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 );
43 use DBM::Deep::Engine;
47 # Setup constants for users to pass to new()
49 sub TYPE_HASH () { DBM::Deep::Engine->SIG_HASH }
50 sub TYPE_ARRAY () { DBM::Deep::Engine->SIG_ARRAY }
58 $proto->_throw_error( "Odd number of parameters to " . (caller(1))[2] );
63 unless ( eval { local $SIG{'__DIE__'}; %{$_[0]} || 1 } ) {
64 $proto->_throw_error( "Not a hashref in args to " . (caller(1))[2] );
69 $args = { file => shift };
77 # Class constructor method for Perl OO interface.
78 # Calls tie() and returns blessed reference to tied hash or array,
79 # providing a hybrid OO/tie interface.
82 my $args = $class->_get_args( @_ );
85 # Check if we want a tied hash or array.
88 if (defined($args->{type}) && $args->{type} eq TYPE_ARRAY) {
89 $class = 'DBM::Deep::Array';
90 require DBM::Deep::Array;
91 tie @$self, $class, %$args;
94 $class = 'DBM::Deep::Hash';
95 require DBM::Deep::Hash;
96 tie %$self, $class, %$args;
99 return bless $self, $class;
102 # This initializer is called from the various TIE* methods. new() calls tie(),
103 # which allows for a single point of entry.
108 $args->{fileobj} = DBM::Deep::File->new( $args )
109 unless exists $args->{fileobj};
111 # locking implicitly enables autoflush
112 if ($args->{locking}) { $args->{autoflush} = 1; }
114 # These are the defaults to be optionally overridden below
117 base_offset => undef,
124 $self->{engine} = DBM::Deep::Engine->new( { %{$args}, obj => $self } );
126 # Grab the parameters we want to use
127 foreach my $param ( keys %$self ) {
128 next unless exists $args->{$param};
129 $self->{$param} = $args->{$param};
132 $self->_engine->setup_fh( $self );
134 $self->{fileobj}->set_db( $self );
141 require DBM::Deep::Hash;
142 return DBM::Deep::Hash->TIEHASH( @_ );
147 require DBM::Deep::Array;
148 return DBM::Deep::Array->TIEARRAY( @_ );
152 my $self = shift->_get_self;
153 return $self->_fileobj->lock( $self, @_ );
157 my $self = shift->_get_self;
158 return $self->_fileobj->unlock( $self, @_ );
162 my $self = shift->_get_self;
163 my ($spot, $value) = @_;
168 elsif ( eval { local $SIG{__DIE__}; $value->isa( 'DBM::Deep' ) } ) {
169 ${$spot} = $value->_repr;
170 $value->_copy_node( ${$spot} );
173 my $r = Scalar::Util::reftype( $value );
174 my $c = Scalar::Util::blessed( $value );
175 if ( $r eq 'ARRAY' ) {
176 ${$spot} = [ @{$value} ];
179 ${$spot} = { %{$value} };
181 ${$spot} = bless ${$spot}, $c
189 die "Must be implemented in a child class\n";
193 die "Must be implemented in a child class\n";
198 # Recursively export into standard Perl hashes and arrays.
200 my $self = shift->_get_self;
202 my $temp = $self->_repr;
205 $self->_copy_node( $temp );
208 # This will always work because $self, after _get_self() is a HASH
209 if ( $self->{parent} ) {
210 my $c = Scalar::Util::blessed(
211 $self->{parent}->get($self->{parent_key})
213 if ( $c && !$c->isa( 'DBM::Deep' ) ) {
223 # Recursively import Perl hash/array structure
225 if (!ref($_[0])) { return; } # Perl calls import() on use -- ignore
227 my $self = shift->_get_self;
230 # struct is not a reference, so just import based on our type
232 $struct = $self->_repr( @_ );
235 return $self->_import( $struct );
240 # Rebuild entire database into new file, then move
241 # it back on top of original.
243 my $self = shift->_get_self;
245 #XXX Need to create a new test for this
246 # if ($self->_fileobj->{links} > 1) {
247 # $self->_throw_error("Cannot optimize: reference count is greater than 1");
250 my $db_temp = DBM::Deep->new(
251 file => $self->_fileobj->{file} . '.tmp',
256 $self->_copy_node( $db_temp );
260 # Attempt to copy user, group and permissions over to new file
262 my @stats = stat($self->_fh);
263 my $perms = $stats[2] & 07777;
266 chown( $uid, $gid, $self->_fileobj->{file} . '.tmp' );
267 chmod( $perms, $self->_fileobj->{file} . '.tmp' );
269 # q.v. perlport for more information on this variable
270 if ( $^O eq 'MSWin32' || $^O eq 'cygwin' ) {
272 # Potential race condition when optmizing on Win32 with locking.
273 # The Windows filesystem requires that the filehandle be closed
274 # before it is overwritten with rename(). This could be redone
278 $self->_fileobj->close;
281 if (!rename $self->_fileobj->{file} . '.tmp', $self->_fileobj->{file}) {
282 unlink $self->_fileobj->{file} . '.tmp';
284 $self->_throw_error("Optimize failed: Cannot copy temp file over original: $!");
288 $self->_fileobj->close;
289 $self->_fileobj->open;
290 $self->_engine->setup_fh( $self );
297 # Make copy of object and return
299 my $self = shift->_get_self;
301 return DBM::Deep->new(
302 type => $self->_type,
303 base_offset => $self->_base_offset,
304 fileobj => $self->_fileobj,
309 my %is_legal_filter = map {
312 store_key store_value
313 fetch_key fetch_value
318 # Setup filter function for storing or fetching the key or value
320 my $self = shift->_get_self;
324 if ( $is_legal_filter{$type} ) {
325 $self->_fileobj->{"filter_$type"} = $func;
334 my $self = shift->_get_self;
335 $self->_fileobj->begin_transaction;
340 my $self = shift->_get_self;
341 $self->_fileobj->end_transaction;
346 my $self = shift->_get_self;
347 $self->_fileobj->commit_transaction;
356 my $self = $_[0]->_get_self;
357 return $self->{engine};
361 my $self = $_[0]->_get_self;
362 return $self->{fileobj};
366 my $self = $_[0]->_get_self;
367 return $self->{type};
371 my $self = $_[0]->_get_self;
372 return $self->{base_offset};
376 my $self = $_[0]->_get_self;
377 return $self->_fileobj->{fh};
385 die "DBM::Deep: $_[1]\n";
390 (O_WRONLY | O_RDWR) & fcntl( $fh, F_GETFL, my $slush = 0);
395 # (O_RDONLY | O_RDWR) & fcntl( $fh, F_GETFL, my $slush = 0);
402 #XXX This if() is redundant
403 if ( my $parent = $self->{parent} ) {
405 while ( $parent->{parent} ) {
407 $parent->_type eq TYPE_HASH
408 ? "\{q{$child->{parent_key}}\}"
409 : "\[$child->{parent_key}\]"
413 $parent = $parent->{parent};
416 $base = "\$db->get( q{$child->{parent_key}} )->" . $base;
419 $base = "\$db->get( q{$child->{parent_key}} )";
427 # Store single hash key/value or array element in database.
429 my $self = shift->_get_self;
430 my ($key, $value, $orig_key) = @_;
433 if ( $^O ne 'MSWin32' && !_is_writable( $self->_fh ) ) {
434 $self->_throw_error( 'Cannot write to a readonly filehandle' );
437 #XXX The second condition needs to disappear
438 if ( defined $orig_key && !( $self->_type eq TYPE_ARRAY && $orig_key eq 'length') ) {
441 my $r = Scalar::Util::reftype( $value ) || '';
442 if ( $r eq 'HASH' ) {
445 elsif ( $r eq 'ARRAY' ) {
448 elsif ( defined $value ) {
455 if ( my $c = Scalar::Util::blessed( $value ) ) {
456 $rhs = "bless $rhs, '$c'";
459 my $lhs = $self->_find_parent;
461 if ( $self->_type eq TYPE_HASH ) {
462 $lhs .= "->\{q{$orig_key}\}";
465 $lhs .= "->\[$orig_key\]";
471 $lhs = "\$db->put(q{$orig_key},$rhs);";
474 $self->_fileobj->audit($lhs);
478 # Request exclusive lock for writing
480 $self->lock( LOCK_EX );
482 my $md5 = $self->_engine->{digest}->($key);
484 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5, { create => 1 } );
486 # User may be storing a hash, in which case we do not want it run
487 # through the filtering system
488 if ( !ref($value) && $self->_fileobj->{filter_store_value} ) {
489 $value = $self->_fileobj->{filter_store_value}->( $value );
493 # Add key/value to bucket list
495 $self->_engine->add_bucket( $tag, $md5, $key, $value, undef, $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) = @_;
509 my $md5 = $self->_engine->{digest}->($key);
512 # Request shared lock for reading
514 $self->lock( LOCK_SH );
516 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );#, { create => 1 } );
517 #XXX This needs to autovivify
524 # Get value from bucket list
526 my $result = $self->_engine->get_bucket_value( $tag, $md5, $orig_key );
530 # Filters only apply to scalar values, so the ref check is making
531 # sure the fetched bucket is a scalar, not a child hash or array.
532 return ($result && !ref($result) && $self->_fileobj->{filter_fetch_value})
533 ? $self->_fileobj->{filter_fetch_value}->($result)
539 # Delete single key/value pair or element given plain key or array index
541 my $self = shift->_get_self;
542 my ($key, $orig_key) = @_;
544 if ( $^O ne 'MSWin32' && !_is_writable( $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->_fileobj->audit( "delete $lhs;" );
554 $self->_fileobj->audit( "\$db->delete('$orig_key');" );
559 # Request exclusive lock for writing
561 $self->lock( LOCK_EX );
563 my $md5 = $self->_engine->{digest}->($key);
565 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
574 my $value = $self->_engine->get_bucket_value( $tag, $md5 );
576 if (defined $value && !ref($value) && $self->_fileobj->{filter_fetch_value}) {
577 $value = $self->_fileobj->{filter_fetch_value}->($value);
580 my $result = $self->_engine->delete_bucket( $tag, $md5, $orig_key );
583 # If this object is an array and the key deleted was on the end of the stack,
584 # decrement the length variable.
594 # Check if a single key or element exists given plain key or array index
596 my $self = shift->_get_self;
599 my $md5 = $self->_engine->{digest}->($key);
602 # Request shared lock for reading
604 $self->lock( LOCK_SH );
606 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
611 # For some reason, the built-in exists() function returns '' for false
617 # Check if bucket exists and return 1 or ''
619 my $result = $self->_engine->bucket_exists( $tag, $md5 ) || '';
628 # Clear all keys from hash, or all elements from array.
630 my $self = shift->_get_self;
632 if ( $^O ne 'MSWin32' && !_is_writable( $self->_fh ) ) {
633 $self->_throw_error( 'Cannot write to a readonly filehandle' );
637 my $lhs = $self->_find_parent;
639 if ( $self->_type eq TYPE_HASH ) {
640 $lhs = '%{' . $lhs . '}';
643 $lhs = '@{' . $lhs . '}';
646 $self->_fileobj->audit( "$lhs = ();" );
650 # Request exclusive lock for writing
652 $self->lock( LOCK_EX );
654 if ( $self->_type eq TYPE_HASH ) {
655 my $key = $self->first_key;
657 my $next_key = $self->next_key( $key );
658 my $md5 = $self->_engine->{digest}->($key);
659 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
660 $self->_engine->delete_bucket( $tag, $md5, $key );
665 my $size = $self->FETCHSIZE;
666 for my $key ( map { pack ( $self->_engine->{long_pack}, $_ ) } 0 .. $size - 1 ) {
667 my $md5 = $self->_engine->{digest}->($key);
668 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
669 $self->_engine->delete_bucket( $tag, $md5, $key );
671 $self->STORESIZE( 0 );
673 #XXX This needs updating to use _release_space
674 # $self->_engine->write_tag(
675 # $self->_base_offset, $self->_type,
676 # chr(0)x$self->_engine->{index_size},
685 # Public method aliases
687 sub put { (shift)->STORE( @_ ) }
688 sub store { (shift)->STORE( @_ ) }
689 sub get { (shift)->FETCH( @_ ) }
690 sub fetch { (shift)->FETCH( @_ ) }
691 sub delete { (shift)->DELETE( @_ ) }
692 sub exists { (shift)->EXISTS( @_ ) }
693 sub clear { (shift)->CLEAR( @_ ) }
700 DBM::Deep - A pure perl multi-level hash/array DBM
705 my $db = DBM::Deep->new( "foo.db" );
707 $db->{key} = 'value';
710 $db->put('key' => 'value');
711 print $db->get('key');
713 # true multi-level support
714 $db->{my_complex} = [
715 'hello', { perl => 'rules' },
719 tie my %db, 'DBM::Deep', 'foo.db';
723 tied(%db)->put('key' => 'value');
724 print tied(%db)->get('key');
728 A unique flat-file database module, written in pure perl. True multi-level
729 hash/array support (unlike MLDBM, which is faked), hybrid OO / tie()
730 interface, cross-platform FTPable files, ACID transactions, and is quite fast.
731 Can handle millions of keys and unlimited levels without significant
732 slow-down. Written from the ground-up in pure perl -- this is NOT a wrapper
733 around a C-based DBM. Out-of-the-box compatibility with Unix, Mac OS X and
736 =head1 VERSION DIFFERENCES
738 B<NOTE>: 0.99_01 and above have significant file format differences from 0.983 and
739 before. There will be a backwards-compatibility layer in 1.00, but that is
740 slated for a later 0.99_x release. This version is B<NOT> backwards compatible
741 with 0.983 and before.
745 Construction can be done OO-style (which is the recommended way), or using
746 Perl's tie() function. Both are examined here.
748 =head2 OO CONSTRUCTION
750 The recommended way to construct a DBM::Deep object is to use the new()
751 method, which gets you a blessed I<and> tied hash (or array) reference.
753 my $db = DBM::Deep->new( "foo.db" );
755 This opens a new database handle, mapped to the file "foo.db". If this
756 file does not exist, it will automatically be created. DB files are
757 opened in "r+" (read/write) mode, and the type of object returned is a
758 hash, unless otherwise specified (see L<OPTIONS> below).
760 You can pass a number of options to the constructor to specify things like
761 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
763 my $db = DBM::Deep->new(
769 Notice that the filename is now specified I<inside> the hash with
770 the "file" parameter, as opposed to being the sole argument to the
771 constructor. This is required if any options are specified.
772 See L<OPTIONS> below for the complete list.
774 You can also start with an array instead of a hash. For this, you must
775 specify the C<type> parameter:
777 my $db = DBM::Deep->new(
779 type => DBM::Deep->TYPE_ARRAY
782 B<Note:> Specifing the C<type> parameter only takes effect when beginning
783 a new DB file. If you create a DBM::Deep object with an existing file, the
784 C<type> will be loaded from the file header, and an error will be thrown if
785 the wrong type is passed in.
787 =head2 TIE CONSTRUCTION
789 Alternately, you can create a DBM::Deep handle by using Perl's built-in
790 tie() function. The object returned from tie() can be used to call methods,
791 such as lock() and unlock(). (That object can be retrieved from the tied
792 variable at any time using tied() - please see L<perltie/> for more info.
795 my $db = tie %hash, "DBM::Deep", "foo.db";
798 my $db = tie @array, "DBM::Deep", "bar.db";
800 As with the OO constructor, you can replace the DB filename parameter with
801 a hash containing one or more options (see L<OPTIONS> just below for the
804 tie %hash, "DBM::Deep", {
812 There are a number of options that can be passed in when constructing your
813 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
819 Filename of the DB file to link the handle to. You can pass a full absolute
820 filesystem path, partial path, or a plain filename if the file is in the
821 current working directory. This is a required parameter (though q.v. fh).
825 If you want, you can pass in the fh instead of the file. This is most useful for doing
828 my $db = DBM::Deep->new( { fh => \*DATA } );
830 You are responsible for making sure that the fh has been opened appropriately for your
831 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
832 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
833 needs to read from the fh.
835 =item * audit_file / audit_fh
837 These are just like file/fh, except for auditing. Please see L</AUDITING> for
842 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
843 not need to set this. However, it's there if you want it.
845 If you pass in fh and do not set this, it will be set appropriately.
849 This parameter specifies what type of object to create, a hash or array. Use
850 one of these two constants:
854 =item * C<DBM::Deep-E<gt>TYPE_HASH>
856 =item * C<DBM::Deep-E<gt>TYPE_ARRAY>.
860 This only takes effect when beginning a new file. This is an optional
861 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
865 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
866 function to lock the database in exclusive mode for writes, and shared mode
867 for reads. Pass any true value to enable. This affects the base DB handle
868 I<and any child hashes or arrays> that use the same DB file. This is an
869 optional parameter, and defaults to 0 (disabled). See L<LOCKING> below for
874 Specifies whether autoflush is to be enabled on the underlying filehandle.
875 This obviously slows down write operations, but is required if you may have
876 multiple processes accessing the same DB file (also consider enable I<locking>).
877 Pass any true value to enable. This is an optional parameter, and defaults to 0
882 If I<autobless> mode is enabled, DBM::Deep will preserve the class something
883 is blessed into, and restores it when fetched. This is an optional parameter, and defaults to 1 (enabled).
885 B<Note:> If you use the OO-interface, you will not be able to call any methods
886 of DBM::Deep on the blessed item. This is considered to be a feature.
890 See L</FILTERS> below.
896 With DBM::Deep you can access your databases using Perl's standard hash/array
897 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
898 treat them as such. DBM::Deep will intercept all reads/writes and direct them
899 to the right place -- the DB file. This has nothing to do with the
900 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
901 using regular hashes and arrays, rather than calling functions like C<get()>
902 and C<put()> (although those work too). It is entirely up to you how to want
903 to access your databases.
907 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
908 or even nested hashes (or arrays) using standard Perl syntax:
910 my $db = DBM::Deep->new( "foo.db" );
912 $db->{mykey} = "myvalue";
914 $db->{myhash}->{subkey} = "subvalue";
916 print $db->{myhash}->{subkey} . "\n";
918 You can even step through hash keys using the normal Perl C<keys()> function:
920 foreach my $key (keys %$db) {
921 print "$key: " . $db->{$key} . "\n";
924 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
925 pushes them onto an array, all before the loop even begins. If you have an
926 extremely large hash, this may exhaust Perl's memory. Instead, consider using
927 Perl's C<each()> function, which pulls keys/values one at a time, using very
930 while (my ($key, $value) = each %$db) {
931 print "$key: $value\n";
934 Please note that when using C<each()>, you should always pass a direct
935 hash reference, not a lookup. Meaning, you should B<never> do this:
938 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
940 This causes an infinite loop, because for each iteration, Perl is calling
941 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
942 it effectively keeps returning the first key over and over again. Instead,
943 assign a temporary variable to C<$db->{foo}>, then pass that to each().
947 As with hashes, you can treat any DBM::Deep object like a normal Perl array
948 reference. This includes inserting, removing and manipulating elements,
949 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
950 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
951 or simply be a nested array reference inside a hash. Example:
953 my $db = DBM::Deep->new(
954 file => "foo-array.db",
955 type => DBM::Deep->TYPE_ARRAY
959 push @$db, "bar", "baz";
962 my $last_elem = pop @$db; # baz
963 my $first_elem = shift @$db; # bah
964 my $second_elem = $db->[1]; # bar
966 my $num_elements = scalar @$db;
970 In addition to the I<tie()> interface, you can also use a standard OO interface
971 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
972 array) has its own methods, but both types share the following common methods:
973 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
974 C<store(> are aliases to C<put()> and C<get()>, respectively.
978 =item * new() / clone()
980 These are the constructor and copy-functions.
982 =item * put() / store()
984 Stores a new hash key/value pair, or sets an array element value. Takes two
985 arguments, the hash key or array index, and the new value. The value can be
986 a scalar, hash ref or array ref. Returns true on success, false on failure.
988 $db->put("foo", "bar"); # for hashes
989 $db->put(1, "bar"); # for arrays
991 =item * get() / fetch()
993 Fetches the value of a hash key or array element. Takes one argument: the hash
994 key or array index. Returns a scalar, hash ref or array ref, depending on the
997 my $value = $db->get("foo"); # for hashes
998 my $value = $db->get(1); # for arrays
1002 Checks if a hash key or array index exists. Takes one argument: the hash key
1003 or array index. Returns true if it exists, false if not.
1005 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
1006 if ($db->exists(1)) { print "yay!\n"; } # for arrays
1010 Deletes one hash key/value pair or array element. Takes one argument: the hash
1011 key or array index. Returns true on success, false if not found. For arrays,
1012 the remaining elements located after the deleted element are NOT moved over.
1013 The deleted element is essentially just undefined, which is exactly how Perl's
1014 internal arrays work. Please note that the space occupied by the deleted
1015 key/value or element is B<not> reused again -- see L<UNUSED SPACE RECOVERY>
1016 below for details and workarounds.
1018 $db->delete("foo"); # for hashes
1019 $db->delete(1); # for arrays
1023 Deletes B<all> hash keys or array elements. Takes no arguments. No return
1024 value. Please note that the space occupied by the deleted keys/values or
1025 elements is B<not> reused again -- see L<UNUSED SPACE RECOVERY> below for
1026 details and workarounds.
1028 $db->clear(); # hashes or arrays
1030 =item * lock() / unlock()
1036 Recover lost disk space. This is important to do, especially if you use
1039 =item * import() / export()
1041 Data going in and out.
1047 For hashes, DBM::Deep supports all the common methods described above, and the
1048 following additional methods: C<first_key()> and C<next_key()>.
1054 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
1055 fetched in an undefined order (which appears random). Takes no arguments,
1056 returns the key as a scalar value.
1058 my $key = $db->first_key();
1062 Returns the "next" key in the hash, given the previous one as the sole argument.
1063 Returns undef if there are no more keys to be fetched.
1065 $key = $db->next_key($key);
1069 Here are some examples of using hashes:
1071 my $db = DBM::Deep->new( "foo.db" );
1073 $db->put("foo", "bar");
1074 print "foo: " . $db->get("foo") . "\n";
1076 $db->put("baz", {}); # new child hash ref
1077 $db->get("baz")->put("buz", "biz");
1078 print "buz: " . $db->get("baz")->get("buz") . "\n";
1080 my $key = $db->first_key();
1082 print "$key: " . $db->get($key) . "\n";
1083 $key = $db->next_key($key);
1086 if ($db->exists("foo")) { $db->delete("foo"); }
1090 For arrays, DBM::Deep supports all the common methods described above, and the
1091 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
1092 C<unshift()> and C<splice()>.
1098 Returns the number of elements in the array. Takes no arguments.
1100 my $len = $db->length();
1104 Adds one or more elements onto the end of the array. Accepts scalars, hash
1105 refs or array refs. No return value.
1107 $db->push("foo", "bar", {});
1111 Fetches the last element in the array, and deletes it. Takes no arguments.
1112 Returns undef if array is empty. Returns the element value.
1114 my $elem = $db->pop();
1118 Fetches the first element in the array, deletes it, then shifts all the
1119 remaining elements over to take up the space. Returns the element value. This
1120 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
1123 my $elem = $db->shift();
1127 Inserts one or more elements onto the beginning of the array, shifting all
1128 existing elements over to make room. Accepts scalars, hash refs or array refs.
1129 No return value. This method is not recommended with large arrays -- see
1130 <LARGE ARRAYS> below for details.
1132 $db->unshift("foo", "bar", {});
1136 Performs exactly like Perl's built-in function of the same name. See L<perldoc
1137 -f splice> for usage -- it is too complicated to document here. This method is
1138 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
1142 Here are some examples of using arrays:
1144 my $db = DBM::Deep->new(
1146 type => DBM::Deep->TYPE_ARRAY
1149 $db->push("bar", "baz");
1150 $db->unshift("foo");
1153 my $len = $db->length();
1154 print "length: $len\n"; # 4
1156 for (my $k=0; $k<$len; $k++) {
1157 print "$k: " . $db->get($k) . "\n";
1160 $db->splice(1, 2, "biz", "baf");
1162 while (my $elem = shift @$db) {
1163 print "shifted: $elem\n";
1168 Enable automatic file locking by passing a true value to the C<locking>
1169 parameter when constructing your DBM::Deep object (see L<SETUP> above).
1171 my $db = DBM::Deep->new(
1176 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
1177 mode for writes, and shared mode for reads. This is required if you have
1178 multiple processes accessing the same database file, to avoid file corruption.
1179 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
1180 NFS> below for more.
1182 =head2 EXPLICIT LOCKING
1184 You can explicitly lock a database, so it remains locked for multiple
1185 transactions. This is done by calling the C<lock()> method, and passing an
1186 optional lock mode argument (defaults to exclusive mode). This is particularly
1187 useful for things like counters, where the current value needs to be fetched,
1188 then incremented, then stored again.
1191 my $counter = $db->get("counter");
1193 $db->put("counter", $counter);
1202 You can pass C<lock()> an optional argument, which specifies which mode to use
1203 (exclusive or shared). Use one of these two constants:
1204 C<DBM::Deep-E<gt>LOCK_EX> or C<DBM::Deep-E<gt>LOCK_SH>. These are passed
1205 directly to C<flock()>, and are the same as the constants defined in Perl's
1208 $db->lock( $db->LOCK_SH );
1212 =head1 IMPORTING/EXPORTING
1214 You can import existing complex structures by calling the C<import()> method,
1215 and export an entire database into an in-memory structure using the C<export()>
1216 method. Both are examined here.
1220 Say you have an existing hash with nested hashes/arrays inside it. Instead of
1221 walking the structure and adding keys/elements to the database as you go,
1222 simply pass a reference to the C<import()> method. This recursively adds
1223 everything to an existing DBM::Deep object for you. Here is an example:
1228 array1 => [ "elem0", "elem1", "elem2" ],
1230 subkey1 => "subvalue1",
1231 subkey2 => "subvalue2"
1235 my $db = DBM::Deep->new( "foo.db" );
1236 $db->import( $struct );
1238 print $db->{key1} . "\n"; # prints "value1"
1240 This recursively imports the entire C<$struct> object into C<$db>, including
1241 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
1242 keys are merged with the existing ones, replacing if they already exist.
1243 The C<import()> method can be called on any database level (not just the base
1244 level), and works with both hash and array DB types.
1246 B<Note:> Make sure your existing structure has no circular references in it.
1247 These will cause an infinite loop when importing. There are plans to fix this
1252 Calling the C<export()> method on an existing DBM::Deep object will return
1253 a reference to a new in-memory copy of the database. The export is done
1254 recursively, so all nested hashes/arrays are all exported to standard Perl
1255 objects. Here is an example:
1257 my $db = DBM::Deep->new( "foo.db" );
1259 $db->{key1} = "value1";
1260 $db->{key2} = "value2";
1262 $db->{hash1}->{subkey1} = "subvalue1";
1263 $db->{hash1}->{subkey2} = "subvalue2";
1265 my $struct = $db->export();
1267 print $struct->{key1} . "\n"; # prints "value1"
1269 This makes a complete copy of the database in memory, and returns a reference
1270 to it. The C<export()> method can be called on any database level (not just
1271 the base level), and works with both hash and array DB types. Be careful of
1272 large databases -- you can store a lot more data in a DBM::Deep object than an
1273 in-memory Perl structure.
1275 B<Note:> Make sure your database has no circular references in it.
1276 These will cause an infinite loop when exporting. There are plans to fix this
1281 DBM::Deep has a number of hooks where you can specify your own Perl function
1282 to perform filtering on incoming or outgoing data. This is a perfect
1283 way to extend the engine, and implement things like real-time compression or
1284 encryption. Filtering applies to the base DB level, and all child hashes /
1285 arrays. Filter hooks can be specified when your DBM::Deep object is first
1286 constructed, or by calling the C<set_filter()> method at any time. There are
1287 four available filter hooks, described below:
1291 =item * filter_store_key
1293 This filter is called whenever a hash key is stored. It
1294 is passed the incoming key, and expected to return a transformed key.
1296 =item * filter_store_value
1298 This filter is called whenever a hash key or array element is stored. It
1299 is passed the incoming value, and expected to return a transformed value.
1301 =item * filter_fetch_key
1303 This filter is called whenever a hash key is fetched (i.e. via
1304 C<first_key()> or C<next_key()>). It is passed the transformed key,
1305 and expected to return the plain key.
1307 =item * filter_fetch_value
1309 This filter is called whenever a hash key or array element is fetched.
1310 It is passed the transformed value, and expected to return the plain value.
1314 Here are the two ways to setup a filter hook:
1316 my $db = DBM::Deep->new(
1318 filter_store_value => \&my_filter_store,
1319 filter_fetch_value => \&my_filter_fetch
1324 $db->set_filter( "filter_store_value", \&my_filter_store );
1325 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
1327 Your filter function will be called only when dealing with SCALAR keys or
1328 values. When nested hashes and arrays are being stored/fetched, filtering
1329 is bypassed. Filters are called as static functions, passed a single SCALAR
1330 argument, and expected to return a single SCALAR value. If you want to
1331 remove a filter, set the function reference to C<undef>:
1333 $db->set_filter( "filter_store_value", undef );
1335 =head2 REAL-TIME ENCRYPTION EXAMPLE
1337 Here is a working example that uses the I<Crypt::Blowfish> module to
1338 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
1339 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
1340 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
1343 use Crypt::Blowfish;
1346 my $cipher = Crypt::CBC->new({
1347 'key' => 'my secret key',
1348 'cipher' => 'Blowfish',
1350 'regenerate_key' => 0,
1351 'padding' => 'space',
1355 my $db = DBM::Deep->new(
1356 file => "foo-encrypt.db",
1357 filter_store_key => \&my_encrypt,
1358 filter_store_value => \&my_encrypt,
1359 filter_fetch_key => \&my_decrypt,
1360 filter_fetch_value => \&my_decrypt,
1363 $db->{key1} = "value1";
1364 $db->{key2} = "value2";
1365 print "key1: " . $db->{key1} . "\n";
1366 print "key2: " . $db->{key2} . "\n";
1372 return $cipher->encrypt( $_[0] );
1375 return $cipher->decrypt( $_[0] );
1378 =head2 REAL-TIME COMPRESSION EXAMPLE
1380 Here is a working example that uses the I<Compress::Zlib> module to do real-time
1381 compression / decompression of keys & values with DBM::Deep Filters.
1382 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
1383 more on I<Compress::Zlib>.
1388 my $db = DBM::Deep->new(
1389 file => "foo-compress.db",
1390 filter_store_key => \&my_compress,
1391 filter_store_value => \&my_compress,
1392 filter_fetch_key => \&my_decompress,
1393 filter_fetch_value => \&my_decompress,
1396 $db->{key1} = "value1";
1397 $db->{key2} = "value2";
1398 print "key1: " . $db->{key1} . "\n";
1399 print "key2: " . $db->{key2} . "\n";
1405 return Compress::Zlib::memGzip( $_[0] ) ;
1408 return Compress::Zlib::memGunzip( $_[0] ) ;
1411 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
1412 actually numerical index numbers, and are not filtered.
1414 =head1 ERROR HANDLING
1416 Most DBM::Deep methods return a true value for success, and call die() on
1417 failure. You can wrap calls in an eval block to catch the die.
1419 my $db = DBM::Deep->new( "foo.db" ); # create hash
1420 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
1422 print $@; # prints error message
1424 =head1 LARGEFILE SUPPORT
1426 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
1427 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
1428 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
1429 by specifying the 'pack_size' parameter when constructing the file.
1432 filename => $filename,
1433 pack_size => 'large',
1436 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
1437 instead of 32-bit longs. After setting these values your DB files have a
1438 theoretical maximum size of 16 XB (exabytes).
1440 You can also use C<pack_size =E<gt> 'small'> in order to use 16-bit file
1443 B<Note:> Changing these values will B<NOT> work for existing database files.
1444 Only change this for new files. Once the value has been set, it is stored in
1445 the file's header and cannot be changed for the life of the file. These
1446 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
1449 B<Note:> We have not personally tested files larger than 2 GB -- all my
1450 systems have only a 32-bit Perl. However, I have received user reports that
1451 this does indeed work!
1453 =head1 LOW-LEVEL ACCESS
1455 If you require low-level access to the underlying filehandle that DBM::Deep uses,
1456 you can call the C<_fh()> method, which returns the handle:
1458 my $fh = $db->_fh();
1460 This method can be called on the root level of the datbase, or any child
1461 hashes or arrays. All levels share a I<root> structure, which contains things
1462 like the filehandle, a reference counter, and all the options specified
1463 when you created the object. You can get access to this file object by
1464 calling the C<_fileobj()> method.
1466 my $file_obj = $db->_fileobj();
1468 This is useful for changing options after the object has already been created,
1469 such as enabling/disabling locking. You can also store your own temporary user
1470 data in this structure (be wary of name collision), which is then accessible from
1471 any child hash or array.
1473 =head1 CUSTOM DIGEST ALGORITHM
1475 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
1476 keys. However you can override this, and use another algorithm (such as SHA-256)
1477 or even write your own. But please note that DBM::Deep currently expects zero
1478 collisions, so your algorithm has to be I<perfect>, so to speak. Collision
1479 detection may be introduced in a later version.
1481 You can specify a custom digest algorithm by passing it into the parameter
1482 list for new(), passing a reference to a subroutine as the 'digest' parameter,
1483 and the length of the algorithm's hashes (in bytes) as the 'hash_size'
1484 parameter. Here is a working example that uses a 256-bit hash from the
1485 I<Digest::SHA256> module. Please see
1486 L<http://search.cpan.org/search?module=Digest::SHA256> for more information.
1491 my $context = Digest::SHA256::new(256);
1493 my $db = DBM::Deep->new(
1494 filename => "foo-sha.db",
1495 digest => \&my_digest,
1499 $db->{key1} = "value1";
1500 $db->{key2} = "value2";
1501 print "key1: " . $db->{key1} . "\n";
1502 print "key2: " . $db->{key2} . "\n";
1508 return substr( $context->hash($_[0]), 0, 32 );
1511 B<Note:> Your returned digest strings must be B<EXACTLY> the number
1512 of bytes you specify in the hash_size parameter (in this case 32).
1514 B<Note:> If you do choose to use a custom digest algorithm, you must set it
1515 every time you access this file. Otherwise, the default (MD5) will be used.
1517 =head1 CIRCULAR REFERENCES
1519 DBM::Deep has B<experimental> support for circular references. Meaning you
1520 can have a nested hash key or array element that points to a parent object.
1521 This relationship is stored in the DB file, and is preserved between sessions.
1524 my $db = DBM::Deep->new( "foo.db" );
1527 $db->{circle} = $db; # ref to self
1529 print $db->{foo} . "\n"; # prints "bar"
1530 print $db->{circle}->{foo} . "\n"; # prints "bar" again
1532 B<Note>: Passing the object to a function that recursively walks the
1533 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
1534 C<export()> methods) will result in an infinite loop. This will be fixed in
1539 New in 0.99_01 is the ability to audit your databases actions. By passing in
1540 audit_file (or audit_fh) to the constructor, all actions will be logged to
1541 that file. The format is one that is suitable for eval'ing against the
1542 database to replay the actions. Please see t/33_audit_trail.t for an example
1547 New in 0.99_01 is ACID transactions. Every DBM::Deep object is completely
1548 transaction-ready - it is not an option you have to turn on. Three new methods
1549 have been added to support them. They are:
1553 =item * begin_work()
1555 This starts a transaction.
1559 This applies the changes done within the transaction to the mainline and ends
1564 This discards the changes done within the transaction to the mainline and ends
1569 Transactions in DBM::Deep are done using the MVCC method, the same method used
1570 by the InnoDB MySQL table type.
1572 =head1 CAVEATS / ISSUES / BUGS
1574 This section describes all the known issues with DBM::Deep. It you have found
1575 something that is not listed here, please send e-mail to L<jhuckaby@cpan.org>.
1577 =head2 UNUSED SPACE RECOVERY
1579 One major caveat with DBM::Deep is that space occupied by existing keys and
1580 values is not recovered when they are deleted. Meaning if you keep deleting
1581 and adding new keys, your file will continuously grow. I am working on this,
1582 but in the meantime you can call the built-in C<optimize()> method from time to
1583 time (perhaps in a crontab or something) to recover all your unused space.
1585 $db->optimize(); # returns true on success
1587 This rebuilds the ENTIRE database into a new file, then moves it on top of
1588 the original. The new file will have no unused space, thus it will take up as
1589 little disk space as possible. Please note that this operation can take
1590 a long time for large files, and you need enough disk space to temporarily hold
1591 2 copies of your DB file. The temporary file is created in the same directory
1592 as the original, named with a ".tmp" extension, and is deleted when the
1593 operation completes. Oh, and if locking is enabled, the DB is automatically
1594 locked for the entire duration of the copy.
1596 B<WARNING:> Only call optimize() on the top-level node of the database, and
1597 make sure there are no child references lying around. DBM::Deep keeps a reference
1598 counter, and if it is greater than 1, optimize() will abort and return undef.
1602 (The reasons given assume a high level of Perl understanding, specifically of
1603 references. You can safely skip this section.)
1605 Currently, the only references supported are HASH and ARRAY. The other reference
1606 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
1612 These are things like filehandles and other sockets. They can't be supported
1613 because it's completely unclear how DBM::Deep should serialize them.
1615 =item * SCALAR / REF
1617 The discussion here refers to the following type of example:
1624 # In some other process ...
1626 my $val = ${ $db->{key1} };
1628 is( $val, 50, "What actually gets stored in the DB file?" );
1630 The problem is one of synchronization. When the variable being referred to
1631 changes value, the reference isn't notified. This means that the new value won't
1632 be stored in the datafile for other processes to read. There is no TIEREF.
1634 It is theoretically possible to store references to values already within a
1635 DBM::Deep object because everything already is synchronized, but the change to
1636 the internals would be quite large. Specifically, DBM::Deep would have to tie
1637 every single value that is stored. This would bloat the RAM footprint of
1638 DBM::Deep at least twofold (if not more) and be a significant performance drain,
1639 all to support a feature that has never been requested.
1643 L<Data::Dump::Streamer/> provides a mechanism for serializing coderefs,
1644 including saving off all closure state. However, just as for SCALAR and REF,
1645 that closure state may change without notifying the DBM::Deep object storing
1650 =head2 FILE CORRUPTION
1652 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
1653 for a 32-bit signature when opened, but other corruption in files can cause
1654 segmentation faults. DBM::Deep may try to seek() past the end of a file, or get
1655 stuck in an infinite loop depending on the level of corruption. File write
1656 operations are not checked for failure (for speed), so if you happen to run
1657 out of disk space, DBM::Deep will probably fail in a bad way. These things will
1658 be addressed in a later version of DBM::Deep.
1662 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
1663 well on local filesystems, but will NOT protect you from file corruption over
1664 NFS. I've heard about setting up your NFS server with a locking daemon, then
1665 using lockf() to lock your files, but your mileage may vary there as well.
1666 From what I understand, there is no real way to do it. However, if you need
1667 access to the underlying filehandle in DBM::Deep for using some other kind of
1668 locking scheme like lockf(), see the L<LOW-LEVEL ACCESS> section above.
1670 =head2 COPYING OBJECTS
1672 Beware of copying tied objects in Perl. Very strange things can happen.
1673 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
1674 returns a new, blessed, tied hash or array to the same level in the DB.
1676 my $copy = $db->clone();
1678 B<Note>: Since clone() here is cloning the object, not the database location, any
1679 modifications to either $db or $copy will be visible to both.
1683 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1684 These functions cause every element in the array to move, which can be murder
1685 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1686 a different location. This will be addressed in the forthcoming version 1.00.
1688 =head2 WRITEONLY FILES
1690 If you pass in a filehandle to new(), you may have opened it in either a readonly or
1691 writeonly mode. STORE will verify that the filehandle is writable. However, there
1692 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
1693 filehandle isn't readable, it's not clear what will happen. So, don't do that.
1695 =head1 CODE COVERAGE
1697 B<Devel::Cover> is used to test the code coverage of the tests. Below is the
1698 B<Devel::Cover> report on this distribution's test suite.
1700 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1701 File stmt bran cond sub pod time total
1702 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1703 blib/lib/DBM/Deep.pm 96.2 89.0 75.0 95.8 89.5 36.0 92.9
1704 blib/lib/DBM/Deep/Array.pm 96.1 88.3 100.0 96.4 100.0 15.9 94.7
1705 blib/lib/DBM/Deep/Engine.pm 96.6 86.6 89.5 100.0 0.0 20.0 91.0
1706 blib/lib/DBM/Deep/File.pm 99.4 88.3 55.6 100.0 0.0 19.6 89.5
1707 blib/lib/DBM/Deep/Hash.pm 98.5 83.3 100.0 100.0 100.0 8.5 96.3
1708 Total 96.9 87.4 81.2 98.0 38.5 100.0 92.1
1709 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1711 =head1 MORE INFORMATION
1713 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1714 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1717 The source code repository is at L<http://svn.perl.org/modules/DBM-Deep>
1721 Rob Kinyon, L<rkinyon@cpan.org>
1723 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1725 Special thanks to Adam Sah and Rich Gaushell! You know why :-)
1729 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1730 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1734 Copyright (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
1735 This is free software, you may use it and distribute it under the
1736 same terms as Perl itself.