 Here is a working example that uses the I<Compress::Zlib> module to do real-time
1338 compression / decompression of keys & values with DBM::Deep Filters.
1339 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
1340 more on I<Compress::Zlib>.
1345 my $db = DBM::Deep->new(
1346 file => "foo-compress.db",
1347 filter_store_key => \&my_compress,
1348 filter_store_value => \&my_compress,
1349 filter_fetch_key => \&my_decompress,
1350 filter_fetch_value => \&my_decompress,
1353 $db->{key1} = "value1";
1354 $db->{key2} = "value2";
1355 print "key1: " . $db->{key1} . "\n";
1356 print "key2: " . $db->{key2} . "\n";
1362 return Compress::Zlib::memGzip( $_[0] ) ;
1365 return Compress::Zlib::memGunzip( $_[0] ) ;
1368 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
1369 actually numerical index numbers, and are not filtered.
1371 =head1 ERROR HANDLING
1373 Most DBM::Deep methods return a true value for success, and call die() on
1374 failure. You can wrap calls in an eval block to catch the die.
1376 my $db = DBM::Deep->new( "foo.db" ); # create hash
1377 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
1379 print $@; # prints error message
1381 =head1 LARGEFILE SUPPORT
1383 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
1384 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
1385 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
1386 by specifying the 'pack_size' parameter when constructing the file.
1389 filename => $filename,
1390 pack_size => 'large',
1393 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
1394 instead of 32-bit longs. After setting these values your DB files have a
1395 theoretical maximum size of 16 XB (exabytes).
1397 You can also use C<pack_size =E<gt> 'small'> in order to use 16-bit file
1400 B<Note:> Changing these values will B<NOT> work for existing database files.
1401 Only change this for new files. Once the value has been set, it is stored in
1402 the file's header and cannot be changed for the life of the file. These
1403 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
1406 B<Note:> We have not personally tested files larger than 2 GB -- all my
1407 systems have only a 32-bit Perl. However, I have received user reports that
1408 this does indeed work!
1410 =head1 LOW-LEVEL ACCESS
1412 If you require low-level access to the underlying filehandle that DBM::Deep uses,
1413 you can call the C<_fh()> method, which returns the handle:
1415 my $fh = $db->_fh();
1417 This method can be called on the root level of the datbase, or any child
1418 hashes or arrays. All levels share a I<root> structure, which contains things
1419 like the filehandle, a reference counter, and all the options specified
1420 when you created the object. You can get access to this file object by
1421 calling the C<_storage()> method.
1423 my $file_obj = $db->_storage();
1425 This is useful for changing options after the object has already been created,
1426 such as enabling/disabling locking. You can also store your own temporary user
1427 data in this structure (be wary of name collision), which is then accessible from
1428 any child hash or array.
1430 =head1 CUSTOM DIGEST ALGORITHM
1432 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
1433 keys. However you can override this, and use another algorithm (such as SHA-256)
1434 or even write your own. But please note that DBM::Deep currently expects zero
1435 collisions, so your algorithm has to be I<perfect>, so to speak. Collision
1436 detection may be introduced in a later version.
1438 You can specify a custom digest algorithm by passing it into the parameter
1439 list for new(), passing a reference to a subroutine as the 'digest' parameter,
1440 and the length of the algorithm's hashes (in bytes) as the 'hash_size'
1441 parameter. Here is a working example that uses a 256-bit hash from the
1442 I<Digest::SHA256> module. Please see
1443 L<http://search.cpan.org/search?module=Digest::SHA256> for more information.
1448 my $context = Digest::SHA256::new(256);
1450 my $db = DBM::Deep->new(
1451 filename => "foo-sha.db",
1452 digest => \&my_digest,
1456 $db->{key1} = "value1";
1457 $db->{key2} = "value2";
1458 print "key1: " . $db->{key1} . "\n";
1459 print "key2: " . $db->{key2} . "\n";
1465 return substr( $context->hash($_[0]), 0, 32 );
1468 B<Note:> Your returned digest strings must be B<EXACTLY> the number
1469 of bytes you specify in the hash_size parameter (in this case 32).
1471 B<Note:> If you do choose to use a custom digest algorithm, you must set it
1472 every time you access this file. Otherwise, the default (MD5) will be used.
1474 =head1 CIRCULAR REFERENCES
1476 DBM::Deep has B<experimental> support for circular references. Meaning you
1477 can have a nested hash key or array element that points to a parent object.
1478 This relationship is stored in the DB file, and is preserved between sessions.
1481 my $db = DBM::Deep->new( "foo.db" );
1484 $db->{circle} = $db; # ref to self
1486 print $db->{foo} . "\n"; # prints "bar"
1487 print $db->{circle}->{foo} . "\n"; # prints "bar" again
1489 B<Note>: Passing the object to a function that recursively walks the
1490 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
1491 C<export()> methods) will result in an infinite loop. This will be fixed in
1496 New in 0.99_01 is the ability to audit your databases actions. By passing in
1497 audit_file (or audit_fh) to the constructor, all actions will be logged to
1498 that file. The format is one that is suitable for eval'ing against the
1499 database to replay the actions. Please see t/33_audit_trail.t for an example
1504 New in 0.99_01 is ACID transactions. Every DBM::Deep object is completely
1505 transaction-ready - it is not an option you have to turn on. Three new methods
1506 have been added to support them. They are:
1510 =item * begin_work()
1512 This starts a transaction.
1516 This applies the changes done within the transaction to the mainline and ends
1521 This discards the changes done within the transaction to the mainline and ends
1526 Transactions in DBM::Deep are done using the MVCC method, the same method used
1527 by the InnoDB MySQL table type.
1529 =head1 CAVEATS / ISSUES / BUGS
1531 This section describes all the known issues with DBM::Deep. It you have found
1532 something that is not listed here, please send e-mail to L<jhuckaby@cpan.org>.
1534 =head2 UNUSED SPACE RECOVERY
1536 One major caveat with DBM::Deep is that space occupied by existing keys and
1537 values is not recovered when they are deleted. Meaning if you keep deleting
1538 and adding new keys, your file will continuously grow. I am working on this,
1539 but in the meantime you can call the built-in C<optimize()> method from time to
1540 time (perhaps in a crontab or something) to recover all your unused space.
1542 $db->optimize(); # returns true on success
1544 This rebuilds the ENTIRE database into a new file, then moves it on top of
1545 the original. The new file will have no unused space, thus it will take up as
1546 little disk space as possible. Please note that this operation can take
1547 a long time for large files, and you need enough disk space to temporarily hold
1548 2 copies of your DB file. The temporary file is created in the same directory
1549 as the original, named with a ".tmp" extension, and is deleted when the
1550 operation completes. Oh, and if locking is enabled, the DB is automatically
1551 locked for the entire duration of the copy.
1553 B<WARNING:> Only call optimize() on the top-level node of the database, and
1554 make sure there are no child references lying around. DBM::Deep keeps a reference
1555 counter, and if it is greater than 1, optimize() will abort and return undef.
1559 (The reasons given assume a high level of Perl understanding, specifically of
1560 references. You can safely skip this section.)
1562 Currently, the only references supported are HASH and ARRAY. The other reference
1563 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
1569 These are things like filehandles and other sockets. They can't be supported
1570 because it's completely unclear how DBM::Deep should serialize them.
1572 =item * SCALAR / REF
1574 The discussion here refers to the following type of example:
1581 # In some other process ...
1583 my $val = ${ $db->{key1} };
1585 is( $val, 50, "What actually gets stored in the DB file?" );
1587 The problem is one of synchronization. When the variable being referred to
1588 changes value, the reference isn't notified. This means that the new value won't
1589 be stored in the datafile for other processes to read. There is no TIEREF.
1591 It is theoretically possible to store references to values already within a
1592 DBM::Deep object because everything already is synchronized, but the change to
1593 the internals would be quite large. Specifically, DBM::Deep would have to tie
1594 every single value that is stored. This would bloat the RAM footprint of
1595 DBM::Deep at least twofold (if not more) and be a significant performance drain,
1596 all to support a feature that has never been requested.
1600 L<Data::Dump::Streamer/> provides a mechanism for serializing coderefs,
1601 including saving off all closure state. However, just as for SCALAR and REF,
1602 that closure state may change without notifying the DBM::Deep object storing
1607 =head2 FILE CORRUPTION
1609 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
1610 for a 32-bit signature when opened, but other corruption in files can cause
1611 segmentation faults. DBM::Deep may try to seek() past the end of a file, or get
1612 stuck in an infinite loop depending on the level of corruption. File write
1613 operations are not checked for failure (for speed), so if you happen to run
1614 out of disk space, DBM::Deep will probably fail in a bad way. These things will
1615 be addressed in a later version of DBM::Deep.
1619 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
1620 well on local filesystems, but will NOT protect you from file corruption over
1621 NFS. I've heard about setting up your NFS server with a locking daemon, then
1622 using lockf() to lock your files, but your mileage may vary there as well.
1623 From what I understand, there is no real way to do it. However, if you need
1624 access to the underlying filehandle in DBM::Deep for using some other kind of
1625 locking scheme like lockf(), see the L<LOW-LEVEL ACCESS> section above.
1627 =head2 COPYING OBJECTS
1629 Beware of copying tied objects in Perl. Very strange things can happen.
1630 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
1631 returns a new, blessed, tied hash or array to the same level in the DB.
1633 my $copy = $db->clone();
1635 B<Note>: Since clone() here is cloning the object, not the database location, any
1636 modifications to either $db or $copy will be visible to both.
1640 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1641 These functions cause every element in the array to move, which can be murder
1642 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1643 a different location. This will be addressed in the forthcoming version 1.00.
1645 =head2 WRITEONLY FILES
1647 If you pass in a filehandle to new(), you may have opened it in either a readonly or
1648 writeonly mode. STORE will verify that the filehandle is writable. However, there
1649 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
1650 filehandle isn't readable, it's not clear what will happen. So, don't do that.
1652 =head1 CODE COVERAGE
1654 B<Devel::Cover> is used to test the code coverage of the tests. Below is the
1655 B<Devel::Cover> report on this distribution's test suite.
1657 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1658 File stmt bran cond sub pod time total
1659 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1660 blib/lib/DBM/Deep.pm 96.2 89.0 75.0 95.8 89.5 36.0 92.9
1661 blib/lib/DBM/Deep/Array.pm 96.1 88.3 100.0 96.4 100.0 15.9 94.7
1662 blib/lib/DBM/Deep/Engine.pm 96.6 86.6 89.5 100.0 0.0 20.0 91.0
1663 blib/lib/DBM/Deep/File.pm 99.4 88.3 55.6 100.0 0.0 19.6 89.5
1664 blib/lib/DBM/Deep/Hash.pm 98.5 83.3 100.0 100.0 100.0 8.5 96.3
1665 Total 96.9 87.4 81.2 98.0 38.5 100.0 92.1
1666 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1668 =head1 MORE INFORMATION
1670 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1671 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1674 The source code repository is at L<http://svn.perl.org/modules/DBM-Deep>
1678 Rob Kinyon, L<rkinyon@cpan.org>
1680 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1682 Special thanks to Adam Sah and Rich Gaushell! You know why :-)
1686 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1687 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1691 Copyright (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
1692 This is free software, you may use it and distribute it under the
1693 same terms as Perl itself.