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 FileHandle::Fmode ();
44 use DBM::Deep::Engine;
48 # Setup constants for users to pass to new()
50 sub TYPE_HASH () { DBM::Deep::Engine->SIG_HASH }
51 sub TYPE_ARRAY () { DBM::Deep::Engine->SIG_ARRAY }
59 $proto->_throw_error( "Odd number of parameters to " . (caller(1))[2] );
64 unless ( eval { local $SIG{'__DIE__'}; %{$_[0]} || 1 } ) {
65 $proto->_throw_error( "Not a hashref in args to " . (caller(1))[2] );
70 $args = { file => shift };
78 # Class constructor method for Perl OO interface.
79 # Calls tie() and returns blessed reference to tied hash or array,
80 # providing a hybrid OO/tie interface.
83 my $args = $class->_get_args( @_ );
86 # Check if we want a tied hash or array.
89 if (defined($args->{type}) && $args->{type} eq TYPE_ARRAY) {
90 $class = 'DBM::Deep::Array';
91 require DBM::Deep::Array;
92 tie @$self, $class, %$args;
95 $class = 'DBM::Deep::Hash';
96 require DBM::Deep::Hash;
97 tie %$self, $class, %$args;
100 return bless $self, $class;
103 # This initializer is called from the various TIE* methods. new() calls tie(),
104 # which allows for a single point of entry.
109 $args->{fileobj} = DBM::Deep::File->new( $args )
110 unless exists $args->{fileobj};
112 # locking implicitly enables autoflush
113 if ($args->{locking}) { $args->{autoflush} = 1; }
115 # These are the defaults to be optionally overridden below
118 base_offset => undef,
125 $self->{engine} = DBM::Deep::Engine->new( { %{$args}, obj => $self } );
127 # Grab the parameters we want to use
128 foreach my $param ( keys %$self ) {
129 next unless exists $args->{$param};
130 $self->{$param} = $args->{$param};
133 $self->_engine->setup_fh( $self );
135 $self->{fileobj}->set_db( $self );
142 require DBM::Deep::Hash;
143 return DBM::Deep::Hash->TIEHASH( @_ );
148 require DBM::Deep::Array;
149 return DBM::Deep::Array->TIEARRAY( @_ );
153 my $self = shift->_get_self;
154 return $self->_fileobj->lock( $self, @_ );
158 my $self = shift->_get_self;
159 return $self->_fileobj->unlock( $self, @_ );
163 my $self = shift->_get_self;
164 my ($spot, $value) = @_;
169 elsif ( eval { local $SIG{__DIE__}; $value->isa( 'DBM::Deep' ) } ) {
170 ${$spot} = $value->_repr;
171 $value->_copy_node( ${$spot} );
174 my $r = Scalar::Util::reftype( $value );
175 my $c = Scalar::Util::blessed( $value );
176 if ( $r eq 'ARRAY' ) {
177 ${$spot} = [ @{$value} ];
180 ${$spot} = { %{$value} };
182 ${$spot} = bless ${$spot}, $c
190 die "Must be implemented in a child class\n";
194 die "Must be implemented in a child class\n";
199 # Recursively export into standard Perl hashes and arrays.
201 my $self = shift->_get_self;
203 my $temp = $self->_repr;
206 $self->_copy_node( $temp );
209 # This will always work because $self, after _get_self() is a HASH
210 if ( $self->{parent} ) {
211 my $c = Scalar::Util::blessed(
212 $self->{parent}->get($self->{parent_key})
214 if ( $c && !$c->isa( 'DBM::Deep' ) ) {
224 # Recursively import Perl hash/array structure
226 if (!ref($_[0])) { return; } # Perl calls import() on use -- ignore
228 my $self = shift->_get_self;
231 # struct is not a reference, so just import based on our type
233 $struct = $self->_repr( @_ );
236 #XXX These are correct, but impossible until the other bug is fixed
239 $self->_import( $struct );
249 #XXX Need to keep track of who has a fh to this file in order to
250 #XXX close them all prior to optimize on Win32/cygwin
253 # Rebuild entire database into new file, then move
254 # it back on top of original.
256 my $self = shift->_get_self;
258 #XXX Need to create a new test for this
259 # if ($self->_fileobj->{links} > 1) {
260 # $self->_throw_error("Cannot optimize: reference count is greater than 1");
263 #XXX Do we have to lock the tempfile?
265 my $db_temp = DBM::Deep->new(
266 file => $self->_fileobj->{file} . '.tmp',
271 $self->_copy_node( $db_temp );
275 # Attempt to copy user, group and permissions over to new file
277 my @stats = stat($self->_fh);
278 my $perms = $stats[2] & 07777;
281 chown( $uid, $gid, $self->_fileobj->{file} . '.tmp' );
282 chmod( $perms, $self->_fileobj->{file} . '.tmp' );
284 # q.v. perlport for more information on this variable
285 if ( $^O eq 'MSWin32' || $^O eq 'cygwin' ) {
287 # Potential race condition when optmizing on Win32 with locking.
288 # The Windows filesystem requires that the filehandle be closed
289 # before it is overwritten with rename(). This could be redone
293 $self->_fileobj->close;
296 if (!rename $self->_fileobj->{file} . '.tmp', $self->_fileobj->{file}) {
297 unlink $self->_fileobj->{file} . '.tmp';
299 $self->_throw_error("Optimize failed: Cannot copy temp file over original: $!");
303 $self->_fileobj->close;
304 $self->_fileobj->open;
305 $self->_engine->setup_fh( $self );
312 # Make copy of object and return
314 my $self = shift->_get_self;
316 return DBM::Deep->new(
317 type => $self->_type,
318 base_offset => $self->_base_offset,
319 fileobj => $self->_fileobj,
324 my %is_legal_filter = map {
327 store_key store_value
328 fetch_key fetch_value
333 # Setup filter function for storing or fetching the key or value
335 my $self = shift->_get_self;
339 if ( $is_legal_filter{$type} ) {
340 $self->_fileobj->{"filter_$type"} = $func;
349 my $self = shift->_get_self;
350 $self->_fileobj->begin_transaction;
355 my $self = shift->_get_self;
356 $self->_fileobj->end_transaction;
361 my $self = shift->_get_self;
362 $self->_fileobj->commit_transaction;
371 my $self = $_[0]->_get_self;
372 return $self->{engine};
376 my $self = $_[0]->_get_self;
377 return $self->{fileobj};
381 my $self = $_[0]->_get_self;
382 return $self->{type};
386 my $self = $_[0]->_get_self;
387 return $self->{base_offset};
391 my $self = $_[0]->_get_self;
392 return $self->_fileobj->{fh};
400 die "DBM::Deep: $_[1]\n";
407 #XXX This if() is redundant
408 if ( my $parent = $self->{parent} ) {
410 while ( $parent->{parent} ) {
412 $parent->_type eq TYPE_HASH
413 ? "\{q{$child->{parent_key}}\}"
414 : "\[$child->{parent_key}\]"
418 $parent = $parent->{parent};
421 $base = "\$db->get( q{$child->{parent_key}} )->" . $base;
424 $base = "\$db->get( q{$child->{parent_key}} )";
432 # Store single hash key/value or array element in database.
434 my $self = shift->_get_self;
435 my ($key, $value, $orig_key) = @_;
438 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
439 $self->_throw_error( 'Cannot write to a readonly filehandle' );
442 #XXX The second condition needs to disappear
443 if ( defined $orig_key && !( $self->_type eq TYPE_ARRAY && $orig_key eq 'length') ) {
446 my $r = Scalar::Util::reftype( $value ) || '';
447 if ( $r eq 'HASH' ) {
450 elsif ( $r eq 'ARRAY' ) {
453 elsif ( defined $value ) {
460 if ( my $c = Scalar::Util::blessed( $value ) ) {
461 $rhs = "bless $rhs, '$c'";
464 my $lhs = $self->_find_parent;
466 if ( $self->_type eq TYPE_HASH ) {
467 $lhs .= "->\{q{$orig_key}\}";
470 $lhs .= "->\[$orig_key\]";
476 $lhs = "\$db->put(q{$orig_key},$rhs);";
479 $self->_fileobj->audit($lhs);
483 # Request exclusive lock for writing
485 $self->lock( LOCK_EX );
487 my $md5 = $self->_engine->{digest}->($key);
489 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5, { create => 1 } );
491 # User may be storing a hash, in which case we do not want it run
492 # through the filtering system
493 if ( !ref($value) && $self->_fileobj->{filter_store_value} ) {
494 $value = $self->_fileobj->{filter_store_value}->( $value );
498 # Add key/value to bucket list
500 $self->_engine->add_bucket( $tag, $md5, $key, $value, undef, $orig_key );
509 # Fetch single value or element given plain key or array index
511 my $self = shift->_get_self;
512 my ($key, $orig_key) = @_;
514 my $md5 = $self->_engine->{digest}->($key);
517 # Request shared lock for reading
519 $self->lock( LOCK_SH );
521 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );#, { create => 1 } );
522 #XXX This needs to autovivify
529 # Get value from bucket list
531 my $result = $self->_engine->get_bucket_value( $tag, $md5, $orig_key );
535 # Filters only apply to scalar values, so the ref check is making
536 # sure the fetched bucket is a scalar, not a child hash or array.
537 return ($result && !ref($result) && $self->_fileobj->{filter_fetch_value})
538 ? $self->_fileobj->{filter_fetch_value}->($result)
544 # Delete single key/value pair or element given plain key or array index
546 my $self = shift->_get_self;
547 my ($key, $orig_key) = @_;
549 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
550 $self->_throw_error( 'Cannot write to a readonly filehandle' );
553 if ( defined $orig_key ) {
554 my $lhs = $self->_find_parent;
556 $self->_fileobj->audit( "delete $lhs;" );
559 $self->_fileobj->audit( "\$db->delete('$orig_key');" );
564 # Request exclusive lock for writing
566 $self->lock( LOCK_EX );
568 my $md5 = $self->_engine->{digest}->($key);
570 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
579 my $value = $self->_engine->get_bucket_value( $tag, $md5 );
581 if (defined $value && !ref($value) && $self->_fileobj->{filter_fetch_value}) {
582 $value = $self->_fileobj->{filter_fetch_value}->($value);
585 my $result = $self->_engine->delete_bucket( $tag, $md5, $orig_key );
588 # If this object is an array and the key deleted was on the end of the stack,
589 # decrement the length variable.
599 # Check if a single key or element exists given plain key or array index
601 my $self = shift->_get_self;
604 my $md5 = $self->_engine->{digest}->($key);
607 # Request shared lock for reading
609 $self->lock( LOCK_SH );
611 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
616 # For some reason, the built-in exists() function returns '' for false
622 # Check if bucket exists and return 1 or ''
624 my $result = $self->_engine->bucket_exists( $tag, $md5 ) || '';
633 # Clear all keys from hash, or all elements from array.
635 my $self = shift->_get_self;
637 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
638 $self->_throw_error( 'Cannot write to a readonly filehandle' );
642 my $lhs = $self->_find_parent;
644 if ( $self->_type eq TYPE_HASH ) {
645 $lhs = '%{' . $lhs . '}';
648 $lhs = '@{' . $lhs . '}';
651 $self->_fileobj->audit( "$lhs = ();" );
655 # Request exclusive lock for writing
657 $self->lock( LOCK_EX );
659 if ( $self->_type eq TYPE_HASH ) {
660 my $key = $self->first_key;
662 my $next_key = $self->next_key( $key );
663 my $md5 = $self->_engine->{digest}->($key);
664 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
665 $self->_engine->delete_bucket( $tag, $md5, $key );
670 my $size = $self->FETCHSIZE;
671 for my $key ( map { pack ( $self->_engine->{long_pack}, $_ ) } 0 .. $size - 1 ) {
672 my $md5 = $self->_engine->{digest}->($key);
673 my $tag = $self->_engine->find_blist( $self->_base_offset, $md5 );
674 $self->_engine->delete_bucket( $tag, $md5, $key );
676 $self->STORESIZE( 0 );
678 #XXX This needs updating to use _release_space
679 # $self->_engine->write_tag(
680 # $self->_base_offset, $self->_type,
681 # chr(0)x$self->_engine->{index_size},
690 # Public method aliases
692 sub put { (shift)->STORE( @_ ) }
693 sub store { (shift)->STORE( @_ ) }
694 sub get { (shift)->FETCH( @_ ) }
695 sub fetch { (shift)->FETCH( @_ ) }
696 sub delete { (shift)->DELETE( @_ ) }
697 sub exists { (shift)->EXISTS( @_ ) }
698 sub clear { (shift)->CLEAR( @_ ) }
705 DBM::Deep - A pure perl multi-level hash/array DBM
710 my $db = DBM::Deep->new( "foo.db" );
712 $db->{key} = 'value';
715 $db->put('key' => 'value');
716 print $db->get('key');
718 # true multi-level support
719 $db->{my_complex} = [
720 'hello', { perl => 'rules' },
724 tie my %db, 'DBM::Deep', 'foo.db';
728 tied(%db)->put('key' => 'value');
729 print tied(%db)->get('key');
733 A unique flat-file database module, written in pure perl. True multi-level
734 hash/array support (unlike MLDBM, which is faked), hybrid OO / tie()
735 interface, cross-platform FTPable files, ACID transactions, and is quite fast.
736 Can handle millions of keys and unlimited levels without significant
737 slow-down. Written from the ground-up in pure perl -- this is NOT a wrapper
738 around a C-based DBM. Out-of-the-box compatibility with Unix, Mac OS X and
741 =head1 VERSION DIFFERENCES
743 B<NOTE>: 0.99_01 and above have significant file format differences from 0.983 and
744 before. There will be a backwards-compatibility layer in 1.00, but that is
745 slated for a later 0.99_x release. This version is B<NOT> backwards compatible
746 with 0.983 and before.
750 Construction can be done OO-style (which is the recommended way), or using
751 Perl's tie() function. Both are examined here.
753 =head2 OO CONSTRUCTION
755 The recommended way to construct a DBM::Deep object is to use the new()
756 method, which gets you a blessed I<and> tied hash (or array) reference.
758 my $db = DBM::Deep->new( "foo.db" );
760 This opens a new database handle, mapped to the file "foo.db". If this
761 file does not exist, it will automatically be created. DB files are
762 opened in "r+" (read/write) mode, and the type of object returned is a
763 hash, unless otherwise specified (see L<OPTIONS> below).
765 You can pass a number of options to the constructor to specify things like
766 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
768 my $db = DBM::Deep->new(
774 Notice that the filename is now specified I<inside> the hash with
775 the "file" parameter, as opposed to being the sole argument to the
776 constructor. This is required if any options are specified.
777 See L<OPTIONS> below for the complete list.
779 You can also start with an array instead of a hash. For this, you must
780 specify the C<type> parameter:
782 my $db = DBM::Deep->new(
784 type => DBM::Deep->TYPE_ARRAY
787 B<Note:> Specifing the C<type> parameter only takes effect when beginning
788 a new DB file. If you create a DBM::Deep object with an existing file, the
789 C<type> will be loaded from the file header, and an error will be thrown if
790 the wrong type is passed in.
792 =head2 TIE CONSTRUCTION
794 Alternately, you can create a DBM::Deep handle by using Perl's built-in
795 tie() function. The object returned from tie() can be used to call methods,
796 such as lock() and unlock(). (That object can be retrieved from the tied
797 variable at any time using tied() - please see L<perltie/> for more info.
800 my $db = tie %hash, "DBM::Deep", "foo.db";
803 my $db = tie @array, "DBM::Deep", "bar.db";
805 As with the OO constructor, you can replace the DB filename parameter with
806 a hash containing one or more options (see L<OPTIONS> just below for the
809 tie %hash, "DBM::Deep", {
817 There are a number of options that can be passed in when constructing your
818 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
824 Filename of the DB file to link the handle to. You can pass a full absolute
825 filesystem path, partial path, or a plain filename if the file is in the
826 current working directory. This is a required parameter (though q.v. fh).
830 If you want, you can pass in the fh instead of the file. This is most useful for doing
833 my $db = DBM::Deep->new( { fh => \*DATA } );
835 You are responsible for making sure that the fh has been opened appropriately for your
836 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
837 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
838 needs to read from the fh.
840 =item * audit_file / audit_fh
842 These are just like file/fh, except for auditing. Please see L</AUDITING> for
847 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
848 not need to set this. However, it's there if you want it.
850 If you pass in fh and do not set this, it will be set appropriately.
854 This parameter specifies what type of object to create, a hash or array. Use
855 one of these two constants:
859 =item * C<DBM::Deep-E<gt>TYPE_HASH>
861 =item * C<DBM::Deep-E<gt>TYPE_ARRAY>.
865 This only takes effect when beginning a new file. This is an optional
866 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
870 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
871 function to lock the database in exclusive mode for writes, and shared mode
872 for reads. Pass any true value to enable. This affects the base DB handle
873 I<and any child hashes or arrays> that use the same DB file. This is an
874 optional parameter, and defaults to 0 (disabled). See L<LOCKING> below for
879 Specifies whether autoflush is to be enabled on the underlying filehandle.
880 This obviously slows down write operations, but is required if you may have
881 multiple processes accessing the same DB file (also consider enable I<locking>).
882 Pass any true value to enable. This is an optional parameter, and defaults to 0
887 If I<autobless> mode is enabled, DBM::Deep will preserve the class something
888 is blessed into, and restores it when fetched. This is an optional parameter, and defaults to 1 (enabled).
890 B<Note:> If you use the OO-interface, you will not be able to call any methods
891 of DBM::Deep on the blessed item. This is considered to be a feature.
895 See L</FILTERS> below.
901 With DBM::Deep you can access your databases using Perl's standard hash/array
902 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
903 treat them as such. DBM::Deep will intercept all reads/writes and direct them
904 to the right place -- the DB file. This has nothing to do with the
905 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
906 using regular hashes and arrays, rather than calling functions like C<get()>
907 and C<put()> (although those work too). It is entirely up to you how to want
908 to access your databases.
912 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
913 or even nested hashes (or arrays) using standard Perl syntax:
915 my $db = DBM::Deep->new( "foo.db" );
917 $db->{mykey} = "myvalue";
919 $db->{myhash}->{subkey} = "subvalue";
921 print $db->{myhash}->{subkey} . "\n";
923 You can even step through hash keys using the normal Perl C<keys()> function:
925 foreach my $key (keys %$db) {
926 print "$key: " . $db->{$key} . "\n";
929 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
930 pushes them onto an array, all before the loop even begins. If you have an
931 extremely large hash, this may exhaust Perl's memory. Instead, consider using
932 Perl's C<each()> function, which pulls keys/values one at a time, using very
935 while (my ($key, $value) = each %$db) {
936 print "$key: $value\n";
939 Please note that when using C<each()>, you should always pass a direct
940 hash reference, not a lookup. Meaning, you should B<never> do this:
943 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
945 This causes an infinite loop, because for each iteration, Perl is calling
946 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
947 it effectively keeps returning the first key over and over again. Instead,
948 assign a temporary variable to C<$db->{foo}>, then pass that to each().
952 As with hashes, you can treat any DBM::Deep object like a normal Perl array
953 reference. This includes inserting, removing and manipulating elements,
954 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
955 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
956 or simply be a nested array reference inside a hash. Example:
958 my $db = DBM::Deep->new(
959 file => "foo-array.db",
960 type => DBM::Deep->TYPE_ARRAY
964 push @$db, "bar", "baz";
967 my $last_elem = pop @$db; # baz
968 my $first_elem = shift @$db; # bah
969 my $second_elem = $db->[1]; # bar
971 my $num_elements = scalar @$db;
975 In addition to the I<tie()> interface, you can also use a standard OO interface
976 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
977 array) has its own methods, but both types share the following common methods:
978 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
979 C<store(> are aliases to C<put()> and C<get()>, respectively.
983 =item * new() / clone()
985 These are the constructor and copy-functions.
987 =item * put() / store()
989 Stores a new hash key/value pair, or sets an array element value. Takes two
990 arguments, the hash key or array index, and the new value. The value can be
991 a scalar, hash ref or array ref. Returns true on success, false on failure.
993 $db->put("foo", "bar"); # for hashes
994 $db->put(1, "bar"); # for arrays
996 =item * get() / fetch()
998 Fetches the value of a hash key or array element. Takes one argument: the hash
999 key or array index. Returns a scalar, hash ref or array ref, depending on the
1002 my $value = $db->get("foo"); # for hashes
1003 my $value = $db->get(1); # for arrays
1007 Checks if a hash key or array index exists. Takes one argument: the hash key
1008 or array index. Returns true if it exists, false if not.
1010 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
1011 if ($db->exists(1)) { print "yay!\n"; } # for arrays
1015 Deletes one hash key/value pair or array element. Takes one argument: the hash
1016 key or array index. Returns true on success, false if not found. For arrays,
1017 the remaining elements located after the deleted element are NOT moved over.
1018 The deleted element is essentially just undefined, which is exactly how Perl's
1019 internal arrays work. Please note that the space occupied by the deleted
1020 key/value or element is B<not> reused again -- see L<UNUSED SPACE RECOVERY>
1021 below for details and workarounds.
1023 $db->delete("foo"); # for hashes
1024 $db->delete(1); # for arrays
1028 Deletes B<all> hash keys or array elements. Takes no arguments. No return
1029 value. Please note that the space occupied by the deleted keys/values or
1030 elements is B<not> reused again -- see L<UNUSED SPACE RECOVERY> below for
1031 details and workarounds.
1033 $db->clear(); # hashes or arrays
1035 =item * lock() / unlock()
1041 Recover lost disk space. This is important to do, especially if you use
1044 =item * import() / export()
1046 Data going in and out.
1052 For hashes, DBM::Deep supports all the common methods described above, and the
1053 following additional methods: C<first_key()> and C<next_key()>.
1059 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
1060 fetched in an undefined order (which appears random). Takes no arguments,
1061 returns the key as a scalar value.
1063 my $key = $db->first_key();
1067 Returns the "next" key in the hash, given the previous one as the sole argument.
1068 Returns undef if there are no more keys to be fetched.
1070 $key = $db->next_key($key);
1074 Here are some examples of using hashes:
1076 my $db = DBM::Deep->new( "foo.db" );
1078 $db->put("foo", "bar");
1079 print "foo: " . $db->get("foo") . "\n";
1081 $db->put("baz", {}); # new child hash ref
1082 $db->get("baz")->put("buz", "biz");
1083 print "buz: " . $db->get("baz")->get("buz") . "\n";
1085 my $key = $db->first_key();
1087 print "$key: " . $db->get($key) . "\n";
1088 $key = $db->next_key($key);
1091 if ($db->exists("foo")) { $db->delete("foo"); }
1095 For arrays, DBM::Deep supports all the common methods described above, and the
1096 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
1097 C<unshift()> and C<splice()>.
1103 Returns the number of elements in the array. Takes no arguments.
1105 my $len = $db->length();
1109 Adds one or more elements onto the end of the array. Accepts scalars, hash
1110 refs or array refs. No return value.
1112 $db->push("foo", "bar", {});
1116 Fetches the last element in the array, and deletes it. Takes no arguments.
1117 Returns undef if array is empty. Returns the element value.
1119 my $elem = $db->pop();
1123 Fetches the first element in the array, deletes it, then shifts all the
1124 remaining elements over to take up the space. Returns the element value. This
1125 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
1128 my $elem = $db->shift();
1132 Inserts one or more elements onto the beginning of the array, shifting all
1133 existing elements over to make room. Accepts scalars, hash refs or array refs.
1134 No return value. This method is not recommended with large arrays -- see
1135 <LARGE ARRAYS> below for details.
1137 $db->unshift("foo", "bar", {});
1141 Performs exactly like Perl's built-in function of the same name. See L<perldoc
1142 -f splice> for usage -- it is too complicated to document here. This method is
1143 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
1147 Here are some examples of using arrays:
1149 my $db = DBM::Deep->new(
1151 type => DBM::Deep->TYPE_ARRAY
1154 $db->push("bar", "baz");
1155 $db->unshift("foo");
1158 my $len = $db->length();
1159 print "length: $len\n"; # 4
1161 for (my $k=0; $k<$len; $k++) {
1162 print "$k: " . $db->get($k) . "\n";
1165 $db->splice(1, 2, "biz", "baf");
1167 while (my $elem = shift @$db) {
1168 print "shifted: $elem\n";
1173 Enable automatic file locking by passing a true value to the C<locking>
1174 parameter when constructing your DBM::Deep object (see L<SETUP> above).
1176 my $db = DBM::Deep->new(
1181 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
1182 mode for writes, and shared mode for reads. This is required if you have
1183 multiple processes accessing the same database file, to avoid file corruption.
1184 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
1185 NFS> below for more.
1187 =head2 EXPLICIT LOCKING
1189 You can explicitly lock a database, so it remains locked for multiple
1190 transactions. This is done by calling the C<lock()> method, and passing an
1191 optional lock mode argument (defaults to exclusive mode). This is particularly
1192 useful for things like counters, where the current value needs to be fetched,
1193 then incremented, then stored again.
1196 my $counter = $db->get("counter");
1198 $db->put("counter", $counter);
1207 You can pass C<lock()> an optional argument, which specifies which mode to use
1208 (exclusive or shared). Use one of these two constants:
1209 C<DBM::Deep-E<gt>LOCK_EX> or C<DBM::Deep-E<gt>LOCK_SH>. These are passed
1210 directly to C<flock()>, and are the same as the constants defined in Perl's
1213 $db->lock( $db->LOCK_SH );
1217 =head1 IMPORTING/EXPORTING
1219 You can import existing complex structures by calling the C<import()> method,
1220 and export an entire database into an in-memory structure using the C<export()>
1221 method. Both are examined here.
1225 Say you have an existing hash with nested hashes/arrays inside it. Instead of
1226 walking the structure and adding keys/elements to the database as you go,
1227 simply pass a reference to the C<import()> method. This recursively adds
1228 everything to an existing DBM::Deep object for you. Here is an example:
1233 array1 => [ "elem0", "elem1", "elem2" ],
1235 subkey1 => "subvalue1",
1236 subkey2 => "subvalue2"
1240 my $db = DBM::Deep->new( "foo.db" );
1241 $db->import( $struct );
1243 print $db->{key1} . "\n"; # prints "value1"
1245 This recursively imports the entire C<$struct> object into C<$db>, including
1246 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
1247 keys are merged with the existing ones, replacing if they already exist.
1248 The C<import()> method can be called on any database level (not just the base
1249 level), and works with both hash and array DB types.
1251 B<Note:> Make sure your existing structure has no circular references in it.
1252 These will cause an infinite loop when importing. There are plans to fix this
1257 Calling the C<export()> method on an existing DBM::Deep object will return
1258 a reference to a new in-memory copy of the database. The export is done
1259 recursively, so all nested hashes/arrays are all exported to standard Perl
1260 objects. Here is an example:
1262 my $db = DBM::Deep->new( "foo.db" );
1264 $db->{key1} = "value1";
1265 $db->{key2} = "value2";
1267 $db->{hash1}->{subkey1} = "subvalue1";
1268 $db->{hash1}->{subkey2} = "subvalue2";
1270 my $struct = $db->export();
1272 print $struct->{key1} . "\n"; # prints "value1"
1274 This makes a complete copy of the database in memory, and returns a reference
1275 to it. The C<export()> method can be called on any database level (not just
1276 the base level), and works with both hash and array DB types. Be careful of
1277 large databases -- you can store a lot more data in a DBM::Deep object than an
1278 in-memory Perl structure.
1280 B<Note:> Make sure your database has no circular references in it.
1281 These will cause an infinite loop when exporting. There are plans to fix this
1286 DBM::Deep has a number of hooks where you can specify your own Perl function
1287 to perform filtering on incoming or outgoing data. This is a perfect
1288 way to extend the engine, and implement things like real-time compression or
1289 encryption. Filtering applies to the base DB level, and all child hashes /
1290 arrays. Filter hooks can be specified when your DBM::Deep object is first
1291 constructed, or by calling the C<set_filter()> method at any time. There are
1292 four available filter hooks, described below:
1296 =item * filter_store_key
1298 This filter is called whenever a hash key is stored. It
1299 is passed the incoming key, and expected to return a transformed key.
1301 =item * filter_store_value
1303 This filter is called whenever a hash key or array element is stored. It
1304 is passed the incoming value, and expected to return a transformed value.
1306 =item * filter_fetch_key
1308 This filter is called whenever a hash key is fetched (i.e. via
1309 C<first_key()> or C<next_key()>). It is passed the transformed key,
1310 and expected to return the plain key.
1312 =item * filter_fetch_value
1314 This filter is called whenever a hash key or array element is fetched.
1315 It is passed the transformed value, and expected to return the plain value.
1319 Here are the two ways to setup a filter hook:
1321 my $db = DBM::Deep->new(
1323 filter_store_value => \&my_filter_store,
1324 filter_fetch_value => \&my_filter_fetch
1329 $db->set_filter( "filter_store_value", \&my_filter_store );
1330 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
1332 Your filter function will be called only when dealing with SCALAR keys or
1333 values. When nested hashes and arrays are being stored/fetched, filtering
1334 is bypassed. Filters are called as static functions, passed a single SCALAR
1335 argument, and expected to return a single SCALAR value. If you want to
1336 remove a filter, set the function reference to C<undef>:
1338 $db->set_filter( "filter_store_value", undef );
1340 =head2 REAL-TIME ENCRYPTION EXAMPLE
1342 Here is a working example that uses the I<Crypt::Blowfish> module to
1343 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
1344 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
1345 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
1348 use Crypt::Blowfish;
1351 my $cipher = Crypt::CBC->new({
1352 'key' => 'my secret key',
1353 'cipher' => 'Blowfish',
1355 'regenerate_key' => 0,
1356 'padding' => 'space',
1360 my $db = DBM::Deep->new(
1361 file => "foo-encrypt.db",
1362 filter_store_key => \&my_encrypt,
1363 filter_store_value => \&my_encrypt,
1364 filter_fetch_key => \&my_decrypt,
1365 filter_fetch_value => \&my_decrypt,
1368 $db->{key1} = "value1";
1369 $db->{key2} = "value2";
1370 print "key1: " . $db->{key1} . "\n";
1371 print "key2: " . $db->{key2} . "\n";
1377 return $cipher->encrypt( $_[0] );
1380 return $cipher->decrypt( $_[0] );
1383 =head2 REAL-TIME COMPRESSION EXAMPLE
1385 Here is a working example that uses the I<Compress::Zlib> module to do real-time
1386 compression / decompression of keys & values with DBM::Deep Filters.
1387 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
1388 more on I<Compress::Zlib>.
1393 my $db = DBM::Deep->new(
1394 file => "foo-compress.db",
1395 filter_store_key => \&my_compress,
1396 filter_store_value => \&my_compress,
1397 filter_fetch_key => \&my_decompress,
1398 filter_fetch_value => \&my_decompress,
1401 $db->{key1} = "value1";
1402 $db->{key2} = "value2";
1403 print "key1: " . $db->{key1} . "\n";
1404 print "key2: " . $db->{key2} . "\n";
1410 return Compress::Zlib::memGzip( $_[0] ) ;
1413 return Compress::Zlib::memGunzip( $_[0] ) ;
1416 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
1417 actually numerical index numbers, and are not filtered.
1419 =head1 ERROR HANDLING
1421 Most DBM::Deep methods return a true value for success, and call die() on
1422 failure. You can wrap calls in an eval block to catch the die.
1424 my $db = DBM::Deep->new( "foo.db" ); # create hash
1425 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
1427 print $@; # prints error message
1429 =head1 LARGEFILE SUPPORT
1431 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
1432 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
1433 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
1434 by specifying the 'pack_size' parameter when constructing the file.
1437 filename => $filename,
1438 pack_size => 'large',
1441 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
1442 instead of 32-bit longs. After setting these values your DB files have a
1443 theoretical maximum size of 16 XB (exabytes).
1445 You can also use C<pack_size =E<gt> 'small'> in order to use 16-bit file
1448 B<Note:> Changing these values will B<NOT> work for existing database files.
1449 Only change this for new files. Once the value has been set, it is stored in
1450 the file's header and cannot be changed for the life of the file. These
1451 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
1454 B<Note:> We have not personally tested files larger than 2 GB -- all my
1455 systems have only a 32-bit Perl. However, I have received user reports that
1456 this does indeed work!
1458 =head1 LOW-LEVEL ACCESS
1460 If you require low-level access to the underlying filehandle that DBM::Deep uses,
1461 you can call the C<_fh()> method, which returns the handle:
1463 my $fh = $db->_fh();
1465 This method can be called on the root level of the datbase, or any child
1466 hashes or arrays. All levels share a I<root> structure, which contains things
1467 like the filehandle, a reference counter, and all the options specified
1468 when you created the object. You can get access to this file object by
1469 calling the C<_fileobj()> method.
1471 my $file_obj = $db->_fileobj();
1473 This is useful for changing options after the object has already been created,
1474 such as enabling/disabling locking. You can also store your own temporary user
1475 data in this structure (be wary of name collision), which is then accessible from
1476 any child hash or array.
1478 =head1 CUSTOM DIGEST ALGORITHM
1480 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
1481 keys. However you can override this, and use another algorithm (such as SHA-256)
1482 or even write your own. But please note that DBM::Deep currently expects zero
1483 collisions, so your algorithm has to be I<perfect>, so to speak. Collision
1484 detection may be introduced in a later version.
1486 You can specify a custom digest algorithm by passing it into the parameter
1487 list for new(), passing a reference to a subroutine as the 'digest' parameter,
1488 and the length of the algorithm's hashes (in bytes) as the 'hash_size'
1489 parameter. Here is a working example that uses a 256-bit hash from the
1490 I<Digest::SHA256> module. Please see
1491 L<http://search.cpan.org/search?module=Digest::SHA256> for more information.
1496 my $context = Digest::SHA256::new(256);
1498 my $db = DBM::Deep->new(
1499 filename => "foo-sha.db",
1500 digest => \&my_digest,
1504 $db->{key1} = "value1";
1505 $db->{key2} = "value2";
1506 print "key1: " . $db->{key1} . "\n";
1507 print "key2: " . $db->{key2} . "\n";
1513 return substr( $context->hash($_[0]), 0, 32 );
1516 B<Note:> Your returned digest strings must be B<EXACTLY> the number
1517 of bytes you specify in the hash_size parameter (in this case 32).
1519 B<Note:> If you do choose to use a custom digest algorithm, you must set it
1520 every time you access this file. Otherwise, the default (MD5) will be used.
1522 =head1 CIRCULAR REFERENCES
1524 DBM::Deep has B<experimental> support for circular references. Meaning you
1525 can have a nested hash key or array element that points to a parent object.
1526 This relationship is stored in the DB file, and is preserved between sessions.
1529 my $db = DBM::Deep->new( "foo.db" );
1532 $db->{circle} = $db; # ref to self
1534 print $db->{foo} . "\n"; # prints "bar"
1535 print $db->{circle}->{foo} . "\n"; # prints "bar" again
1537 B<Note>: Passing the object to a function that recursively walks the
1538 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
1539 C<export()> methods) will result in an infinite loop. This will be fixed in
1544 New in 0.99_01 is the ability to audit your databases actions. By passing in
1545 audit_file (or audit_fh) to the constructor, all actions will be logged to
1546 that file. The format is one that is suitable for eval'ing against the
1547 database to replay the actions. Please see t/33_audit_trail.t for an example
1552 New in 0.99_01 is ACID transactions. Every DBM::Deep object is completely
1553 transaction-ready - it is not an option you have to turn on. Three new methods
1554 have been added to support them. They are:
1558 =item * begin_work()
1560 This starts a transaction.
1564 This applies the changes done within the transaction to the mainline and ends
1569 This discards the changes done within the transaction to the mainline and ends
1574 Transactions in DBM::Deep are done using the MVCC method, the same method used
1575 by the InnoDB MySQL table type.
1577 =head1 CAVEATS / ISSUES / BUGS
1579 This section describes all the known issues with DBM::Deep. It you have found
1580 something that is not listed here, please send e-mail to L<jhuckaby@cpan.org>.
1582 =head2 UNUSED SPACE RECOVERY
1584 One major caveat with DBM::Deep is that space occupied by existing keys and
1585 values is not recovered when they are deleted. Meaning if you keep deleting
1586 and adding new keys, your file will continuously grow. I am working on this,
1587 but in the meantime you can call the built-in C<optimize()> method from time to
1588 time (perhaps in a crontab or something) to recover all your unused space.
1590 $db->optimize(); # returns true on success
1592 This rebuilds the ENTIRE database into a new file, then moves it on top of
1593 the original. The new file will have no unused space, thus it will take up as
1594 little disk space as possible. Please note that this operation can take
1595 a long time for large files, and you need enough disk space to temporarily hold
1596 2 copies of your DB file. The temporary file is created in the same directory
1597 as the original, named with a ".tmp" extension, and is deleted when the
1598 operation completes. Oh, and if locking is enabled, the DB is automatically
1599 locked for the entire duration of the copy.
1601 B<WARNING:> Only call optimize() on the top-level node of the database, and
1602 make sure there are no child references lying around. DBM::Deep keeps a reference
1603 counter, and if it is greater than 1, optimize() will abort and return undef.
1607 (The reasons given assume a high level of Perl understanding, specifically of
1608 references. You can safely skip this section.)
1610 Currently, the only references supported are HASH and ARRAY. The other reference
1611 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
1617 These are things like filehandles and other sockets. They can't be supported
1618 because it's completely unclear how DBM::Deep should serialize them.
1620 =item * SCALAR / REF
1622 The discussion here refers to the following type of example:
1629 # In some other process ...
1631 my $val = ${ $db->{key1} };
1633 is( $val, 50, "What actually gets stored in the DB file?" );
1635 The problem is one of synchronization. When the variable being referred to
1636 changes value, the reference isn't notified. This means that the new value won't
1637 be stored in the datafile for other processes to read. There is no TIEREF.
1639 It is theoretically possible to store references to values already within a
1640 DBM::Deep object because everything already is synchronized, but the change to
1641 the internals would be quite large. Specifically, DBM::Deep would have to tie
1642 every single value that is stored. This would bloat the RAM footprint of
1643 DBM::Deep at least twofold (if not more) and be a significant performance drain,
1644 all to support a feature that has never been requested.
1648 L<Data::Dump::Streamer/> provides a mechanism for serializing coderefs,
1649 including saving off all closure state. However, just as for SCALAR and REF,
1650 that closure state may change without notifying the DBM::Deep object storing
1655 =head2 FILE CORRUPTION
1657 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
1658 for a 32-bit signature when opened, but other corruption in files can cause
1659 segmentation faults. DBM::Deep may try to seek() past the end of a file, or get
1660 stuck in an infinite loop depending on the level of corruption. File write
1661 operations are not checked for failure (for speed), so if you happen to run
1662 out of disk space, DBM::Deep will probably fail in a bad way. These things will
1663 be addressed in a later version of DBM::Deep.
1667 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
1668 well on local filesystems, but will NOT protect you from file corruption over
1669 NFS. I've heard about setting up your NFS server with a locking daemon, then
1670 using lockf() to lock your files, but your mileage may vary there as well.
1671 From what I understand, there is no real way to do it. However, if you need
1672 access to the underlying filehandle in DBM::Deep for using some other kind of
1673 locking scheme like lockf(), see the L<LOW-LEVEL ACCESS> section above.
1675 =head2 COPYING OBJECTS
1677 Beware of copying tied objects in Perl. Very strange things can happen.
1678 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
1679 returns a new, blessed, tied hash or array to the same level in the DB.
1681 my $copy = $db->clone();
1683 B<Note>: Since clone() here is cloning the object, not the database location, any
1684 modifications to either $db or $copy will be visible to both.
1688 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1689 These functions cause every element in the array to move, which can be murder
1690 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1691 a different location. This will be addressed in the forthcoming version 1.00.
1693 =head2 WRITEONLY FILES
1695 If you pass in a filehandle to new(), you may have opened it in either a readonly or
1696 writeonly mode. STORE will verify that the filehandle is writable. However, there
1697 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
1698 filehandle isn't readable, it's not clear what will happen. So, don't do that.
1700 =head1 CODE COVERAGE
1702 B<Devel::Cover> is used to test the code coverage of the tests. Below is the
1703 B<Devel::Cover> report on this distribution's test suite.
1705 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1706 File stmt bran cond sub pod time total
1707 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1708 blib/lib/DBM/Deep.pm 96.2 89.0 75.0 95.8 89.5 36.0 92.9
1709 blib/lib/DBM/Deep/Array.pm 96.1 88.3 100.0 96.4 100.0 15.9 94.7
1710 blib/lib/DBM/Deep/Engine.pm 96.6 86.6 89.5 100.0 0.0 20.0 91.0
1711 blib/lib/DBM/Deep/File.pm 99.4 88.3 55.6 100.0 0.0 19.6 89.5
1712 blib/lib/DBM/Deep/Hash.pm 98.5 83.3 100.0 100.0 100.0 8.5 96.3
1713 Total 96.9 87.4 81.2 98.0 38.5 100.0 92.1
1714 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1716 =head1 MORE INFORMATION
1718 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1719 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1722 The source code repository is at L<http://svn.perl.org/modules/DBM-Deep>
1726 Rob Kinyon, L<rkinyon@cpan.org>
1728 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1730 Special thanks to Adam Sah and Rich Gaushell! You know why :-)
1734 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1735 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1739 Copyright (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
1740 This is free software, you may use it and distribute it under the
1741 same terms as Perl itself.