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::Engine3;
50 # Setup constants for users to pass to new()
52 sub TYPE_HASH () { DBM::Deep::Engine3->SIG_HASH }
53 sub TYPE_ARRAY () { DBM::Deep::Engine3->SIG_ARRAY }
55 # This is used in all the children of this class in their TIE<type> methods.
62 $proto->_throw_error( "Odd number of parameters to " . (caller(1))[2] );
67 unless ( eval { local $SIG{'__DIE__'}; %{$_[0]} || 1 } ) {
68 $proto->_throw_error( "Not a hashref in args to " . (caller(1))[2] );
73 $args = { file => shift };
81 # Class constructor method for Perl OO interface.
82 # Calls tie() and returns blessed reference to tied hash or array,
83 # providing a hybrid OO/tie interface.
86 my $args = $class->_get_args( @_ );
89 # Check if we want a tied hash or array.
92 if (defined($args->{type}) && $args->{type} eq TYPE_ARRAY) {
93 $class = 'DBM::Deep::Array';
94 require DBM::Deep::Array;
95 tie @$self, $class, %$args;
98 $class = 'DBM::Deep::Hash';
99 require DBM::Deep::Hash;
100 tie %$self, $class, %$args;
103 return bless $self, $class;
106 # This initializer is called from the various TIE* methods. new() calls tie(),
107 # which allows for a single point of entry.
112 $args->{storage} = DBM::Deep::File->new( $args )
113 unless exists $args->{storage};
115 # locking implicitly enables autoflush
116 if ($args->{locking}) { $args->{autoflush} = 1; }
118 # These are the defaults to be optionally overridden below
121 base_offset => undef,
130 $args->{engine} = DBM::Deep::Engine3->new( { %{$args}, obj => $self } )
131 unless exists $args->{engine};
133 # Grab the parameters we want to use
134 foreach my $param ( keys %$self ) {
135 next unless exists $args->{$param};
136 $self->{$param} = $args->{$param};
140 local $SIG{'__DIE__'};
143 $self->_engine->setup_fh( $self );
144 $self->_storage->set_inode;
148 eval { local $SIG{'__DIE__'}; $self->unlock; };
157 require DBM::Deep::Hash;
158 return DBM::Deep::Hash->TIEHASH( @_ );
163 require DBM::Deep::Array;
164 return DBM::Deep::Array->TIEARRAY( @_ );
168 my $self = shift->_get_self;
169 return $self->_storage->lock( $self, @_ );
173 my $self = shift->_get_self;
174 return $self->_storage->unlock( $self, @_ );
178 my $self = shift->_get_self;
179 my ($spot, $value) = @_;
184 elsif ( eval { local $SIG{__DIE__}; $value->isa( 'DBM::Deep' ) } ) {
185 ${$spot} = $value->_repr;
186 $value->_copy_node( ${$spot} );
189 my $r = Scalar::Util::reftype( $value );
190 my $c = Scalar::Util::blessed( $value );
191 if ( $r eq 'ARRAY' ) {
192 ${$spot} = [ @{$value} ];
195 ${$spot} = { %{$value} };
197 ${$spot} = bless ${$spot}, $c
205 die "Must be implemented in a child class\n";
209 die "Must be implemented in a child class\n";
214 # Recursively export into standard Perl hashes and arrays.
216 my $self = shift->_get_self;
218 my $temp = $self->_repr;
221 $self->_copy_node( $temp );
224 my $classname = $self->_engine->get_classname( $self );
225 if ( defined $classname ) {
226 bless $temp, $classname;
234 # Recursively import Perl hash/array structure
236 if (!ref($_[0])) { return; } # Perl calls import() on use -- ignore
238 my $self = shift->_get_self;
241 # struct is not a reference, so just import based on our type
243 $struct = $self->_repr( @_ );
246 #XXX This isn't the best solution. Better would be to use Data::Walker,
247 #XXX but that's a lot more thinking than I want to do right now.
250 $self->_import( _clone_data( $struct ) );
260 #XXX Need to keep track of who has a fh to this file in order to
261 #XXX close them all prior to optimize on Win32/cygwin
264 # Rebuild entire database into new file, then move
265 # it back on top of original.
267 my $self = shift->_get_self;
269 #XXX Need to create a new test for this
270 # if ($self->_storage->{links} > 1) {
271 # $self->_throw_error("Cannot optimize: reference count is greater than 1");
274 #XXX Do we have to lock the tempfile?
276 my $db_temp = DBM::Deep->new(
277 file => $self->_storage->{file} . '.tmp',
282 $self->_copy_node( $db_temp );
286 # Attempt to copy user, group and permissions over to new file
288 my @stats = stat($self->_fh);
289 my $perms = $stats[2] & 07777;
292 chown( $uid, $gid, $self->_storage->{file} . '.tmp' );
293 chmod( $perms, $self->_storage->{file} . '.tmp' );
295 # q.v. perlport for more information on this variable
296 if ( $^O eq 'MSWin32' || $^O eq 'cygwin' ) {
298 # Potential race condition when optmizing on Win32 with locking.
299 # The Windows filesystem requires that the filehandle be closed
300 # before it is overwritten with rename(). This could be redone
304 $self->_storage->close;
307 if (!rename $self->_storage->{file} . '.tmp', $self->_storage->{file}) {
308 unlink $self->_storage->{file} . '.tmp';
310 $self->_throw_error("Optimize failed: Cannot copy temp file over original: $!");
314 $self->_storage->close;
315 $self->_storage->open;
316 $self->_engine->setup_fh( $self );
323 # Make copy of object and return
325 my $self = shift->_get_self;
327 return DBM::Deep->new(
328 type => $self->_type,
329 base_offset => $self->_base_offset,
330 storage => $self->_storage,
331 engine => $self->_engine,
332 parent => $self->{parent},
333 parent_key => $self->{parent_key},
338 my %is_legal_filter = map {
341 store_key store_value
342 fetch_key fetch_value
347 # Setup filter function for storing or fetching the key or value
349 my $self = shift->_get_self;
353 if ( $is_legal_filter{$type} ) {
354 $self->_storage->{"filter_$type"} = $func;
363 my $self = shift->_get_self;
364 return $self->_engine->begin_work( $self, @_ );
368 my $self = shift->_get_self;
369 return $self->_engine->rollback( $self, @_ );
373 my $self = shift->_get_self;
374 return $self->_engine->commit( $self, @_ );
382 my $self = $_[0]->_get_self;
383 return $self->{engine};
387 my $self = $_[0]->_get_self;
388 return $self->{storage};
392 my $self = $_[0]->_get_self;
393 return $self->{type};
397 my $self = $_[0]->_get_self;
398 return $self->{base_offset};
402 my $self = $_[0]->_get_self;
403 return $self->_storage->{fh};
411 die "DBM::Deep: $_[1]\n";
418 #XXX This if() is redundant
419 if ( my $parent = $self->{parent} ) {
421 while ( $parent->{parent} ) {
423 $parent->_type eq TYPE_HASH
424 ? "\{q{$child->{parent_key}}\}"
425 : "\[$child->{parent_key}\]"
429 $parent = $parent->{parent};
433 $base = "\$db->get( q{$child->{parent_key}} )->" . $base;
436 $base = "\$db->get( q{$child->{parent_key}} )";
444 # Store single hash key/value or array element in database.
446 my $self = shift->_get_self;
447 my ($key, $value, $orig_key) = @_;
448 $orig_key = $key unless defined $orig_key;
450 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
451 $self->_throw_error( 'Cannot write to a readonly filehandle' );
454 #XXX The second condition needs to disappear
455 if ( !( $self->_type eq TYPE_ARRAY && $orig_key eq 'length') ) {
458 my $r = Scalar::Util::reftype( $value ) || '';
459 if ( $r eq 'HASH' ) {
462 elsif ( $r eq 'ARRAY' ) {
465 elsif ( defined $value ) {
472 if ( my $c = Scalar::Util::blessed( $value ) ) {
473 $rhs = "bless $rhs, '$c'";
476 my $lhs = $self->_find_parent;
478 if ( $self->_type eq TYPE_HASH ) {
479 $lhs .= "->\{q{$orig_key}\}";
482 $lhs .= "->\[$orig_key\]";
488 $lhs = "\$db->put(q{$orig_key},$rhs);";
491 $self->_storage->audit($lhs);
495 # Request exclusive lock for writing
497 $self->lock( LOCK_EX );
499 # User may be storing a complex value, in which case we do not want it run
500 # through the filtering system.
501 if ( !ref($value) && $self->_storage->{filter_store_value} ) {
502 $value = $self->_storage->{filter_store_value}->( $value );
505 $self->_engine->write_value( $self, $key, $value, $orig_key );
514 # Fetch single value or element given plain key or array index
516 my $self = shift->_get_self;
517 my ($key, $orig_key) = @_;
518 $orig_key = $key unless defined $orig_key;
521 # Request shared lock for reading
523 $self->lock( LOCK_SH );
525 my $result = $self->_engine->read_value( $self, $key, $orig_key );
529 # Filters only apply to scalar values, so the ref check is making
530 # sure the fetched bucket is a scalar, not a child hash or array.
531 return ($result && !ref($result) && $self->_storage->{filter_fetch_value})
532 ? $self->_storage->{filter_fetch_value}->($result)
538 # Delete single key/value pair or element given plain key or array index
540 my $self = shift->_get_self;
541 my ($key, $orig_key) = @_;
542 $orig_key = $key unless defined $orig_key;
544 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
545 $self->_throw_error( 'Cannot write to a readonly filehandle' );
548 if ( defined $orig_key ) {
549 my $lhs = $self->_find_parent;
551 $self->_storage->audit( "delete $lhs;" );
554 $self->_storage->audit( "\$db->delete('$orig_key');" );
559 # Request exclusive lock for writing
561 $self->lock( LOCK_EX );
566 my $value = $self->_engine->delete_key( $self, $key, $orig_key );
568 if (defined $value && !ref($value) && $self->_storage->{filter_fetch_value}) {
569 $value = $self->_storage->{filter_fetch_value}->($value);
579 # Check if a single key or element exists given plain key or array index
581 my $self = shift->_get_self;
585 # Request shared lock for reading
587 $self->lock( LOCK_SH );
589 my $result = $self->_engine->key_exists( $self, $key );
598 # Clear all keys from hash, or all elements from array.
600 my $self = shift->_get_self;
602 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
603 $self->_throw_error( 'Cannot write to a readonly filehandle' );
607 my $lhs = $self->_find_parent;
609 if ( $self->_type eq TYPE_HASH ) {
610 $lhs = '%{' . $lhs . '}';
613 $lhs = '@{' . $lhs . '}';
616 $self->_storage->audit( "$lhs = ();" );
620 # Request exclusive lock for writing
622 $self->lock( LOCK_EX );
624 if ( $self->_type eq TYPE_HASH ) {
625 my $key = $self->first_key;
627 # Retrieve the key before deleting because we depend on next_key
628 my $next_key = $self->next_key( $key );
629 $self->_engine->delete_key( $self, $key, $key );
634 my $size = $self->FETCHSIZE;
635 for my $key ( 0 .. $size - 1 ) {
636 $self->_engine->delete_key( $self, $key, $key );
638 $self->STORESIZE( 0 );
640 #XXX This needs updating to use _release_space
641 # $self->_engine->write_tag(
642 # $self->_base_offset, $self->_type,
643 # chr(0)x$self->_engine->{index_size},
652 # Public method aliases
654 sub put { (shift)->STORE( @_ ) }
655 sub store { (shift)->STORE( @_ ) }
656 sub get { (shift)->FETCH( @_ ) }
657 sub fetch { (shift)->FETCH( @_ ) }
658 sub delete { (shift)->DELETE( @_ ) }
659 sub exists { (shift)->EXISTS( @_ ) }
660 sub clear { (shift)->CLEAR( @_ ) }
667 DBM::Deep - A pure perl multi-level hash/array DBM
672 my $db = DBM::Deep->new( "foo.db" );
674 $db->{key} = 'value';
677 $db->put('key' => 'value');
678 print $db->get('key');
680 # true multi-level support
681 $db->{my_complex} = [
682 'hello', { perl => 'rules' },
686 tie my %db, 'DBM::Deep', 'foo.db';
690 tied(%db)->put('key' => 'value');
691 print tied(%db)->get('key');
695 A unique flat-file database module, written in pure perl. True multi-level
696 hash/array support (unlike MLDBM, which is faked), hybrid OO / tie()
697 interface, cross-platform FTPable files, ACID transactions, and is quite fast.
698 Can handle millions of keys and unlimited levels without significant
699 slow-down. Written from the ground-up in pure perl -- this is NOT a wrapper
700 around a C-based DBM. Out-of-the-box compatibility with Unix, Mac OS X and
703 =head1 VERSION DIFFERENCES
705 B<NOTE>: 0.99_01 and above have significant file format differences from 0.983 and
706 before. There will be a backwards-compatibility layer in 1.00, but that is
707 slated for a later 0.99_x release. This version is B<NOT> backwards compatible
708 with 0.983 and before.
712 Construction can be done OO-style (which is the recommended way), or using
713 Perl's tie() function. Both are examined here.
715 =head2 OO CONSTRUCTION
717 The recommended way to construct a DBM::Deep object is to use the new()
718 method, which gets you a blessed I<and> tied hash (or array) reference.
720 my $db = DBM::Deep->new( "foo.db" );
722 This opens a new database handle, mapped to the file "foo.db". If this
723 file does not exist, it will automatically be created. DB files are
724 opened in "r+" (read/write) mode, and the type of object returned is a
725 hash, unless otherwise specified (see L<OPTIONS> below).
727 You can pass a number of options to the constructor to specify things like
728 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
730 my $db = DBM::Deep->new(
736 Notice that the filename is now specified I<inside> the hash with
737 the "file" parameter, as opposed to being the sole argument to the
738 constructor. This is required if any options are specified.
739 See L<OPTIONS> below for the complete list.
741 You can also start with an array instead of a hash. For this, you must
742 specify the C<type> parameter:
744 my $db = DBM::Deep->new(
746 type => DBM::Deep->TYPE_ARRAY
749 B<Note:> Specifing the C<type> parameter only takes effect when beginning
750 a new DB file. If you create a DBM::Deep object with an existing file, the
751 C<type> will be loaded from the file header, and an error will be thrown if
752 the wrong type is passed in.
754 =head2 TIE CONSTRUCTION
756 Alternately, you can create a DBM::Deep handle by using Perl's built-in
757 tie() function. The object returned from tie() can be used to call methods,
758 such as lock() and unlock(). (That object can be retrieved from the tied
759 variable at any time using tied() - please see L<perltie/> for more info.
762 my $db = tie %hash, "DBM::Deep", "foo.db";
765 my $db = tie @array, "DBM::Deep", "bar.db";
767 As with the OO constructor, you can replace the DB filename parameter with
768 a hash containing one or more options (see L<OPTIONS> just below for the
771 tie %hash, "DBM::Deep", {
779 There are a number of options that can be passed in when constructing your
780 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
786 Filename of the DB file to link the handle to. You can pass a full absolute
787 filesystem path, partial path, or a plain filename if the file is in the
788 current working directory. This is a required parameter (though q.v. fh).
792 If you want, you can pass in the fh instead of the file. This is most useful for doing
795 my $db = DBM::Deep->new( { fh => \*DATA } );
797 You are responsible for making sure that the fh has been opened appropriately for your
798 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
799 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
800 needs to read from the fh.
802 =item * audit_file / audit_fh
804 These are just like file/fh, except for auditing. Please see L</AUDITING> for
809 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
810 not need to set this. However, it's there if you want it.
812 If you pass in fh and do not set this, it will be set appropriately.
816 This parameter specifies what type of object to create, a hash or array. Use
817 one of these two constants:
821 =item * C<DBM::Deep-E<gt>TYPE_HASH>
823 =item * C<DBM::Deep-E<gt>TYPE_ARRAY>.
827 This only takes effect when beginning a new file. This is an optional
828 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
832 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
833 function to lock the database in exclusive mode for writes, and shared mode
834 for reads. Pass any true value to enable. This affects the base DB handle
835 I<and any child hashes or arrays> that use the same DB file. This is an
836 optional parameter, and defaults to 0 (disabled). See L<LOCKING> below for
841 Specifies whether autoflush is to be enabled on the underlying filehandle.
842 This obviously slows down write operations, but is required if you may have
843 multiple processes accessing the same DB file (also consider enable I<locking>).
844 Pass any true value to enable. This is an optional parameter, and defaults to 0
849 If I<autobless> mode is enabled, DBM::Deep will preserve the class something
850 is blessed into, and restores it when fetched. This is an optional parameter, and defaults to 1 (enabled).
852 B<Note:> If you use the OO-interface, you will not be able to call any methods
853 of DBM::Deep on the blessed item. This is considered to be a feature.
857 See L</FILTERS> below.
863 With DBM::Deep you can access your databases using Perl's standard hash/array
864 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
865 treat them as such. DBM::Deep will intercept all reads/writes and direct them
866 to the right place -- the DB file. This has nothing to do with the
867 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
868 using regular hashes and arrays, rather than calling functions like C<get()>
869 and C<put()> (although those work too). It is entirely up to you how to want
870 to access your databases.
874 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
875 or even nested hashes (or arrays) using standard Perl syntax:
877 my $db = DBM::Deep->new( "foo.db" );
879 $db->{mykey} = "myvalue";
881 $db->{myhash}->{subkey} = "subvalue";
883 print $db->{myhash}->{subkey} . "\n";
885 You can even step through hash keys using the normal Perl C<keys()> function:
887 foreach my $key (keys %$db) {
888 print "$key: " . $db->{$key} . "\n";
891 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
892 pushes them onto an array, all before the loop even begins. If you have an
893 extremely large hash, this may exhaust Perl's memory. Instead, consider using
894 Perl's C<each()> function, which pulls keys/values one at a time, using very
897 while (my ($key, $value) = each %$db) {
898 print "$key: $value\n";
901 Please note that when using C<each()>, you should always pass a direct
902 hash reference, not a lookup. Meaning, you should B<never> do this:
905 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
907 This causes an infinite loop, because for each iteration, Perl is calling
908 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
909 it effectively keeps returning the first key over and over again. Instead,
910 assign a temporary variable to C<$db->{foo}>, then pass that to each().
914 As with hashes, you can treat any DBM::Deep object like a normal Perl array
915 reference. This includes inserting, removing and manipulating elements,
916 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
917 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
918 or simply be a nested array reference inside a hash. Example:
920 my $db = DBM::Deep->new(
921 file => "foo-array.db",
922 type => DBM::Deep->TYPE_ARRAY
926 push @$db, "bar", "baz";
929 my $last_elem = pop @$db; # baz
930 my $first_elem = shift @$db; # bah
931 my $second_elem = $db->[1]; # bar
933 my $num_elements = scalar @$db;
937 In addition to the I<tie()> interface, you can also use a standard OO interface
938 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
939 array) has its own methods, but both types share the following common methods:
940 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
941 C<store(> are aliases to C<put()> and C<get()>, respectively.
945 =item * new() / clone()
947 These are the constructor and copy-functions.
949 =item * put() / store()
951 Stores a new hash key/value pair, or sets an array element value. Takes two
952 arguments, the hash key or array index, and the new value. The value can be
953 a scalar, hash ref or array ref. Returns true on success, false on failure.
955 $db->put("foo", "bar"); # for hashes
956 $db->put(1, "bar"); # for arrays
958 =item * get() / fetch()
960 Fetches the value of a hash key or array element. Takes one argument: the hash
961 key or array index. Returns a scalar, hash ref or array ref, depending on the
964 my $value = $db->get("foo"); # for hashes
965 my $value = $db->get(1); # for arrays
969 Checks if a hash key or array index exists. Takes one argument: the hash key
970 or array index. Returns true if it exists, false if not.
972 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
973 if ($db->exists(1)) { print "yay!\n"; } # for arrays
977 Deletes one hash key/value pair or array element. Takes one argument: the hash
978 key or array index. Returns true on success, false if not found. For arrays,
979 the remaining elements located after the deleted element are NOT moved over.
980 The deleted element is essentially just undefined, which is exactly how Perl's
981 internal arrays work. Please note that the space occupied by the deleted
982 key/value or element is B<not> reused again -- see L<UNUSED SPACE RECOVERY>
983 below for details and workarounds.
985 $db->delete("foo"); # for hashes
986 $db->delete(1); # for arrays
990 Deletes B<all> hash keys or array elements. Takes no arguments. No return
991 value. Please note that the space occupied by the deleted keys/values or
992 elements is B<not> reused again -- see L<UNUSED SPACE RECOVERY> below for
993 details and workarounds.
995 $db->clear(); # hashes or arrays
997 =item * lock() / unlock()
1003 Recover lost disk space. This is important to do, especially if you use
1006 =item * import() / export()
1008 Data going in and out.
1014 For hashes, DBM::Deep supports all the common methods described above, and the
1015 following additional methods: C<first_key()> and C<next_key()>.
1021 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
1022 fetched in an undefined order (which appears random). Takes no arguments,
1023 returns the key as a scalar value.
1025 my $key = $db->first_key();
1029 Returns the "next" key in the hash, given the previous one as the sole argument.
1030 Returns undef if there are no more keys to be fetched.
1032 $key = $db->next_key($key);
1036 Here are some examples of using hashes:
1038 my $db = DBM::Deep->new( "foo.db" );
1040 $db->put("foo", "bar");
1041 print "foo: " . $db->get("foo") . "\n";
1043 $db->put("baz", {}); # new child hash ref
1044 $db->get("baz")->put("buz", "biz");
1045 print "buz: " . $db->get("baz")->get("buz") . "\n";
1047 my $key = $db->first_key();
1049 print "$key: " . $db->get($key) . "\n";
1050 $key = $db->next_key($key);
1053 if ($db->exists("foo")) { $db->delete("foo"); }
1057 For arrays, DBM::Deep supports all the common methods described above, and the
1058 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
1059 C<unshift()> and C<splice()>.
1065 Returns the number of elements in the array. Takes no arguments.
1067 my $len = $db->length();
1071 Adds one or more elements onto the end of the array. Accepts scalars, hash
1072 refs or array refs. No return value.
1074 $db->push("foo", "bar", {});
1078 Fetches the last element in the array, and deletes it. Takes no arguments.
1079 Returns undef if array is empty. Returns the element value.
1081 my $elem = $db->pop();
1085 Fetches the first element in the array, deletes it, then shifts all the
1086 remaining elements over to take up the space. Returns the element value. This
1087 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
1090 my $elem = $db->shift();
1094 Inserts one or more elements onto the beginning of the array, shifting all
1095 existing elements over to make room. Accepts scalars, hash refs or array refs.
1096 No return value. This method is not recommended with large arrays -- see
1097 <LARGE ARRAYS> below for details.
1099 $db->unshift("foo", "bar", {});
1103 Performs exactly like Perl's built-in function of the same name. See L<perldoc
1104 -f splice> for usage -- it is too complicated to document here. This method is
1105 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
1109 Here are some examples of using arrays:
1111 my $db = DBM::Deep->new(
1113 type => DBM::Deep->TYPE_ARRAY
1116 $db->push("bar", "baz");
1117 $db->unshift("foo");
1120 my $len = $db->length();
1121 print "length: $len\n"; # 4
1123 for (my $k=0; $k<$len; $k++) {
1124 print "$k: " . $db->get($k) . "\n";
1127 $db->splice(1, 2, "biz", "baf");
1129 while (my $elem = shift @$db) {
1130 print "shifted: $elem\n";
1135 Enable automatic file locking by passing a true value to the C<locking>
1136 parameter when constructing your DBM::Deep object (see L<SETUP> above).
1138 my $db = DBM::Deep->new(
1143 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
1144 mode for writes, and shared mode for reads. This is required if you have
1145 multiple processes accessing the same database file, to avoid file corruption.
1146 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
1147 NFS> below for more.
1149 =head2 EXPLICIT LOCKING
1151 You can explicitly lock a database, so it remains locked for multiple
1152 transactions. This is done by calling the C<lock()> method, and passing an
1153 optional lock mode argument (defaults to exclusive mode). This is particularly
1154 useful for things like counters, where the current value needs to be fetched,
1155 then incremented, then stored again.
1158 my $counter = $db->get("counter");
1160 $db->put("counter", $counter);
1169 You can pass C<lock()> an optional argument, which specifies which mode to use
1170 (exclusive or shared). Use one of these two constants:
1171 C<DBM::Deep-E<gt>LOCK_EX> or C<DBM::Deep-E<gt>LOCK_SH>. These are passed
1172 directly to C<flock()>, and are the same as the constants defined in Perl's
1175 $db->lock( $db->LOCK_SH );
1179 =head1 IMPORTING/EXPORTING
1181 You can import existing complex structures by calling the C<import()> method,
1182 and export an entire database into an in-memory structure using the C<export()>
1183 method. Both are examined here.
1187 Say you have an existing hash with nested hashes/arrays inside it. Instead of
1188 walking the structure and adding keys/elements to the database as you go,
1189 simply pass a reference to the C<import()> method. This recursively adds
1190 everything to an existing DBM::Deep object for you. Here is an example:
1195 array1 => [ "elem0", "elem1", "elem2" ],
1197 subkey1 => "subvalue1",
1198 subkey2 => "subvalue2"
1202 my $db = DBM::Deep->new( "foo.db" );
1203 $db->import( $struct );
1205 print $db->{key1} . "\n"; # prints "value1"
1207 This recursively imports the entire C<$struct> object into C<$db>, including
1208 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
1209 keys are merged with the existing ones, replacing if they already exist.
1210 The C<import()> method can be called on any database level (not just the base
1211 level), and works with both hash and array DB types.
1213 B<Note:> Make sure your existing structure has no circular references in it.
1214 These will cause an infinite loop when importing. There are plans to fix this
1219 Calling the C<export()> method on an existing DBM::Deep object will return
1220 a reference to a new in-memory copy of the database. The export is done
1221 recursively, so all nested hashes/arrays are all exported to standard Perl
1222 objects. Here is an example:
1224 my $db = DBM::Deep->new( "foo.db" );
1226 $db->{key1} = "value1";
1227 $db->{key2} = "value2";
1229 $db->{hash1}->{subkey1} = "subvalue1";
1230 $db->{hash1}->{subkey2} = "subvalue2";
1232 my $struct = $db->export();
1234 print $struct->{key1} . "\n"; # prints "value1"
1236 This makes a complete copy of the database in memory, and returns a reference
1237 to it. The C<export()> method can be called on any database level (not just
1238 the base level), and works with both hash and array DB types. Be careful of
1239 large databases -- you can store a lot more data in a DBM::Deep object than an
1240 in-memory Perl structure.
1242 B<Note:> Make sure your database has no circular references in it.
1243 These will cause an infinite loop when exporting. There are plans to fix this
1248 DBM::Deep has a number of hooks where you can specify your own Perl function
1249 to perform filtering on incoming or outgoing data. This is a perfect
1250 way to extend the engine, and implement things like real-time compression or
1251 encryption. Filtering applies to the base DB level, and all child hashes /
1252 arrays. Filter hooks can be specified when your DBM::Deep object is first
1253 constructed, or by calling the C<set_filter()> method at any time. There are
1254 four available filter hooks, described below:
1258 =item * filter_store_key
1260 This filter is called whenever a hash key is stored. It
1261 is passed the incoming key, and expected to return a transformed key.
1263 =item * filter_store_value
1265 This filter is called whenever a hash key or array element is stored. It
1266 is passed the incoming value, and expected to return a transformed value.
1268 =item * filter_fetch_key
1270 This filter is called whenever a hash key is fetched (i.e. via
1271 C<first_key()> or C<next_key()>). It is passed the transformed key,
1272 and expected to return the plain key.
1274 =item * filter_fetch_value
1276 This filter is called whenever a hash key or array element is fetched.
1277 It is passed the transformed value, and expected to return the plain value.
1281 Here are the two ways to setup a filter hook:
1283 my $db = DBM::Deep->new(
1285 filter_store_value => \&my_filter_store,
1286 filter_fetch_value => \&my_filter_fetch
1291 $db->set_filter( "filter_store_value", \&my_filter_store );
1292 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
1294 Your filter function will be called only when dealing with SCALAR keys or
1295 values. When nested hashes and arrays are being stored/fetched, filtering
1296 is bypassed. Filters are called as static functions, passed a single SCALAR
1297 argument, and expected to return a single SCALAR value. If you want to
1298 remove a filter, set the function reference to C<undef>:
1300 $db->set_filter( "filter_store_value", undef );
1302 =head2 REAL-TIME ENCRYPTION EXAMPLE
1304 Here is a working example that uses the I<Crypt::Blowfish> module to
1305 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
1306 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
1307 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
1310 use Crypt::Blowfish;
1313 my $cipher = Crypt::CBC->new({
1314 'key' => 'my secret key',
1315 'cipher' => 'Blowfish',
1317 'regenerate_key' => 0,
1318 'padding' => 'space',
1322 my $db = DBM::Deep->new(
1323 file => "foo-encrypt.db",
1324 filter_store_key => \&my_encrypt,
1325 filter_store_value => \&my_encrypt,
1326 filter_fetch_key => \&my_decrypt,
1327 filter_fetch_value => \&my_decrypt,
1330 $db->{key1} = "value1";
1331 $db->{key2} = "value2";
1332 print "key1: " . $db->{key1} . "\n";
1333 print "key2: " . $db->{key2} . "\n";
1339 return $cipher->encrypt( $_[0] );
1342 return $cipher->decrypt( $_[0] );
1345 =head2 REAL-TIME COMPRESSION EXAMPLE
1347 Here is a working example that uses the I<Compress::Zlib> module to do real-time
1348 compression / decompression of keys & values with DBM::Deep Filters.
1349 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
1350 more on I<Compress::Zlib>.
1355 my $db = DBM::Deep->new(
1356 file => "foo-compress.db",
1357 filter_store_key => \&my_compress,
1358 filter_store_value => \&my_compress,
1359 filter_fetch_key => \&my_decompress,
1360 filter_fetch_value => \&my_decompress,
1363 $db->{key1} = "value1";
1364 $db->{key2} = "value2";
1365 print "key1: " . $db->{key1} . "\n";
1366 print "key2: " . $db->{key2} . "\n";
1372 return Compress::Zlib::memGzip( $_[0] ) ;
1375 return Compress::Zlib::memGunzip( $_[0] ) ;
1378 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
1379 actually numerical index numbers, and are not filtered.
1381 =head1 ERROR HANDLING
1383 Most DBM::Deep methods return a true value for success, and call die() on
1384 failure. You can wrap calls in an eval block to catch the die.
1386 my $db = DBM::Deep->new( "foo.db" ); # create hash
1387 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
1389 print $@; # prints error message
1391 =head1 LARGEFILE SUPPORT
1393 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
1394 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
1395 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
1396 by specifying the 'pack_size' parameter when constructing the file.
1399 filename => $filename,
1400 pack_size => 'large',
1403 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
1404 instead of 32-bit longs. After setting these values your DB files have a
1405 theoretical maximum size of 16 XB (exabytes).
1407 You can also use C<pack_size =E<gt> 'small'> in order to use 16-bit file
1410 B<Note:> Changing these values will B<NOT> work for existing database files.
1411 Only change this for new files. Once the value has been set, it is stored in
1412 the file's header and cannot be changed for the life of the file. These
1413 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
1416 B<Note:> We have not personally tested files larger than 2 GB -- all my
1417 systems have only a 32-bit Perl. However, I have received user reports that
1418 this does indeed work!
1420 =head1 LOW-LEVEL ACCESS
1422 If you require low-level access to the underlying filehandle that DBM::Deep uses,
1423 you can call the C<_fh()> method, which returns the handle:
1425 my $fh = $db->_fh();
1427 This method can be called on the root level of the datbase, or any child
1428 hashes or arrays. All levels share a I<root> structure, which contains things
1429 like the filehandle, a reference counter, and all the options specified
1430 when you created the object. You can get access to this file object by
1431 calling the C<_storage()> method.
1433 my $file_obj = $db->_storage();
1435 This is useful for changing options after the object has already been created,
1436 such as enabling/disabling locking. You can also store your own temporary user
1437 data in this structure (be wary of name collision), which is then accessible from
1438 any child hash or array.
1440 =head1 CUSTOM DIGEST ALGORITHM
1442 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
1443 keys. However you can override this, and use another algorithm (such as SHA-256)
1444 or even write your own. But please note that DBM::Deep currently expects zero
1445 collisions, so your algorithm has to be I<perfect>, so to speak. Collision
1446 detection may be introduced in a later version.
1448 You can specify a custom digest algorithm by passing it into the parameter
1449 list for new(), passing a reference to a subroutine as the 'digest' parameter,
1450 and the length of the algorithm's hashes (in bytes) as the 'hash_size'
1451 parameter. Here is a working example that uses a 256-bit hash from the
1452 I<Digest::SHA256> module. Please see
1453 L<http://search.cpan.org/search?module=Digest::SHA256> for more information.
1458 my $context = Digest::SHA256::new(256);
1460 my $db = DBM::Deep->new(
1461 filename => "foo-sha.db",
1462 digest => \&my_digest,
1466 $db->{key1} = "value1";
1467 $db->{key2} = "value2";
1468 print "key1: " . $db->{key1} . "\n";
1469 print "key2: " . $db->{key2} . "\n";
1475 return substr( $context->hash($_[0]), 0, 32 );
1478 B<Note:> Your returned digest strings must be B<EXACTLY> the number
1479 of bytes you specify in the hash_size parameter (in this case 32).
1481 B<Note:> If you do choose to use a custom digest algorithm, you must set it
1482 every time you access this file. Otherwise, the default (MD5) will be used.
1484 =head1 CIRCULAR REFERENCES
1486 DBM::Deep has B<experimental> support for circular references. Meaning you
1487 can have a nested hash key or array element that points to a parent object.
1488 This relationship is stored in the DB file, and is preserved between sessions.
1491 my $db = DBM::Deep->new( "foo.db" );
1494 $db->{circle} = $db; # ref to self
1496 print $db->{foo} . "\n"; # prints "bar"
1497 print $db->{circle}->{foo} . "\n"; # prints "bar" again
1499 B<Note>: Passing the object to a function that recursively walks the
1500 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
1501 C<export()> methods) will result in an infinite loop. This will be fixed in
1506 New in 0.99_01 is the ability to audit your databases actions. By passing in
1507 audit_file (or audit_fh) to the constructor, all actions will be logged to
1508 that file. The format is one that is suitable for eval'ing against the
1509 database to replay the actions. Please see t/33_audit_trail.t for an example
1514 New in 0.99_01 is ACID transactions. Every DBM::Deep object is completely
1515 transaction-ready - it is not an option you have to turn on. Three new methods
1516 have been added to support them. They are:
1520 =item * begin_work()
1522 This starts a transaction.
1526 This applies the changes done within the transaction to the mainline and ends
1531 This discards the changes done within the transaction to the mainline and ends
1536 Transactions in DBM::Deep are done using the MVCC method, the same method used
1537 by the InnoDB MySQL table type.
1539 =head1 CAVEATS / ISSUES / BUGS
1541 This section describes all the known issues with DBM::Deep. It you have found
1542 something that is not listed here, please send e-mail to L<jhuckaby@cpan.org>.
1544 =head2 UNUSED SPACE RECOVERY
1546 One major caveat with DBM::Deep is that space occupied by existing keys and
1547 values is not recovered when they are deleted. Meaning if you keep deleting
1548 and adding new keys, your file will continuously grow. I am working on this,
1549 but in the meantime you can call the built-in C<optimize()> method from time to
1550 time (perhaps in a crontab or something) to recover all your unused space.
1552 $db->optimize(); # returns true on success
1554 This rebuilds the ENTIRE database into a new file, then moves it on top of
1555 the original. The new file will have no unused space, thus it will take up as
1556 little disk space as possible. Please note that this operation can take
1557 a long time for large files, and you need enough disk space to temporarily hold
1558 2 copies of your DB file. The temporary file is created in the same directory
1559 as the original, named with a ".tmp" extension, and is deleted when the
1560 operation completes. Oh, and if locking is enabled, the DB is automatically
1561 locked for the entire duration of the copy.
1563 B<WARNING:> Only call optimize() on the top-level node of the database, and
1564 make sure there are no child references lying around. DBM::Deep keeps a reference
1565 counter, and if it is greater than 1, optimize() will abort and return undef.
1569 (The reasons given assume a high level of Perl understanding, specifically of
1570 references. You can safely skip this section.)
1572 Currently, the only references supported are HASH and ARRAY. The other reference
1573 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
1579 These are things like filehandles and other sockets. They can't be supported
1580 because it's completely unclear how DBM::Deep should serialize them.
1582 =item * SCALAR / REF
1584 The discussion here refers to the following type of example:
1591 # In some other process ...
1593 my $val = ${ $db->{key1} };
1595 is( $val, 50, "What actually gets stored in the DB file?" );
1597 The problem is one of synchronization. When the variable being referred to
1598 changes value, the reference isn't notified. This means that the new value won't
1599 be stored in the datafile for other processes to read. There is no TIEREF.
1601 It is theoretically possible to store references to values already within a
1602 DBM::Deep object because everything already is synchronized, but the change to
1603 the internals would be quite large. Specifically, DBM::Deep would have to tie
1604 every single value that is stored. This would bloat the RAM footprint of
1605 DBM::Deep at least twofold (if not more) and be a significant performance drain,
1606 all to support a feature that has never been requested.
1610 L<Data::Dump::Streamer/> provides a mechanism for serializing coderefs,
1611 including saving off all closure state. However, just as for SCALAR and REF,
1612 that closure state may change without notifying the DBM::Deep object storing
1617 =head2 FILE CORRUPTION
1619 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
1620 for a 32-bit signature when opened, but other corruption in files can cause
1621 segmentation faults. DBM::Deep may try to seek() past the end of a file, or get
1622 stuck in an infinite loop depending on the level of corruption. File write
1623 operations are not checked for failure (for speed), so if you happen to run
1624 out of disk space, DBM::Deep will probably fail in a bad way. These things will
1625 be addressed in a later version of DBM::Deep.
1629 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
1630 well on local filesystems, but will NOT protect you from file corruption over
1631 NFS. I've heard about setting up your NFS server with a locking daemon, then
1632 using lockf() to lock your files, but your mileage may vary there as well.
1633 From what I understand, there is no real way to do it. However, if you need
1634 access to the underlying filehandle in DBM::Deep for using some other kind of
1635 locking scheme like lockf(), see the L<LOW-LEVEL ACCESS> section above.
1637 =head2 COPYING OBJECTS
1639 Beware of copying tied objects in Perl. Very strange things can happen.
1640 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
1641 returns a new, blessed, tied hash or array to the same level in the DB.
1643 my $copy = $db->clone();
1645 B<Note>: Since clone() here is cloning the object, not the database location, any
1646 modifications to either $db or $copy will be visible to both.
1650 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1651 These functions cause every element in the array to move, which can be murder
1652 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1653 a different location. This will be addressed in the forthcoming version 1.00.
1655 =head2 WRITEONLY FILES
1657 If you pass in a filehandle to new(), you may have opened it in either a readonly or
1658 writeonly mode. STORE will verify that the filehandle is writable. However, there
1659 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
1660 filehandle isn't readable, it's not clear what will happen. So, don't do that.
1662 =head1 CODE COVERAGE
1664 B<Devel::Cover> is used to test the code coverage of the tests. Below is the
1665 B<Devel::Cover> report on this distribution's test suite.
1667 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1668 File stmt bran cond sub pod time total
1669 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1670 blib/lib/DBM/Deep.pm 96.2 89.0 75.0 95.8 89.5 36.0 92.9
1671 blib/lib/DBM/Deep/Array.pm 96.1 88.3 100.0 96.4 100.0 15.9 94.7
1672 blib/lib/DBM/Deep/Engine.pm 96.6 86.6 89.5 100.0 0.0 20.0 91.0
1673 blib/lib/DBM/Deep/File.pm 99.4 88.3 55.6 100.0 0.0 19.6 89.5
1674 blib/lib/DBM/Deep/Hash.pm 98.5 83.3 100.0 100.0 100.0 8.5 96.3
1675 Total 96.9 87.4 81.2 98.0 38.5 100.0 92.1
1676 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1678 =head1 MORE INFORMATION
1680 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1681 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1684 The source code repository is at L<http://svn.perl.org/modules/DBM-Deep>
1688 Rob Kinyon, L<rkinyon@cpan.org>
1690 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1692 Special thanks to Adam Sah and Rich Gaushell! You know why :-)
1696 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1697 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1701 Copyright (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
1702 This is free software, you may use it and distribute it under the
1703 same terms as Perl itself.