3 DBM::Deep - A pure perl multi-level hash/array DBM that supports transactions
8 my $db = DBM::Deep->new( "foo.db" );
13 $db->put('key' => 'value');
14 print $db->get('key');
16 # true multi-level support
18 'hello', { perl => 'rules' },
29 tie my %db, 'DBM::Deep', 'foo.db';
33 tied(%db)->put('key' => 'value');
34 print tied(%db)->get('key');
38 A unique flat-file database module, written in pure perl. True multi-level
39 hash/array support (unlike MLDBM, which is faked), hybrid OO / tie()
40 interface, cross-platform FTPable files, ACID transactions, and is quite fast.
41 Can handle millions of keys and unlimited levels without significant
42 slow-down. Written from the ground-up in pure perl -- this is NOT a wrapper
43 around a C-based DBM. Out-of-the-box compatibility with Unix, Mac OS X and
46 =head1 VERSION DIFFERENCES
48 B<NOTE>: 0.99_03 has significant file format differences from prior versions.
49 THere will be a backwards-compatibility layer in 1.00, but that is slated for
50 a later 0.99_x release. This version is B<NOT> backwards compatible with any
51 other release of DBM::Deep.
53 B<NOTE>: 0.99_01 and above have significant file format differences from 0.983 and
54 before. There will be a backwards-compatibility layer in 1.00, but that is
55 slated for a later 0.99_x release. This version is B<NOT> backwards compatible
56 with 0.983 and before.
60 Construction can be done OO-style (which is the recommended way), or using
61 Perl's tie() function. Both are examined here.
63 =head2 OO Construction
65 The recommended way to construct a DBM::Deep object is to use the new()
66 method, which gets you a blessed I<and> tied hash (or array) reference.
68 my $db = DBM::Deep->new( "foo.db" );
70 This opens a new database handle, mapped to the file "foo.db". If this
71 file does not exist, it will automatically be created. DB files are
72 opened in "r+" (read/write) mode, and the type of object returned is a
73 hash, unless otherwise specified (see L<OPTIONS> below).
75 You can pass a number of options to the constructor to specify things like
76 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
78 my $db = DBM::Deep->new(
84 Notice that the filename is now specified I<inside> the hash with
85 the "file" parameter, as opposed to being the sole argument to the
86 constructor. This is required if any options are specified.
87 See L<OPTIONS> below for the complete list.
89 You can also start with an array instead of a hash. For this, you must
90 specify the C<type> parameter:
92 my $db = DBM::Deep->new(
94 type => DBM::Deep->TYPE_ARRAY
97 B<Note:> Specifing the C<type> parameter only takes effect when beginning
98 a new DB file. If you create a DBM::Deep object with an existing file, the
99 C<type> will be loaded from the file header, and an error will be thrown if
100 the wrong type is passed in.
102 =head2 Tie Construction
104 Alternately, you can create a DBM::Deep handle by using Perl's built-in
105 tie() function. The object returned from tie() can be used to call methods,
106 such as lock() and unlock(). (That object can be retrieved from the tied
107 variable at any time using tied() - please see L<perltie/> for more info.
110 my $db = tie %hash, "DBM::Deep", "foo.db";
113 my $db = tie @array, "DBM::Deep", "bar.db";
115 As with the OO constructor, you can replace the DB filename parameter with
116 a hash containing one or more options (see L<OPTIONS> just below for the
119 tie %hash, "DBM::Deep", {
127 There are a number of options that can be passed in when constructing your
128 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
134 Filename of the DB file to link the handle to. You can pass a full absolute
135 filesystem path, partial path, or a plain filename if the file is in the
136 current working directory. This is a required parameter (though q.v. fh).
140 If you want, you can pass in the fh instead of the file. This is most useful for doing
143 my $db = DBM::Deep->new( { fh => \*DATA } );
145 You are responsible for making sure that the fh has been opened appropriately for your
146 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
147 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
148 needs to read from the fh.
152 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
153 not need to set this. However, it's there if you want it.
155 If you pass in fh and do not set this, it will be set appropriately.
159 This parameter specifies what type of object to create, a hash or array. Use
160 one of these two constants:
164 =item * C<DBM::Deep-E<gt>TYPE_HASH>
166 =item * C<DBM::Deep-E<gt>TYPE_ARRAY>.
170 This only takes effect when beginning a new file. This is an optional
171 parameter, and defaults to C<DBM::Deep-E<gt>TYPE_HASH>.
175 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
176 function to lock the database in exclusive mode for writes, and shared mode
177 for reads. Pass any true value to enable. This affects the base DB handle
178 I<and any child hashes or arrays> that use the same DB file. This is an
179 optional parameter, and defaults to 1 (enabled). See L<LOCKING> below for
184 Specifies whether autoflush is to be enabled on the underlying filehandle.
185 This obviously slows down write operations, but is required if you may have
186 multiple processes accessing the same DB file (also consider enable I<locking>).
187 Pass any true value to enable. This is an optional parameter, and defaults to 1
192 See L</FILTERS> below.
196 The following parameters may be specified in the constructor the first time the
197 datafile is created. However, they will be stored in the header of the file and
198 cannot be overridden by subsequent openings of the file - the values will be set
199 from the values stored in the datafile's header.
205 This is the maximum number of transactions that can be running at one time. The
206 default is two - the HEAD and one for imports. The minimum is two and the
207 maximum is 255. The more transactions, the larger and quicker the datafile grows.
209 See L</TRANSACTIONS> below.
213 This is the number of entries that can be added before a reindexing. The larger
214 this number is made, the larger a file gets, but the better performance you will
215 have. The default and minimum number this can be is 16. There is no maximum, but
216 more than 32 isn't recommended.
220 This is the size of the file pointer used throughout the file. The valid values
227 This uses 2-byte offsets, allowing for a maximum file size of 65K
229 =item * medium (default)
231 This uses 4-byte offsets, allowing for a maximum file size of 2G.
235 This uses 8-byte offsets, allowing for a maximum file size of 16XB (exabytes).
239 See L</LARGEFILE SUPPORT> for more information.
245 With DBM::Deep you can access your databases using Perl's standard hash/array
246 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
247 treat them as such. DBM::Deep will intercept all reads/writes and direct them
248 to the right place -- the DB file. This has nothing to do with the
249 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
250 using regular hashes and arrays, rather than calling functions like C<get()>
251 and C<put()> (although those work too). It is entirely up to you how to want
252 to access your databases.
256 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
257 or even nested hashes (or arrays) using standard Perl syntax:
259 my $db = DBM::Deep->new( "foo.db" );
261 $db->{mykey} = "myvalue";
263 $db->{myhash}->{subkey} = "subvalue";
265 print $db->{myhash}->{subkey} . "\n";
267 You can even step through hash keys using the normal Perl C<keys()> function:
269 foreach my $key (keys %$db) {
270 print "$key: " . $db->{$key} . "\n";
273 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
274 pushes them onto an array, all before the loop even begins. If you have an
275 extremely large hash, this may exhaust Perl's memory. Instead, consider using
276 Perl's C<each()> function, which pulls keys/values one at a time, using very
279 while (my ($key, $value) = each %$db) {
280 print "$key: $value\n";
283 Please note that when using C<each()>, you should always pass a direct
284 hash reference, not a lookup. Meaning, you should B<never> do this:
287 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
289 This causes an infinite loop, because for each iteration, Perl is calling
290 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
291 it effectively keeps returning the first key over and over again. Instead,
292 assign a temporary variable to C<$db->{foo}>, then pass that to each().
296 As with hashes, you can treat any DBM::Deep object like a normal Perl array
297 reference. This includes inserting, removing and manipulating elements,
298 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
299 The object must have first been created using type C<DBM::Deep-E<gt>TYPE_ARRAY>,
300 or simply be a nested array reference inside a hash. Example:
302 my $db = DBM::Deep->new(
303 file => "foo-array.db",
304 type => DBM::Deep->TYPE_ARRAY
308 push @$db, "bar", "baz";
311 my $last_elem = pop @$db; # baz
312 my $first_elem = shift @$db; # bah
313 my $second_elem = $db->[1]; # bar
315 my $num_elements = scalar @$db;
319 In addition to the I<tie()> interface, you can also use a standard OO interface
320 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
321 array) has its own methods, but both types share the following common methods:
322 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
323 C<store(> are aliases to C<put()> and C<get()>, respectively.
327 =item * new() / clone()
329 These are the constructor and copy-functions.
331 =item * put() / store()
333 Stores a new hash key/value pair, or sets an array element value. Takes two
334 arguments, the hash key or array index, and the new value. The value can be
335 a scalar, hash ref or array ref. Returns true on success, false on failure.
337 $db->put("foo", "bar"); # for hashes
338 $db->put(1, "bar"); # for arrays
340 =item * get() / fetch()
342 Fetches the value of a hash key or array element. Takes one argument: the hash
343 key or array index. Returns a scalar, hash ref or array ref, depending on the
346 my $value = $db->get("foo"); # for hashes
347 my $value = $db->get(1); # for arrays
351 Checks if a hash key or array index exists. Takes one argument: the hash key
352 or array index. Returns true if it exists, false if not.
354 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
355 if ($db->exists(1)) { print "yay!\n"; } # for arrays
359 Deletes one hash key/value pair or array element. Takes one argument: the hash
360 key or array index. Returns true on success, false if not found. For arrays,
361 the remaining elements located after the deleted element are NOT moved over.
362 The deleted element is essentially just undefined, which is exactly how Perl's
363 internal arrays work.
365 $db->delete("foo"); # for hashes
366 $db->delete(1); # for arrays
370 Deletes B<all> hash keys or array elements. Takes no arguments. No return
373 $db->clear(); # hashes or arrays
375 =item * lock() / unlock()
381 Recover lost disk space. This is important to do, especially if you use
384 =item * import() / export()
386 Data going in and out.
388 =item * begin_work() / commit() / rollback()
390 These are the transactional functions. L</TRANSACTIONS> for more information.
396 For hashes, DBM::Deep supports all the common methods described above, and the
397 following additional methods: C<first_key()> and C<next_key()>.
403 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
404 fetched in an undefined order (which appears random). Takes no arguments,
405 returns the key as a scalar value.
407 my $key = $db->first_key();
411 Returns the "next" key in the hash, given the previous one as the sole argument.
412 Returns undef if there are no more keys to be fetched.
414 $key = $db->next_key($key);
418 Here are some examples of using hashes:
420 my $db = DBM::Deep->new( "foo.db" );
422 $db->put("foo", "bar");
423 print "foo: " . $db->get("foo") . "\n";
425 $db->put("baz", {}); # new child hash ref
426 $db->get("baz")->put("buz", "biz");
427 print "buz: " . $db->get("baz")->get("buz") . "\n";
429 my $key = $db->first_key();
431 print "$key: " . $db->get($key) . "\n";
432 $key = $db->next_key($key);
435 if ($db->exists("foo")) { $db->delete("foo"); }
439 For arrays, DBM::Deep supports all the common methods described above, and the
440 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
441 C<unshift()> and C<splice()>.
447 Returns the number of elements in the array. Takes no arguments.
449 my $len = $db->length();
453 Adds one or more elements onto the end of the array. Accepts scalars, hash
454 refs or array refs. No return value.
456 $db->push("foo", "bar", {});
460 Fetches the last element in the array, and deletes it. Takes no arguments.
461 Returns undef if array is empty. Returns the element value.
463 my $elem = $db->pop();
467 Fetches the first element in the array, deletes it, then shifts all the
468 remaining elements over to take up the space. Returns the element value. This
469 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
472 my $elem = $db->shift();
476 Inserts one or more elements onto the beginning of the array, shifting all
477 existing elements over to make room. Accepts scalars, hash refs or array refs.
478 No return value. This method is not recommended with large arrays -- see
479 <LARGE ARRAYS> below for details.
481 $db->unshift("foo", "bar", {});
485 Performs exactly like Perl's built-in function of the same name. See L<perldoc
486 -f splice> for usage -- it is too complicated to document here. This method is
487 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
491 Here are some examples of using arrays:
493 my $db = DBM::Deep->new(
495 type => DBM::Deep->TYPE_ARRAY
498 $db->push("bar", "baz");
502 my $len = $db->length();
503 print "length: $len\n"; # 4
505 for (my $k=0; $k<$len; $k++) {
506 print "$k: " . $db->get($k) . "\n";
509 $db->splice(1, 2, "biz", "baf");
511 while (my $elem = shift @$db) {
512 print "shifted: $elem\n";
517 Enable or disable automatic file locking by passing a boolean value to the
518 C<locking> parameter when constructing your DBM::Deep object (see L<SETUP>
521 my $db = DBM::Deep->new(
526 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
527 mode for writes, and shared mode for reads. This is required if you have
528 multiple processes accessing the same database file, to avoid file corruption.
529 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
532 =head2 Explicit Locking
534 You can explicitly lock a database, so it remains locked for multiple
535 actions. This is done by calling the C<lock()> method, and passing an
536 optional lock mode argument (defaults to exclusive mode). This is particularly
537 useful for things like counters, where the current value needs to be fetched,
538 then incremented, then stored again.
541 my $counter = $db->get("counter");
543 $db->put("counter", $counter);
552 You can pass C<lock()> an optional argument, which specifies which mode to use
553 (exclusive or shared). Use one of these two constants:
554 C<DBM::Deep-E<gt>LOCK_EX> or C<DBM::Deep-E<gt>LOCK_SH>. These are passed
555 directly to C<flock()>, and are the same as the constants defined in Perl's
558 $db->lock( $db->LOCK_SH );
562 =head1 IMPORTING/EXPORTING
564 You can import existing complex structures by calling the C<import()> method,
565 and export an entire database into an in-memory structure using the C<export()>
566 method. Both are examined here.
570 Say you have an existing hash with nested hashes/arrays inside it. Instead of
571 walking the structure and adding keys/elements to the database as you go,
572 simply pass a reference to the C<import()> method. This recursively adds
573 everything to an existing DBM::Deep object for you. Here is an example:
578 array1 => [ "elem0", "elem1", "elem2" ],
580 subkey1 => "subvalue1",
581 subkey2 => "subvalue2"
585 my $db = DBM::Deep->new( "foo.db" );
586 $db->import( $struct );
588 print $db->{key1} . "\n"; # prints "value1"
590 This recursively imports the entire C<$struct> object into C<$db>, including
591 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
592 keys are merged with the existing ones, replacing if they already exist.
593 The C<import()> method can be called on any database level (not just the base
594 level), and works with both hash and array DB types.
596 B<Note:> Make sure your existing structure has no circular references in it.
597 These will cause an infinite loop when importing. There are plans to fix this
600 B<Note:> With the addition of transactions, importing is performed within a
601 transaction, then immediately committed upon success (and rolled back upon
602 failre). As a result, you cannot call C<import()> from within a transaction.
603 This restriction will be lifted when subtransactions are added in a future
608 Calling the C<export()> method on an existing DBM::Deep object will return
609 a reference to a new in-memory copy of the database. The export is done
610 recursively, so all nested hashes/arrays are all exported to standard Perl
611 objects. Here is an example:
613 my $db = DBM::Deep->new( "foo.db" );
615 $db->{key1} = "value1";
616 $db->{key2} = "value2";
618 $db->{hash1}->{subkey1} = "subvalue1";
619 $db->{hash1}->{subkey2} = "subvalue2";
621 my $struct = $db->export();
623 print $struct->{key1} . "\n"; # prints "value1"
625 This makes a complete copy of the database in memory, and returns a reference
626 to it. The C<export()> method can be called on any database level (not just
627 the base level), and works with both hash and array DB types. Be careful of
628 large databases -- you can store a lot more data in a DBM::Deep object than an
629 in-memory Perl structure.
631 B<Note:> Make sure your database has no circular references in it.
632 These will cause an infinite loop when exporting. There are plans to fix this
637 DBM::Deep has a number of hooks where you can specify your own Perl function
638 to perform filtering on incoming or outgoing data. This is a perfect
639 way to extend the engine, and implement things like real-time compression or
640 encryption. Filtering applies to the base DB level, and all child hashes /
641 arrays. Filter hooks can be specified when your DBM::Deep object is first
642 constructed, or by calling the C<set_filter()> method at any time. There are
643 four available filter hooks, described below:
647 =item * filter_store_key
649 This filter is called whenever a hash key is stored. It
650 is passed the incoming key, and expected to return a transformed key.
652 =item * filter_store_value
654 This filter is called whenever a hash key or array element is stored. It
655 is passed the incoming value, and expected to return a transformed value.
657 =item * filter_fetch_key
659 This filter is called whenever a hash key is fetched (i.e. via
660 C<first_key()> or C<next_key()>). It is passed the transformed key,
661 and expected to return the plain key.
663 =item * filter_fetch_value
665 This filter is called whenever a hash key or array element is fetched.
666 It is passed the transformed value, and expected to return the plain value.
670 Here are the two ways to setup a filter hook:
672 my $db = DBM::Deep->new(
674 filter_store_value => \&my_filter_store,
675 filter_fetch_value => \&my_filter_fetch
680 $db->set_filter( "filter_store_value", \&my_filter_store );
681 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
683 Your filter function will be called only when dealing with SCALAR keys or
684 values. When nested hashes and arrays are being stored/fetched, filtering
685 is bypassed. Filters are called as static functions, passed a single SCALAR
686 argument, and expected to return a single SCALAR value. If you want to
687 remove a filter, set the function reference to C<undef>:
689 $db->set_filter( "filter_store_value", undef );
691 =head2 Real-time Encryption Example
693 Here is a working example that uses the I<Crypt::Blowfish> module to
694 do real-time encryption / decryption of keys & values with DBM::Deep Filters.
695 Please visit L<http://search.cpan.org/search?module=Crypt::Blowfish> for more
696 on I<Crypt::Blowfish>. You'll also need the I<Crypt::CBC> module.
702 my $cipher = Crypt::CBC->new({
703 'key' => 'my secret key',
704 'cipher' => 'Blowfish',
706 'regenerate_key' => 0,
707 'padding' => 'space',
711 my $db = DBM::Deep->new(
712 file => "foo-encrypt.db",
713 filter_store_key => \&my_encrypt,
714 filter_store_value => \&my_encrypt,
715 filter_fetch_key => \&my_decrypt,
716 filter_fetch_value => \&my_decrypt,
719 $db->{key1} = "value1";
720 $db->{key2} = "value2";
721 print "key1: " . $db->{key1} . "\n";
722 print "key2: " . $db->{key2} . "\n";
728 return $cipher->encrypt( $_[0] );
731 return $cipher->decrypt( $_[0] );
734 =head2 Real-time Compression Example
736 Here is a working example that uses the I<Compress::Zlib> module to do real-time
737 compression / decompression of keys & values with DBM::Deep Filters.
738 Please visit L<http://search.cpan.org/search?module=Compress::Zlib> for
739 more on I<Compress::Zlib>.
744 my $db = DBM::Deep->new(
745 file => "foo-compress.db",
746 filter_store_key => \&my_compress,
747 filter_store_value => \&my_compress,
748 filter_fetch_key => \&my_decompress,
749 filter_fetch_value => \&my_decompress,
752 $db->{key1} = "value1";
753 $db->{key2} = "value2";
754 print "key1: " . $db->{key1} . "\n";
755 print "key2: " . $db->{key2} . "\n";
761 return Compress::Zlib::memGzip( $_[0] ) ;
764 return Compress::Zlib::memGunzip( $_[0] ) ;
767 B<Note:> Filtering of keys only applies to hashes. Array "keys" are
768 actually numerical index numbers, and are not filtered.
770 =head1 ERROR HANDLING
772 Most DBM::Deep methods return a true value for success, and call die() on
773 failure. You can wrap calls in an eval block to catch the die.
775 my $db = DBM::Deep->new( "foo.db" ); # create hash
776 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
778 print $@; # prints error message
780 =head1 LARGEFILE SUPPORT
782 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
783 and 64-bit support, you I<may> be able to create databases larger than 2 GB.
784 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
785 by specifying the 'pack_size' parameter when constructing the file.
788 filename => $filename,
789 pack_size => 'large',
792 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
793 instead of 32-bit longs. After setting these values your DB files have a
794 theoretical maximum size of 16 XB (exabytes).
796 You can also use C<pack_size =E<gt> 'small'> in order to use 16-bit file
799 B<Note:> Changing these values will B<NOT> work for existing database files.
800 Only change this for new files. Once the value has been set, it is stored in
801 the file's header and cannot be changed for the life of the file. These
802 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
805 B<Note:> We have not personally tested files larger than 2 GB -- all my
806 systems have only a 32-bit Perl. However, I have received user reports that
807 this does indeed work!
809 =head1 LOW-LEVEL ACCESS
811 If you require low-level access to the underlying filehandle that DBM::Deep uses,
812 you can call the C<_fh()> method, which returns the handle:
816 This method can be called on the root level of the datbase, or any child
817 hashes or arrays. All levels share a I<root> structure, which contains things
818 like the filehandle, a reference counter, and all the options specified
819 when you created the object. You can get access to this file object by
820 calling the C<_storage()> method.
822 my $file_obj = $db->_storage();
824 This is useful for changing options after the object has already been created,
825 such as enabling/disabling locking. You can also store your own temporary user
826 data in this structure (be wary of name collision), which is then accessible from
827 any child hash or array.
829 =head1 CUSTOM DIGEST ALGORITHM
831 DBM::Deep by default uses the I<Message Digest 5> (MD5) algorithm for hashing
832 keys. However you can override this, and use another algorithm (such as SHA-256)
833 or even write your own. But please note that DBM::Deep currently expects zero
834 collisions, so your algorithm has to be I<perfect>, so to speak. Collision
835 detection may be introduced in a later version.
837 You can specify a custom digest algorithm by passing it into the parameter
838 list for new(), passing a reference to a subroutine as the 'digest' parameter,
839 and the length of the algorithm's hashes (in bytes) as the 'hash_size'
840 parameter. Here is a working example that uses a 256-bit hash from the
841 I<Digest::SHA256> module. Please see
842 L<http://search.cpan.org/search?module=Digest::SHA256> for more information.
847 my $context = Digest::SHA256::new(256);
849 my $db = DBM::Deep->new(
850 filename => "foo-sha.db",
851 digest => \&my_digest,
855 $db->{key1} = "value1";
856 $db->{key2} = "value2";
857 print "key1: " . $db->{key1} . "\n";
858 print "key2: " . $db->{key2} . "\n";
864 return substr( $context->hash($_[0]), 0, 32 );
867 B<Note:> Your returned digest strings must be B<EXACTLY> the number
868 of bytes you specify in the hash_size parameter (in this case 32).
870 B<Note:> If you do choose to use a custom digest algorithm, you must set it
871 every time you access this file. Otherwise, the default (MD5) will be used.
873 =head1 CIRCULAR REFERENCES
875 B<NOTE>: DBM::Deep 0.99_03 has turned off circular references pending
876 evaluation of some edge cases. I hope to be able to re-enable circular
877 references in a future version after 1.00. This means that circular references
878 are B<NO LONGER> available.
880 DBM::Deep has B<experimental> support for circular references. Meaning you
881 can have a nested hash key or array element that points to a parent object.
882 This relationship is stored in the DB file, and is preserved between sessions.
885 my $db = DBM::Deep->new( "foo.db" );
888 $db->{circle} = $db; # ref to self
890 print $db->{foo} . "\n"; # prints "bar"
891 print $db->{circle}->{foo} . "\n"; # prints "bar" again
893 B<Note>: Passing the object to a function that recursively walks the
894 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
895 C<export()> methods) will result in an infinite loop. This will be fixed in
900 New in 0.99_01 is ACID transactions. Every DBM::Deep object is completely
901 transaction-ready - it is not an option you have to turn on. You do have to
902 specify how many transactions may run simultaneously (q.v. L</num_txns>).
904 Three new methods have been added to support them. They are:
910 This starts a transaction.
914 This applies the changes done within the transaction to the mainline and ends
919 This discards the changes done within the transaction to the mainline and ends
924 Transactions in DBM::Deep are done using a variant of the MVCC method, the
925 same method used by the InnoDB MySQL engine.
927 =head2 Software-Transactional Memory
929 The addition of transactions to this module provides the basis for STM within
930 Perl 5. Contention is resolved using a default last-write-wins. Currently,
931 this default cannot be changed, but it will be addressed in a future version.
935 Because DBM::Deep is a conncurrent datastore, every change is flushed to disk
936 immediately and every read goes to disk. This means that DBM::Deep functions
937 at the speed of disk (generally 10-20ms) vs. the speed of RAM (generally
938 50-70ns), or at least 150-200x slower than the comparable in-memory
939 datastructure in Perl.
941 There are several techniques you can use to speed up how DBM::Deep functions.
945 =item * Put it on a ramdisk
947 The easiest and quickest mechanism to making DBM::Deep run faster is to create
948 a ramdisk and locate the DBM::Deep file there. Doing this as an option may
949 become a feature of DBM::Deep, assuming there is a good ramdisk wrapper on CPAN.
951 =item * Work at the tightest level possible
953 It is much faster to assign the level of your db that you are working with to
954 an intermediate variable than to re-look it up every time. Thus
957 while ( my ($k, $v) = each %{$db->{foo}{bar}{baz}} ) {
962 my $x = $db->{foo}{bar}{baz};
963 while ( my ($k, $v) = each %$x ) {
967 =item * Make your file as tight as possible
969 If you know that you are not going to use more than 65K in your database,
970 consider using the C<pack_size =E<gt> 'small'> option. This will instruct
971 DBM::Deep to use 16bit addresses, meaning that the seek times will be less.
977 The following are items that are planned to be added in future releases. These
978 are separate from the L<CAVEATS, ISSUES & BUGS> below.
980 =head2 Sub-Transactions
982 Right now, you cannot run a transaction within a transaction. Removing this
983 restriction is technically straightforward, but the combinatorial explosion of
984 possible usecases hurts my head. If this is something you want to see
985 immediately, please submit many testcases.
989 If a user is willing to assert upon opening the file that this process will be
990 the only consumer of that datafile, then there are a number of caching
991 possibilities that can be taken advantage of. This does, however, mean that
992 DBM::Deep is more vulnerable to losing data due to unflushed changes. It also
993 means a much larger in-memory footprint. As such, it's not clear exactly how
994 this should be done. Suggestions are welcome.
998 The techniques used in DBM::Deep simply require a seekable contiguous
999 datastore. This could just as easily be a large string as a file. By using
1000 substr, the STM capabilities of DBM::Deep could be used within a
1001 single-process. I have no idea how I'd specify this, though. Suggestions are
1004 =head2 Importing using Data::Walker
1006 Right now, importing is done using C<Clone::clone()> to make a complete copy
1007 in memory, then tying that copy. It would be much better to use
1008 L<Data::Walker/> to walk the data structure instead, particularly in the case
1009 of large datastructures.
1011 =head2 Different contention resolution mechanisms
1013 Currently, the only contention resolution mechanism is last-write-wins. This
1014 is the mechanism used by most RDBMSes and should be good enough for most uses.
1015 For advanced uses of STM, other contention mechanisms will be needed. If you
1016 have an idea of how you'd like to see contention resolution in DBM::Deep,
1019 =head1 CAVEATS, ISSUES & BUGS
1021 This section describes all the known issues with DBM::Deep. These are issues
1022 that are either intractable or depend on some feature within Perl working
1023 exactly right. It you have found something that is not listed below, please
1024 send an e-mail to L<rkinyon@cpan.org>. Likewise, if you think you know of a
1025 way around one of these issues, please let me know.
1029 (The following assumes a high level of Perl understanding, specifically of
1030 references. Most users can safely skip this section.)
1032 Currently, the only references supported are HASH and ARRAY. The other reference
1033 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
1039 These are things like filehandles and other sockets. They can't be supported
1040 because it's completely unclear how DBM::Deep should serialize them.
1042 =item * SCALAR / REF
1044 The discussion here refers to the following type of example:
1051 # In some other process ...
1053 my $val = ${ $db->{key1} };
1055 is( $val, 50, "What actually gets stored in the DB file?" );
1057 The problem is one of synchronization. When the variable being referred to
1058 changes value, the reference isn't notified, which is kind of the point of
1059 references. This means that the new value won't be stored in the datafile for
1060 other processes to read. There is no TIEREF.
1062 It is theoretically possible to store references to values already within a
1063 DBM::Deep object because everything already is synchronized, but the change to
1064 the internals would be quite large. Specifically, DBM::Deep would have to tie
1065 every single value that is stored. This would bloat the RAM footprint of
1066 DBM::Deep at least twofold (if not more) and be a significant performance drain,
1067 all to support a feature that has never been requested.
1071 L<Data::Dump::Streamer/> provides a mechanism for serializing coderefs,
1072 including saving off all closure state. This would allow for DBM::Deep to
1073 store the code for a subroutine. Then, whenever the subroutine is read, the
1074 code could be C<eval()>'ed into being. However, just as for SCALAR and REF,
1075 that closure state may change without notifying the DBM::Deep object storing
1076 the reference. Again, this would generally be considered a feature.
1080 =head2 File corruption
1082 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
1083 for a 32-bit signature when opened, but any other form of corruption in the
1084 datafile can cause segmentation faults. DBM::Deep may try to C<seek()> past
1085 the end of a file, or get stuck in an infinite loop depending on the level and
1086 type of corruption. File write operations are not checked for failure (for
1087 speed), so if you happen to run out of disk space, DBM::Deep will probably fail in
1088 a bad way. These things will be addressed in a later version of DBM::Deep.
1092 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
1093 well on local filesystems, but will NOT protect you from file corruption over
1094 NFS. I've heard about setting up your NFS server with a locking daemon, then
1095 using C<lockf()> to lock your files, but your mileage may vary there as well.
1096 From what I understand, there is no real way to do it. However, if you need
1097 access to the underlying filehandle in DBM::Deep for using some other kind of
1098 locking scheme like C<lockf()>, see the L<LOW-LEVEL ACCESS> section above.
1100 =head2 Copying Objects
1102 Beware of copying tied objects in Perl. Very strange things can happen.
1103 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
1104 returns a new, blessed and tied hash or array to the same level in the DB.
1106 my $copy = $db->clone();
1108 B<Note>: Since clone() here is cloning the object, not the database location, any
1109 modifications to either $db or $copy will be visible to both.
1113 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1114 These functions cause every element in the array to move, which can be murder
1115 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1116 a different location. This will be addressed in a future version.
1118 =head2 Writeonly Files
1120 If you pass in a filehandle to new(), you may have opened it in either a readonly or
1121 writeonly mode. STORE will verify that the filehandle is writable. However, there
1122 doesn't seem to be a good way to determine if a filehandle is readable. And, if the
1123 filehandle isn't readable, it's not clear what will happen. So, don't do that.
1125 =head2 Assignments Within Transactions
1127 The following will I<not> work as one might expect:
1135 is( $x->{a}, 1 ); # This will fail!
1137 The problem is that the moment a reference used as the rvalue to a DBM::Deep
1138 object's lvalue, it becomes tied itself. This is so that future changes to
1139 C<$x> can be tracked within the DBM::Deep file and is considered to be a
1140 feature. By the time the rollback occurs, there is no knowledge that there had
1141 been an C<$x> or what memory location to assign an C<export()> to.
1143 B<NOTE:> This does not affect importing because imports do a walk over the
1144 reference to be imported in order to explicitly leave it untied.
1146 =head1 CODE COVERAGE
1148 B<Devel::Cover> is used to test the code coverage of the tests. Below is the
1149 B<Devel::Cover> report on this distribution's test suite.
1151 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1152 File stmt bran cond sub pod time total
1153 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1154 blib/lib/DBM/Deep.pm 96.8 87.9 90.5 100.0 89.5 4.5 95.2
1155 blib/lib/DBM/Deep/Array.pm 100.0 94.3 100.0 100.0 100.0 4.9 98.7
1156 blib/lib/DBM/Deep/Engine.pm 96.9 85.2 79.7 100.0 0.0 58.2 90.3
1157 blib/lib/DBM/Deep/File.pm 99.0 88.9 77.8 100.0 0.0 30.0 90.3
1158 blib/lib/DBM/Deep/Hash.pm 100.0 100.0 100.0 100.0 100.0 2.4 100.0
1159 Total 97.6 87.9 84.0 100.0 32.1 100.0 92.8
1160 ---------------------------- ------ ------ ------ ------ ------ ------ ------
1162 =head1 MORE INFORMATION
1164 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1165 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1168 The source code repository is at L<http://svn.perl.org/modules/DBM-Deep>
1170 =head1 MAINTAINER(S)
1172 Rob Kinyon, L<rkinyon@cpan.org>
1174 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1178 The following have contributed greatly to make DBM::Deep what it is today:
1182 =item * Adam Sah and Rich Gaushell
1184 =item * Stonehenge for sponsoring the 1.00 release
1186 =item * Dan Golden and others at YAPC::NA 2006 for helping me design through transactions.
1192 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1193 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1197 Copyright (c) 2007 Rob Kinyon. All Rights Reserved.
1198 This is free software, you may use it and distribute it under the
1199 same terms as Perl itself.