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->{fileobj} = DBM::Deep::File->new( $args )
112 unless exists $args->{fileobj};
114 # locking implicitly enables autoflush
115 if ($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->_fileobj->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->_fileobj->lock( $self, @_ );
160 my $self = shift->_get_self;
161 return $self->_fileobj->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->_fileobj->{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->_fileobj->{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->_fileobj->{file} . '.tmp' );
285 chmod( $perms, $self->_fileobj->{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->_fileobj->close;
299 if (!rename $self->_fileobj->{file} . '.tmp', $self->_fileobj->{file}) {
300 unlink $self->_fileobj->{file} . '.tmp';
302 $self->_throw_error("Optimize failed: Cannot copy temp file over original: $!");
306 $self->_fileobj->close;
307 $self->_fileobj->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 fileobj => $self->_fileobj,
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 $self->_fileobj->{"filter_$type"} = $func;
354 my $self = shift->_get_self;
355 return $self->_fileobj->begin_transaction;
359 my $self = shift->_get_self;
360 return $self->_fileobj->end_transaction;
364 my $self = shift->_get_self;
365 return $self->_fileobj->commit_transaction;
373 my $self = $_[0]->_get_self;
374 return $self->{engine};
378 my $self = $_[0]->_get_self;
379 return $self->{fileobj};
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->_fileobj->{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 $self->_fileobj->audit($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->_fileobj->{filter_store_value} ) {
492 $value = $self->_fileobj->{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;
510 my $md5 = $self->_engine->apply_digest($key);
513 # Request shared lock for reading
515 $self->lock( LOCK_SH );
517 my $result = $self->_engine->read_value( $self->_base_offset, $key, $orig_key );
521 # Filters only apply to scalar values, so the ref check is making
522 # sure the fetched bucket is a scalar, not a child hash or array.
523 return ($result && !ref($result) && $self->_fileobj->{filter_fetch_value})
524 ? $self->_fileobj->{filter_fetch_value}->($result)
530 # Delete single key/value pair or element given plain key or array index
532 my $self = shift->_get_self;
533 my ($key, $orig_key) = @_;
534 $orig_key = $key unless defined $orig_key;
536 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
537 $self->_throw_error( 'Cannot write to a readonly filehandle' );
540 if ( defined $orig_key ) {
541 my $lhs = $self->_find_parent;
543 $self->_fileobj->audit( "delete $lhs;" );
546 $self->_fileobj->audit( "\$db->delete('$orig_key');" );
551 # Request exclusive lock for writing
553 $self->lock( LOCK_EX );
555 my $md5 = $self->_engine->apply_digest($key);
557 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
566 my $value = $self->_engine->get_bucket_value( $tag, $md5 );
568 if (defined $value && !ref($value) && $self->_fileobj->{filter_fetch_value}) {
569 $value = $self->_fileobj->{filter_fetch_value}->($value);
572 my $result = $self->_engine->delete_bucket( $tag, $md5, $orig_key );
575 # If this object is an array and the key deleted was on the end of the stack,
576 # decrement the length variable.
586 # Check if a single key or element exists given plain key or array index
588 my $self = shift->_get_self;
591 my $md5 = $self->_engine->apply_digest($key);
594 # Request shared lock for reading
596 $self->lock( LOCK_SH );
598 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
603 # For some reason, the built-in exists() function returns '' for false
609 # Check if bucket exists and return 1 or ''
611 my $result = $self->_engine->bucket_exists( $tag, $md5 ) || '';
620 # Clear all keys from hash, or all elements from array.
622 my $self = shift->_get_self;
624 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
625 $self->_throw_error( 'Cannot write to a readonly filehandle' );
629 my $lhs = $self->_find_parent;
631 if ( $self->_type eq TYPE_HASH ) {
632 $lhs = '%{' . $lhs . '}';
635 $lhs = '@{' . $lhs . '}';
638 $self->_fileobj->audit( "$lhs = ();" );
642 # Request exclusive lock for writing
644 $self->lock( LOCK_EX );
646 if ( $self->_type eq TYPE_HASH ) {
647 my $key = $self->first_key;
649 my $next_key = $self->next_key( $key );
650 my $md5 = $self->_engine->apply_digest($key);
651 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
652 $self->_engine->delete_bucket( $tag, $md5, $key );
657 my $size = $self->FETCHSIZE;
658 for my $key ( 0 .. $size - 1 ) {
659 my $md5 = $self->_engine->apply_digest($key);
660 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
661 $self->_engine->delete_bucket( $tag, $md5, $key );
663 $self->STORESIZE( 0 );
665 #XXX This needs updating to use _release_space
666 # $self->_engine->write_tag(
667 # $self->_base_offset, $self->_type,
668 # chr(0)x$self->_engine->{index_size},
677 # Public method aliases
679 sub put { (shift)->STORE( @_ ) }
680 sub store { (shift)->STORE( @_ ) }
681 sub get { (shift)->FETCH( @_ ) }
682 sub fetch { (shift)->FETCH( @_ ) }
683 sub delete { (shift)->DELETE( @_ ) }
684 sub exists { (shift)->EXISTS( @_ ) }
685 sub clear { (shift)->CLEAR( @_ ) }
692 DBM::Deep - A pure perl multi-level hash/array DBM
697 my $db = DBM::Deep->new( "foo.db" );
699 $db->{key} = 'value';
702 $db->put('key' => 'value');
703 print $db->get('key');
705 # true multi-level support
706 $db->{my_complex} = [
707 'hello', { perl => 'rules' },
711 tie my %db, 'DBM::Deep', 'foo.db';
715 tied(%db)->put('key' => 'value');
716 print tied(%db)->get('key');
720 A unique flat-file database module, written in pure perl. True multi-level
721 hash/array support (unlike MLDBM, which is faked), hybrid OO / tie()
722 interface, cross-platform FTPable files, ACID transactions, and is quite fast.
723 Can handle millions of keys and unlimited levels without significant
724 slow-down. Written from the ground-up in pure perl -- this is NOT a wrapper
725 around a C-based DBM. Out-of-the-box compatibility with Unix, Mac OS X and
728 =head1 VERSION DIFFERENCES
730 B<NOTE>: 0.99_01 and above have significant file format differences from 0.983 and
731 before. There will be a backwards-compatibility layer in 1.00, but that is
732 slated for a later 0.99_x release. This version is B<NOT> backwards compatible
733 with 0.983 and before.
737 Construction can be done OO-style (which is the recommended way), or using
738 Perl's tie() function. Both are examined here.
740 =head2 OO CONSTRUCTION
742 The recommended way to construct a DBM::Deep object is to use the new()
743 method, which gets you a blessed I<and> tied hash (or array) reference.
745 my $db = DBM::Deep->new( "foo.db" );
747 This opens a new database handle, mapped to the file "foo.db". If this
748 file does not exist, it will automatically be created. DB files are
749 opened in "r+" (read/write) mode, and the type of object returned is a
750 hash, unless otherwise specified (see L<OPTIONS> below).
752 You can pass a number of options to the constructor to specify things like
753 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
755 my $db = DBM::Deep->new(
761 Notice that the filename is now specified I<inside> the hash with
762 the "file" parameter, as opposed to being the sole argument to the
763 constructor. This is required if any options are specified.
764 See L<OPTIONS> below for the complete list.
766 You can also start with an array instead of a hash. For this, you must
767 specify the C<type> parameter:
769 my $db = DBM::Deep->new(
771 type => DBM::Deep->TYPE_ARRAY
774 B<Note:> Specifing the C<type> parameter only takes effect when beginning
775 a new DB file. If you create a DBM::Deep object with an existing file, the
776 C<type> will be loaded from the file header, and an error will be thrown if
777 the wrong type is passed in.
779 =head2 TIE CONSTRUCTION
781 Alternately, you can create a DBM::Deep handle by using Perl's built-in
782 tie() function. The object returned from tie() can be used to call methods,
783 such as lock() and unlock(). (That object can be retrieved from the tied
784 variable at any time using tied() - please see L<perltie/> for more info.
787 my $db = tie %hash, "DBM::Deep", "foo.db";
790 my $db = tie @array, "DBM::Deep", "bar.db";
792 As with the OO constructor, you can replace the DB filename parameter with
793 a hash containing one or more options (see L<OPTIONS> just below for the
796 tie %hash, "DBM::Deep", {
804 There are a number of options that can be passed in when constructing your
805 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
811 Filename of the DB file to link the handle to. You can pass a full absolute
812 filesystem path, partial path, or a plain filename if the file is in the
813 current working directory. This is a required parameter (though q.v. fh).
817 If you want, you can pass in the fh instead of the file. This is most useful for doing
820 my $db = DBM::Deep->new( { fh => \*DATA } );
822 You are responsible for making sure that the fh has been opened appropriately for your
823 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
824 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
825 needs to read from the fh.
827 =item * audit_file / audit_fh
829 These are just like file/fh, except for auditing. Please see L</AUDITING> for
834 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
835 not need to set this. However, it's there if you want it.
837 If you pass in fh and do not set this, it will be set appropriately.
841 This parameter specifies what type of object to create, a hash or array. Use
842 one of these two constants:
846 =item * C<DBM::Deep-E<gt>TYPE_HASH>
848 =item * C<DBM::Deep-E<gt>TYPE_ARRAY>.
852 This only takes effect when beginning a new file. This is an optional
853 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
857 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
858 function to lock the database in exclusive mode for writes, and shared mode
859 for reads. Pass any true value to enable. This affects the base DB handle
860 I<and any child hashes or arrays> that use the same DB file. This is an
861 optional parameter, and defaults to 0 (disabled). See L<LOCKING> below for
866 Specifies whether autoflush is to be enabled on the underlying filehandle.
867 This obviously slows down write operations, but is required if you may have
868 multiple processes accessing the same DB file (also consider enable I<locking>).
869 Pass any true value to enable. This is an optional parameter, and defaults to 0
874 If I<autobless> mode is enabled, DBM::Deep will preserve the class something
875 is blessed into, and restores it when fetched. This is an optional parameter, and defaults to 1 (enabled).
877 B<Note:> If you use the OO-interface, you will not be able to call any methods
878 of DBM::Deep on the blessed item. This is considered to be a feature.
882 See L</FILTERS> below.
888 With DBM::Deep you can access your databases using Perl's standard hash/array
889 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
890 treat them as such. DBM::Deep will intercept all reads/writes and direct them
891 to the right place -- the DB file. This has nothing to do with the
892 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
893 using regular hashes and arrays, rather than calling functions like C<get()>
894 and C<put()> (although those work too). It is entirely up to you how to want
895 to access your databases.
899 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
900 or even nested hashes (or arrays) using standard Perl syntax:
902 my $db = DBM::Deep->new( "foo.db" );
904 $db->{mykey} = "myvalue";
906 $db->{myhash}->{subkey} = "subvalue";
908 print $db->{myhash}->{subkey} . "\n";
910 You can even step through hash keys using the normal Perl C<keys()> function:
912 foreach my $key (keys %$db) {
913 print "$key: " . $db->{$key} . "\n";
916 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
917 pushes them onto an array, all before the loop even begins. If you have an
918 extremely large hash, this may exhaust Perl's memory. Instead, consider using
919 Perl's C<each()> function, which pulls keys/values one at a time, using very
922 while (my ($key, $value) = each %$db) {
923 print "$key: $value\n";
926 Please note that when using C<each()>, you should always pass a direct
927 hash reference, not a lookup. Meaning, you should B<never> do this:
930 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
932 This causes an infinite loop, because for each iteration, Perl is calling
933 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
934 it effectively keeps returning the first key over and over again. Instead,
935 assign a temporary variable to C<$db->{foo}>, then pass that to each().
939 As with hashes, you can treat any DBM::Deep object like a normal Perl array
940 reference. This includes inserting, removing and manipulating elements,
941 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
942 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
943 or simply be a nested array reference inside a hash. Example:
945 my $db = DBM::Deep->new(
946 file => "foo-array.db",
947 type => DBM::Deep->TYPE_ARRAY
951 push @$db, "bar", "baz";
954 my $last_elem = pop @$db; # baz
955 my $first_elem = shift @$db; # bah
956 my $second_elem = $db->[1]; # bar
958 my $num_elements = scalar @$db;
962 In addition to the I<tie()> interface, you can also use a standard OO interface
963 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
964 array) has its own methods, but both types share the following common methods:
965 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
966 C<store(> are aliases to C<put()> and C<get()>, respectively.
970 =item * new() / clone()
972 These are the constructor and copy-functions.
974 =item * put() / store()
976 Stores a new hash key/value pair, or sets an array element value. Takes two
977 arguments, the hash key or array index, and the new value. The value can be
978 a scalar, hash ref or array ref. Returns true on success, false on failure.
980 $db->put("foo", "bar"); # for hashes
981 $db->put(1, "bar"); # for arrays
983 =item * get() / fetch()
985 Fetches the value of a hash key or array element. Takes one argument: the hash
986 key or array index. Returns a scalar, hash ref or array ref, depending on the
989 my $value = $db->get("foo"); # for hashes
990 my $value = $db->get(1); # for arrays
994 Checks if a hash key or array index exists. Takes one argument: the hash key
995 or array index. Returns true if it exists, false if not.
997 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
998 if ($db->exists(1)) { print "yay!\n"; } # for arrays
1002 Deletes one hash key/value pair or array element. Takes one argument: the hash
1003 key or array index. Returns true on success, false if not found. For arrays,
1004 the remaining elements located after the deleted element are NOT moved over.
1005 The deleted element is essentially just undefined, which is exactly how Perl's
1006 internal arrays work. Please note that the space occupied by the deleted
1007 key/value or element is B<not> reused again -- see L<UNUSED SPACE RECOVERY>
1008 below for details and workarounds.
1010 $db->delete("foo"); # for hashes
1011 $db->delete(1); # for arrays
1015 Deletes B<all> hash keys or array elements. Takes no arguments. No return
1016 value. Please note that the space occupied by the deleted keys/values or
1017 elements is B<not> reused again -- see L<UNUSED SPACE RECOVERY> below for
1018 details and workarounds.
1020 $db->clear(); # hashes or arrays
1022 =item * lock() / unlock()
1028 Recover lost disk space. This is important to do, especially if you use
1031 =item * import() / export()
1033 Data going in and out.
1039 For hashes, DBM::Deep supports all the common methods described above, and the
1040 following additional methods: C<first_key()> and C<next_key()>.
1046 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
1047 fetched in an undefined order (which appears random). Takes no arguments,
1048 returns the key as a scalar value.
1050 my $key = $db->first_key();
1054 Returns the "next" key in the hash, given the previous one as the sole argument.
1055 Returns undef if there are no more keys to be fetched.
1057 $key = $db->next_key($key);
1061 Here are some examples of using hashes:
1063 my $db = DBM::Deep->new( "foo.db" );
1065 $db->put("foo", "bar");
1066 print "foo: " . $db->get("foo") . "\n";
1068 $db->put("baz", {}); # new child hash ref
1069 $db->get("baz")->put("buz", "biz");
1070 print "buz: " . $db->get("baz")->get("buz") . "\n";
1072 my $key = $db->first_key();
1074 print "$key: " . $db->get($key) . "\n";
1075 $key = $db->next_key($key);
1078 if ($db->exists("foo")) { $db->delete("foo"); }
1082 For arrays, DBM::Deep supports all the common methods described above, and the
1083 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
1084 C<unshift()> and C<splice()>.
1090 Returns the number of elements in the array. Takes no arguments.
1092 my $len = $db->length();
1096 Adds one or more elements onto the end of the array. Accepts scalars, hash
1097 refs or array refs. No return value.
1099 $db->push("foo", "bar", {});
1103 Fetches the last element in the array, and deletes it. Takes no arguments.
1104 Returns undef if array is empty. Returns the element value.
1106 my $elem = $db->pop();
1110 Fetches the first element in the array, deletes it, then shifts all the
1111 remaining elements over to take up the space. Returns the element value. This
1112 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
1115 my $elem = $db->shift();
1119 Inserts one or more elements onto the beginning of the array, shifting all
1120 existing elements over to make room. Accepts scalars, hash refs or array refs.
1121 No return value. This method is not recommended with large arrays -- see
1122 <LARGE ARRAYS> below for details.
1124 $db->unshift("foo", "bar", {});
1128 Performs exactly like Perl's built-in function of the same name. See L<perldoc
1129 -f splice> for usage -- it is too complicated to document here. This method is
1130 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
1134 Here are some examples of using arrays:
1136 my $db = DBM::Deep->new(
1138 type => DBM::Deep->TYPE_ARRAY
1141 $db->push("bar", "baz");
1142 $db->unshift("foo");
1145 my $len = $db->length();
1146 print "length: $len\n"; # 4
1148 for (my $k=0; $k<$len; $k++) {
1149 print "$k: " . $db->get($k) . "\n";
1152 $db->splice(1, 2, "biz", "baf");
1154 while (my $elem = shift @$db) {
1155 print "shifted: $elem\n";
1160 Enable automatic file locking by passing a true value to the C<locking>
1161 parameter when constructing your DBM::Deep object (see L<SETUP> above).
1163 my $db = DBM::Deep->new(
1168 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
1169 mode for writes, and shared mode for reads. This is required if you have
1170 multiple processes accessing the same database file, to avoid file corruption.
1171 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
1172 NFS> below for more.
1174 =head2 EXPLICIT LOCKING
1176 You can explicitly lock a database, so it remains locked for multiple
1177 transactions. This is done by calling the C<lock()> method, and passing an
1178 optional lock mode argument (defaults to exclusive mode). This is particularly
1179 useful for things like counters, where the current value needs to be fetched,
1180 then incremented, then stored again.
1183 my $counter = $db->get("counter");
1185 $db->put("counter", $counter);
1194 You can pass C<lock()> an optional argument, which specifies which mode to use
1195 (exclusive or shared). Use one of these two constants:
1196 C<DBM::Deep-E<gt>LOCK_EX> or C<DBM::Deep-E<gt>LOCK_SH>. These are passed
1197 directly to C<flock()>, and are the same as the constants defined in Perl's
1200 $db->lock( $db->LOCK_SH );
1204 =head1 IMPORTING/EXPORTING
1206 You can import existing complex structures by calling the C<import()> method,
1207 and export an entire database into an in-memory structure using the C<export()>
1208 method. Both are examined here.
1212 Say you have an existing hash with nested hashes/arrays inside it. Instead of
1213 walking the structure and adding keys/elements to the database as you go,
1214 simply pass a reference to the C<import()> method. This recursively adds
1215 everything to an existing DBM::Deep object for you. Here is an example:
1220 array1 => [ "elem0", "elem1", "elem2" ],
1222 subkey1 => "subvalue1",
1223 subkey2 => "subvalue2"
1227 my $db = DBM::Deep->new( "foo.db" );
1228 $db->import( $struct );
1230 print $db->{key1} . "\n"; # prints "value1"
1232 This recursively imports the entire C<$struct> object into C<$db>, including
1233 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
1234 keys are merged with the existing ones, replacing if they already exist.
1235 The C<import()> method can be called on any database level (not just the base
1236 level), and works with both hash and array DB types.
1238 B<Note:> Make sure your existing structure has no circular references in it.
1239 These will cause an infinite loop when importing. There are plans to fix this
1244 Calling the C<export()> method on an existing DBM::Deep object will return
1245 a reference to a new in-memory copy of the database. The export is done
1246 recursively, so all nested hashes/arrays are all exported to standard Perl
1247 objects. Here is an example:
1249 my $db = DBM::Deep->new( "foo.db" );
1251 $db->{key1} = "value1";
1252 $db->{key2} = "value2";
1254 $db->{hash1}->{subkey1} = "subvalue1";
1255 $db->{hash1}->{subkey2} = "subvalue2";
1257 my $struct = $db->export();
1259 print $struct->{key1} . "\n"; # prints "value1"
1261 This makes a complete copy of the database in memory, and returns a reference
1262 to it. The C<export()> method can be called on any database level (not just
1263 the base level), and works with both hash and array DB types. Be careful of
1264 large databases -- you can store a lot more data in a DBM::Deep object than an
1265 in-memory Perl structure.
1267 B<Note:> Make sure your database has no circular references in it.
1268 These will cause an infinite loop when exporting. There are plans to fix this
1273 DBM::Deep has a number of hooks where you can specify your own Perl function
1274 to perform filtering on incoming or outgoing data. This is a perfect
1275 way to extend the engine, and implement things like real-time compression or
1276 encryption. Filtering applies to the base DB level, and all child hashes /
1277 arrays. Filter hooks can be specified when your DBM::Deep object is first
1278 constructed, or by calling the C<set_filter()> method at any time. There are
1279 four available filter hooks, described below:
1283 =item * filter_store_key
1285 This filter is called whenever a hash key is stored. It
1286 is passed the incoming key, and expected to return a transformed key.
1288 =item * filter_store_value
1290 This filter is called whenever a hash key or array element is stored. It
1291 is passed the incoming value, and expected to return a transformed value.
1293 =item * filter_fetch_key
1295 This filter is called whenever a hash key is fetched (i.e. via
1296 C<first_key()> or C<next_key()>). It is passed the transformed key,
1297 and expected to return the plain key.
1299 =item * filter_fetch_value
1301 This filter is called whenever a hash key or array element is fetched.
1302 It is passed the transformed value, and expected to return the plain value.
1306 Here are the two ways to setup a filter hook:
1308 my $db = DBM::Deep->new(
1310 filter_store_value => \&my_filter_store,
1311 filter_fetch_value => \&my_filter_fetch
1316 $db->set_filter( "filter_store_value", \&my_filter_store );
1317 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
1319 Your filter function will be called only when dealing with SCALAR keys or
1320 values. When nested hashes and arrays are being stored/fetched, filtering
1321 is bypassed. Filters are called as static functions, passed a single SCALAR
1322 argument, and expected to return a single SCALAR value. If you want to
1323 remove a filter, set the function reference to C<undef>:
1325 $db->set_filter( "filter_store_value", undef );
1327 =head2 REAL-TIME ENCRYPTION EXAMPLE
1329 Here is a working example that uses the I<Crypt::Blowfish> module to
1330 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
1331 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
1332 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
1335 use Crypt::Blowfish;
1338 my $cipher = Crypt::CBC->new({
1339 'key' => 'my secret key',
1340 'cipher' => 'Blowfish',
1342 'regenerate_key' => 0,
1343 'padding' => 'space',
1347 my $db = DBM::Deep->new(
1348 file => "foo-encrypt.db",
1349 filter_store_key => \&my_encrypt,
1350 filter_store_value => \&my_encrypt,
1351 filter_fetch_key => \&my_decrypt,
1352 filter_fetch_value => \&my_decrypt,
1355 $db->{key1} = "value1";
1356 $db->{key2} = "value2";
1357 print "key1: " . $db->{key1} . "\n";
1358 print "key2: " . $db->{key2} . "\n";
1364 return $cipher->encrypt( $_[0] );
1367 return $cipher->decrypt( $_[0] );
1370 =head2 REAL-TIME COMPRESSION EXAMPLE
1372 Here is a working example that uses the I<Compress::Zlib> module to do real-time
1373 compression / decompression of keys & values with DBM::Deep Filters.
1374 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
1375 more on I<Compress::Zlib>.
1380 my $db = DBM::Deep->new(
1381 file => "foo-compress.db",
1382 filter_store_key => \&my_compress,
1383 filter_store_value => \&my_compress,
1384 filter_fetch_key => \&my_decompress,
1385 filter_fetch_value => \&my_decompress,
1388 $db->{key1} = "value1";
1389 $db->{key2} = "value2";
1390 print "key1: " . $db->{key1} . "\n";
1391 print "key2: " . $db->{key2} . "\n";
1397 return Compress::Zlib::memGzip( $_[0] ) ;
1400 return Compress::Zlib::memGunzip( $_[0] ) ;
1403 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
1404 actually numerical index numbers, and are not filtered.
1406 =head1 ERROR HANDLING
1408 Most DBM::Deep methods return a true value for success, and call die() on
1409 failure. You can wrap calls in an eval block to catch the die.
1411 my $db = DBM::Deep->new( "foo.db" ); # create hash
1412 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
1414 print $@; # prints error message
1416 =head1 LARGEFILE SUPPORT
1418 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
1419 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
1420 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
1421 by specifying the 'pack_size' parameter when constructing the file.
1424 filename => $filename,
1425 pack_size => 'large',
1428 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
1429 instead of 32-bit longs. After setting these values your DB files have a
1430 theoretical maximum size of 16 XB (exabytes).
1432 You can also use C<pack_size =E<gt> 'small'> in order to use 16-bit file
1435 B<Note:> Changing these values will B<NOT> work for existing database files.
1436 Only change this for new files. Once the value has been set, it is stored in
1437 the file's header and cannot be changed for the life of the file. These
1438 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
1441 B<Note:> We have not personally tested files larger than 2 GB -- all my
1442 systems have only a 32-bit Perl. However, I have received user reports that
1443 this does indeed work!
1445 =head1 LOW-LEVEL ACCESS
1447 If you require low-level access to the underlying filehandle that DBM::Deep uses,
1448 you can call the C<_fh()> method, which returns the handle:
1450 my $fh = $db->_fh();
1452 This method can be called on the root level of the datbase, or any child
1453 hashes or arrays. All levels share a I<root> structure, which contains things
1454 like the filehandle, a reference counter, and all the options specified
1455 when you created the object. You can get access to this file object by
1456 calling the C<_fileobj()> method.
1458 my $file_obj = $db->_fileobj();
1460 This is useful for changing options after the object has already been created,
1461 such as enabling/disabling locking. You can also store your own temporary user
1462 data in this structure (be wary of name collision), which is then accessible from
1463 any child hash or array.
1465 =head1 CUSTOM DIGEST ALGORITHM
1467 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
1468 keys. However you can override this, and use another algorithm (such as SHA-256)
1469 or even write your own. But please note that DBM::Deep currently expects zero
1470 collisions, so your algorithm has to be I<perfect>, so to speak. Collision
1471 detection may be introduced in a later version.
1473 You can specify a custom digest algorithm by passing it into the parameter
1474 list for new(), passing a reference to a subroutine as the 'digest' parameter,
1475 and the length of the algorithm's hashes (in bytes) as the 'hash_size'
1476 parameter. Here is a working example that uses a 256-bit hash from the
1477 I<Digest::SHA256> module. Please see
1478 L<http://search.cpan.org/search?module=Digest::SHA256> for more information.
1483 my $context = Digest::SHA256::new(256);
1485 my $db = DBM::Deep->new(
1486 filename => "foo-sha.db",
1487 digest => \&my_digest,
1491 $db->{key1} = "value1";
1492 $db->{key2} = "value2";
1493 print "key1: " . $db->{key1} . "\n";
1494 print "key2: " . $db->{key2} . "\n";
1500 return substr( $context->hash($_[0]), 0, 32 );
1503 B<Note:> Your returned digest strings must be B<EXACTLY> the number
1504 of bytes you specify in the hash_size parameter (in this case 32).
1506 B<Note:> If you do choose to use a custom digest algorithm, you must set it
1507 every time you access this file. Otherwise, the default (MD5) will be used.
1509 =head1 CIRCULAR REFERENCES
1511 DBM::Deep has B<experimental> support for circular references. Meaning you
1512 can have a nested hash key or array element that points to a parent object.
1513 This relationship is stored in the DB file, and is preserved between sessions.
1516 my $db = DBM::Deep->new( "foo.db" );
1519 $db->{circle} = $db; # ref to self
1521 print $db->{foo} . "\n"; # prints "bar"
1522 print $db->{circle}->{foo} . "\n"; # prints "bar" again
1524 B<Note>: Passing the object to a function that recursively walks the
1525 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
1526 C<export()> methods) will result in an infinite loop. This will be fixed in
1531 New in 0.99_01 is the ability to audit your databases actions. By passing in
1532 audit_file (or audit_fh) to the constructor, all actions will be logged to
1533 that file. The format is one that is suitable for eval'ing against the
1534 database to replay the actions. Please see t/33_audit_trail.t for an example
1539 New in 0.99_01 is ACID transactions. Every DBM::Deep object is completely
1540 transaction-ready - it is not an option you have to turn on. Three new methods
1541 have been added to support them. They are:
1545 =item * begin_work()
1547 This starts a transaction.
1551 This applies the changes done within the transaction to the mainline and ends
1556 This discards the changes done within the transaction to the mainline and ends
1561 Transactions in DBM::Deep are done using the MVCC method, the same method used
1562 by the InnoDB MySQL table type.
1564 =head1 CAVEATS / ISSUES / BUGS
1566 This section describes all the known issues with DBM::Deep. It you have found
1567 something that is not listed here, please send e-mail to L<jhuckaby@cpan.org>.
1569 =head2 UNUSED SPACE RECOVERY
1571 One major caveat with DBM::Deep is that space occupied by existing keys and
1572 values is not recovered when they are deleted. Meaning if you keep deleting
1573 and adding new keys, your file will continuously grow. I am working on this,
1574 but in the meantime you can call the built-in C<optimize()> method from time to
1575 time (perhaps in a crontab or something) to recover all your unused space.
1577 $db->optimize(); # returns true on success
1579 This rebuilds the ENTIRE database into a new file, then moves it on top of
1580 the original. The new file will have no unused space, thus it will take up as
1581 little disk space as possible. Please note that this operation can take
1582 a long time for large files, and you need enough disk space to temporarily hold
1583 2 copies of your DB file. The temporary file is created in the same directory
1584 as the original, named with a ".tmp" extension, and is deleted when the
1585 operation completes. Oh, and if locking is enabled, the DB is automatically
1586 locked for the entire duration of the copy.
1588 B<WARNING:> Only call optimize() on the top-level node of the database, and
1589 make sure there are no child references lying around. DBM::Deep keeps a reference
1590 counter, and if it is greater than 1, optimize() will abort and return undef.
1594 (The reasons given assume a high level of Perl understanding, specifically of
1595 references. You can safely skip this section.)
1597 Currently, the only references supported are HASH and ARRAY. The other reference
1598 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
1604 These are things like filehandles and other sockets. They can't be supported
1605 because it's completely unclear how DBM::Deep should serialize them.
1607 =item * SCALAR / REF
1609 The discussion here refers to the following type of example:
1616 # In some other process ...
1618 my $val = ${ $db->{key1} };
1620 is( $val, 50, "What actually gets stored in the DB file?" );
1622 The problem is one of synchronization. When the variable being referred to
1623 changes value, the reference isn't notified. This means that the new value won't
1624 be stored in the datafile for other processes to read. There is no TIEREF.
1626 It is theoretically possible to store references to values already within a
1627 DBM::Deep object because everything already is synchronized, but the change to
1628 the internals would be quite large. Specifically, DBM::Deep would have to tie
1629 every single value that is stored. This would bloat the RAM footprint of
1630 DBM::Deep at least twofold (if not more) and be a significant performance drain,
1631 all to support a feature that has never been requested.
1635 L<Data::Dump::Streamer/> provides a mechanism for serializing coderefs,
1636 including saving off all closure state. However, just as for SCALAR and REF,
1637 that closure state may change without notifying the DBM::Deep object storing
1642 =head2 FILE CORRUPTION
1644 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
1645 for a 32-bit signature when opened, but other corruption in files can cause
1646 segmentation faults. DBM::Deep may try to seek() past the end of a file, or get
1647 stuck in an infinite loop depending on the level of corruption. File write
1648 operations are not checked for failure (for speed), so if you happen to run
1649 out of disk space, DBM::Deep will probably fail in a bad way. These things will
1650 be addressed in a later version of DBM::Deep.
1654 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
1655 well on local filesystems, but will NOT protect you from file corruption over
1656 NFS. I've heard about setting up your NFS server with a locking daemon, then
1657 using lockf() to lock your files, but your mileage may vary there as well.
1658 From what I understand, there is no real way to do it. However, if you need
1659 access to the underlying filehandle in DBM::Deep for using some other kind of
1660 locking scheme like lockf(), see the L<LOW-LEVEL ACCESS> section above.
1662 =head2 COPYING OBJECTS
1664 Beware of copying tied objects in Perl. Very strange things can happen.
1665 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
1666 returns a new, blessed, tied hash or array to the same level in the DB.
1668 my $copy = $db->clone();
1670 B<Note>: Since clone() here is cloning the object, not the database location, any
1671 modifications to either $db or $copy will be visible to both.
1675 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1676 These functions cause every element in the array to move, which can be murder
1677 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1678 a different location. This will be addressed in the forthcoming version 1.00.
1680 =head2 WRITEONLY FILES
1682 If you pass in a filehandle to new(), you may have opened it in either a readonly or
1683 writeonly mode. STORE will verify that the filehandle is writable. However, there
1684 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
1685 filehandle isn't readable, it's not clear what will happen. So, don't do that.
1687 =head1 CODE COVERAGE
1689 B<Devel::Cover> is used to test the code coverage of the tests. Below is the
1690 B<Devel::Cover> report on this distribution's test suite.
1692 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1693 File stmt bran cond sub pod time total
1694 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1695 blib/lib/DBM/Deep.pm 96.2 89.0 75.0 95.8 89.5 36.0 92.9
1696 blib/lib/DBM/Deep/Array.pm 96.1 88.3 100.0 96.4 100.0 15.9 94.7
1697 blib/lib/DBM/Deep/Engine.pm 96.6 86.6 89.5 100.0 0.0 20.0 91.0
1698 blib/lib/DBM/Deep/File.pm 99.4 88.3 55.6 100.0 0.0 19.6 89.5
1699 blib/lib/DBM/Deep/Hash.pm 98.5 83.3 100.0 100.0 100.0 8.5 96.3
1700 Total 96.9 87.4 81.2 98.0 38.5 100.0 92.1
1701 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1703 =head1 MORE INFORMATION
1705 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1706 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1709 The source code repository is at L<http://svn.perl.org/modules/DBM-Deep>
1713 Rob Kinyon, L<rkinyon@cpan.org>
1715 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1717 Special thanks to Adam Sah and Rich Gaushell! You know why :-)
1721 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1722 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1726 Copyright (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
1727 This is free software, you may use it and distribute it under the
1728 same terms as Perl itself.