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,
327 my %is_legal_filter = map {
330 store_key store_value
331 fetch_key fetch_value
336 # Setup filter function for storing or fetching the key or value
338 my $self = shift->_get_self;
342 if ( $is_legal_filter{$type} ) {
343 $self->_fileobj->{"filter_$type"} = $func;
352 my $self = shift->_get_self;
353 $self->_fileobj->begin_transaction;
358 my $self = shift->_get_self;
359 $self->_fileobj->end_transaction;
364 my $self = shift->_get_self;
365 $self->_fileobj->commit_transaction;
374 my $self = $_[0]->_get_self;
375 return $self->{engine};
379 my $self = $_[0]->_get_self;
380 return $self->{fileobj};
384 my $self = $_[0]->_get_self;
385 return $self->{type};
389 my $self = $_[0]->_get_self;
390 return $self->{base_offset};
394 my $self = $_[0]->_get_self;
395 return $self->_fileobj->{fh};
403 die "DBM::Deep: $_[1]\n";
410 #XXX This if() is redundant
411 if ( my $parent = $self->{parent} ) {
413 while ( $parent->{parent} ) {
415 $parent->_type eq TYPE_HASH
416 ? "\{q{$child->{parent_key}}\}"
417 : "\[$child->{parent_key}\]"
421 $parent = $parent->{parent};
424 $base = "\$db->get( q{$child->{parent_key}} )->" . $base;
427 $base = "\$db->get( q{$child->{parent_key}} )";
435 # Store single hash key/value or array element in database.
437 my $self = shift->_get_self;
438 my ($key, $value, $orig_key) = @_;
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 ( defined $orig_key && !( $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 my $md5 = $self->_engine->{digest}->($key);
491 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5, { create => 1 } );
493 # User may be storing a hash, in which case we do not want it run
494 # through the filtering system
495 if ( !ref($value) && $self->_fileobj->{filter_store_value} ) {
496 $value = $self->_fileobj->{filter_store_value}->( $value );
500 # Add key/value to bucket list
502 $self->_engine->add_bucket( $tag, $md5, $key, $value, undef, $orig_key );
511 # Fetch single value or element given plain key or array index
513 my $self = shift->_get_self;
514 my ($key, $orig_key) = @_;
516 my $md5 = $self->_engine->{digest}->($key);
519 # Request shared lock for reading
521 $self->lock( LOCK_SH );
523 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );#, { create => 1 } );
524 #XXX This needs to autovivify
531 # Get value from bucket list
533 my $result = $self->_engine->get_bucket_value( $tag, $md5, $orig_key );
537 # Filters only apply to scalar values, so the ref check is making
538 # sure the fetched bucket is a scalar, not a child hash or array.
539 return ($result && !ref($result) && $self->_fileobj->{filter_fetch_value})
540 ? $self->_fileobj->{filter_fetch_value}->($result)
546 # Delete single key/value pair or element given plain key or array index
548 my $self = shift->_get_self;
549 my ($key, $orig_key) = @_;
551 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
552 $self->_throw_error( 'Cannot write to a readonly filehandle' );
555 if ( defined $orig_key ) {
556 my $lhs = $self->_find_parent;
558 $self->_fileobj->audit( "delete $lhs;" );
561 $self->_fileobj->audit( "\$db->delete('$orig_key');" );
566 # Request exclusive lock for writing
568 $self->lock( LOCK_EX );
570 my $md5 = $self->_engine->{digest}->($key);
572 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
581 my $value = $self->_engine->get_bucket_value( $tag, $md5 );
583 if (defined $value && !ref($value) && $self->_fileobj->{filter_fetch_value}) {
584 $value = $self->_fileobj->{filter_fetch_value}->($value);
587 my $result = $self->_engine->delete_bucket( $tag, $md5, $orig_key );
590 # If this object is an array and the key deleted was on the end of the stack,
591 # decrement the length variable.
601 # Check if a single key or element exists given plain key or array index
603 my $self = shift->_get_self;
606 my $md5 = $self->_engine->{digest}->($key);
609 # Request shared lock for reading
611 $self->lock( LOCK_SH );
613 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
618 # For some reason, the built-in exists() function returns '' for false
624 # Check if bucket exists and return 1 or ''
626 my $result = $self->_engine->bucket_exists( $tag, $md5 ) || '';
635 # Clear all keys from hash, or all elements from array.
637 my $self = shift->_get_self;
639 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
640 $self->_throw_error( 'Cannot write to a readonly filehandle' );
644 my $lhs = $self->_find_parent;
646 if ( $self->_type eq TYPE_HASH ) {
647 $lhs = '%{' . $lhs . '}';
650 $lhs = '@{' . $lhs . '}';
653 $self->_fileobj->audit( "$lhs = ();" );
657 # Request exclusive lock for writing
659 $self->lock( LOCK_EX );
661 if ( $self->_type eq TYPE_HASH ) {
662 my $key = $self->first_key;
664 my $next_key = $self->next_key( $key );
665 my $md5 = $self->_engine->{digest}->($key);
666 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
667 $self->_engine->delete_bucket( $tag, $md5, $key );
672 my $size = $self->FETCHSIZE;
673 for my $key ( map { pack ( $self->_engine->{long_pack}, $_ ) } 0 .. $size - 1 ) {
674 my $md5 = $self->_engine->{digest}->($key);
675 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
676 $self->_engine->delete_bucket( $tag, $md5, $key );
678 $self->STORESIZE( 0 );
680 #XXX This needs updating to use _release_space
681 # $self->_engine->write_tag(
682 # $self->_base_offset, $self->_type,
683 # chr(0)x$self->_engine->{index_size},
692 # Public method aliases
694 sub put { (shift)->STORE( @_ ) }
695 sub store { (shift)->STORE( @_ ) }
696 sub get { (shift)->FETCH( @_ ) }
697 sub fetch { (shift)->FETCH( @_ ) }
698 sub delete { (shift)->DELETE( @_ ) }
699 sub exists { (shift)->EXISTS( @_ ) }
700 sub clear { (shift)->CLEAR( @_ ) }
707 DBM::Deep - A pure perl multi-level hash/array DBM
712 my $db = DBM::Deep->new( "foo.db" );
714 $db->{key} = 'value';
717 $db->put('key' => 'value');
718 print $db->get('key');
720 # true multi-level support
721 $db->{my_complex} = [
722 'hello', { perl => 'rules' },
726 tie my %db, 'DBM::Deep', 'foo.db';
730 tied(%db)->put('key' => 'value');
731 print tied(%db)->get('key');
735 A unique flat-file database module, written in pure perl. True multi-level
736 hash/array support (unlike MLDBM, which is faked), hybrid OO / tie()
737 interface, cross-platform FTPable files, ACID transactions, and is quite fast.
738 Can handle millions of keys and unlimited levels without significant
739 slow-down. Written from the ground-up in pure perl -- this is NOT a wrapper
740 around a C-based DBM. Out-of-the-box compatibility with Unix, Mac OS X and
743 =head1 VERSION DIFFERENCES
745 B<NOTE>: 0.99_01 and above have significant file format differences from 0.983 and
746 before. There will be a backwards-compatibility layer in 1.00, but that is
747 slated for a later 0.99_x release. This version is B<NOT> backwards compatible
748 with 0.983 and before.
752 Construction can be done OO-style (which is the recommended way), or using
753 Perl's tie() function. Both are examined here.
755 =head2 OO CONSTRUCTION
757 The recommended way to construct a DBM::Deep object is to use the new()
758 method, which gets you a blessed I<and> tied hash (or array) reference.
760 my $db = DBM::Deep->new( "foo.db" );
762 This opens a new database handle, mapped to the file "foo.db". If this
763 file does not exist, it will automatically be created. DB files are
764 opened in "r+" (read/write) mode, and the type of object returned is a
765 hash, unless otherwise specified (see L<OPTIONS> below).
767 You can pass a number of options to the constructor to specify things like
768 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
770 my $db = DBM::Deep->new(
776 Notice that the filename is now specified I<inside> the hash with
777 the "file" parameter, as opposed to being the sole argument to the
778 constructor. This is required if any options are specified.
779 See L<OPTIONS> below for the complete list.
781 You can also start with an array instead of a hash. For this, you must
782 specify the C<type> parameter:
784 my $db = DBM::Deep->new(
786 type => DBM::Deep->TYPE_ARRAY
789 B<Note:> Specifing the C<type> parameter only takes effect when beginning
790 a new DB file. If you create a DBM::Deep object with an existing file, the
791 C<type> will be loaded from the file header, and an error will be thrown if
792 the wrong type is passed in.
794 =head2 TIE CONSTRUCTION
796 Alternately, you can create a DBM::Deep handle by using Perl's built-in
797 tie() function. The object returned from tie() can be used to call methods,
798 such as lock() and unlock(). (That object can be retrieved from the tied
799 variable at any time using tied() - please see L<perltie/> for more info.
802 my $db = tie %hash, "DBM::Deep", "foo.db";
805 my $db = tie @array, "DBM::Deep", "bar.db";
807 As with the OO constructor, you can replace the DB filename parameter with
808 a hash containing one or more options (see L<OPTIONS> just below for the
811 tie %hash, "DBM::Deep", {
819 There are a number of options that can be passed in when constructing your
820 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
826 Filename of the DB file to link the handle to. You can pass a full absolute
827 filesystem path, partial path, or a plain filename if the file is in the
828 current working directory. This is a required parameter (though q.v. fh).
832 If you want, you can pass in the fh instead of the file. This is most useful for doing
835 my $db = DBM::Deep->new( { fh => \*DATA } );
837 You are responsible for making sure that the fh has been opened appropriately for your
838 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
839 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
840 needs to read from the fh.
842 =item * audit_file / audit_fh
844 These are just like file/fh, except for auditing. Please see L</AUDITING> for
849 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
850 not need to set this. However, it's there if you want it.
852 If you pass in fh and do not set this, it will be set appropriately.
856 This parameter specifies what type of object to create, a hash or array. Use
857 one of these two constants:
861 =item * C<DBM::Deep-E<gt>TYPE_HASH>
863 =item * C<DBM::Deep-E<gt>TYPE_ARRAY>.
867 This only takes effect when beginning a new file. This is an optional
868 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
872 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
873 function to lock the database in exclusive mode for writes, and shared mode
874 for reads. Pass any true value to enable. This affects the base DB handle
875 I<and any child hashes or arrays> that use the same DB file. This is an
876 optional parameter, and defaults to 0 (disabled). See L<LOCKING> below for
881 Specifies whether autoflush is to be enabled on the underlying filehandle.
882 This obviously slows down write operations, but is required if you may have
883 multiple processes accessing the same DB file (also consider enable I<locking>).
884 Pass any true value to enable. This is an optional parameter, and defaults to 0
889 If I<autobless> mode is enabled, DBM::Deep will preserve the class something
890 is blessed into, and restores it when fetched. This is an optional parameter, and defaults to 1 (enabled).
892 B<Note:> If you use the OO-interface, you will not be able to call any methods
893 of DBM::Deep on the blessed item. This is considered to be a feature.
897 See L</FILTERS> below.
903 With DBM::Deep you can access your databases using Perl's standard hash/array
904 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
905 treat them as such. DBM::Deep will intercept all reads/writes and direct them
906 to the right place -- the DB file. This has nothing to do with the
907 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
908 using regular hashes and arrays, rather than calling functions like C<get()>
909 and C<put()> (although those work too). It is entirely up to you how to want
910 to access your databases.
914 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
915 or even nested hashes (or arrays) using standard Perl syntax:
917 my $db = DBM::Deep->new( "foo.db" );
919 $db->{mykey} = "myvalue";
921 $db->{myhash}->{subkey} = "subvalue";
923 print $db->{myhash}->{subkey} . "\n";
925 You can even step through hash keys using the normal Perl C<keys()> function:
927 foreach my $key (keys %$db) {
928 print "$key: " . $db->{$key} . "\n";
931 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
932 pushes them onto an array, all before the loop even begins. If you have an
933 extremely large hash, this may exhaust Perl's memory. Instead, consider using
934 Perl's C<each()> function, which pulls keys/values one at a time, using very
937 while (my ($key, $value) = each %$db) {
938 print "$key: $value\n";
941 Please note that when using C<each()>, you should always pass a direct
942 hash reference, not a lookup. Meaning, you should B<never> do this:
945 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
947 This causes an infinite loop, because for each iteration, Perl is calling
948 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
949 it effectively keeps returning the first key over and over again. Instead,
950 assign a temporary variable to C<$db->{foo}>, then pass that to each().
954 As with hashes, you can treat any DBM::Deep object like a normal Perl array
955 reference. This includes inserting, removing and manipulating elements,
956 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
957 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
958 or simply be a nested array reference inside a hash. Example:
960 my $db = DBM::Deep->new(
961 file => "foo-array.db",
962 type => DBM::Deep->TYPE_ARRAY
966 push @$db, "bar", "baz";
969 my $last_elem = pop @$db; # baz
970 my $first_elem = shift @$db; # bah
971 my $second_elem = $db->[1]; # bar
973 my $num_elements = scalar @$db;
977 In addition to the I<tie()> interface, you can also use a standard OO interface
978 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
979 array) has its own methods, but both types share the following common methods:
980 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
981 C<store(> are aliases to C<put()> and C<get()>, respectively.
985 =item * new() / clone()
987 These are the constructor and copy-functions.
989 =item * put() / store()
991 Stores a new hash key/value pair, or sets an array element value. Takes two
992 arguments, the hash key or array index, and the new value. The value can be
993 a scalar, hash ref or array ref. Returns true on success, false on failure.
995 $db->put("foo", "bar"); # for hashes
996 $db->put(1, "bar"); # for arrays
998 =item * get() / fetch()
1000 Fetches the value of a hash key or array element. Takes one argument: the hash
1001 key or array index. Returns a scalar, hash ref or array ref, depending on the
1004 my $value = $db->get("foo"); # for hashes
1005 my $value = $db->get(1); # for arrays
1009 Checks if a hash key or array index exists. Takes one argument: the hash key
1010 or array index. Returns true if it exists, false if not.
1012 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
1013 if ($db->exists(1)) { print "yay!\n"; } # for arrays
1017 Deletes one hash key/value pair or array element. Takes one argument: the hash
1018 key or array index. Returns true on success, false if not found. For arrays,
1019 the remaining elements located after the deleted element are NOT moved over.
1020 The deleted element is essentially just undefined, which is exactly how Perl's
1021 internal arrays work. Please note that the space occupied by the deleted
1022 key/value or element is B<not> reused again -- see L<UNUSED SPACE RECOVERY>
1023 below for details and workarounds.
1025 $db->delete("foo"); # for hashes
1026 $db->delete(1); # for arrays
1030 Deletes B<all> hash keys or array elements. Takes no arguments. No return
1031 value. Please note that the space occupied by the deleted keys/values or
1032 elements is B<not> reused again -- see L<UNUSED SPACE RECOVERY> below for
1033 details and workarounds.
1035 $db->clear(); # hashes or arrays
1037 =item * lock() / unlock()
1043 Recover lost disk space. This is important to do, especially if you use
1046 =item * import() / export()
1048 Data going in and out.
1054 For hashes, DBM::Deep supports all the common methods described above, and the
1055 following additional methods: C<first_key()> and C<next_key()>.
1061 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
1062 fetched in an undefined order (which appears random). Takes no arguments,
1063 returns the key as a scalar value.
1065 my $key = $db->first_key();
1069 Returns the "next" key in the hash, given the previous one as the sole argument.
1070 Returns undef if there are no more keys to be fetched.
1072 $key = $db->next_key($key);
1076 Here are some examples of using hashes:
1078 my $db = DBM::Deep->new( "foo.db" );
1080 $db->put("foo", "bar");
1081 print "foo: " . $db->get("foo") . "\n";
1083 $db->put("baz", {}); # new child hash ref
1084 $db->get("baz")->put("buz", "biz");
1085 print "buz: " . $db->get("baz")->get("buz") . "\n";
1087 my $key = $db->first_key();
1089 print "$key: " . $db->get($key) . "\n";
1090 $key = $db->next_key($key);
1093 if ($db->exists("foo")) { $db->delete("foo"); }
1097 For arrays, DBM::Deep supports all the common methods described above, and the
1098 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
1099 C<unshift()> and C<splice()>.
1105 Returns the number of elements in the array. Takes no arguments.
1107 my $len = $db->length();
1111 Adds one or more elements onto the end of the array. Accepts scalars, hash
1112 refs or array refs. No return value.
1114 $db->push("foo", "bar", {});
1118 Fetches the last element in the array, and deletes it. Takes no arguments.
1119 Returns undef if array is empty. Returns the element value.
1121 my $elem = $db->pop();
1125 Fetches the first element in the array, deletes it, then shifts all the
1126 remaining elements over to take up the space. Returns the element value. This
1127 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
1130 my $elem = $db->shift();
1134 Inserts one or more elements onto the beginning of the array, shifting all
1135 existing elements over to make room. Accepts scalars, hash refs or array refs.
1136 No return value. This method is not recommended with large arrays -- see
1137 <LARGE ARRAYS> below for details.
1139 $db->unshift("foo", "bar", {});
1143 Performs exactly like Perl's built-in function of the same name. See L<perldoc
1144 -f splice> for usage -- it is too complicated to document here. This method is
1145 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
1149 Here are some examples of using arrays:
1151 my $db = DBM::Deep->new(
1153 type => DBM::Deep->TYPE_ARRAY
1156 $db->push("bar", "baz");
1157 $db->unshift("foo");
1160 my $len = $db->length();
1161 print "length: $len\n"; # 4
1163 for (my $k=0; $k<$len; $k++) {
1164 print "$k: " . $db->get($k) . "\n";
1167 $db->splice(1, 2, "biz", "baf");
1169 while (my $elem = shift @$db) {
1170 print "shifted: $elem\n";
1175 Enable automatic file locking by passing a true value to the C<locking>
1176 parameter when constructing your DBM::Deep object (see L<SETUP> above).
1178 my $db = DBM::Deep->new(
1183 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
1184 mode for writes, and shared mode for reads. This is required if you have
1185 multiple processes accessing the same database file, to avoid file corruption.
1186 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
1187 NFS> below for more.
1189 =head2 EXPLICIT LOCKING
1191 You can explicitly lock a database, so it remains locked for multiple
1192 transactions. This is done by calling the C<lock()> method, and passing an
1193 optional lock mode argument (defaults to exclusive mode). This is particularly
1194 useful for things like counters, where the current value needs to be fetched,
1195 then incremented, then stored again.
1198 my $counter = $db->get("counter");
1200 $db->put("counter", $counter);
1209 You can pass C<lock()> an optional argument, which specifies which mode to use
1210 (exclusive or shared). Use one of these two constants:
1211 C<DBM::Deep-E<gt>LOCK_EX> or C<DBM::Deep-E<gt>LOCK_SH>. These are passed
1212 directly to C<flock()>, and are the same as the constants defined in Perl's
1215 $db->lock( $db->LOCK_SH );
1219 =head1 IMPORTING/EXPORTING
1221 You can import existing complex structures by calling the C<import()> method,
1222 and export an entire database into an in-memory structure using the C<export()>
1223 method. Both are examined here.
1227 Say you have an existing hash with nested hashes/arrays inside it. Instead of
1228 walking the structure and adding keys/elements to the database as you go,
1229 simply pass a reference to the C<import()> method. This recursively adds
1230 everything to an existing DBM::Deep object for you. Here is an example:
1235 array1 => [ "elem0", "elem1", "elem2" ],
1237 subkey1 => "subvalue1",
1238 subkey2 => "subvalue2"
1242 my $db = DBM::Deep->new( "foo.db" );
1243 $db->import( $struct );
1245 print $db->{key1} . "\n"; # prints "value1"
1247 This recursively imports the entire C<$struct> object into C<$db>, including
1248 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
1249 keys are merged with the existing ones, replacing if they already exist.
1250 The C<import()> method can be called on any database level (not just the base
1251 level), and works with both hash and array DB types.
1253 B<Note:> Make sure your existing structure has no circular references in it.
1254 These will cause an infinite loop when importing. There are plans to fix this
1259 Calling the C<export()> method on an existing DBM::Deep object will return
1260 a reference to a new in-memory copy of the database. The export is done
1261 recursively, so all nested hashes/arrays are all exported to standard Perl
1262 objects. Here is an example:
1264 my $db = DBM::Deep->new( "foo.db" );
1266 $db->{key1} = "value1";
1267 $db->{key2} = "value2";
1269 $db->{hash1}->{subkey1} = "subvalue1";
1270 $db->{hash1}->{subkey2} = "subvalue2";
1272 my $struct = $db->export();
1274 print $struct->{key1} . "\n"; # prints "value1"
1276 This makes a complete copy of the database in memory, and returns a reference
1277 to it. The C<export()> method can be called on any database level (not just
1278 the base level), and works with both hash and array DB types. Be careful of
1279 large databases -- you can store a lot more data in a DBM::Deep object than an
1280 in-memory Perl structure.
1282 B<Note:> Make sure your database has no circular references in it.
1283 These will cause an infinite loop when exporting. There are plans to fix this
1288 DBM::Deep has a number of hooks where you can specify your own Perl function
1289 to perform filtering on incoming or outgoing data. This is a perfect
1290 way to extend the engine, and implement things like real-time compression or
1291 encryption. Filtering applies to the base DB level, and all child hashes /
1292 arrays. Filter hooks can be specified when your DBM::Deep object is first
1293 constructed, or by calling the C<set_filter()> method at any time. There are
1294 four available filter hooks, described below:
1298 =item * filter_store_key
1300 This filter is called whenever a hash key is stored. It
1301 is passed the incoming key, and expected to return a transformed key.
1303 =item * filter_store_value
1305 This filter is called whenever a hash key or array element is stored. It
1306 is passed the incoming value, and expected to return a transformed value.
1308 =item * filter_fetch_key
1310 This filter is called whenever a hash key is fetched (i.e. via
1311 C<first_key()> or C<next_key()>). It is passed the transformed key,
1312 and expected to return the plain key.
1314 =item * filter_fetch_value
1316 This filter is called whenever a hash key or array element is fetched.
1317 It is passed the transformed value, and expected to return the plain value.
1321 Here are the two ways to setup a filter hook:
1323 my $db = DBM::Deep->new(
1325 filter_store_value => \&my_filter_store,
1326 filter_fetch_value => \&my_filter_fetch
1331 $db->set_filter( "filter_store_value", \&my_filter_store );
1332 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
1334 Your filter function will be called only when dealing with SCALAR keys or
1335 values. When nested hashes and arrays are being stored/fetched, filtering
1336 is bypassed. Filters are called as static functions, passed a single SCALAR
1337 argument, and expected to return a single SCALAR value. If you want to
1338 remove a filter, set the function reference to C<undef>:
1340 $db->set_filter( "filter_store_value", undef );
1342 =head2 REAL-TIME ENCRYPTION EXAMPLE
1344 Here is a working example that uses the I<Crypt::Blowfish> module to
1345 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
1346 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
1347 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
1350 use Crypt::Blowfish;
1353 my $cipher = Crypt::CBC->new({
1354 'key' => 'my secret key',
1355 'cipher' => 'Blowfish',
1357 'regenerate_key' => 0,
1358 'padding' => 'space',
1362 my $db = DBM::Deep->new(
1363 file => "foo-encrypt.db",
1364 filter_store_key => \&my_encrypt,
1365 filter_store_value => \&my_encrypt,
1366 filter_fetch_key => \&my_decrypt,
1367 filter_fetch_value => \&my_decrypt,
1370 $db->{key1} = "value1";
1371 $db->{key2} = "value2";
1372 print "key1: " . $db->{key1} . "\n";
1373 print "key2: " . $db->{key2} . "\n";
1379 return $cipher->encrypt( $_[0] );
1382 return $cipher->decrypt( $_[0] );
1385 =head2 REAL-TIME COMPRESSION EXAMPLE
1387 Here is a working example that uses the I<Compress::Zlib> module to do real-time
1388 compression / decompression of keys & values with DBM::Deep Filters.
1389 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
1390 more on I<Compress::Zlib>.
1395 my $db = DBM::Deep->new(
1396 file => "foo-compress.db",
1397 filter_store_key => \&my_compress,
1398 filter_store_value => \&my_compress,
1399 filter_fetch_key => \&my_decompress,
1400 filter_fetch_value => \&my_decompress,
1403 $db->{key1} = "value1";
1404 $db->{key2} = "value2";
1405 print "key1: " . $db->{key1} . "\n";
1406 print "key2: " . $db->{key2} . "\n";
1412 return Compress::Zlib::memGzip( $_[0] ) ;
1415 return Compress::Zlib::memGunzip( $_[0] ) ;
1418 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
1419 actually numerical index numbers, and are not filtered.
1421 =head1 ERROR HANDLING
1423 Most DBM::Deep methods return a true value for success, and call die() on
1424 failure. You can wrap calls in an eval block to catch the die.
1426 my $db = DBM::Deep->new( "foo.db" ); # create hash
1427 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
1429 print $@; # prints error message
1431 =head1 LARGEFILE SUPPORT
1433 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
1434 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
1435 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
1436 by specifying the 'pack_size' parameter when constructing the file.
1439 filename => $filename,
1440 pack_size => 'large',
1443 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
1444 instead of 32-bit longs. After setting these values your DB files have a
1445 theoretical maximum size of 16 XB (exabytes).
1447 You can also use C<pack_size =E<gt> 'small'> in order to use 16-bit file
1450 B<Note:> Changing these values will B<NOT> work for existing database files.
1451 Only change this for new files. Once the value has been set, it is stored in
1452 the file's header and cannot be changed for the life of the file. These
1453 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
1456 B<Note:> We have not personally tested files larger than 2 GB -- all my
1457 systems have only a 32-bit Perl. However, I have received user reports that
1458 this does indeed work!
1460 =head1 LOW-LEVEL ACCESS
1462 If you require low-level access to the underlying filehandle that DBM::Deep uses,
1463 you can call the C<_fh()> method, which returns the handle:
1465 my $fh = $db->_fh();
1467 This method can be called on the root level of the datbase, or any child
1468 hashes or arrays. All levels share a I<root> structure, which contains things
1469 like the filehandle, a reference counter, and all the options specified
1470 when you created the object. You can get access to this file object by
1471 calling the C<_fileobj()> method.
1473 my $file_obj = $db->_fileobj();
1475 This is useful for changing options after the object has already been created,
1476 such as enabling/disabling locking. You can also store your own temporary user
1477 data in this structure (be wary of name collision), which is then accessible from
1478 any child hash or array.
1480 =head1 CUSTOM DIGEST ALGORITHM
1482 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
1483 keys. However you can override this, and use another algorithm (such as SHA-256)
1484 or even write your own. But please note that DBM::Deep currently expects zero
1485 collisions, so your algorithm has to be I<perfect>, so to speak. Collision
1486 detection may be introduced in a later version.
1488 You can specify a custom digest algorithm by passing it into the parameter
1489 list for new(), passing a reference to a subroutine as the 'digest' parameter,
1490 and the length of the algorithm's hashes (in bytes) as the 'hash_size'
1491 parameter. Here is a working example that uses a 256-bit hash from the
1492 I<Digest::SHA256> module. Please see
1493 L<http://search.cpan.org/search?module=Digest::SHA256> for more information.
1498 my $context = Digest::SHA256::new(256);
1500 my $db = DBM::Deep->new(
1501 filename => "foo-sha.db",
1502 digest => \&my_digest,
1506 $db->{key1} = "value1";
1507 $db->{key2} = "value2";
1508 print "key1: " . $db->{key1} . "\n";
1509 print "key2: " . $db->{key2} . "\n";
1515 return substr( $context->hash($_[0]), 0, 32 );
1518 B<Note:> Your returned digest strings must be B<EXACTLY> the number
1519 of bytes you specify in the hash_size parameter (in this case 32).
1521 B<Note:> If you do choose to use a custom digest algorithm, you must set it
1522 every time you access this file. Otherwise, the default (MD5) will be used.
1524 =head1 CIRCULAR REFERENCES
1526 DBM::Deep has B<experimental> support for circular references. Meaning you
1527 can have a nested hash key or array element that points to a parent object.
1528 This relationship is stored in the DB file, and is preserved between sessions.
1531 my $db = DBM::Deep->new( "foo.db" );
1534 $db->{circle} = $db; # ref to self
1536 print $db->{foo} . "\n"; # prints "bar"
1537 print $db->{circle}->{foo} . "\n"; # prints "bar" again
1539 B<Note>: Passing the object to a function that recursively walks the
1540 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
1541 C<export()> methods) will result in an infinite loop. This will be fixed in
1546 New in 0.99_01 is the ability to audit your databases actions. By passing in
1547 audit_file (or audit_fh) to the constructor, all actions will be logged to
1548 that file. The format is one that is suitable for eval'ing against the
1549 database to replay the actions. Please see t/33_audit_trail.t for an example
1554 New in 0.99_01 is ACID transactions. Every DBM::Deep object is completely
1555 transaction-ready - it is not an option you have to turn on. Three new methods
1556 have been added to support them. They are:
1560 =item * begin_work()
1562 This starts a transaction.
1566 This applies the changes done within the transaction to the mainline and ends
1571 This discards the changes done within the transaction to the mainline and ends
1576 Transactions in DBM::Deep are done using the MVCC method, the same method used
1577 by the InnoDB MySQL table type.
1579 =head1 CAVEATS / ISSUES / BUGS
1581 This section describes all the known issues with DBM::Deep. It you have found
1582 something that is not listed here, please send e-mail to L<jhuckaby@cpan.org>.
1584 =head2 UNUSED SPACE RECOVERY
1586 One major caveat with DBM::Deep is that space occupied by existing keys and
1587 values is not recovered when they are deleted. Meaning if you keep deleting
1588 and adding new keys, your file will continuously grow. I am working on this,
1589 but in the meantime you can call the built-in C<optimize()> method from time to
1590 time (perhaps in a crontab or something) to recover all your unused space.
1592 $db->optimize(); # returns true on success
1594 This rebuilds the ENTIRE database into a new file, then moves it on top of
1595 the original. The new file will have no unused space, thus it will take up as
1596 little disk space as possible. Please note that this operation can take
1597 a long time for large files, and you need enough disk space to temporarily hold
1598 2 copies of your DB file. The temporary file is created in the same directory
1599 as the original, named with a ".tmp" extension, and is deleted when the
1600 operation completes. Oh, and if locking is enabled, the DB is automatically
1601 locked for the entire duration of the copy.
1603 B<WARNING:> Only call optimize() on the top-level node of the database, and
1604 make sure there are no child references lying around. DBM::Deep keeps a reference
1605 counter, and if it is greater than 1, optimize() will abort and return undef.
1609 (The reasons given assume a high level of Perl understanding, specifically of
1610 references. You can safely skip this section.)
1612 Currently, the only references supported are HASH and ARRAY. The other reference
1613 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
1619 These are things like filehandles and other sockets. They can't be supported
1620 because it's completely unclear how DBM::Deep should serialize them.
1622 =item * SCALAR / REF
1624 The discussion here refers to the following type of example:
1631 # In some other process ...
1633 my $val = ${ $db->{key1} };
1635 is( $val, 50, "What actually gets stored in the DB file?" );
1637 The problem is one of synchronization. When the variable being referred to
1638 changes value, the reference isn't notified. This means that the new value won't
1639 be stored in the datafile for other processes to read. There is no TIEREF.
1641 It is theoretically possible to store references to values already within a
1642 DBM::Deep object because everything already is synchronized, but the change to
1643 the internals would be quite large. Specifically, DBM::Deep would have to tie
1644 every single value that is stored. This would bloat the RAM footprint of
1645 DBM::Deep at least twofold (if not more) and be a significant performance drain,
1646 all to support a feature that has never been requested.
1650 L<Data::Dump::Streamer/> provides a mechanism for serializing coderefs,
1651 including saving off all closure state. However, just as for SCALAR and REF,
1652 that closure state may change without notifying the DBM::Deep object storing
1657 =head2 FILE CORRUPTION
1659 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
1660 for a 32-bit signature when opened, but other corruption in files can cause
1661 segmentation faults. DBM::Deep may try to seek() past the end of a file, or get
1662 stuck in an infinite loop depending on the level of corruption. File write
1663 operations are not checked for failure (for speed), so if you happen to run
1664 out of disk space, DBM::Deep will probably fail in a bad way. These things will
1665 be addressed in a later version of DBM::Deep.
1669 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
1670 well on local filesystems, but will NOT protect you from file corruption over
1671 NFS. I've heard about setting up your NFS server with a locking daemon, then
1672 using lockf() to lock your files, but your mileage may vary there as well.
1673 From what I understand, there is no real way to do it. However, if you need
1674 access to the underlying filehandle in DBM::Deep for using some other kind of
1675 locking scheme like lockf(), see the L<LOW-LEVEL ACCESS> section above.
1677 =head2 COPYING OBJECTS
1679 Beware of copying tied objects in Perl. Very strange things can happen.
1680 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
1681 returns a new, blessed, tied hash or array to the same level in the DB.
1683 my $copy = $db->clone();
1685 B<Note>: Since clone() here is cloning the object, not the database location, any
1686 modifications to either $db or $copy will be visible to both.
1690 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1691 These functions cause every element in the array to move, which can be murder
1692 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1693 a different location. This will be addressed in the forthcoming version 1.00.
1695 =head2 WRITEONLY FILES
1697 If you pass in a filehandle to new(), you may have opened it in either a readonly or
1698 writeonly mode. STORE will verify that the filehandle is writable. However, there
1699 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
1700 filehandle isn't readable, it's not clear what will happen. So, don't do that.
1702 =head1 CODE COVERAGE
1704 B<Devel::Cover> is used to test the code coverage of the tests. Below is the
1705 B<Devel::Cover> report on this distribution's test suite.
1707 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1708 File stmt bran cond sub pod time total
1709 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1710 blib/lib/DBM/Deep.pm 96.2 89.0 75.0 95.8 89.5 36.0 92.9
1711 blib/lib/DBM/Deep/Array.pm 96.1 88.3 100.0 96.4 100.0 15.9 94.7
1712 blib/lib/DBM/Deep/Engine.pm 96.6 86.6 89.5 100.0 0.0 20.0 91.0
1713 blib/lib/DBM/Deep/File.pm 99.4 88.3 55.6 100.0 0.0 19.6 89.5
1714 blib/lib/DBM/Deep/Hash.pm 98.5 83.3 100.0 100.0 100.0 8.5 96.3
1715 Total 96.9 87.4 81.2 98.0 38.5 100.0 92.1
1716 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1718 =head1 MORE INFORMATION
1720 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1721 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1724 The source code repository is at L<http://svn.perl.org/modules/DBM-Deep>
1728 Rob Kinyon, L<rkinyon@cpan.org>
1730 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1732 Special thanks to Adam Sah and Rich Gaushell! You know why :-)
1736 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1737 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1741 Copyright (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
1742 This is free software, you may use it and distribute it under the
1743 same terms as Perl itself.