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.
34 use Fcntl qw( :DEFAULT :flock :seek );
38 use DBM::Deep::Engine;
40 use vars qw( $VERSION );
41 $VERSION = q(0.99_01);
44 # Set to 4 and 'N' for 32-bit offset tags (default). Theoretical limit of 4 GB per file.
45 # (Perl must be compiled with largefile support for files > 2 GB)
47 # Set to 8 and 'Q' for 64-bit offsets. Theoretical limit of 16 XB per file.
48 # (Perl must be compiled with largefile and 64-bit long support)
54 # Set to 4 and 'N' for 32-bit data length prefixes. Limit of 4 GB for each key/value.
55 # Upgrading this is possible (see above) but probably not necessary. If you need
56 # more than 4 GB for a single key or value, this module is really not for you :-)
58 #my $DATA_LENGTH_SIZE = 4;
59 #my $DATA_LENGTH_PACK = 'N';
60 our ($LONG_SIZE, $LONG_PACK, $DATA_LENGTH_SIZE, $DATA_LENGTH_PACK);
63 # Maximum number of buckets per list before another level of indexing is done.
64 # Increase this value for slightly greater speed, but larger database files.
65 # DO NOT decrease this value below 16, due to risk of recursive reindex overrun.
70 # Better not adjust anything below here, unless you're me :-)
74 # Setup digest function for keys
76 our ($DIGEST_FUNC, $HASH_SIZE);
77 #my $DIGEST_FUNC = \&Digest::MD5::md5;
80 # Precalculate index and bucket sizes based on values above.
83 our ($INDEX_SIZE, $BUCKET_SIZE, $BUCKET_LIST_SIZE);
90 # Setup file and tag signatures. These should never change.
92 sub SIG_FILE () { 'DPDB' }
93 sub SIG_HASH () { 'H' }
94 sub SIG_ARRAY () { 'A' }
95 sub SIG_SCALAR () { 'S' }
96 sub SIG_NULL () { 'N' }
97 sub SIG_DATA () { 'D' }
98 sub SIG_INDEX () { 'I' }
99 sub SIG_BLIST () { 'B' }
100 sub SIG_SIZE () { 1 }
103 # Setup constants for users to pass to new()
105 sub TYPE_HASH () { SIG_HASH }
106 sub TYPE_ARRAY () { SIG_ARRAY }
107 sub TYPE_SCALAR () { SIG_SCALAR }
113 if (scalar(@_) > 1) {
115 $proto->_throw_error( "Odd number of parameters to " . (caller(1))[2] );
119 elsif ( ref $_[0] ) {
120 unless ( eval { local $SIG{'__DIE__'}; %{$_[0]} || 1 } ) {
121 $proto->_throw_error( "Not a hashref in args to " . (caller(1))[2] );
126 $args = { file => shift };
134 # Class constructor method for Perl OO interface.
135 # Calls tie() and returns blessed reference to tied hash or array,
136 # providing a hybrid OO/tie interface.
139 my $args = $class->_get_args( @_ );
142 # Check if we want a tied hash or array.
145 if (defined($args->{type}) && $args->{type} eq TYPE_ARRAY) {
146 $class = 'DBM::Deep::Array';
147 require DBM::Deep::Array;
148 tie @$self, $class, %$args;
151 $class = 'DBM::Deep::Hash';
152 require DBM::Deep::Hash;
153 tie %$self, $class, %$args;
156 return bless $self, $class;
161 # Setup $self and bless into this class.
166 # These are the defaults to be optionally overridden below
169 base_offset => length(SIG_FILE),
170 engine => 'DBM::Deep::Engine',
173 foreach my $param ( keys %$self ) {
174 next unless exists $args->{$param};
175 $self->{$param} = delete $args->{$param}
178 # locking implicitly enables autoflush
179 if ($args->{locking}) { $args->{autoflush} = 1; }
181 $self->{root} = exists $args->{root}
183 : DBM::Deep::_::Root->new( $args );
185 if (!defined($self->_fh)) { $self->{engine}->open( $self ); }
192 require DBM::Deep::Hash;
193 return DBM::Deep::Hash->TIEHASH( @_ );
198 require DBM::Deep::Array;
199 return DBM::Deep::Array->TIEARRAY( @_ );
202 #XXX Unneeded now ...
208 # Adds one key/value pair to bucket list, given offset, MD5 digest of key,
209 # plain (undigested) key and value.
212 my ($tag, $md5, $plain_key, $value) = @_;
213 my $keys = $tag->{content};
217 my $root = $self->_root;
219 my $is_dbm_deep = eval { local $SIG{'__DIE__'}; $value->isa( 'DBM::Deep' ) };
220 my $internal_ref = $is_dbm_deep && ($value->_root eq $root);
225 # Iterate through buckets, seeing if this is a new entry or a replace.
227 for (my $i=0; $i<$MAX_BUCKETS; $i++) {
228 my $subloc = unpack($LONG_PACK, substr($keys, ($i * $BUCKET_SIZE) + $HASH_SIZE, $LONG_SIZE));
231 # Found empty bucket (end of list). Populate and exit loop.
235 $location = $internal_ref
236 ? $value->_base_offset
239 seek($fh, $tag->{offset} + ($i * $BUCKET_SIZE) + $root->{file_offset}, SEEK_SET);
240 print( $fh $md5 . pack($LONG_PACK, $location) );
244 my $key = substr($keys, $i * $BUCKET_SIZE, $HASH_SIZE);
247 # Found existing bucket with same key. Replace with new value.
252 $location = $value->_base_offset;
253 seek($fh, $tag->{offset} + ($i * $BUCKET_SIZE) + $root->{file_offset}, SEEK_SET);
254 print( $fh $md5 . pack($LONG_PACK, $location) );
258 seek($fh, $subloc + SIG_SIZE + $root->{file_offset}, SEEK_SET);
260 read( $fh, $size, $DATA_LENGTH_SIZE); $size = unpack($DATA_LENGTH_PACK, $size);
263 # If value is a hash, array, or raw value with equal or less size, we can
264 # reuse the same content area of the database. Otherwise, we have to create
265 # a new content area at the EOF.
268 my $r = Scalar::Util::reftype( $value ) || '';
269 if ( $r eq 'HASH' || $r eq 'ARRAY' ) {
270 $actual_length = $INDEX_SIZE;
272 # if autobless is enabled, must also take into consideration
273 # the class name, as it is stored along with key/value.
274 if ( $root->{autobless} ) {
275 my $value_class = Scalar::Util::blessed($value);
276 if ( defined $value_class && !$value->isa('DBM::Deep') ) {
277 $actual_length += length($value_class);
281 else { $actual_length = length($value); }
283 if ($actual_length <= $size) {
287 $location = $root->{end};
288 seek($fh, $tag->{offset} + ($i * $BUCKET_SIZE) + $HASH_SIZE + $root->{file_offset}, SEEK_SET);
289 print( $fh pack($LONG_PACK, $location) );
297 # If this is an internal reference, return now.
298 # No need to write value or plain key
305 # If bucket didn't fit into list, split into a new index level
308 seek($fh, $tag->{ref_loc} + $root->{file_offset}, SEEK_SET);
309 print( $fh pack($LONG_PACK, $root->{end}) );
311 my $index_tag = $self->{engine}->create_tag($self, $root->{end}, SIG_INDEX, chr(0) x $INDEX_SIZE);
314 $keys .= $md5 . pack($LONG_PACK, 0);
316 for (my $i=0; $i<=$MAX_BUCKETS; $i++) {
317 my $key = substr($keys, $i * $BUCKET_SIZE, $HASH_SIZE);
319 my $old_subloc = unpack($LONG_PACK, substr($keys, ($i * $BUCKET_SIZE) + $HASH_SIZE, $LONG_SIZE));
320 my $num = ord(substr($key, $tag->{ch} + 1, 1));
322 if ($offsets[$num]) {
323 my $offset = $offsets[$num] + SIG_SIZE + $DATA_LENGTH_SIZE;
324 seek($fh, $offset + $root->{file_offset}, SEEK_SET);
326 read( $fh, $subkeys, $BUCKET_LIST_SIZE);
328 for (my $k=0; $k<$MAX_BUCKETS; $k++) {
329 my $subloc = unpack($LONG_PACK, substr($subkeys, ($k * $BUCKET_SIZE) + $HASH_SIZE, $LONG_SIZE));
331 seek($fh, $offset + ($k * $BUCKET_SIZE) + $root->{file_offset}, SEEK_SET);
332 print( $fh $key . pack($LONG_PACK, $old_subloc || $root->{end}) );
338 $offsets[$num] = $root->{end};
339 seek($fh, $index_tag->{offset} + ($num * $LONG_SIZE) + $root->{file_offset}, SEEK_SET);
340 print( $fh pack($LONG_PACK, $root->{end}) );
342 my $blist_tag = $self->{engine}->create_tag($self, $root->{end}, SIG_BLIST, chr(0) x $BUCKET_LIST_SIZE);
344 seek($fh, $blist_tag->{offset} + $root->{file_offset}, SEEK_SET);
345 print( $fh $key . pack($LONG_PACK, $old_subloc || $root->{end}) );
350 $location ||= $root->{end};
351 } # re-index bucket list
354 # Seek to content area and store signature, value and plaintext key
358 seek($fh, $location + $root->{file_offset}, SEEK_SET);
361 # Write signature based on content type, set content length and write actual value.
363 my $r = Scalar::Util::reftype($value) || '';
365 print( $fh TYPE_HASH );
366 print( $fh pack($DATA_LENGTH_PACK, $INDEX_SIZE) . chr(0) x $INDEX_SIZE );
367 $content_length = $INDEX_SIZE;
369 elsif ($r eq 'ARRAY') {
370 print( $fh TYPE_ARRAY );
371 print( $fh pack($DATA_LENGTH_PACK, $INDEX_SIZE) . chr(0) x $INDEX_SIZE );
372 $content_length = $INDEX_SIZE;
374 elsif (!defined($value)) {
375 print( $fh SIG_NULL );
376 print( $fh pack($DATA_LENGTH_PACK, 0) );
380 print( $fh SIG_DATA );
381 print( $fh pack($DATA_LENGTH_PACK, length($value)) . $value );
382 $content_length = length($value);
386 # Plain key is stored AFTER value, as keys are typically fetched less often.
388 print( $fh pack($DATA_LENGTH_PACK, length($plain_key)) . $plain_key );
391 # If value is blessed, preserve class name
393 if ( $root->{autobless} ) {
394 my $value_class = Scalar::Util::blessed($value);
395 if ( defined $value_class && $value_class ne 'DBM::Deep' ) {
397 # Blessed ref -- will restore later
400 print( $fh pack($DATA_LENGTH_PACK, length($value_class)) . $value_class );
401 $content_length += 1;
402 $content_length += $DATA_LENGTH_SIZE + length($value_class);
406 $content_length += 1;
411 # If this is a new content area, advance EOF counter
413 if ($location == $root->{end}) {
414 $root->{end} += SIG_SIZE;
415 $root->{end} += $DATA_LENGTH_SIZE + $content_length;
416 $root->{end} += $DATA_LENGTH_SIZE + length($plain_key);
420 # If content is a hash or array, create new child DBM::Deep object and
421 # pass each key or element to it.
424 my $branch = DBM::Deep->new(
426 base_offset => $location,
429 foreach my $key (keys %{$value}) {
430 $branch->STORE( $key, $value->{$key} );
433 elsif ($r eq 'ARRAY') {
434 my $branch = DBM::Deep->new(
436 base_offset => $location,
440 foreach my $element (@{$value}) {
441 $branch->STORE( $index, $element );
449 return $self->_throw_error("Fatal error: indexing failed -- possibly due to corruption in file");
452 sub _get_bucket_value {
454 # Fetch single value given tag and MD5 digested key.
457 my ($tag, $md5) = @_;
458 my $keys = $tag->{content};
463 # Iterate through buckets, looking for a key match
466 for (my $i=0; $i<$MAX_BUCKETS; $i++) {
467 my $key = substr($keys, $i * $BUCKET_SIZE, $HASH_SIZE);
468 my $subloc = unpack($LONG_PACK, substr($keys, ($i * $BUCKET_SIZE) + $HASH_SIZE, $LONG_SIZE));
472 # Hit end of list, no match
477 if ( $md5 ne $key ) {
482 # Found match -- seek to offset and read signature
485 seek($fh, $subloc + $self->_root->{file_offset}, SEEK_SET);
486 read( $fh, $signature, SIG_SIZE);
489 # If value is a hash or array, return new DBM::Deep object with correct offset
491 if (($signature eq TYPE_HASH) || ($signature eq TYPE_ARRAY)) {
492 my $obj = DBM::Deep->new(
494 base_offset => $subloc,
498 if ($self->_root->{autobless}) {
500 # Skip over value and plain key to see if object needs
503 seek($fh, $DATA_LENGTH_SIZE + $INDEX_SIZE, SEEK_CUR);
506 read( $fh, $size, $DATA_LENGTH_SIZE); $size = unpack($DATA_LENGTH_PACK, $size);
507 if ($size) { seek($fh, $size, SEEK_CUR); }
510 read( $fh, $bless_bit, 1);
511 if (ord($bless_bit)) {
513 # Yes, object needs to be re-blessed
516 read( $fh, $size, $DATA_LENGTH_SIZE); $size = unpack($DATA_LENGTH_PACK, $size);
517 if ($size) { read( $fh, $class_name, $size); }
518 if ($class_name) { $obj = bless( $obj, $class_name ); }
526 # Otherwise return actual value
528 elsif ($signature eq SIG_DATA) {
531 read( $fh, $size, $DATA_LENGTH_SIZE); $size = unpack($DATA_LENGTH_PACK, $size);
532 if ($size) { read( $fh, $value, $size); }
537 # Key exists, but content is null
547 # Delete single key/value pair given tag and MD5 digested key.
550 my ($tag, $md5) = @_;
551 my $keys = $tag->{content};
556 # Iterate through buckets, looking for a key match
559 for (my $i=0; $i<$MAX_BUCKETS; $i++) {
560 my $key = substr($keys, $i * $BUCKET_SIZE, $HASH_SIZE);
561 my $subloc = unpack($LONG_PACK, substr($keys, ($i * $BUCKET_SIZE) + $HASH_SIZE, $LONG_SIZE));
565 # Hit end of list, no match
570 if ( $md5 ne $key ) {
575 # Matched key -- delete bucket and return
577 seek($fh, $tag->{offset} + ($i * $BUCKET_SIZE) + $self->_root->{file_offset}, SEEK_SET);
578 print( $fh substr($keys, ($i+1) * $BUCKET_SIZE ) );
579 print( $fh chr(0) x $BUCKET_SIZE );
589 # Check existence of single key given tag and MD5 digested key.
592 my ($tag, $md5) = @_;
593 my $keys = $tag->{content};
596 # Iterate through buckets, looking for a key match
599 for (my $i=0; $i<$MAX_BUCKETS; $i++) {
600 my $key = substr($keys, $i * $BUCKET_SIZE, $HASH_SIZE);
601 my $subloc = unpack($LONG_PACK, substr($keys, ($i * $BUCKET_SIZE) + $HASH_SIZE, $LONG_SIZE));
605 # Hit end of list, no match
610 if ( $md5 ne $key ) {
615 # Matched key -- return true
623 sub _find_bucket_list {
625 # Locate offset for bucket list, given digested key
631 # Locate offset for bucket list using digest index system
634 my $tag = $self->{engine}->load_tag($self, $self->_base_offset);
635 if (!$tag) { return; }
637 while ($tag->{signature} ne SIG_BLIST) {
638 $tag = $self->{engine}->index_lookup($self, $tag, ord(substr($md5, $ch, 1)));
639 if (!$tag) { return; }
646 sub _traverse_index {
648 # Scan index and recursively step into deeper levels, looking for next key.
650 my ($self, $offset, $ch, $force_return_next) = @_;
651 $force_return_next = undef unless $force_return_next;
653 my $tag = $self->{engine}->load_tag($self, $offset );
657 if ($tag->{signature} ne SIG_BLIST) {
658 my $content = $tag->{content};
660 if ($self->{return_next}) { $start = 0; }
661 else { $start = ord(substr($self->{prev_md5}, $ch, 1)); }
663 for (my $index = $start; $index < 256; $index++) {
664 my $subloc = unpack($LONG_PACK, substr($content, $index * $LONG_SIZE, $LONG_SIZE) );
666 my $result = $self->_traverse_index( $subloc, $ch + 1, $force_return_next );
667 if (defined($result)) { return $result; }
671 $self->{return_next} = 1;
674 elsif ($tag->{signature} eq SIG_BLIST) {
675 my $keys = $tag->{content};
676 if ($force_return_next) { $self->{return_next} = 1; }
679 # Iterate through buckets, looking for a key match
681 for (my $i=0; $i<$MAX_BUCKETS; $i++) {
682 my $key = substr($keys, $i * $BUCKET_SIZE, $HASH_SIZE);
683 my $subloc = unpack($LONG_PACK, substr($keys, ($i * $BUCKET_SIZE) + $HASH_SIZE, $LONG_SIZE));
687 # End of bucket list -- return to outer loop
689 $self->{return_next} = 1;
692 elsif ($key eq $self->{prev_md5}) {
694 # Located previous key -- return next one found
696 $self->{return_next} = 1;
699 elsif ($self->{return_next}) {
701 # Seek to bucket location and skip over signature
703 seek($fh, $subloc + SIG_SIZE + $self->_root->{file_offset}, SEEK_SET);
706 # Skip over value to get to plain key
709 read( $fh, $size, $DATA_LENGTH_SIZE); $size = unpack($DATA_LENGTH_PACK, $size);
710 if ($size) { seek($fh, $size, SEEK_CUR); }
713 # Read in plain key and return as scalar
716 read( $fh, $size, $DATA_LENGTH_SIZE); $size = unpack($DATA_LENGTH_PACK, $size);
717 if ($size) { read( $fh, $plain_key, $size); }
723 $self->{return_next} = 1;
724 } # tag is a bucket list
731 # Locate next key, given digested previous one
733 my $self = $_[0]->_get_self;
735 $self->{prev_md5} = $_[1] ? $_[1] : undef;
736 $self->{return_next} = 0;
739 # If the previous key was not specifed, start at the top and
740 # return the first one found.
742 if (!$self->{prev_md5}) {
743 $self->{prev_md5} = chr(0) x $HASH_SIZE;
744 $self->{return_next} = 1;
747 return $self->_traverse_index( $self->_base_offset, 0 );
752 # If db locking is set, flock() the db file. If called multiple
753 # times before unlock(), then the same number of unlocks() must
754 # be called before the lock is released.
756 my $self = $_[0]->_get_self;
758 $type = LOCK_EX unless defined $type;
760 if (!defined($self->_fh)) { return; }
762 if ($self->_root->{locking}) {
763 if (!$self->_root->{locked}) {
764 flock($self->_fh, $type);
766 # refresh end counter in case file has changed size
767 my @stats = stat($self->_root->{file});
768 $self->_root->{end} = $stats[7];
770 # double-check file inode, in case another process
771 # has optimize()d our file while we were waiting.
772 if ($stats[1] != $self->_root->{inode}) {
773 $self->{engine}->open( $self ); # re-open
774 flock($self->_fh, $type); # re-lock
775 $self->_root->{end} = (stat($self->_fh))[7]; # re-end
778 $self->_root->{locked}++;
788 # If db locking is set, unlock the db file. See note in lock()
789 # regarding calling lock() multiple times.
791 my $self = $_[0]->_get_self;
793 if (!defined($self->_fh)) { return; }
795 if ($self->_root->{locking} && $self->_root->{locked} > 0) {
796 $self->_root->{locked}--;
797 if (!$self->_root->{locked}) { flock($self->_fh, LOCK_UN); }
806 my $self = shift->_get_self;
807 my ($spot, $value) = @_;
812 elsif ( eval { local $SIG{__DIE__}; $value->isa( 'DBM::Deep' ) } ) {
813 my $type = $value->_type;
814 ${$spot} = $type eq TYPE_HASH ? {} : [];
815 $value->_copy_node( ${$spot} );
818 my $r = Scalar::Util::reftype( $value );
819 my $c = Scalar::Util::blessed( $value );
820 if ( $r eq 'ARRAY' ) {
821 ${$spot} = [ @{$value} ];
824 ${$spot} = { %{$value} };
826 ${$spot} = bless ${$spot}, $c
835 # Copy single level of keys or elements to new DB handle.
836 # Recurse for nested structures
838 my $self = shift->_get_self;
841 if ($self->_type eq TYPE_HASH) {
842 my $key = $self->first_key();
844 my $value = $self->get($key);
845 $self->_copy_value( \$db_temp->{$key}, $value );
846 $key = $self->next_key($key);
850 my $length = $self->length();
851 for (my $index = 0; $index < $length; $index++) {
852 my $value = $self->get($index);
853 $self->_copy_value( \$db_temp->[$index], $value );
862 # Recursively export into standard Perl hashes and arrays.
864 my $self = $_[0]->_get_self;
867 if ($self->_type eq TYPE_HASH) { $temp = {}; }
868 elsif ($self->_type eq TYPE_ARRAY) { $temp = []; }
871 $self->_copy_node( $temp );
879 # Recursively import Perl hash/array structure
881 #XXX This use of ref() seems to be ok
882 if (!ref($_[0])) { return; } # Perl calls import() on use -- ignore
884 my $self = $_[0]->_get_self;
887 #XXX This use of ref() seems to be ok
890 # struct is not a reference, so just import based on our type
894 if ($self->_type eq TYPE_HASH) { $struct = {@_}; }
895 elsif ($self->_type eq TYPE_ARRAY) { $struct = [@_]; }
898 my $r = Scalar::Util::reftype($struct) || '';
899 if ($r eq "HASH" && $self->_type eq TYPE_HASH) {
900 foreach my $key (keys %$struct) { $self->put($key, $struct->{$key}); }
902 elsif ($r eq "ARRAY" && $self->_type eq TYPE_ARRAY) {
903 $self->push( @$struct );
906 return $self->_throw_error("Cannot import: type mismatch");
914 # Rebuild entire database into new file, then move
915 # it back on top of original.
917 my $self = $_[0]->_get_self;
919 #XXX Need to create a new test for this
920 # if ($self->_root->{links} > 1) {
921 # return $self->_throw_error("Cannot optimize: reference count is greater than 1");
924 my $db_temp = DBM::Deep->new(
925 file => $self->_root->{file} . '.tmp',
929 return $self->_throw_error("Cannot optimize: failed to open temp file: $!");
933 $self->_copy_node( $db_temp );
937 # Attempt to copy user, group and permissions over to new file
939 my @stats = stat($self->_fh);
940 my $perms = $stats[2] & 07777;
943 chown( $uid, $gid, $self->_root->{file} . '.tmp' );
944 chmod( $perms, $self->_root->{file} . '.tmp' );
946 # q.v. perlport for more information on this variable
947 if ( $^O eq 'MSWin32' || $^O eq 'cygwin' ) {
949 # Potential race condition when optmizing on Win32 with locking.
950 # The Windows filesystem requires that the filehandle be closed
951 # before it is overwritten with rename(). This could be redone
955 $self->{engine}->close( $self );
958 if (!rename $self->_root->{file} . '.tmp', $self->_root->{file}) {
959 unlink $self->_root->{file} . '.tmp';
961 return $self->_throw_error("Optimize failed: Cannot copy temp file over original: $!");
965 $self->{engine}->close( $self );
966 $self->{engine}->open( $self );
973 # Make copy of object and return
975 my $self = $_[0]->_get_self;
977 return DBM::Deep->new(
978 type => $self->_type,
979 base_offset => $self->_base_offset,
985 my %is_legal_filter = map {
988 store_key store_value
989 fetch_key fetch_value
994 # Setup filter function for storing or fetching the key or value
996 my $self = $_[0]->_get_self;
998 my $func = $_[2] ? $_[2] : undef;
1000 if ( $is_legal_filter{$type} ) {
1001 $self->_root->{"filter_$type"} = $func;
1015 # Get access to the root structure
1017 my $self = $_[0]->_get_self;
1018 return $self->{root};
1023 # Get access to the raw fh
1025 #XXX It will be useful, though, when we split out HASH and ARRAY
1026 my $self = $_[0]->_get_self;
1027 return $self->_root->{fh};
1032 # Get type of current node (TYPE_HASH or TYPE_ARRAY)
1034 my $self = $_[0]->_get_self;
1035 return $self->{type};
1040 # Get base_offset of current node (TYPE_HASH or TYPE_ARRAY)
1042 my $self = $_[0]->_get_self;
1043 return $self->{base_offset};
1051 die "DBM::Deep: $_[1]\n";
1054 sub _precalc_sizes {
1056 # Precalculate index, bucket and bucket list sizes
1059 #XXX I don't like this ...
1060 set_pack() unless defined $LONG_SIZE;
1062 $INDEX_SIZE = 256 * $LONG_SIZE;
1063 $BUCKET_SIZE = $HASH_SIZE + $LONG_SIZE;
1064 $BUCKET_LIST_SIZE = $MAX_BUCKETS * $BUCKET_SIZE;
1069 # Set pack/unpack modes (see file header for more)
1071 my ($long_s, $long_p, $data_s, $data_p) = @_;
1073 $LONG_SIZE = $long_s ? $long_s : 4;
1074 $LONG_PACK = $long_p ? $long_p : 'N';
1076 $DATA_LENGTH_SIZE = $data_s ? $data_s : 4;
1077 $DATA_LENGTH_PACK = $data_p ? $data_p : 'N';
1084 # Set key digest function (default is MD5)
1086 my ($digest_func, $hash_size) = @_;
1088 $DIGEST_FUNC = $digest_func ? $digest_func : \&Digest::MD5::md5;
1089 $HASH_SIZE = $hash_size ? $hash_size : 16;
1096 (O_WRONLY | O_RDWR) & fcntl( $fh, F_GETFL, my $slush = 0);
1101 # (O_RDONLY | O_RDWR) & fcntl( $fh, F_GETFL, my $slush = 0);
1105 # tie() methods (hashes and arrays)
1110 # Store single hash key/value or array element in database.
1112 my $self = $_[0]->_get_self;
1115 # User may be storing a hash, in which case we do not want it run
1116 # through the filtering system
1117 my $value = ($self->_root->{filter_store_value} && !ref($_[2]))
1118 ? $self->_root->{filter_store_value}->($_[2])
1121 my $md5 = $DIGEST_FUNC->($key);
1123 unless ( _is_writable( $self->_fh ) ) {
1124 $self->_throw_error( 'Cannot write to a readonly filehandle' );
1128 # Request exclusive lock for writing
1130 $self->lock( LOCK_EX );
1132 my $fh = $self->_fh;
1135 # Locate offset for bucket list using digest index system
1137 my $tag = $self->{engine}->load_tag($self, $self->_base_offset);
1139 $tag = $self->{engine}->create_tag($self, $self->_base_offset, SIG_INDEX, chr(0) x $INDEX_SIZE);
1143 while ($tag->{signature} ne SIG_BLIST) {
1144 my $num = ord(substr($md5, $ch, 1));
1146 my $ref_loc = $tag->{offset} + ($num * $LONG_SIZE);
1147 my $new_tag = $self->{engine}->index_lookup($self, $tag, $num);
1150 seek($fh, $ref_loc + $self->_root->{file_offset}, SEEK_SET);
1151 print( $fh pack($LONG_PACK, $self->_root->{end}) );
1153 $tag = $self->{engine}->create_tag($self, $self->_root->{end}, SIG_BLIST, chr(0) x $BUCKET_LIST_SIZE);
1155 $tag->{ref_loc} = $ref_loc;
1163 $tag->{ref_loc} = $ref_loc;
1170 # Add key/value to bucket list
1172 my $result = $self->_add_bucket( $tag, $md5, $key, $value );
1181 # Fetch single value or element given plain key or array index
1183 my $self = shift->_get_self;
1186 my $md5 = $DIGEST_FUNC->($key);
1189 # Request shared lock for reading
1191 $self->lock( LOCK_SH );
1193 my $tag = $self->_find_bucket_list( $md5 );
1200 # Get value from bucket list
1202 my $result = $self->_get_bucket_value( $tag, $md5 );
1206 #XXX What is ref() checking here?
1207 #YYY Filters only apply on scalar values, so the ref check is making
1208 #YYY sure the fetched bucket is a scalar, not a child hash or array.
1209 return ($result && !ref($result) && $self->_root->{filter_fetch_value})
1210 ? $self->_root->{filter_fetch_value}->($result)
1216 # Delete single key/value pair or element given plain key or array index
1218 my $self = $_[0]->_get_self;
1221 my $md5 = $DIGEST_FUNC->($key);
1224 # Request exclusive lock for writing
1226 $self->lock( LOCK_EX );
1228 my $tag = $self->_find_bucket_list( $md5 );
1237 my $value = $self->_get_bucket_value( $tag, $md5 );
1238 if ($value && !ref($value) && $self->_root->{filter_fetch_value}) {
1239 $value = $self->_root->{filter_fetch_value}->($value);
1242 my $result = $self->_delete_bucket( $tag, $md5 );
1245 # If this object is an array and the key deleted was on the end of the stack,
1246 # decrement the length variable.
1256 # Check if a single key or element exists given plain key or array index
1258 my $self = $_[0]->_get_self;
1261 my $md5 = $DIGEST_FUNC->($key);
1264 # Request shared lock for reading
1266 $self->lock( LOCK_SH );
1268 my $tag = $self->_find_bucket_list( $md5 );
1271 # For some reason, the built-in exists() function returns '' for false
1279 # Check if bucket exists and return 1 or ''
1281 my $result = $self->_bucket_exists( $tag, $md5 ) || '';
1290 # Clear all keys from hash, or all elements from array.
1292 my $self = $_[0]->_get_self;
1295 # Request exclusive lock for writing
1297 $self->lock( LOCK_EX );
1299 my $fh = $self->_fh;
1301 seek($fh, $self->_base_offset + $self->_root->{file_offset}, SEEK_SET);
1307 $self->{engine}->create_tag($self, $self->_base_offset, $self->_type, chr(0) x $INDEX_SIZE);
1315 # Public method aliases
1317 sub put { (shift)->STORE( @_ ) }
1318 sub store { (shift)->STORE( @_ ) }
1319 sub get { (shift)->FETCH( @_ ) }
1320 sub fetch { (shift)->FETCH( @_ ) }
1321 sub delete { (shift)->DELETE( @_ ) }
1322 sub exists { (shift)->EXISTS( @_ ) }
1323 sub clear { (shift)->CLEAR( @_ ) }
1325 package DBM::Deep::_::Root;
1339 filter_store_key => undef,
1340 filter_store_value => undef,
1341 filter_fetch_key => undef,
1342 filter_fetch_value => undef,
1348 if ( $self->{fh} && !$self->{file_offset} ) {
1349 $self->{file_offset} = tell( $self->{fh} );
1357 return unless $self;
1359 close $self->{fh} if $self->{fh};
1370 DBM::Deep - A pure perl multi-level hash/array DBM
1375 my $db = DBM::Deep->new( "foo.db" );
1377 $db->{key} = 'value'; # tie() style
1380 $db->put('key' => 'value'); # OO style
1381 print $db->get('key');
1383 # true multi-level support
1384 $db->{my_complex} = [
1385 'hello', { perl => 'rules' },
1391 A unique flat-file database module, written in pure perl. True
1392 multi-level hash/array support (unlike MLDBM, which is faked), hybrid
1393 OO / tie() interface, cross-platform FTPable files, and quite fast. Can
1394 handle millions of keys and unlimited hash levels without significant
1395 slow-down. Written from the ground-up in pure perl -- this is NOT a
1396 wrapper around a C-based DBM. Out-of-the-box compatibility with Unix,
1397 Mac OS X and Windows.
1401 Hopefully you are using Perl's excellent CPAN module, which will download
1402 and install the module for you. If not, get the tarball, and run these
1414 Construction can be done OO-style (which is the recommended way), or using
1415 Perl's tie() function. Both are examined here.
1417 =head2 OO CONSTRUCTION
1419 The recommended way to construct a DBM::Deep object is to use the new()
1420 method, which gets you a blessed, tied hash or array reference.
1422 my $db = DBM::Deep->new( "foo.db" );
1424 This opens a new database handle, mapped to the file "foo.db". If this
1425 file does not exist, it will automatically be created. DB files are
1426 opened in "r+" (read/write) mode, and the type of object returned is a
1427 hash, unless otherwise specified (see L<OPTIONS> below).
1429 You can pass a number of options to the constructor to specify things like
1430 locking, autoflush, etc. This is done by passing an inline hash:
1432 my $db = DBM::Deep->new(
1438 Notice that the filename is now specified I<inside> the hash with
1439 the "file" parameter, as opposed to being the sole argument to the
1440 constructor. This is required if any options are specified.
1441 See L<OPTIONS> below for the complete list.
1445 You can also start with an array instead of a hash. For this, you must
1446 specify the C<type> parameter:
1448 my $db = DBM::Deep->new(
1450 type => DBM::Deep->TYPE_ARRAY
1453 B<Note:> Specifing the C<type> parameter only takes effect when beginning
1454 a new DB file. If you create a DBM::Deep object with an existing file, the
1455 C<type> will be loaded from the file header, and an error will be thrown if
1456 the wrong type is passed in.
1458 =head2 TIE CONSTRUCTION
1460 Alternately, you can create a DBM::Deep handle by using Perl's built-in
1461 tie() function. The object returned from tie() can be used to call methods,
1462 such as lock() and unlock(), but cannot be used to assign to the DBM::Deep
1463 file (as expected with most tie'd objects).
1466 my $db = tie %hash, "DBM::Deep", "foo.db";
1469 my $db = tie @array, "DBM::Deep", "bar.db";
1471 As with the OO constructor, you can replace the DB filename parameter with
1472 a hash containing one or more options (see L<OPTIONS> just below for the
1475 tie %hash, "DBM::Deep", {
1483 There are a number of options that can be passed in when constructing your
1484 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
1490 Filename of the DB file to link the handle to. You can pass a full absolute
1491 filesystem path, partial path, or a plain filename if the file is in the
1492 current working directory. This is a required parameter (though q.v. fh).
1496 If you want, you can pass in the fh instead of the file. This is most useful for doing
1499 my $db = DBM::Deep->new( { fh => \*DATA } );
1501 You are responsible for making sure that the fh has been opened appropriately for your
1502 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
1503 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
1504 needs to read from the fh.
1508 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
1509 not need to set this. However, it's there if you want it.
1511 If you pass in fh and do not set this, it will be set appropriately.
1515 This parameter specifies what type of object to create, a hash or array. Use
1516 one of these two constants: C<DBM::Deep-E<gt>TYPE_HASH> or C<DBM::Deep-E<gt>TYPE_ARRAY>.
1517 This only takes effect when beginning a new file. This is an optional
1518 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
1522 Specifies whether locking is to be enabled. DBM::Deep uses Perl's Fnctl flock()
1523 function to lock the database in exclusive mode for writes, and shared mode for
1524 reads. Pass any true value to enable. This affects the base DB handle I<and
1525 any child hashes or arrays> that use the same DB file. This is an optional
1526 parameter, and defaults to 0 (disabled). See L<LOCKING> below for more.
1530 Specifies whether autoflush is to be enabled on the underlying filehandle.
1531 This obviously slows down write operations, but is required if you may have
1532 multiple processes accessing the same DB file (also consider enable I<locking>).
1533 Pass any true value to enable. This is an optional parameter, and defaults to 0
1538 If I<autobless> mode is enabled, DBM::Deep will preserve blessed hashes, and
1539 restore them when fetched. This is an B<experimental> feature, and does have
1540 side-effects. Basically, when hashes are re-blessed into their original
1541 classes, they are no longer blessed into the DBM::Deep class! So you won't be
1542 able to call any DBM::Deep methods on them. You have been warned.
1543 This is an optional parameter, and defaults to 0 (disabled).
1547 See L<FILTERS> below.
1551 Setting I<debug> mode will make all errors non-fatal, dump them out to
1552 STDERR, and continue on. This is for debugging purposes only, and probably
1553 not what you want. This is an optional parameter, and defaults to 0 (disabled).
1555 B<NOTE>: This parameter is considered deprecated and should not be used anymore.
1559 =head1 TIE INTERFACE
1561 With DBM::Deep you can access your databases using Perl's standard hash/array
1562 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
1563 treat them as such. DBM::Deep will intercept all reads/writes and direct them
1564 to the right place -- the DB file. This has nothing to do with the
1565 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
1566 using regular hashes and arrays, rather than calling functions like C<get()>
1567 and C<put()> (although those work too). It is entirely up to you how to want
1568 to access your databases.
1572 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
1573 or even nested hashes (or arrays) using standard Perl syntax:
1575 my $db = DBM::Deep->new( "foo.db" );
1577 $db->{mykey} = "myvalue";
1579 $db->{myhash}->{subkey} = "subvalue";
1581 print $db->{myhash}->{subkey} . "\n";
1583 You can even step through hash keys using the normal Perl C<keys()> function:
1585 foreach my $key (keys %$db) {
1586 print "$key: " . $db->{$key} . "\n";
1589 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
1590 pushes them onto an array, all before the loop even begins. If you have an
1591 extra large hash, this may exhaust Perl's memory. Instead, consider using
1592 Perl's C<each()> function, which pulls keys/values one at a time, using very
1595 while (my ($key, $value) = each %$db) {
1596 print "$key: $value\n";
1599 Please note that when using C<each()>, you should always pass a direct
1600 hash reference, not a lookup. Meaning, you should B<never> do this:
1603 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
1605 This causes an infinite loop, because for each iteration, Perl is calling
1606 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
1607 it effectively keeps returning the first key over and over again. Instead,
1608 assign a temporary variable to C<$db->{foo}>, then pass that to each().
1612 As with hashes, you can treat any DBM::Deep object like a normal Perl array
1613 reference. This includes inserting, removing and manipulating elements,
1614 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
1615 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
1616 or simply be a nested array reference inside a hash. Example:
1618 my $db = DBM::Deep->new(
1619 file => "foo-array.db",
1620 type => DBM::Deep->TYPE_ARRAY
1624 push @$db, "bar", "baz";
1625 unshift @$db, "bah";
1627 my $last_elem = pop @$db; # baz
1628 my $first_elem = shift @$db; # bah
1629 my $second_elem = $db->[1]; # bar
1631 my $num_elements = scalar @$db;
1635 In addition to the I<tie()> interface, you can also use a standard OO interface
1636 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
1637 array) has its own methods, but both types share the following common methods:
1638 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>.
1642 =item * new() / clone()
1644 These are the constructor and copy-functions.
1646 =item * put() / store()
1648 Stores a new hash key/value pair, or sets an array element value. Takes two
1649 arguments, the hash key or array index, and the new value. The value can be
1650 a scalar, hash ref or array ref. Returns true on success, false on failure.
1652 $db->put("foo", "bar"); # for hashes
1653 $db->put(1, "bar"); # for arrays
1655 =item * get() / fetch()
1657 Fetches the value of a hash key or array element. Takes one argument: the hash
1658 key or array index. Returns a scalar, hash ref or array ref, depending on the
1661 my $value = $db->get("foo"); # for hashes
1662 my $value = $db->get(1); # for arrays
1666 Checks if a hash key or array index exists. Takes one argument: the hash key
1667 or array index. Returns true if it exists, false if not.
1669 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
1670 if ($db->exists(1)) { print "yay!\n"; } # for arrays
1674 Deletes one hash key/value pair or array element. Takes one argument: the hash
1675 key or array index. Returns true on success, false if not found. For arrays,
1676 the remaining elements located after the deleted element are NOT moved over.
1677 The deleted element is essentially just undefined, which is exactly how Perl's
1678 internal arrays work. Please note that the space occupied by the deleted
1679 key/value or element is B<not> reused again -- see L<UNUSED SPACE RECOVERY>
1680 below for details and workarounds.
1682 $db->delete("foo"); # for hashes
1683 $db->delete(1); # for arrays
1687 Deletes B<all> hash keys or array elements. Takes no arguments. No return
1688 value. Please note that the space occupied by the deleted keys/values or
1689 elements is B<not> reused again -- see L<UNUSED SPACE RECOVERY> below for
1690 details and workarounds.
1692 $db->clear(); # hashes or arrays
1694 =item * lock() / unlock()
1700 Recover lost disk space.
1702 =item * import() / export()
1704 Data going in and out.
1706 =item * set_digest() / set_pack() / set_filter()
1708 q.v. adjusting the interal parameters.
1714 For hashes, DBM::Deep supports all the common methods described above, and the
1715 following additional methods: C<first_key()> and C<next_key()>.
1721 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
1722 fetched in an undefined order (which appears random). Takes no arguments,
1723 returns the key as a scalar value.
1725 my $key = $db->first_key();
1729 Returns the "next" key in the hash, given the previous one as the sole argument.
1730 Returns undef if there are no more keys to be fetched.
1732 $key = $db->next_key($key);
1736 Here are some examples of using hashes:
1738 my $db = DBM::Deep->new( "foo.db" );
1740 $db->put("foo", "bar");
1741 print "foo: " . $db->get("foo") . "\n";
1743 $db->put("baz", {}); # new child hash ref
1744 $db->get("baz")->put("buz", "biz");
1745 print "buz: " . $db->get("baz")->get("buz") . "\n";
1747 my $key = $db->first_key();
1749 print "$key: " . $db->get($key) . "\n";
1750 $key = $db->next_key($key);
1753 if ($db->exists("foo")) { $db->delete("foo"); }
1757 For arrays, DBM::Deep supports all the common methods described above, and the
1758 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
1759 C<unshift()> and C<splice()>.
1765 Returns the number of elements in the array. Takes no arguments.
1767 my $len = $db->length();
1771 Adds one or more elements onto the end of the array. Accepts scalars, hash
1772 refs or array refs. No return value.
1774 $db->push("foo", "bar", {});
1778 Fetches the last element in the array, and deletes it. Takes no arguments.
1779 Returns undef if array is empty. Returns the element value.
1781 my $elem = $db->pop();
1785 Fetches the first element in the array, deletes it, then shifts all the
1786 remaining elements over to take up the space. Returns the element value. This
1787 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
1790 my $elem = $db->shift();
1794 Inserts one or more elements onto the beginning of the array, shifting all
1795 existing elements over to make room. Accepts scalars, hash refs or array refs.
1796 No return value. This method is not recommended with large arrays -- see
1797 <LARGE ARRAYS> below for details.
1799 $db->unshift("foo", "bar", {});
1803 Performs exactly like Perl's built-in function of the same name. See L<perldoc
1804 -f splice> for usage -- it is too complicated to document here. This method is
1805 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
1809 Here are some examples of using arrays:
1811 my $db = DBM::Deep->new(
1813 type => DBM::Deep->TYPE_ARRAY
1816 $db->push("bar", "baz");
1817 $db->unshift("foo");
1820 my $len = $db->length();
1821 print "length: $len\n"; # 4
1823 for (my $k=0; $k<$len; $k++) {
1824 print "$k: " . $db->get($k) . "\n";
1827 $db->splice(1, 2, "biz", "baf");
1829 while (my $elem = shift @$db) {
1830 print "shifted: $elem\n";
1835 Enable automatic file locking by passing a true value to the C<locking>
1836 parameter when constructing your DBM::Deep object (see L<SETUP> above).
1838 my $db = DBM::Deep->new(
1843 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
1844 mode for writes, and shared mode for reads. This is required if you have
1845 multiple processes accessing the same database file, to avoid file corruption.
1846 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
1847 NFS> below for more.
1849 =head2 EXPLICIT LOCKING
1851 You can explicitly lock a database, so it remains locked for multiple
1852 transactions. This is done by calling the C<lock()> method, and passing an
1853 optional lock mode argument (defaults to exclusive mode). This is particularly
1854 useful for things like counters, where the current value needs to be fetched,
1855 then incremented, then stored again.
1858 my $counter = $db->get("counter");
1860 $db->put("counter", $counter);
1869 You can pass C<lock()> an optional argument, which specifies which mode to use
1870 (exclusive or shared). Use one of these two constants: C<DBM::Deep-E<gt>LOCK_EX>
1871 or C<DBM::Deep-E<gt>LOCK_SH>. These are passed directly to C<flock()>, and are the
1872 same as the constants defined in Perl's C<Fcntl> module.
1874 $db->lock( DBM::Deep->LOCK_SH );
1878 =head1 IMPORTING/EXPORTING
1880 You can import existing complex structures by calling the C<import()> method,
1881 and export an entire database into an in-memory structure using the C<export()>
1882 method. Both are examined here.
1886 Say you have an existing hash with nested hashes/arrays inside it. Instead of
1887 walking the structure and adding keys/elements to the database as you go,
1888 simply pass a reference to the C<import()> method. This recursively adds
1889 everything to an existing DBM::Deep object for you. Here is an example:
1894 array1 => [ "elem0", "elem1", "elem2" ],
1896 subkey1 => "subvalue1",
1897 subkey2 => "subvalue2"
1901 my $db = DBM::Deep->new( "foo.db" );
1902 $db->import( $struct );
1904 print $db->{key1} . "\n"; # prints "value1"
1906 This recursively imports the entire C<$struct> object into C<$db>, including
1907 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
1908 keys are merged with the existing ones, replacing if they already exist.
1909 The C<import()> method can be called on any database level (not just the base
1910 level), and works with both hash and array DB types.
1912 B<Note:> Make sure your existing structure has no circular references in it.
1913 These will cause an infinite loop when importing.
1917 Calling the C<export()> method on an existing DBM::Deep object will return
1918 a reference to a new in-memory copy of the database. The export is done
1919 recursively, so all nested hashes/arrays are all exported to standard Perl
1920 objects. Here is an example:
1922 my $db = DBM::Deep->new( "foo.db" );
1924 $db->{key1} = "value1";
1925 $db->{key2} = "value2";
1927 $db->{hash1}->{subkey1} = "subvalue1";
1928 $db->{hash1}->{subkey2} = "subvalue2";
1930 my $struct = $db->export();
1932 print $struct->{key1} . "\n"; # prints "value1"
1934 This makes a complete copy of the database in memory, and returns a reference
1935 to it. The C<export()> method can be called on any database level (not just
1936 the base level), and works with both hash and array DB types. Be careful of
1937 large databases -- you can store a lot more data in a DBM::Deep object than an
1938 in-memory Perl structure.
1940 B<Note:> Make sure your database has no circular references in it.
1941 These will cause an infinite loop when exporting.
1945 DBM::Deep has a number of hooks where you can specify your own Perl function
1946 to perform filtering on incoming or outgoing data. This is a perfect
1947 way to extend the engine, and implement things like real-time compression or
1948 encryption. Filtering applies to the base DB level, and all child hashes /
1949 arrays. Filter hooks can be specified when your DBM::Deep object is first
1950 constructed, or by calling the C<set_filter()> method at any time. There are
1951 four available filter hooks, described below:
1955 =item * filter_store_key
1957 This filter is called whenever a hash key is stored. It
1958 is passed the incoming key, and expected to return a transformed key.
1960 =item * filter_store_value
1962 This filter is called whenever a hash key or array element is stored. It
1963 is passed the incoming value, and expected to return a transformed value.
1965 =item * filter_fetch_key
1967 This filter is called whenever a hash key is fetched (i.e. via
1968 C<first_key()> or C<next_key()>). It is passed the transformed key,
1969 and expected to return the plain key.
1971 =item * filter_fetch_value
1973 This filter is called whenever a hash key or array element is fetched.
1974 It is passed the transformed value, and expected to return the plain value.
1978 Here are the two ways to setup a filter hook:
1980 my $db = DBM::Deep->new(
1982 filter_store_value => \&my_filter_store,
1983 filter_fetch_value => \&my_filter_fetch
1988 $db->set_filter( "filter_store_value", \&my_filter_store );
1989 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
1991 Your filter function will be called only when dealing with SCALAR keys or
1992 values. When nested hashes and arrays are being stored/fetched, filtering
1993 is bypassed. Filters are called as static functions, passed a single SCALAR
1994 argument, and expected to return a single SCALAR value. If you want to
1995 remove a filter, set the function reference to C<undef>:
1997 $db->set_filter( "filter_store_value", undef );
1999 =head2 REAL-TIME ENCRYPTION EXAMPLE
2001 Here is a working example that uses the I<Crypt::Blowfish> module to
2002 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
2003 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
2004 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
2007 use Crypt::Blowfish;
2010 my $cipher = Crypt::CBC->new({
2011 'key' => 'my secret key',
2012 'cipher' => 'Blowfish',
2014 'regenerate_key' => 0,
2015 'padding' => 'space',
2019 my $db = DBM::Deep->new(
2020 file => "foo-encrypt.db",
2021 filter_store_key => \&my_encrypt,
2022 filter_store_value => \&my_encrypt,
2023 filter_fetch_key => \&my_decrypt,
2024 filter_fetch_value => \&my_decrypt,
2027 $db->{key1} = "value1";
2028 $db->{key2} = "value2";
2029 print "key1: " . $db->{key1} . "\n";
2030 print "key2: " . $db->{key2} . "\n";
2036 return $cipher->encrypt( $_[0] );
2039 return $cipher->decrypt( $_[0] );
2042 =head2 REAL-TIME COMPRESSION EXAMPLE
2044 Here is a working example that uses the I<Compress::Zlib> module to do real-time
2045 compression / decompression of keys & values with DBM::Deep Filters.
2046 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
2047 more on I<Compress::Zlib>.
2052 my $db = DBM::Deep->new(
2053 file => "foo-compress.db",
2054 filter_store_key => \&my_compress,
2055 filter_store_value => \&my_compress,
2056 filter_fetch_key => \&my_decompress,
2057 filter_fetch_value => \&my_decompress,
2060 $db->{key1} = "value1";
2061 $db->{key2} = "value2";
2062 print "key1: " . $db->{key1} . "\n";
2063 print "key2: " . $db->{key2} . "\n";
2069 return Compress::Zlib::memGzip( $_[0] ) ;
2072 return Compress::Zlib::memGunzip( $_[0] ) ;
2075 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
2076 actually numerical index numbers, and are not filtered.
2078 =head1 ERROR HANDLING
2080 Most DBM::Deep methods return a true value for success, and call die() on
2081 failure. You can wrap calls in an eval block to catch the die.
2083 my $db = DBM::Deep->new( "foo.db" ); # create hash
2084 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
2086 print $@; # prints error message
2088 =head1 LARGEFILE SUPPORT
2090 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
2091 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
2092 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
2093 by calling the static C<set_pack()> method before you do anything else.
2095 DBM::Deep::set_pack(8, 'Q');
2097 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
2098 instead of 32-bit longs. After setting these values your DB files have a
2099 theoretical maximum size of 16 XB (exabytes).
2101 B<Note:> Changing these values will B<NOT> work for existing database files.
2102 Only change this for new files, and make sure it stays set consistently
2103 throughout the file's life. If you do set these values, you can no longer
2104 access 32-bit DB files. You can, however, call C<set_pack(4, 'N')> to change
2105 back to 32-bit mode.
2107 B<Note:> I have not personally tested files > 2 GB -- all my systems have
2108 only a 32-bit Perl. However, I have received user reports that this does
2111 =head1 LOW-LEVEL ACCESS
2113 If you require low-level access to the underlying filehandle that DBM::Deep uses,
2114 you can call the C<_fh()> method, which returns the handle:
2116 my $fh = $db->_fh();
2118 This method can be called on the root level of the datbase, or any child
2119 hashes or arrays. All levels share a I<root> structure, which contains things
2120 like the filehandle, a reference counter, and all the options specified
2121 when you created the object. You can get access to this root structure by
2122 calling the C<root()> method.
2124 my $root = $db->_root();
2126 This is useful for changing options after the object has already been created,
2127 such as enabling/disabling locking, or debug modes. You can also
2128 store your own temporary user data in this structure (be wary of name
2129 collision), which is then accessible from any child hash or array.
2131 =head1 CUSTOM DIGEST ALGORITHM
2133 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
2134 keys. However you can override this, and use another algorithm (such as SHA-256)
2135 or even write your own. But please note that DBM::Deep currently expects zero
2136 collisions, so your algorithm has to be I<perfect>, so to speak.
2137 Collision detection may be introduced in a later version.
2141 You can specify a custom digest algorithm by calling the static C<set_digest()>
2142 function, passing a reference to a subroutine, and the length of the algorithm's
2143 hashes (in bytes). This is a global static function, which affects ALL DBM::Deep
2144 objects. Here is a working example that uses a 256-bit hash from the
2145 I<Digest::SHA256> module. Please see
2146 L<http://search.cpan.org/search?module=Digest::SHA256> for more.
2151 my $context = Digest::SHA256::new(256);
2153 DBM::Deep::set_digest( \&my_digest, 32 );
2155 my $db = DBM::Deep->new( "foo-sha.db" );
2157 $db->{key1} = "value1";
2158 $db->{key2} = "value2";
2159 print "key1: " . $db->{key1} . "\n";
2160 print "key2: " . $db->{key2} . "\n";
2166 return substr( $context->hash($_[0]), 0, 32 );
2169 B<Note:> Your returned digest strings must be B<EXACTLY> the number
2170 of bytes you specify in the C<set_digest()> function (in this case 32).
2172 =head1 CIRCULAR REFERENCES
2174 DBM::Deep has B<experimental> support for circular references. Meaning you
2175 can have a nested hash key or array element that points to a parent object.
2176 This relationship is stored in the DB file, and is preserved between sessions.
2179 my $db = DBM::Deep->new( "foo.db" );
2182 $db->{circle} = $db; # ref to self
2184 print $db->{foo} . "\n"; # prints "foo"
2185 print $db->{circle}->{foo} . "\n"; # prints "foo" again
2187 One catch is, passing the object to a function that recursively walks the
2188 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
2189 C<export()> methods) will result in an infinite loop. The other catch is,
2190 if you fetch the I<key> of a circular reference (i.e. using the C<first_key()>
2191 or C<next_key()> methods), you will get the I<target object's key>, not the
2192 ref's key. This gets even more interesting with the above example, where
2193 the I<circle> key points to the base DB object, which technically doesn't
2194 have a key. So I made DBM::Deep return "[base]" as the key name in that
2197 =head1 CAVEATS / ISSUES / BUGS
2199 This section describes all the known issues with DBM::Deep. It you have found
2200 something that is not listed here, please send e-mail to L<jhuckaby@cpan.org>.
2202 =head2 UNUSED SPACE RECOVERY
2204 One major caveat with DBM::Deep is that space occupied by existing keys and
2205 values is not recovered when they are deleted. Meaning if you keep deleting
2206 and adding new keys, your file will continuously grow. I am working on this,
2207 but in the meantime you can call the built-in C<optimize()> method from time to
2208 time (perhaps in a crontab or something) to recover all your unused space.
2210 $db->optimize(); # returns true on success
2212 This rebuilds the ENTIRE database into a new file, then moves it on top of
2213 the original. The new file will have no unused space, thus it will take up as
2214 little disk space as possible. Please note that this operation can take
2215 a long time for large files, and you need enough disk space to temporarily hold
2216 2 copies of your DB file. The temporary file is created in the same directory
2217 as the original, named with a ".tmp" extension, and is deleted when the
2218 operation completes. Oh, and if locking is enabled, the DB is automatically
2219 locked for the entire duration of the copy.
2221 B<WARNING:> Only call optimize() on the top-level node of the database, and
2222 make sure there are no child references lying around. DBM::Deep keeps a reference
2223 counter, and if it is greater than 1, optimize() will abort and return undef.
2225 =head2 AUTOVIVIFICATION
2227 Unfortunately, autovivification doesn't work with tied hashes. This appears to
2228 be a bug in Perl's tie() system, as I<Jakob Schmidt> encountered the very same
2229 issue with his I<DWH_FIle> module (see L<http://search.cpan.org/search?module=DWH_File>),
2230 and it is also mentioned in the BUGS section for the I<MLDBM> module <see
2231 L<http://search.cpan.org/search?module=MLDBM>). Basically, on a new db file,
2234 $db->{foo}->{bar} = "hello";
2236 Since "foo" doesn't exist, you cannot add "bar" to it. You end up with "foo"
2237 being an empty hash. Try this instead, which works fine:
2239 $db->{foo} = { bar => "hello" };
2241 As of Perl 5.8.7, this bug still exists. I have walked very carefully through
2242 the execution path, and Perl indeed passes an empty hash to the STORE() method.
2243 Probably a bug in Perl.
2245 =head2 FILE CORRUPTION
2247 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
2248 for a 32-bit signature when opened, but other corruption in files can cause
2249 segmentation faults. DBM::Deep may try to seek() past the end of a file, or get
2250 stuck in an infinite loop depending on the level of corruption. File write
2251 operations are not checked for failure (for speed), so if you happen to run
2252 out of disk space, DBM::Deep will probably fail in a bad way. These things will
2253 be addressed in a later version of DBM::Deep.
2257 Beware of using DB files over NFS. DBM::Deep uses flock(), which works well on local
2258 filesystems, but will NOT protect you from file corruption over NFS. I've heard
2259 about setting up your NFS server with a locking daemon, then using lockf() to
2260 lock your files, but your mileage may vary there as well. From what I
2261 understand, there is no real way to do it. However, if you need access to the
2262 underlying filehandle in DBM::Deep for using some other kind of locking scheme like
2263 lockf(), see the L<LOW-LEVEL ACCESS> section above.
2265 =head2 COPYING OBJECTS
2267 Beware of copying tied objects in Perl. Very strange things can happen.
2268 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
2269 returns a new, blessed, tied hash or array to the same level in the DB.
2271 my $copy = $db->clone();
2273 B<Note>: Since clone() here is cloning the object, not the database location, any
2274 modifications to either $db or $copy will be visible in both.
2278 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
2279 These functions cause every element in the array to move, which can be murder
2280 on DBM::Deep, as every element has to be fetched from disk, then stored again in
2281 a different location. This will be addressed in the forthcoming version 1.00.
2283 =head2 WRITEONLY FILES
2285 If you pass in a filehandle to new(), you may have opened it in either a readonly or
2286 writeonly mode. STORE will verify that the filehandle is writable. However, there
2287 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
2288 filehandle isn't readable, it's not clear what will happen. So, don't do that.
2292 This section discusses DBM::Deep's speed and memory usage.
2296 Obviously, DBM::Deep isn't going to be as fast as some C-based DBMs, such as
2297 the almighty I<BerkeleyDB>. But it makes up for it in features like true
2298 multi-level hash/array support, and cross-platform FTPable files. Even so,
2299 DBM::Deep is still pretty fast, and the speed stays fairly consistent, even
2300 with huge databases. Here is some test data:
2302 Adding 1,000,000 keys to new DB file...
2304 At 100 keys, avg. speed is 2,703 keys/sec
2305 At 200 keys, avg. speed is 2,642 keys/sec
2306 At 300 keys, avg. speed is 2,598 keys/sec
2307 At 400 keys, avg. speed is 2,578 keys/sec
2308 At 500 keys, avg. speed is 2,722 keys/sec
2309 At 600 keys, avg. speed is 2,628 keys/sec
2310 At 700 keys, avg. speed is 2,700 keys/sec
2311 At 800 keys, avg. speed is 2,607 keys/sec
2312 At 900 keys, avg. speed is 2,190 keys/sec
2313 At 1,000 keys, avg. speed is 2,570 keys/sec
2314 At 2,000 keys, avg. speed is 2,417 keys/sec
2315 At 3,000 keys, avg. speed is 1,982 keys/sec
2316 At 4,000 keys, avg. speed is 1,568 keys/sec
2317 At 5,000 keys, avg. speed is 1,533 keys/sec
2318 At 6,000 keys, avg. speed is 1,787 keys/sec
2319 At 7,000 keys, avg. speed is 1,977 keys/sec
2320 At 8,000 keys, avg. speed is 2,028 keys/sec
2321 At 9,000 keys, avg. speed is 2,077 keys/sec
2322 At 10,000 keys, avg. speed is 2,031 keys/sec
2323 At 20,000 keys, avg. speed is 1,970 keys/sec
2324 At 30,000 keys, avg. speed is 2,050 keys/sec
2325 At 40,000 keys, avg. speed is 2,073 keys/sec
2326 At 50,000 keys, avg. speed is 1,973 keys/sec
2327 At 60,000 keys, avg. speed is 1,914 keys/sec
2328 At 70,000 keys, avg. speed is 2,091 keys/sec
2329 At 80,000 keys, avg. speed is 2,103 keys/sec
2330 At 90,000 keys, avg. speed is 1,886 keys/sec
2331 At 100,000 keys, avg. speed is 1,970 keys/sec
2332 At 200,000 keys, avg. speed is 2,053 keys/sec
2333 At 300,000 keys, avg. speed is 1,697 keys/sec
2334 At 400,000 keys, avg. speed is 1,838 keys/sec
2335 At 500,000 keys, avg. speed is 1,941 keys/sec
2336 At 600,000 keys, avg. speed is 1,930 keys/sec
2337 At 700,000 keys, avg. speed is 1,735 keys/sec
2338 At 800,000 keys, avg. speed is 1,795 keys/sec
2339 At 900,000 keys, avg. speed is 1,221 keys/sec
2340 At 1,000,000 keys, avg. speed is 1,077 keys/sec
2342 This test was performed on a PowerMac G4 1gHz running Mac OS X 10.3.2 & Perl
2343 5.8.1, with an 80GB Ultra ATA/100 HD spinning at 7200RPM. The hash keys and
2344 values were between 6 - 12 chars in length. The DB file ended up at 210MB.
2345 Run time was 12 min 3 sec.
2349 One of the great things about DBM::Deep is that it uses very little memory.
2350 Even with huge databases (1,000,000+ keys) you will not see much increased
2351 memory on your process. DBM::Deep relies solely on the filesystem for storing
2352 and fetching data. Here is output from I</usr/bin/top> before even opening a
2355 PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND
2356 22831 root 11 0 2716 2716 1296 R 0.0 0.2 0:07 perl
2358 Basically the process is taking 2,716K of memory. And here is the same
2359 process after storing and fetching 1,000,000 keys:
2361 PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND
2362 22831 root 14 0 2772 2772 1328 R 0.0 0.2 13:32 perl
2364 Notice the memory usage increased by only 56K. Test was performed on a 700mHz
2365 x86 box running Linux RedHat 7.2 & Perl 5.6.1.
2367 =head1 DB FILE FORMAT
2369 In case you were interested in the underlying DB file format, it is documented
2370 here in this section. You don't need to know this to use the module, it's just
2371 included for reference.
2375 DBM::Deep files always start with a 32-bit signature to identify the file type.
2376 This is at offset 0. The signature is "DPDB" in network byte order. This is
2377 checked for when the file is opened and an error will be thrown if it's not found.
2381 The DBM::Deep file is in a I<tagged format>, meaning each section of the file
2382 has a standard header containing the type of data, the length of data, and then
2383 the data itself. The type is a single character (1 byte), the length is a
2384 32-bit unsigned long in network byte order, and the data is, well, the data.
2385 Here is how it unfolds:
2389 Immediately after the 32-bit file signature is the I<Master Index> record.
2390 This is a standard tag header followed by 1024 bytes (in 32-bit mode) or 2048
2391 bytes (in 64-bit mode) of data. The type is I<H> for hash or I<A> for array,
2392 depending on how the DBM::Deep object was constructed.
2394 The index works by looking at a I<MD5 Hash> of the hash key (or array index
2395 number). The first 8-bit char of the MD5 signature is the offset into the
2396 index, multipled by 4 in 32-bit mode, or 8 in 64-bit mode. The value of the
2397 index element is a file offset of the next tag for the key/element in question,
2398 which is usually a I<Bucket List> tag (see below).
2400 The next tag I<could> be another index, depending on how many keys/elements
2401 exist. See L<RE-INDEXING> below for details.
2405 A I<Bucket List> is a collection of 16 MD5 hashes for keys/elements, plus
2406 file offsets to where the actual data is stored. It starts with a standard
2407 tag header, with type I<B>, and a data size of 320 bytes in 32-bit mode, or
2408 384 bytes in 64-bit mode. Each MD5 hash is stored in full (16 bytes), plus
2409 the 32-bit or 64-bit file offset for the I<Bucket> containing the actual data.
2410 When the list fills up, a I<Re-Index> operation is performed (See
2411 L<RE-INDEXING> below).
2415 A I<Bucket> is a tag containing a key/value pair (in hash mode), or a
2416 index/value pair (in array mode). It starts with a standard tag header with
2417 type I<D> for scalar data (string, binary, etc.), or it could be a nested
2418 hash (type I<H>) or array (type I<A>). The value comes just after the tag
2419 header. The size reported in the tag header is only for the value, but then,
2420 just after the value is another size (32-bit unsigned long) and then the plain
2421 key itself. Since the value is likely to be fetched more often than the plain
2422 key, I figured it would be I<slightly> faster to store the value first.
2424 If the type is I<H> (hash) or I<A> (array), the value is another I<Master Index>
2425 record for the nested structure, where the process begins all over again.
2429 After a I<Bucket List> grows to 16 records, its allocated space in the file is
2430 exhausted. Then, when another key/element comes in, the list is converted to a
2431 new index record. However, this index will look at the next char in the MD5
2432 hash, and arrange new Bucket List pointers accordingly. This process is called
2433 I<Re-Indexing>. Basically, a new index tag is created at the file EOF, and all
2434 17 (16 + new one) keys/elements are removed from the old Bucket List and
2435 inserted into the new index. Several new Bucket Lists are created in the
2436 process, as a new MD5 char from the key is being examined (it is unlikely that
2437 the keys will all share the same next char of their MD5s).
2439 Because of the way the I<MD5> algorithm works, it is impossible to tell exactly
2440 when the Bucket Lists will turn into indexes, but the first round tends to
2441 happen right around 4,000 keys. You will see a I<slight> decrease in
2442 performance here, but it picks back up pretty quick (see L<SPEED> above). Then
2443 it takes B<a lot> more keys to exhaust the next level of Bucket Lists. It's
2444 right around 900,000 keys. This process can continue nearly indefinitely --
2445 right up until the point the I<MD5> signatures start colliding with each other,
2446 and this is B<EXTREMELY> rare -- like winning the lottery 5 times in a row AND
2447 getting struck by lightning while you are walking to cash in your tickets.
2448 Theoretically, since I<MD5> hashes are 128-bit values, you I<could> have up to
2449 340,282,366,921,000,000,000,000,000,000,000,000,000 keys/elements (I believe
2450 this is 340 unodecillion, but don't quote me).
2454 When a new key/element is stored, the key (or index number) is first run through
2455 I<Digest::MD5> to get a 128-bit signature (example, in hex:
2456 b05783b0773d894396d475ced9d2f4f6). Then, the I<Master Index> record is checked
2457 for the first char of the signature (in this case I<b0>). If it does not exist,
2458 a new I<Bucket List> is created for our key (and the next 15 future keys that
2459 happen to also have I<b> as their first MD5 char). The entire MD5 is written
2460 to the I<Bucket List> along with the offset of the new I<Bucket> record (EOF at
2461 this point, unless we are replacing an existing I<Bucket>), where the actual
2462 data will be stored.
2466 Fetching an existing key/element involves getting a I<Digest::MD5> of the key
2467 (or index number), then walking along the indexes. If there are enough
2468 keys/elements in this DB level, there might be nested indexes, each linked to
2469 a particular char of the MD5. Finally, a I<Bucket List> is pointed to, which
2470 contains up to 16 full MD5 hashes. Each is checked for equality to the key in
2471 question. If we found a match, the I<Bucket> tag is loaded, where the value and
2472 plain key are stored.
2474 Fetching the plain key occurs when calling the I<first_key()> and I<next_key()>
2475 methods. In this process the indexes are walked systematically, and each key
2476 fetched in increasing MD5 order (which is why it appears random). Once the
2477 I<Bucket> is found, the value is skipped and the plain key returned instead.
2478 B<Note:> Do not count on keys being fetched as if the MD5 hashes were
2479 alphabetically sorted. This only happens on an index-level -- as soon as the
2480 I<Bucket Lists> are hit, the keys will come out in the order they went in --
2481 so it's pretty much undefined how the keys will come out -- just like Perl's
2484 =head1 CODE COVERAGE
2486 We use B<Devel::Cover> to test the code coverage of our tests, below is the
2487 B<Devel::Cover> report on this module's test suite.
2489 ---------------------------- ------ ------ ------ ------ ------ ------ ------
2490 File stmt bran cond sub pod time total
2491 ---------------------------- ------ ------ ------ ------ ------ ------ ------
2492 blib/lib/DBM/Deep.pm 95.2 83.8 70.0 98.2 100.0 58.0 91.0
2493 blib/lib/DBM/Deep/Array.pm 100.0 91.1 100.0 100.0 n/a 26.7 98.0
2494 blib/lib/DBM/Deep/Hash.pm 95.3 80.0 100.0 100.0 n/a 15.3 92.4
2495 Total 96.2 84.8 74.4 98.8 100.0 100.0 92.4
2496 ---------------------------- ------ ------ ------ ------ ------ ------ ------
2498 =head1 MORE INFORMATION
2500 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
2501 or send email to L<DBM-Deep@googlegroups.com>.
2505 Joseph Huckaby, L<jhuckaby@cpan.org>
2507 Rob Kinyon, L<rkinyon@cpan.org>
2509 Special thanks to Adam Sah and Rich Gaushell! You know why :-)
2513 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
2514 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
2518 Copyright (c) 2002-2006 Joseph Huckaby. All Rights Reserved.
2519 This is free software, you may use it and distribute it under the
2520 same terms as Perl itself.