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( :flock );
41 use Clone::Any '_clone_data';
43 use FileHandle::Fmode ();
46 use DBM::Deep::Engine3;
50 # Setup constants for users to pass to new()
52 sub TYPE_HASH () { DBM::Deep::Engine3->SIG_HASH }
53 sub TYPE_ARRAY () { DBM::Deep::Engine3->SIG_ARRAY }
55 # This is used in all the children of this class in their TIE<type> methods.
62 $proto->_throw_error( "Odd number of parameters to " . (caller(1))[2] );
67 unless ( eval { local $SIG{'__DIE__'}; %{$_[0]} || 1 } ) {
68 $proto->_throw_error( "Not a hashref in args to " . (caller(1))[2] );
73 $args = { file => shift };
81 # Class constructor method for Perl OO interface.
82 # Calls tie() and returns blessed reference to tied hash or array,
83 # providing a hybrid OO/tie interface.
86 my $args = $class->_get_args( @_ );
89 # Check if we want a tied hash or array.
92 if (defined($args->{type}) && $args->{type} eq TYPE_ARRAY) {
93 $class = 'DBM::Deep::Array';
94 require DBM::Deep::Array;
95 tie @$self, $class, %$args;
98 $class = 'DBM::Deep::Hash';
99 require DBM::Deep::Hash;
100 tie %$self, $class, %$args;
103 return bless $self, $class;
106 # This initializer is called from the various TIE* methods. new() calls tie(),
107 # which allows for a single point of entry.
112 $args->{storage} = DBM::Deep::File->new( $args )
113 unless exists $args->{storage};
115 # locking implicitly enables autoflush
116 if ($args->{locking}) { $args->{autoflush} = 1; }
118 # These are the defaults to be optionally overridden below
121 base_offset => undef,
127 $args->{engine} = DBM::Deep::Engine3->new( { %{$args}, obj => $self } )
128 unless exists $args->{engine};
130 # Grab the parameters we want to use
131 foreach my $param ( keys %$self ) {
132 next unless exists $args->{$param};
133 $self->{$param} = $args->{$param};
137 local $SIG{'__DIE__'};
140 $self->_engine->setup_fh( $self );
141 $self->_storage->set_inode;
145 eval { local $SIG{'__DIE__'}; $self->unlock; };
154 require DBM::Deep::Hash;
155 return DBM::Deep::Hash->TIEHASH( @_ );
160 require DBM::Deep::Array;
161 return DBM::Deep::Array->TIEARRAY( @_ );
165 my $self = shift->_get_self;
166 return $self->_storage->lock( $self, @_ );
170 my $self = shift->_get_self;
171 return $self->_storage->unlock( $self, @_ );
175 my $self = shift->_get_self;
176 my ($spot, $value) = @_;
181 elsif ( eval { local $SIG{__DIE__}; $value->isa( 'DBM::Deep' ) } ) {
182 ${$spot} = $value->_repr;
183 $value->_copy_node( ${$spot} );
186 my $r = Scalar::Util::reftype( $value );
187 my $c = Scalar::Util::blessed( $value );
188 if ( $r eq 'ARRAY' ) {
189 ${$spot} = [ @{$value} ];
192 ${$spot} = { %{$value} };
194 ${$spot} = bless ${$spot}, $c
202 die "Must be implemented in a child class\n";
206 die "Must be implemented in a child class\n";
211 # Recursively export into standard Perl hashes and arrays.
213 my $self = shift->_get_self;
215 my $temp = $self->_repr;
218 $self->_copy_node( $temp );
221 my $classname = $self->_engine->get_classname( $self );
222 if ( defined $classname ) {
223 bless $temp, $classname;
231 # Recursively import Perl hash/array structure
233 if (!ref($_[0])) { return; } # Perl calls import() on use -- ignore
235 my $self = shift->_get_self;
238 # struct is not a reference, so just import based on our type
240 $struct = $self->_repr( @_ );
243 #XXX This isn't the best solution. Better would be to use Data::Walker,
244 #XXX but that's a lot more thinking than I want to do right now.
247 $self->_import( _clone_data( $struct ) );
257 #XXX Need to keep track of who has a fh to this file in order to
258 #XXX close them all prior to optimize on Win32/cygwin
261 # Rebuild entire database into new file, then move
262 # it back on top of original.
264 my $self = shift->_get_self;
266 #XXX Need to create a new test for this
267 # if ($self->_storage->{links} > 1) {
268 # $self->_throw_error("Cannot optimize: reference count is greater than 1");
271 #XXX Do we have to lock the tempfile?
273 my $db_temp = DBM::Deep->new(
274 file => $self->_storage->{file} . '.tmp',
279 $self->_copy_node( $db_temp );
283 # Attempt to copy user, group and permissions over to new file
285 my @stats = stat($self->_fh);
286 my $perms = $stats[2] & 07777;
289 chown( $uid, $gid, $self->_storage->{file} . '.tmp' );
290 chmod( $perms, $self->_storage->{file} . '.tmp' );
292 # q.v. perlport for more information on this variable
293 if ( $^O eq 'MSWin32' || $^O eq 'cygwin' ) {
295 # Potential race condition when optmizing on Win32 with locking.
296 # The Windows filesystem requires that the filehandle be closed
297 # before it is overwritten with rename(). This could be redone
301 $self->_storage->close;
304 if (!rename $self->_storage->{file} . '.tmp', $self->_storage->{file}) {
305 unlink $self->_storage->{file} . '.tmp';
307 $self->_throw_error("Optimize failed: Cannot copy temp file over original: $!");
311 $self->_storage->close;
312 $self->_storage->open;
313 $self->_engine->setup_fh( $self );
320 # Make copy of object and return
322 my $self = shift->_get_self;
324 return DBM::Deep->new(
325 type => $self->_type,
326 base_offset => $self->_base_offset,
327 storage => $self->_storage,
328 engine => $self->_engine,
333 my %is_legal_filter = map {
336 store_key store_value
337 fetch_key fetch_value
342 # Setup filter function for storing or fetching the key or value
344 my $self = shift->_get_self;
348 if ( $is_legal_filter{$type} ) {
349 $self->_storage->{"filter_$type"} = $func;
358 my $self = shift->_get_self;
359 return $self->_engine->begin_work( $self, @_ );
363 my $self = shift->_get_self;
364 return $self->_engine->rollback( $self, @_ );
368 my $self = shift->_get_self;
369 return $self->_engine->commit( $self, @_ );
377 my $self = $_[0]->_get_self;
378 return $self->{engine};
382 my $self = $_[0]->_get_self;
383 return $self->{storage};
387 my $self = $_[0]->_get_self;
388 return $self->{type};
392 my $self = $_[0]->_get_self;
393 return $self->{base_offset};
397 my $self = $_[0]->_get_self;
398 return $self->_storage->{fh};
406 die "DBM::Deep: $_[1]\n";
411 # Store single hash key/value or array element in database.
413 my $self = shift->_get_self;
414 my ($key, $value, $orig_key) = @_;
415 $orig_key = $key unless defined $orig_key;
417 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
418 $self->_throw_error( 'Cannot write to a readonly filehandle' );
422 # Request exclusive lock for writing
424 $self->lock( LOCK_EX );
426 # User may be storing a complex value, in which case we do not want it run
427 # through the filtering system.
428 if ( !ref($value) && $self->_storage->{filter_store_value} ) {
429 $value = $self->_storage->{filter_store_value}->( $value );
432 $self->_engine->write_value( $self, $key, $value, $orig_key );
441 # Fetch single value or element given plain key or array index
443 my $self = shift->_get_self;
444 my ($key, $orig_key) = @_;
445 $orig_key = $key unless defined $orig_key;
448 # Request shared lock for reading
450 $self->lock( LOCK_SH );
452 my $result = $self->_engine->read_value( $self, $key, $orig_key );
456 # Filters only apply to scalar values, so the ref check is making
457 # sure the fetched bucket is a scalar, not a child hash or array.
458 return ($result && !ref($result) && $self->_storage->{filter_fetch_value})
459 ? $self->_storage->{filter_fetch_value}->($result)
465 # Delete single key/value pair or element given plain key or array index
467 my $self = shift->_get_self;
468 my ($key, $orig_key) = @_;
469 $orig_key = $key unless defined $orig_key;
471 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
472 $self->_throw_error( 'Cannot write to a readonly filehandle' );
476 # Request exclusive lock for writing
478 $self->lock( LOCK_EX );
483 my $value = $self->_engine->delete_key( $self, $key, $orig_key );
485 if (defined $value && !ref($value) && $self->_storage->{filter_fetch_value}) {
486 $value = $self->_storage->{filter_fetch_value}->($value);
496 # Check if a single key or element exists given plain key or array index
498 my $self = shift->_get_self;
502 # Request shared lock for reading
504 $self->lock( LOCK_SH );
506 my $result = $self->_engine->key_exists( $self, $key );
515 # Clear all keys from hash, or all elements from array.
517 my $self = shift->_get_self;
519 if ( !FileHandle::Fmode::is_W( $self->_fh ) ) {
520 $self->_throw_error( 'Cannot write to a readonly filehandle' );
524 # Request exclusive lock for writing
526 $self->lock( LOCK_EX );
528 if ( $self->_type eq TYPE_HASH ) {
529 my $key = $self->first_key;
531 # Retrieve the key before deleting because we depend on next_key
532 my $next_key = $self->next_key( $key );
533 $self->_engine->delete_key( $self, $key, $key );
538 my $size = $self->FETCHSIZE;
539 for my $key ( 0 .. $size - 1 ) {
540 $self->_engine->delete_key( $self, $key, $key );
542 $self->STORESIZE( 0 );
544 #XXX This needs updating to use _release_space
545 # $self->_engine->write_tag(
546 # $self->_base_offset, $self->_type,
547 # chr(0)x$self->_engine->{index_size},
556 # Public method aliases
558 sub put { (shift)->STORE( @_ ) }
559 sub store { (shift)->STORE( @_ ) }
560 sub get { (shift)->FETCH( @_ ) }
561 sub fetch { (shift)->FETCH( @_ ) }
562 sub delete { (shift)->DELETE( @_ ) }
563 sub exists { (shift)->EXISTS( @_ ) }
564 sub clear { (shift)->CLEAR( @_ ) }
571 DBM::Deep - A pure perl multi-level hash/array DBM
576 my $db = DBM::Deep->new( "foo.db" );
578 $db->{key} = 'value';
581 $db->put('key' => 'value');
582 print $db->get('key');
584 # true multi-level support
585 $db->{my_complex} = [
586 'hello', { perl => 'rules' },
590 tie my %db, 'DBM::Deep', 'foo.db';
594 tied(%db)->put('key' => 'value');
595 print tied(%db)->get('key');
599 A unique flat-file database module, written in pure perl. True multi-level
600 hash/array support (unlike MLDBM, which is faked), hybrid OO / tie()
601 interface, cross-platform FTPable files, ACID transactions, and is quite fast.
602 Can handle millions of keys and unlimited levels without significant
603 slow-down. Written from the ground-up in pure perl -- this is NOT a wrapper
604 around a C-based DBM. Out-of-the-box compatibility with Unix, Mac OS X and
607 =head1 VERSION DIFFERENCES
609 B<NOTE>: 0.99_01 and above have significant file format differences from 0.983 and
610 before. There will be a backwards-compatibility layer in 1.00, but that is
611 slated for a later 0.99_x release. This version is B<NOT> backwards compatible
612 with 0.983 and before.
616 Construction can be done OO-style (which is the recommended way), or using
617 Perl's tie() function. Both are examined here.
619 =head2 OO CONSTRUCTION
621 The recommended way to construct a DBM::Deep object is to use the new()
622 method, which gets you a blessed I<and> tied hash (or array) reference.
624 my $db = DBM::Deep->new( "foo.db" );
626 This opens a new database handle, mapped to the file "foo.db". If this
627 file does not exist, it will automatically be created. DB files are
628 opened in "r+" (read/write) mode, and the type of object returned is a
629 hash, unless otherwise specified (see L<OPTIONS> below).
631 You can pass a number of options to the constructor to specify things like
632 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
634 my $db = DBM::Deep->new(
640 Notice that the filename is now specified I<inside> the hash with
641 the "file" parameter, as opposed to being the sole argument to the
642 constructor. This is required if any options are specified.
643 See L<OPTIONS> below for the complete list.
645 You can also start with an array instead of a hash. For this, you must
646 specify the C<type> parameter:
648 my $db = DBM::Deep->new(
650 type => DBM::Deep->TYPE_ARRAY
653 B<Note:> Specifing the C<type> parameter only takes effect when beginning
654 a new DB file. If you create a DBM::Deep object with an existing file, the
655 C<type> will be loaded from the file header, and an error will be thrown if
656 the wrong type is passed in.
658 =head2 TIE CONSTRUCTION
660 Alternately, you can create a DBM::Deep handle by using Perl's built-in
661 tie() function. The object returned from tie() can be used to call methods,
662 such as lock() and unlock(). (That object can be retrieved from the tied
663 variable at any time using tied() - please see L<perltie/> for more info.
666 my $db = tie %hash, "DBM::Deep", "foo.db";
669 my $db = tie @array, "DBM::Deep", "bar.db";
671 As with the OO constructor, you can replace the DB filename parameter with
672 a hash containing one or more options (see L<OPTIONS> just below for the
675 tie %hash, "DBM::Deep", {
683 There are a number of options that can be passed in when constructing your
684 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
690 Filename of the DB file to link the handle to. You can pass a full absolute
691 filesystem path, partial path, or a plain filename if the file is in the
692 current working directory. This is a required parameter (though q.v. fh).
696 If you want, you can pass in the fh instead of the file. This is most useful for doing
699 my $db = DBM::Deep->new( { fh => \*DATA } );
701 You are responsible for making sure that the fh has been opened appropriately for your
702 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
703 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
704 needs to read from the fh.
708 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
709 not need to set this. However, it's there if you want it.
711 If you pass in fh and do not set this, it will be set appropriately.
715 This parameter specifies what type of object to create, a hash or array. Use
716 one of these two constants:
720 =item * C<DBM::Deep-E<gt>TYPE_HASH>
722 =item * C<DBM::Deep-E<gt>TYPE_ARRAY>.
726 This only takes effect when beginning a new file. This is an optional
727 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
731 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
732 function to lock the database in exclusive mode for writes, and shared mode
733 for reads. Pass any true value to enable. This affects the base DB handle
734 I<and any child hashes or arrays> that use the same DB file. This is an
735 optional parameter, and defaults to 0 (disabled). See L<LOCKING> below for
740 Specifies whether autoflush is to be enabled on the underlying filehandle.
741 This obviously slows down write operations, but is required if you may have
742 multiple processes accessing the same DB file (also consider enable I<locking>).
743 Pass any true value to enable. This is an optional parameter, and defaults to 0
748 If I<autobless> mode is enabled, DBM::Deep will preserve the class something
749 is blessed into, and restores it when fetched. This is an optional parameter, and defaults to 1 (enabled).
751 B<Note:> If you use the OO-interface, you will not be able to call any methods
752 of DBM::Deep on the blessed item. This is considered to be a feature.
756 See L</FILTERS> below.
762 With DBM::Deep you can access your databases using Perl's standard hash/array
763 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
764 treat them as such. DBM::Deep will intercept all reads/writes and direct them
765 to the right place -- the DB file. This has nothing to do with the
766 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
767 using regular hashes and arrays, rather than calling functions like C<get()>
768 and C<put()> (although those work too). It is entirely up to you how to want
769 to access your databases.
773 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
774 or even nested hashes (or arrays) using standard Perl syntax:
776 my $db = DBM::Deep->new( "foo.db" );
778 $db->{mykey} = "myvalue";
780 $db->{myhash}->{subkey} = "subvalue";
782 print $db->{myhash}->{subkey} . "\n";
784 You can even step through hash keys using the normal Perl C<keys()> function:
786 foreach my $key (keys %$db) {
787 print "$key: " . $db->{$key} . "\n";
790 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
791 pushes them onto an array, all before the loop even begins. If you have an
792 extremely large hash, this may exhaust Perl's memory. Instead, consider using
793 Perl's C<each()> function, which pulls keys/values one at a time, using very
796 while (my ($key, $value) = each %$db) {
797 print "$key: $value\n";
800 Please note that when using C<each()>, you should always pass a direct
801 hash reference, not a lookup. Meaning, you should B<never> do this:
804 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
806 This causes an infinite loop, because for each iteration, Perl is calling
807 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
808 it effectively keeps returning the first key over and over again. Instead,
809 assign a temporary variable to C<$db->{foo}>, then pass that to each().
813 As with hashes, you can treat any DBM::Deep object like a normal Perl array
814 reference. This includes inserting, removing and manipulating elements,
815 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
816 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
817 or simply be a nested array reference inside a hash. Example:
819 my $db = DBM::Deep->new(
820 file => "foo-array.db",
821 type => DBM::Deep->TYPE_ARRAY
825 push @$db, "bar", "baz";
828 my $last_elem = pop @$db; # baz
829 my $first_elem = shift @$db; # bah
830 my $second_elem = $db->[1]; # bar
832 my $num_elements = scalar @$db;
836 In addition to the I<tie()> interface, you can also use a standard OO interface
837 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
838 array) has its own methods, but both types share the following common methods:
839 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
840 C<store(> are aliases to C<put()> and C<get()>, respectively.
844 =item * new() / clone()
846 These are the constructor and copy-functions.
848 =item * put() / store()
850 Stores a new hash key/value pair, or sets an array element value. Takes two
851 arguments, the hash key or array index, and the new value. The value can be
852 a scalar, hash ref or array ref. Returns true on success, false on failure.
854 $db->put("foo", "bar"); # for hashes
855 $db->put(1, "bar"); # for arrays
857 =item * get() / fetch()
859 Fetches the value of a hash key or array element. Takes one argument: the hash
860 key or array index. Returns a scalar, hash ref or array ref, depending on the
863 my $value = $db->get("foo"); # for hashes
864 my $value = $db->get(1); # for arrays
868 Checks if a hash key or array index exists. Takes one argument: the hash key
869 or array index. Returns true if it exists, false if not.
871 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
872 if ($db->exists(1)) { print "yay!\n"; } # for arrays
876 Deletes one hash key/value pair or array element. Takes one argument: the hash
877 key or array index. Returns true on success, false if not found. For arrays,
878 the remaining elements located after the deleted element are NOT moved over.
879 The deleted element is essentially just undefined, which is exactly how Perl's
880 internal arrays work. Please note that the space occupied by the deleted
881 key/value or element is B<not> reused again -- see L<UNUSED SPACE RECOVERY>
882 below for details and workarounds.
884 $db->delete("foo"); # for hashes
885 $db->delete(1); # for arrays
889 Deletes B<all> hash keys or array elements. Takes no arguments. No return
890 value. Please note that the space occupied by the deleted keys/values or
891 elements is B<not> reused again -- see L<UNUSED SPACE RECOVERY> below for
892 details and workarounds.
894 $db->clear(); # hashes or arrays
896 =item * lock() / unlock()
902 Recover lost disk space. This is important to do, especially if you use
905 =item * import() / export()
907 Data going in and out.
913 For hashes, DBM::Deep supports all the common methods described above, and the
914 following additional methods: C<first_key()> and C<next_key()>.
920 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
921 fetched in an undefined order (which appears random). Takes no arguments,
922 returns the key as a scalar value.
924 my $key = $db->first_key();
928 Returns the "next" key in the hash, given the previous one as the sole argument.
929 Returns undef if there are no more keys to be fetched.
931 $key = $db->next_key($key);
935 Here are some examples of using hashes:
937 my $db = DBM::Deep->new( "foo.db" );
939 $db->put("foo", "bar");
940 print "foo: " . $db->get("foo") . "\n";
942 $db->put("baz", {}); # new child hash ref
943 $db->get("baz")->put("buz", "biz");
944 print "buz: " . $db->get("baz")->get("buz") . "\n";
946 my $key = $db->first_key();
948 print "$key: " . $db->get($key) . "\n";
949 $key = $db->next_key($key);
952 if ($db->exists("foo")) { $db->delete("foo"); }
956 For arrays, DBM::Deep supports all the common methods described above, and the
957 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
958 C<unshift()> and C<splice()>.
964 Returns the number of elements in the array. Takes no arguments.
966 my $len = $db->length();
970 Adds one or more elements onto the end of the array. Accepts scalars, hash
971 refs or array refs. No return value.
973 $db->push("foo", "bar", {});
977 Fetches the last element in the array, and deletes it. Takes no arguments.
978 Returns undef if array is empty. Returns the element value.
980 my $elem = $db->pop();
984 Fetches the first element in the array, deletes it, then shifts all the
985 remaining elements over to take up the space. Returns the element value. This
986 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
989 my $elem = $db->shift();
993 Inserts one or more elements onto the beginning of the array, shifting all
994 existing elements over to make room. Accepts scalars, hash refs or array refs.
995 No return value. This method is not recommended with large arrays -- see
996 <LARGE ARRAYS> below for details.
998 $db->unshift("foo", "bar", {});
1002 Performs exactly like Perl's built-in function of the same name. See L<perldoc
1003 -f splice> for usage -- it is too complicated to document here. This method is
1004 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
1008 Here are some examples of using arrays:
1010 my $db = DBM::Deep->new(
1012 type => DBM::Deep->TYPE_ARRAY
1015 $db->push("bar", "baz");
1016 $db->unshift("foo");
1019 my $len = $db->length();
1020 print "length: $len\n"; # 4
1022 for (my $k=0; $k<$len; $k++) {
1023 print "$k: " . $db->get($k) . "\n";
1026 $db->splice(1, 2, "biz", "baf");
1028 while (my $elem = shift @$db) {
1029 print "shifted: $elem\n";
1034 Enable automatic file locking by passing a true value to the C<locking>
1035 parameter when constructing your DBM::Deep object (see L<SETUP> above).
1037 my $db = DBM::Deep->new(
1042 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
1043 mode for writes, and shared mode for reads. This is required if you have
1044 multiple processes accessing the same database file, to avoid file corruption.
1045 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
1046 NFS> below for more.
1048 =head2 EXPLICIT LOCKING
1050 You can explicitly lock a database, so it remains locked for multiple
1051 transactions. This is done by calling the C<lock()> method, and passing an
1052 optional lock mode argument (defaults to exclusive mode). This is particularly
1053 useful for things like counters, where the current value needs to be fetched,
1054 then incremented, then stored again.
1057 my $counter = $db->get("counter");
1059 $db->put("counter", $counter);
1068 You can pass C<lock()> an optional argument, which specifies which mode to use
1069 (exclusive or shared). Use one of these two constants:
1070 C<DBM::Deep-E<gt>LOCK_EX> or C<DBM::Deep-E<gt>LOCK_SH>. These are passed
1071 directly to C<flock()>, and are the same as the constants defined in Perl's
1074 $db->lock( $db->LOCK_SH );
1078 =head1 IMPORTING/EXPORTING
1080 You can import existing complex structures by calling the C<import()> method,
1081 and export an entire database into an in-memory structure using the C<export()>
1082 method. Both are examined here.
1086 Say you have an existing hash with nested hashes/arrays inside it. Instead of
1087 walking the structure and adding keys/elements to the database as you go,
1088 simply pass a reference to the C<import()> method. This recursively adds
1089 everything to an existing DBM::Deep object for you. Here is an example:
1094 array1 => [ "elem0", "elem1", "elem2" ],
1096 subkey1 => "subvalue1",
1097 subkey2 => "subvalue2"
1101 my $db = DBM::Deep->new( "foo.db" );
1102 $db->import( $struct );
1104 print $db->{key1} . "\n"; # prints "value1"
1106 This recursively imports the entire C<$struct> object into C<$db>, including
1107 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
1108 keys are merged with the existing ones, replacing if they already exist.
1109 The C<import()> method can be called on any database level (not just the base
1110 level), and works with both hash and array DB types.
1112 B<Note:> Make sure your existing structure has no circular references in it.
1113 These will cause an infinite loop when importing. There are plans to fix this
1118 Calling the C<export()> method on an existing DBM::Deep object will return
1119 a reference to a new in-memory copy of the database. The export is done
1120 recursively, so all nested hashes/arrays are all exported to standard Perl
1121 objects. Here is an example:
1123 my $db = DBM::Deep->new( "foo.db" );
1125 $db->{key1} = "value1";
1126 $db->{key2} = "value2";
1128 $db->{hash1}->{subkey1} = "subvalue1";
1129 $db->{hash1}->{subkey2} = "subvalue2";
1131 my $struct = $db->export();
1133 print $struct->{key1} . "\n"; # prints "value1"
1135 This makes a complete copy of the database in memory, and returns a reference
1136 to it. The C<export()> method can be called on any database level (not just
1137 the base level), and works with both hash and array DB types. Be careful of
1138 large databases -- you can store a lot more data in a DBM::Deep object than an
1139 in-memory Perl structure.
1141 B<Note:> Make sure your database has no circular references in it.
1142 These will cause an infinite loop when exporting. There are plans to fix this
1147 DBM::Deep has a number of hooks where you can specify your own Perl function
1148 to perform filtering on incoming or outgoing data. This is a perfect
1149 way to extend the engine, and implement things like real-time compression or
1150 encryption. Filtering applies to the base DB level, and all child hashes /
1151 arrays. Filter hooks can be specified when your DBM::Deep object is first
1152 constructed, or by calling the C<set_filter()> method at any time. There are
1153 four available filter hooks, described below:
1157 =item * filter_store_key
1159 This filter is called whenever a hash key is stored. It
1160 is passed the incoming key, and expected to return a transformed key.
1162 =item * filter_store_value
1164 This filter is called whenever a hash key or array element is stored. It
1165 is passed the incoming value, and expected to return a transformed value.
1167 =item * filter_fetch_key
1169 This filter is called whenever a hash key is fetched (i.e. via
1170 C<first_key()> or C<next_key()>). It is passed the transformed key,
1171 and expected to return the plain key.
1173 =item * filter_fetch_value
1175 This filter is called whenever a hash key or array element is fetched.
1176 It is passed the transformed value, and expected to return the plain value.
1180 Here are the two ways to setup a filter hook:
1182 my $db = DBM::Deep->new(
1184 filter_store_value => \&my_filter_store,
1185 filter_fetch_value => \&my_filter_fetch
1190 $db->set_filter( "filter_store_value", \&my_filter_store );
1191 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
1193 Your filter function will be called only when dealing with SCALAR keys or
1194 values. When nested hashes and arrays are being stored/fetched, filtering
1195 is bypassed. Filters are called as static functions, passed a single SCALAR
1196 argument, and expected to return a single SCALAR value. If you want to
1197 remove a filter, set the function reference to C<undef>:
1199 $db->set_filter( "filter_store_value", undef );
1201 =head2 REAL-TIME ENCRYPTION EXAMPLE
1203 Here is a working example that uses the I<Crypt::Blowfish> module to
1204 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
1205 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
1206 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
1209 use Crypt::Blowfish;
1212 my $cipher = Crypt::CBC->new({
1213 'key' => 'my secret key',
1214 'cipher' => 'Blowfish',
1216 'regenerate_key' => 0,
1217 'padding' => 'space',
1221 my $db = DBM::Deep->new(
1222 file => "foo-encrypt.db",
1223 filter_store_key => \&my_encrypt,
1224 filter_store_value => \&my_encrypt,
1225 filter_fetch_key => \&my_decrypt,
1226 filter_fetch_value => \&my_decrypt,
1229 $db->{key1} = "value1";
1230 $db->{key2} = "value2";
1231 print "key1: " . $db->{key1} . "\n";
1232 print "key2: " . $db->{key2} . "\n";
1238 return $cipher->encrypt( $_[0] );
1241 return $cipher->decrypt( $_[0] );
1244 =head2 REAL-TIME COMPRESSION EXAMPLE
1246 Here is a working example that uses the I<Compress::Zlib> module to do real-time
1247 compression / decompression of keys & values with DBM::Deep Filters.
1248 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
1249 more on I<Compress::Zlib>.
1254 my $db = DBM::Deep->new(
1255 file => "foo-compress.db",
1256 filter_store_key => \&my_compress,
1257 filter_store_value => \&my_compress,
1258 filter_fetch_key => \&my_decompress,
1259 filter_fetch_value => \&my_decompress,
1262 $db->{key1} = "value1";
1263 $db->{key2} = "value2";
1264 print "key1: " . $db->{key1} . "\n";
1265 print "key2: " . $db->{key2} . "\n";
1271 return Compress::Zlib::memGzip( $_[0] ) ;
1274 return Compress::Zlib::memGunzip( $_[0] ) ;
1277 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
1278 actually numerical index numbers, and are not filtered.
1280 =head1 ERROR HANDLING
1282 Most DBM::Deep methods return a true value for success, and call die() on
1283 failure. You can wrap calls in an eval block to catch the die.
1285 my $db = DBM::Deep->new( "foo.db" ); # create hash
1286 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
1288 print $@; # prints error message
1290 =head1 LARGEFILE SUPPORT
1292 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
1293 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
1294 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
1295 by specifying the 'pack_size' parameter when constructing the file.
1298 filename => $filename,
1299 pack_size => 'large',
1302 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
1303 instead of 32-bit longs. After setting these values your DB files have a
1304 theoretical maximum size of 16 XB (exabytes).
1306 You can also use C<pack_size =E<gt> 'small'> in order to use 16-bit file
1309 B<Note:> Changing these values will B<NOT> work for existing database files.
1310 Only change this for new files. Once the value has been set, it is stored in
1311 the file's header and cannot be changed for the life of the file. These
1312 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
1315 B<Note:> We have not personally tested files larger than 2 GB -- all my
1316 systems have only a 32-bit Perl. However, I have received user reports that
1317 this does indeed work!
1319 =head1 LOW-LEVEL ACCESS
1321 If you require low-level access to the underlying filehandle that DBM::Deep uses,
1322 you can call the C<_fh()> method, which returns the handle:
1324 my $fh = $db->_fh();
1326 This method can be called on the root level of the datbase, or any child
1327 hashes or arrays. All levels share a I<root> structure, which contains things
1328 like the filehandle, a reference counter, and all the options specified
1329 when you created the object. You can get access to this file object by
1330 calling the C<_storage()> method.
1332 my $file_obj = $db->_storage();
1334 This is useful for changing options after the object has already been created,
1335 such as enabling/disabling locking. You can also store your own temporary user
1336 data in this structure (be wary of name collision), which is then accessible from
1337 any child hash or array.
1339 =head1 CUSTOM DIGEST ALGORITHM
1341 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
1342 keys. However you can override this, and use another algorithm (such as SHA-256)
1343 or even write your own. But please note that DBM::Deep currently expects zero
1344 collisions, so your algorithm has to be I<perfect>, so to speak. Collision
1345 detection may be introduced in a later version.
1347 You can specify a custom digest algorithm by passing it into the parameter
1348 list for new(), passing a reference to a subroutine as the 'digest' parameter,
1349 and the length of the algorithm's hashes (in bytes) as the 'hash_size'
1350 parameter. Here is a working example that uses a 256-bit hash from the
1351 I<Digest::SHA256> module. Please see
1352 L<http://search.cpan.org/search?module=Digest::SHA256> for more information.
1357 my $context = Digest::SHA256::new(256);
1359 my $db = DBM::Deep->new(
1360 filename => "foo-sha.db",
1361 digest => \&my_digest,
1365 $db->{key1} = "value1";
1366 $db->{key2} = "value2";
1367 print "key1: " . $db->{key1} . "\n";
1368 print "key2: " . $db->{key2} . "\n";
1374 return substr( $context->hash($_[0]), 0, 32 );
1377 B<Note:> Your returned digest strings must be B<EXACTLY> the number
1378 of bytes you specify in the hash_size parameter (in this case 32).
1380 B<Note:> If you do choose to use a custom digest algorithm, you must set it
1381 every time you access this file. Otherwise, the default (MD5) will be used.
1383 =head1 CIRCULAR REFERENCES
1385 DBM::Deep has B<experimental> support for circular references. Meaning you
1386 can have a nested hash key or array element that points to a parent object.
1387 This relationship is stored in the DB file, and is preserved between sessions.
1390 my $db = DBM::Deep->new( "foo.db" );
1393 $db->{circle} = $db; # ref to self
1395 print $db->{foo} . "\n"; # prints "bar"
1396 print $db->{circle}->{foo} . "\n"; # prints "bar" again
1398 B<Note>: Passing the object to a function that recursively walks the
1399 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
1400 C<export()> methods) will result in an infinite loop. This will be fixed in
1405 New in 0.99_01 is ACID transactions. Every DBM::Deep object is completely
1406 transaction-ready - it is not an option you have to turn on. Three new methods
1407 have been added to support them. They are:
1411 =item * begin_work()
1413 This starts a transaction.
1417 This applies the changes done within the transaction to the mainline and ends
1422 This discards the changes done within the transaction to the mainline and ends
1427 Transactions in DBM::Deep are done using the MVCC method, the same method used
1428 by the InnoDB MySQL table type.
1430 =head1 CAVEATS / ISSUES / BUGS
1432 This section describes all the known issues with DBM::Deep. It you have found
1433 something that is not listed here, please send e-mail to L<jhuckaby@cpan.org>.
1435 =head2 UNUSED SPACE RECOVERY
1437 One major caveat with DBM::Deep is that space occupied by existing keys and
1438 values is not recovered when they are deleted. Meaning if you keep deleting
1439 and adding new keys, your file will continuously grow. I am working on this,
1440 but in the meantime you can call the built-in C<optimize()> method from time to
1441 time (perhaps in a crontab or something) to recover all your unused space.
1443 $db->optimize(); # returns true on success
1445 This rebuilds the ENTIRE database into a new file, then moves it on top of
1446 the original. The new file will have no unused space, thus it will take up as
1447 little disk space as possible. Please note that this operation can take
1448 a long time for large files, and you need enough disk space to temporarily hold
1449 2 copies of your DB file. The temporary file is created in the same directory
1450 as the original, named with a ".tmp" extension, and is deleted when the
1451 operation completes. Oh, and if locking is enabled, the DB is automatically
1452 locked for the entire duration of the copy.
1454 B<WARNING:> Only call optimize() on the top-level node of the database, and
1455 make sure there are no child references lying around. DBM::Deep keeps a reference
1456 counter, and if it is greater than 1, optimize() will abort and return undef.
1460 (The reasons given assume a high level of Perl understanding, specifically of
1461 references. You can safely skip this section.)
1463 Currently, the only references supported are HASH and ARRAY. The other reference
1464 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
1470 These are things like filehandles and other sockets. They can't be supported
1471 because it's completely unclear how DBM::Deep should serialize them.
1473 =item * SCALAR / REF
1475 The discussion here refers to the following type of example:
1482 # In some other process ...
1484 my $val = ${ $db->{key1} };
1486 is( $val, 50, "What actually gets stored in the DB file?" );
1488 The problem is one of synchronization. When the variable being referred to
1489 changes value, the reference isn't notified. This means that the new value won't
1490 be stored in the datafile for other processes to read. There is no TIEREF.
1492 It is theoretically possible to store references to values already within a
1493 DBM::Deep object because everything already is synchronized, but the change to
1494 the internals would be quite large. Specifically, DBM::Deep would have to tie
1495 every single value that is stored. This would bloat the RAM footprint of
1496 DBM::Deep at least twofold (if not more) and be a significant performance drain,
1497 all to support a feature that has never been requested.
1501 L<Data::Dump::Streamer/> provides a mechanism for serializing coderefs,
1502 including saving off all closure state. However, just as for SCALAR and REF,
1503 that closure state may change without notifying the DBM::Deep object storing
1508 =head2 FILE CORRUPTION
1510 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
1511 for a 32-bit signature when opened, but other corruption in files can cause
1512 segmentation faults. DBM::Deep may try to seek() past the end of a file, or get
1513 stuck in an infinite loop depending on the level of corruption. File write
1514 operations are not checked for failure (for speed), so if you happen to run
1515 out of disk space, DBM::Deep will probably fail in a bad way. These things will
1516 be addressed in a later version of DBM::Deep.
1520 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
1521 well on local filesystems, but will NOT protect you from file corruption over
1522 NFS. I've heard about setting up your NFS server with a locking daemon, then
1523 using lockf() to lock your files, but your mileage may vary there as well.
1524 From what I understand, there is no real way to do it. However, if you need
1525 access to the underlying filehandle in DBM::Deep for using some other kind of
1526 locking scheme like lockf(), see the L<LOW-LEVEL ACCESS> section above.
1528 =head2 COPYING OBJECTS
1530 Beware of copying tied objects in Perl. Very strange things can happen.
1531 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
1532 returns a new, blessed, tied hash or array to the same level in the DB.
1534 my $copy = $db->clone();
1536 B<Note>: Since clone() here is cloning the object, not the database location, any
1537 modifications to either $db or $copy will be visible to both.
1541 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1542 These functions cause every element in the array to move, which can be murder
1543 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1544 a different location. This will be addressed in the forthcoming version 1.00.
1546 =head2 WRITEONLY FILES
1548 If you pass in a filehandle to new(), you may have opened it in either a readonly or
1549 writeonly mode. STORE will verify that the filehandle is writable. However, there
1550 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
1551 filehandle isn't readable, it's not clear what will happen. So, don't do that.
1553 =head1 CODE COVERAGE
1555 B<Devel::Cover> is used to test the code coverage of the tests. Below is the
1556 B<Devel::Cover> report on this distribution's test suite.
1558 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1559 File stmt bran cond sub pod time total
1560 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1561 blib/lib/DBM/Deep.pm 96.2 89.0 75.0 95.8 89.5 36.0 92.9
1562 blib/lib/DBM/Deep/Array.pm 96.1 88.3 100.0 96.4 100.0 15.9 94.7
1563 blib/lib/DBM/Deep/Engine.pm 96.6 86.6 89.5 100.0 0.0 20.0 91.0
1564 blib/lib/DBM/Deep/File.pm 99.4 88.3 55.6 100.0 0.0 19.6 89.5
1565 blib/lib/DBM/Deep/Hash.pm 98.5 83.3 100.0 100.0 100.0 8.5 96.3
1566 Total 96.9 87.4 81.2 98.0 38.5 100.0 92.1
1567 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1569 =head1 MORE INFORMATION
1571 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1572 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1575 The source code repository is at L<http://svn.perl.org/modules/DBM-Deep>
1579 Rob Kinyon, L<rkinyon@cpan.org>
1581 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1583 Special thanks to Adam Sah and Rich Gaushell! You know why :-)
1587 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1588 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1592 Copyright (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
1593 This is free software, you may use it and distribute it under the
1594 same terms as Perl itself.