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>: 1.0000 has significant file format differences from prior versions.
49 THere is a backwards-compatibility layer at C<utils/upgrade_db.pl>. Files
50 created by 1.0000 or higher are B<NOT> compatible with scripts using prior
55 Construction can be done OO-style (which is the recommended way), or using
56 Perl's tie() function. Both are examined here.
58 =head2 OO Construction
60 The recommended way to construct a DBM::Deep object is to use the new()
61 method, which gets you a blessed I<and> tied hash (or array) reference.
63 my $db = DBM::Deep->new( "foo.db" );
65 This opens a new database handle, mapped to the file "foo.db". If this
66 file does not exist, it will automatically be created. DB files are
67 opened in "r+" (read/write) mode, and the type of object returned is a
68 hash, unless otherwise specified (see L<OPTIONS> below).
70 You can pass a number of options to the constructor to specify things like
71 locking, autoflush, etc. This is done by passing an inline hash (or hashref):
73 my $db = DBM::Deep->new(
79 Notice that the filename is now specified I<inside> the hash with
80 the "file" parameter, as opposed to being the sole argument to the
81 constructor. This is required if any options are specified.
82 See L<OPTIONS> below for the complete list.
84 You can also start with an array instead of a hash. For this, you must
85 specify the C<type> parameter:
87 my $db = DBM::Deep->new(
89 type => DBM::Deep->TYPE_ARRAY
92 B<Note:> Specifing the C<type> parameter only takes effect when beginning
93 a new DB file. If you create a DBM::Deep object with an existing file, the
94 C<type> will be loaded from the file header, and an error will be thrown if
95 the wrong type is passed in.
97 =head2 Tie Construction
99 Alternately, you can create a DBM::Deep handle by using Perl's built-in
100 tie() function. The object returned from tie() can be used to call methods,
101 such as lock() and unlock(). (That object can be retrieved from the tied
102 variable at any time using tied() - please see L<perltie> for more info.
105 my $db = tie %hash, "DBM::Deep", "foo.db";
108 my $db = tie @array, "DBM::Deep", "bar.db";
110 As with the OO constructor, you can replace the DB filename parameter with
111 a hash containing one or more options (see L<OPTIONS> just below for the
114 tie %hash, "DBM::Deep", {
122 There are a number of options that can be passed in when constructing your
123 DBM::Deep objects. These apply to both the OO- and tie- based approaches.
129 Filename of the DB file to link the handle to. You can pass a full absolute
130 filesystem path, partial path, or a plain filename if the file is in the
131 current working directory. This is a required parameter (though q.v. fh).
135 If you want, you can pass in the fh instead of the file. This is most useful for doing
138 my $db = DBM::Deep->new( { fh => \*DATA } );
140 You are responsible for making sure that the fh has been opened appropriately for your
141 needs. If you open it read-only and attempt to write, an exception will be thrown. If you
142 open it write-only or append-only, an exception will be thrown immediately as DBM::Deep
143 needs to read from the fh.
147 This is the offset within the file that the DBM::Deep db starts. Most of the time, you will
148 not need to set this. However, it's there if you want it.
150 If you pass in fh and do not set this, it will be set appropriately.
154 This parameter specifies what type of object to create, a hash or array. Use
155 one of these two constants:
159 =item * C<<DBM::Deep->TYPE_HASH>>
161 =item * C<<DBM::Deep->TYPE_ARRAY>>
165 This only takes effect when beginning a new file. This is an optional
166 parameter, and defaults to C<<DBM::Deep->TYPE_HASH>>.
170 Specifies whether locking is to be enabled. DBM::Deep uses Perl's flock()
171 function to lock the database in exclusive mode for writes, and shared mode
172 for reads. Pass any true value to enable. This affects the base DB handle
173 I<and any child hashes or arrays> that use the same DB file. This is an
174 optional parameter, and defaults to 1 (enabled). See L<LOCKING> below for
179 Specifies whether autoflush is to be enabled on the underlying filehandle.
180 This obviously slows down write operations, but is required if you may have
181 multiple processes accessing the same DB file (also consider enable I<locking>).
182 Pass any true value to enable. This is an optional parameter, and defaults to 1
187 See L</FILTERS> below.
191 The following parameters may be specified in the constructor the first time the
192 datafile is created. However, they will be stored in the header of the file and
193 cannot be overridden by subsequent openings of the file - the values will be set
194 from the values stored in the datafile's header.
200 This is the number of transactions that can be running at one time. The
201 default is one - the HEAD. The minimum is one and the maximum is 255. The more
202 transactions, the larger and quicker the datafile grows.
204 See L</TRANSACTIONS> below.
208 This is the number of entries that can be added before a reindexing. The larger
209 this number is made, the larger a file gets, but the better performance you will
210 have. The default and minimum number this can be is 16. The maximum is 256, but
211 more than 64 isn't recommended.
213 =item * data_sector_size
215 This is the size in bytes of a given data sector. Data sectors will chain, so
216 a value of any size can be stored. However, chaining is expensive in terms of
217 time. Setting this value to something close to the expected common length of
218 your scalars will improve your performance. If it is too small, your file will
219 have a lot of chaining. If it is too large, your file will have a lot of dead
222 The default for this is 64 bytes. The minimum value is 32 and the maximum is
225 B<Note:> There are between 6 and 10 bytes taken up in each data sector for
226 bookkeeping. (It's 4 + the number of bytes in your L</pack_size>.) This is
227 included within the data_sector_size, thus the effective value is 6-10 bytes
228 less than what you specified.
232 This is the size of the file pointer used throughout the file. The valid values
239 This uses 2-byte offsets, allowing for a maximum file size of 65 KB.
241 =item * medium (default)
243 This uses 4-byte offsets, allowing for a maximum file size of 4 GB.
247 This uses 8-byte offsets, allowing for a maximum file size of 16 XB
248 (exabytes). This can only be enabled if your Perl is compiled for 64-bit.
252 See L</LARGEFILE SUPPORT> for more information.
258 With DBM::Deep you can access your databases using Perl's standard hash/array
259 syntax. Because all DBM::Deep objects are I<tied> to hashes or arrays, you can
260 treat them as such. DBM::Deep will intercept all reads/writes and direct them
261 to the right place -- the DB file. This has nothing to do with the
262 L<TIE CONSTRUCTION> section above. This simply tells you how to use DBM::Deep
263 using regular hashes and arrays, rather than calling functions like C<get()>
264 and C<put()> (although those work too). It is entirely up to you how to want
265 to access your databases.
269 You can treat any DBM::Deep object like a normal Perl hash reference. Add keys,
270 or even nested hashes (or arrays) using standard Perl syntax:
272 my $db = DBM::Deep->new( "foo.db" );
274 $db->{mykey} = "myvalue";
276 $db->{myhash}->{subkey} = "subvalue";
278 print $db->{myhash}->{subkey} . "\n";
280 You can even step through hash keys using the normal Perl C<keys()> function:
282 foreach my $key (keys %$db) {
283 print "$key: " . $db->{$key} . "\n";
286 Remember that Perl's C<keys()> function extracts I<every> key from the hash and
287 pushes them onto an array, all before the loop even begins. If you have an
288 extremely large hash, this may exhaust Perl's memory. Instead, consider using
289 Perl's C<each()> function, which pulls keys/values one at a time, using very
292 while (my ($key, $value) = each %$db) {
293 print "$key: $value\n";
296 Please note that when using C<each()>, you should always pass a direct
297 hash reference, not a lookup. Meaning, you should B<never> do this:
300 while (my ($key, $value) = each %{$db->{foo}}) { # BAD
302 This causes an infinite loop, because for each iteration, Perl is calling
303 FETCH() on the $db handle, resulting in a "new" hash for foo every time, so
304 it effectively keeps returning the first key over and over again. Instead,
305 assign a temporary variable to C<$db->{foo}>, then pass that to each().
309 As with hashes, you can treat any DBM::Deep object like a normal Perl array
310 reference. This includes inserting, removing and manipulating elements,
311 and the C<push()>, C<pop()>, C<shift()>, C<unshift()> and C<splice()> functions.
312 The object must have first been created using type C<<DBM::Deep->TYPE_ARRAY>>,
313 or simply be a nested array reference inside a hash. Example:
315 my $db = DBM::Deep->new(
316 file => "foo-array.db",
317 type => DBM::Deep->TYPE_ARRAY
321 push @$db, "bar", "baz";
324 my $last_elem = pop @$db; # baz
325 my $first_elem = shift @$db; # bah
326 my $second_elem = $db->[1]; # bar
328 my $num_elements = scalar @$db;
332 In addition to the I<tie()> interface, you can also use a standard OO interface
333 to manipulate all aspects of DBM::Deep databases. Each type of object (hash or
334 array) has its own methods, but both types share the following common methods:
335 C<put()>, C<get()>, C<exists()>, C<delete()> and C<clear()>. C<fetch()> and
336 C<store(> are aliases to C<put()> and C<get()>, respectively.
340 =item * new() / clone()
342 These are the constructor and copy-functions.
344 =item * put() / store()
346 Stores a new hash key/value pair, or sets an array element value. Takes two
347 arguments, the hash key or array index, and the new value. The value can be
348 a scalar, hash ref or array ref. Returns true on success, false on failure.
350 $db->put("foo", "bar"); # for hashes
351 $db->put(1, "bar"); # for arrays
353 =item * get() / fetch()
355 Fetches the value of a hash key or array element. Takes one argument: the hash
356 key or array index. Returns a scalar, hash ref or array ref, depending on the
359 my $value = $db->get("foo"); # for hashes
360 my $value = $db->get(1); # for arrays
364 Checks if a hash key or array index exists. Takes one argument: the hash key
365 or array index. Returns true if it exists, false if not.
367 if ($db->exists("foo")) { print "yay!\n"; } # for hashes
368 if ($db->exists(1)) { print "yay!\n"; } # for arrays
372 Deletes one hash key/value pair or array element. Takes one argument: the hash
373 key or array index. Returns true on success, false if not found. For arrays,
374 the remaining elements located after the deleted element are NOT moved over.
375 The deleted element is essentially just undefined, which is exactly how Perl's
376 internal arrays work.
378 $db->delete("foo"); # for hashes
379 $db->delete(1); # for arrays
383 Deletes B<all> hash keys or array elements. Takes no arguments. No return
386 $db->clear(); # hashes or arrays
388 =item * lock() / unlock() / lock_exclusive() / lock_shared()
390 q.v. L</LOCKING> for more info.
394 This will compress the datafile so that it takes up as little space as possible.
395 There is a freespace manager so that when space is freed up, it is used before
396 extending the size of the datafile. But, that freespace just sits in the datafile
397 unless C<optimize()> is called.
401 Unlike simple assignment, C<import()> does not tie the right-hand side. Instead,
402 a copy of your data is put into the DB. C<import()> takes either an arrayref (if
403 your DB is an array) or a hashref (if your DB is a hash). C<import()> will die
404 if anything else is passed in.
408 This returns a complete copy of the data structure at the point you do the export.
409 This copy is in RAM, not on disk like the DB is.
411 =item * begin_work() / commit() / rollback()
413 These are the transactional functions. L</TRANSACTIONS> for more information.
415 =item * supports( $option )
417 This returns a boolean depending on if this instance of DBM::Dep supports
418 that feature. C<$option> can be one of:
430 For hashes, DBM::Deep supports all the common methods described above, and the
431 following additional methods: C<first_key()> and C<next_key()>.
437 Returns the "first" key in the hash. As with built-in Perl hashes, keys are
438 fetched in an undefined order (which appears random). Takes no arguments,
439 returns the key as a scalar value.
441 my $key = $db->first_key();
445 Returns the "next" key in the hash, given the previous one as the sole argument.
446 Returns undef if there are no more keys to be fetched.
448 $key = $db->next_key($key);
452 Here are some examples of using hashes:
454 my $db = DBM::Deep->new( "foo.db" );
456 $db->put("foo", "bar");
457 print "foo: " . $db->get("foo") . "\n";
459 $db->put("baz", {}); # new child hash ref
460 $db->get("baz")->put("buz", "biz");
461 print "buz: " . $db->get("baz")->get("buz") . "\n";
463 my $key = $db->first_key();
465 print "$key: " . $db->get($key) . "\n";
466 $key = $db->next_key($key);
469 if ($db->exists("foo")) { $db->delete("foo"); }
473 For arrays, DBM::Deep supports all the common methods described above, and the
474 following additional methods: C<length()>, C<push()>, C<pop()>, C<shift()>,
475 C<unshift()> and C<splice()>.
481 Returns the number of elements in the array. Takes no arguments.
483 my $len = $db->length();
487 Adds one or more elements onto the end of the array. Accepts scalars, hash
488 refs or array refs. No return value.
490 $db->push("foo", "bar", {});
494 Fetches the last element in the array, and deletes it. Takes no arguments.
495 Returns undef if array is empty. Returns the element value.
497 my $elem = $db->pop();
501 Fetches the first element in the array, deletes it, then shifts all the
502 remaining elements over to take up the space. Returns the element value. This
503 method is not recommended with large arrays -- see L<LARGE ARRAYS> below for
506 my $elem = $db->shift();
510 Inserts one or more elements onto the beginning of the array, shifting all
511 existing elements over to make room. Accepts scalars, hash refs or array refs.
512 No return value. This method is not recommended with large arrays -- see
513 <LARGE ARRAYS> below for details.
515 $db->unshift("foo", "bar", {});
519 Performs exactly like Perl's built-in function of the same name. See L<perldoc
520 -f splice> for usage -- it is too complicated to document here. This method is
521 not recommended with large arrays -- see L<LARGE ARRAYS> below for details.
525 Here are some examples of using arrays:
527 my $db = DBM::Deep->new(
529 type => DBM::Deep->TYPE_ARRAY
532 $db->push("bar", "baz");
536 my $len = $db->length();
537 print "length: $len\n"; # 4
539 for (my $k=0; $k<$len; $k++) {
540 print "$k: " . $db->get($k) . "\n";
543 $db->splice(1, 2, "biz", "baf");
545 while (my $elem = shift @$db) {
546 print "shifted: $elem\n";
551 Enable or disable automatic file locking by passing a boolean value to the
552 C<locking> parameter when constructing your DBM::Deep object (see L<SETUP>
555 my $db = DBM::Deep->new(
560 This causes DBM::Deep to C<flock()> the underlying filehandle with exclusive
561 mode for writes, and shared mode for reads. This is required if you have
562 multiple processes accessing the same database file, to avoid file corruption.
563 Please note that C<flock()> does NOT work for files over NFS. See L<DB OVER
566 =head2 Explicit Locking
568 You can explicitly lock a database, so it remains locked for multiple
569 actions. This is done by calling the C<lock_exclusive()> method (for when you
570 want to write) or the C<lock_shared()> method (for when you want to read).
571 This is particularly useful for things like counters, where the current value
572 needs to be fetched, then incremented, then stored again.
574 $db->lock_exclusive();
575 my $counter = $db->get("counter");
577 $db->put("counter", $counter);
582 $db->lock_exclusive();
588 Due to Win32 actually enforcing the read-only status of a shared lock, all
589 locks on Win32 and cygwin are exclusive. This is because of how autovivification
590 currently works. Hopefully, this will go away in a future release.
592 =head1 IMPORTING/EXPORTING
594 You can import existing complex structures by calling the C<import()> method,
595 and export an entire database into an in-memory structure using the C<export()>
596 method. Both are examined here.
600 Say you have an existing hash with nested hashes/arrays inside it. Instead of
601 walking the structure and adding keys/elements to the database as you go,
602 simply pass a reference to the C<import()> method. This recursively adds
603 everything to an existing DBM::Deep object for you. Here is an example:
608 array1 => [ "elem0", "elem1", "elem2" ],
610 subkey1 => "subvalue1",
611 subkey2 => "subvalue2"
615 my $db = DBM::Deep->new( "foo.db" );
616 $db->import( $struct );
618 print $db->{key1} . "\n"; # prints "value1"
620 This recursively imports the entire C<$struct> object into C<$db>, including
621 all nested hashes and arrays. If the DBM::Deep object contains exsiting data,
622 keys are merged with the existing ones, replacing if they already exist.
623 The C<import()> method can be called on any database level (not just the base
624 level), and works with both hash and array DB types.
626 B<Note:> Make sure your existing structure has no circular references in it.
627 These will cause an infinite loop when importing. There are plans to fix this
632 Calling the C<export()> method on an existing DBM::Deep object will return
633 a reference to a new in-memory copy of the database. The export is done
634 recursively, so all nested hashes/arrays are all exported to standard Perl
635 objects. Here is an example:
637 my $db = DBM::Deep->new( "foo.db" );
639 $db->{key1} = "value1";
640 $db->{key2} = "value2";
642 $db->{hash1}->{subkey1} = "subvalue1";
643 $db->{hash1}->{subkey2} = "subvalue2";
645 my $struct = $db->export();
647 print $struct->{key1} . "\n"; # prints "value1"
649 This makes a complete copy of the database in memory, and returns a reference
650 to it. The C<export()> method can be called on any database level (not just
651 the base level), and works with both hash and array DB types. Be careful of
652 large databases -- you can store a lot more data in a DBM::Deep object than an
653 in-memory Perl structure.
655 B<Note:> Make sure your database has no circular references in it.
656 These will cause an infinite loop when exporting. There are plans to fix this
661 DBM::Deep has a number of hooks where you can specify your own Perl function
662 to perform filtering on incoming or outgoing data. This is a perfect
663 way to extend the engine, and implement things like real-time compression or
664 encryption. Filtering applies to the base DB level, and all child hashes /
665 arrays. Filter hooks can be specified when your DBM::Deep object is first
666 constructed, or by calling the C<set_filter()> method at any time. There are
667 four available filter hooks.
671 This method takes two paramters - the filter type and the filter subreference.
676 =item * filter_store_key
678 This filter is called whenever a hash key is stored. It
679 is passed the incoming key, and expected to return a transformed key.
681 =item * filter_store_value
683 This filter is called whenever a hash key or array element is stored. It
684 is passed the incoming value, and expected to return a transformed value.
686 =item * filter_fetch_key
688 This filter is called whenever a hash key is fetched (i.e. via
689 C<first_key()> or C<next_key()>). It is passed the transformed key,
690 and expected to return the plain key.
692 =item * filter_fetch_value
694 This filter is called whenever a hash key or array element is fetched.
695 It is passed the transformed value, and expected to return the plain value.
699 Here are the two ways to setup a filter hook:
701 my $db = DBM::Deep->new(
703 filter_store_value => \&my_filter_store,
704 filter_fetch_value => \&my_filter_fetch
709 $db->set_filter( "filter_store_value", \&my_filter_store );
710 $db->set_filter( "filter_fetch_value", \&my_filter_fetch );
712 Your filter function will be called only when dealing with SCALAR keys or
713 values. When nested hashes and arrays are being stored/fetched, filtering
714 is bypassed. Filters are called as static functions, passed a single SCALAR
715 argument, and expected to return a single SCALAR value. If you want to
716 remove a filter, set the function reference to C<undef>:
718 $db->set_filter( "filter_store_value", undef );
722 Please read L<DBM::Deep::Manual> for examples of filters.
724 =head1 ERROR HANDLING
726 Most DBM::Deep methods return a true value for success, and call die() on
727 failure. You can wrap calls in an eval block to catch the die.
729 my $db = DBM::Deep->new( "foo.db" ); # create hash
730 eval { $db->push("foo"); }; # ILLEGAL -- push is array-only call
732 print $@; # prints error message
734 =head1 LARGEFILE SUPPORT
736 If you have a 64-bit system, and your Perl is compiled with both LARGEFILE
737 and 64-bit support, you I<may> be able to create databases larger than 4 GB.
738 DBM::Deep by default uses 32-bit file offset tags, but these can be changed
739 by specifying the 'pack_size' parameter when constructing the file.
743 pack_size => 'large',
746 This tells DBM::Deep to pack all file offsets with 8-byte (64-bit) quad words
747 instead of 32-bit longs. After setting these values your DB files have a
748 theoretical maximum size of 16 XB (exabytes).
750 You can also use C<<pack_size => 'small'>> in order to use 16-bit file
753 B<Note:> Changing these values will B<NOT> work for existing database files.
754 Only change this for new files. Once the value has been set, it is stored in
755 the file's header and cannot be changed for the life of the file. These
756 parameters are per-file, meaning you can access 32-bit and 64-bit files, as
759 B<Note:> We have not personally tested files larger than 4 GB -- all our
760 systems have only a 32-bit Perl. However, we have received user reports that
761 this does indeed work.
763 =head1 LOW-LEVEL ACCESS
765 If you require low-level access to the underlying filehandle that DBM::Deep uses,
766 you can call the C<_fh()> method, which returns the handle:
770 This method can be called on the root level of the datbase, or any child
771 hashes or arrays. All levels share a I<root> structure, which contains things
772 like the filehandle, a reference counter, and all the options specified
773 when you created the object. You can get access to this file object by
774 calling the C<_storage()> method.
776 my $file_obj = $db->_storage();
778 This is useful for changing options after the object has already been created,
779 such as enabling/disabling locking. You can also store your own temporary user
780 data in this structure (be wary of name collision), which is then accessible from
781 any child hash or array.
783 =head1 CIRCULAR REFERENCES
785 DBM::Deep has full support for circular references. Meaning you
786 can have a nested hash key or array element that points to a parent object.
787 This relationship is stored in the DB file, and is preserved between sessions.
790 my $db = DBM::Deep->new( "foo.db" );
793 $db->{circle} = $db; # ref to self
795 print $db->{foo} . "\n"; # prints "bar"
796 print $db->{circle}->{foo} . "\n"; # prints "bar" again
798 This also works as expected with array and hash references. So, the following
801 $db->{foo} = [ 1 .. 3 ];
802 $db->{bar} = $db->{foo};
804 push @{$db->{foo}}, 42;
805 is( $db->{bar}[-1], 42 ); # Passes
807 This, however, does I<not> extend to assignments from one DB file to another.
808 So, the following will throw an error:
810 my $db1 = DBM::Deep->new( "foo.db" );
811 my $db2 = DBM::Deep->new( "bar.db" );
814 $db2->{foo} = $db1->{foo}; # dies
816 B<Note>: Passing the object to a function that recursively walks the
817 object tree (such as I<Data::Dumper> or even the built-in C<optimize()> or
818 C<export()> methods) will result in an infinite loop. This will be fixed in
819 a future release by adding singleton support.
823 As of 1.0000, DBM::Deep hass ACID transactions. Every DBM::Deep object is completely
824 transaction-ready - it is not an option you have to turn on. You do have to
825 specify how many transactions may run simultaneously (q.v. L</num_txns>).
827 Three new methods have been added to support them. They are:
833 This starts a transaction.
837 This applies the changes done within the transaction to the mainline and ends
842 This discards the changes done within the transaction to the mainline and ends
847 Transactions in DBM::Deep are done using a variant of the MVCC method, the
848 same method used by the InnoDB MySQL engine.
852 As of 1.0000, the file format has changed. Furthermore, DBM::Deep is now
853 designed to potentially change file format between point-releases, if needed to
854 support a requested feature. To aid in this, a migration script is provided
855 within the CPAN distribution called C<utils/upgrade_db.pl>.
857 B<NOTE:> This script is not installed onto your system because it carries a copy
858 of every version prior to the current version.
862 The following are items that are planned to be added in future releases. These
863 are separate from the L<CAVEATS, ISSUES & BUGS> below.
865 =head2 Sub-Transactions
867 Right now, you cannot run a transaction within a transaction. Removing this
868 restriction is technically straightforward, but the combinatorial explosion of
869 possible usecases hurts my head. If this is something you want to see
870 immediately, please submit many testcases.
874 If a client is willing to assert upon opening the file that this process will be
875 the only consumer of that datafile, then there are a number of caching
876 possibilities that can be taken advantage of. This does, however, mean that
877 DBM::Deep is more vulnerable to losing data due to unflushed changes. It also
878 means a much larger in-memory footprint. As such, it's not clear exactly how
879 this should be done. Suggestions are welcome.
883 The techniques used in DBM::Deep simply require a seekable contiguous
884 datastore. This could just as easily be a large string as a file. By using
885 substr, the STM capabilities of DBM::Deep could be used within a
886 single-process. I have no idea how I'd specify this, though. Suggestions are
889 =head2 Different contention resolution mechanisms
891 Currently, the only contention resolution mechanism is last-write-wins. This
892 is the mechanism used by most RDBMSes and should be good enough for most uses.
893 For advanced uses of STM, other contention mechanisms will be needed. If you
894 have an idea of how you'd like to see contention resolution in DBM::Deep,
897 =head1 CAVEATS, ISSUES & BUGS
899 This section describes all the known issues with DBM::Deep. These are issues
900 that are either intractable or depend on some feature within Perl working
901 exactly right. It you have found something that is not listed below, please
902 send an e-mail to L<rkinyon@cpan.org>. Likewise, if you think you know of a
903 way around one of these issues, please let me know.
907 (The following assumes a high level of Perl understanding, specifically of
908 references. Most users can safely skip this section.)
910 Currently, the only references supported are HASH and ARRAY. The other reference
911 types (SCALAR, CODE, GLOB, and REF) cannot be supported for various reasons.
917 These are things like filehandles and other sockets. They can't be supported
918 because it's completely unclear how DBM::Deep should serialize them.
922 The discussion here refers to the following type of example:
929 # In some other process ...
931 my $val = ${ $db->{key1} };
933 is( $val, 50, "What actually gets stored in the DB file?" );
935 The problem is one of synchronization. When the variable being referred to
936 changes value, the reference isn't notified, which is kind of the point of
937 references. This means that the new value won't be stored in the datafile for
938 other processes to read. There is no TIEREF.
940 It is theoretically possible to store references to values already within a
941 DBM::Deep object because everything already is synchronized, but the change to
942 the internals would be quite large. Specifically, DBM::Deep would have to tie
943 every single value that is stored. This would bloat the RAM footprint of
944 DBM::Deep at least twofold (if not more) and be a significant performance drain,
945 all to support a feature that has never been requested.
949 L<Data::Dump::Streamer> provides a mechanism for serializing coderefs,
950 including saving off all closure state. This would allow for DBM::Deep to
951 store the code for a subroutine. Then, whenever the subroutine is read, the
952 code could be C<eval()>'ed into being. However, just as for SCALAR and REF,
953 that closure state may change without notifying the DBM::Deep object storing
954 the reference. Again, this would generally be considered a feature.
958 =head2 External references and transactions
960 If you do C<<my $x = $db->{foo};>>, then start a transaction, $x will be
961 referencing the database from outside the transaction. A fix for this (and other
962 issues with how external references into the database) is being looked into. This
963 is the skipped set of tests in t/39_singletons.t and a related issue is the focus
964 of t/37_delete_edge_cases.t
966 =head2 File corruption
968 The current level of error handling in DBM::Deep is minimal. Files I<are> checked
969 for a 32-bit signature when opened, but any other form of corruption in the
970 datafile can cause segmentation faults. DBM::Deep may try to C<seek()> past
971 the end of a file, or get stuck in an infinite loop depending on the level and
972 type of corruption. File write operations are not checked for failure (for
973 speed), so if you happen to run out of disk space, DBM::Deep will probably fail in
974 a bad way. These things will be addressed in a later version of DBM::Deep.
978 Beware of using DBM::Deep files over NFS. DBM::Deep uses flock(), which works
979 well on local filesystems, but will NOT protect you from file corruption over
980 NFS. I've heard about setting up your NFS server with a locking daemon, then
981 using C<lockf()> to lock your files, but your mileage may vary there as well.
982 From what I understand, there is no real way to do it. However, if you need
983 access to the underlying filehandle in DBM::Deep for using some other kind of
984 locking scheme like C<lockf()>, see the L<LOW-LEVEL ACCESS> section above.
986 =head2 Copying Objects
988 Beware of copying tied objects in Perl. Very strange things can happen.
989 Instead, use DBM::Deep's C<clone()> method which safely copies the object and
990 returns a new, blessed and tied hash or array to the same level in the DB.
992 my $copy = $db->clone();
994 B<Note>: Since clone() here is cloning the object, not the database location, any
995 modifications to either $db or $copy will be visible to both.
999 Beware of using C<shift()>, C<unshift()> or C<splice()> with large arrays.
1000 These functions cause every element in the array to move, which can be murder
1001 on DBM::Deep, as every element has to be fetched from disk, then stored again in
1002 a different location. This will be addressed in a future version.
1004 This has been somewhat addressed so that the cost is constant, regardless of
1005 what is stored at those locations. So, small arrays with huge data structures in
1006 them are faster. But, large arrays are still large.
1008 =head2 Writeonly Files
1010 If you pass in a filehandle to new(), you may have opened it in either a
1011 readonly or writeonly mode. STORE will verify that the filehandle is writable.
1012 However, there doesn't seem to be a good way to determine if a filehandle is
1013 readable. And, if the filehandle isn't readable, it's not clear what will
1014 happen. So, don't do that.
1016 =head2 Assignments Within Transactions
1018 The following will I<not> work as one might expect:
1026 is( $x->{a}, 1 ); # This will fail!
1028 The problem is that the moment a reference used as the rvalue to a DBM::Deep
1029 object's lvalue, it becomes tied itself. This is so that future changes to
1030 C<$x> can be tracked within the DBM::Deep file and is considered to be a
1031 feature. By the time the rollback occurs, there is no knowledge that there had
1032 been an C<$x> or what memory location to assign an C<export()> to.
1034 B<NOTE:> This does not affect importing because imports do a walk over the
1035 reference to be imported in order to explicitly leave it untied.
1037 =head1 CODE COVERAGE
1039 L<Devel::Cover> is used to test the code coverage of the tests. Below is the
1040 L<Devel::Cover> report on this distribution's test suite.
1042 ------------------------------------------ ------ ------ ------ ------ ------
1043 File stmt bran cond sub total
1044 ------------------------------------------ ------ ------ ------ ------ ------
1045 blib/lib/DBM/Deep.pm 97.2 90.9 83.3 100.0 95.4
1046 blib/lib/DBM/Deep/Array.pm 100.0 95.7 100.0 100.0 99.0
1047 blib/lib/DBM/Deep/Engine.pm 95.6 84.7 81.6 98.4 92.5
1048 blib/lib/DBM/Deep/File.pm 97.2 81.6 66.7 100.0 91.9
1049 blib/lib/DBM/Deep/Hash.pm 100.0 100.0 100.0 100.0 100.0
1050 Total 96.7 87.5 82.2 99.2 94.1
1051 ------------------------------------------ ------ ------ ------ ------ ------
1053 =head1 MORE INFORMATION
1055 Check out the DBM::Deep Google Group at L<http://groups.google.com/group/DBM-Deep>
1056 or send email to L<DBM-Deep@googlegroups.com>. You can also visit #dbm-deep on
1059 The source code repository is at L<http://github.com/robkinyon/dbm-deep>
1063 Rob Kinyon, L<rkinyon@cpan.org>
1065 Originally written by Joseph Huckaby, L<jhuckaby@cpan.org>
1069 Stonehenge Consulting (L<http://www.stonehenge.com/>) sponsored the
1070 developement of transactions and freespace management, leading to the 1.0000
1071 release. A great debt of gratitude goes out to them for their continuing
1072 leadership in and support of the Perl community.
1076 The following have contributed greatly to make DBM::Deep what it is today:
1080 =item * Adam Sah and Rich Gaushell for innumerable contributions early on.
1082 =item * Dan Golden and others at YAPC::NA 2006 for helping me design through transactions.
1088 perltie(1), Tie::Hash(3), Digest::MD5(3), Fcntl(3), flock(2), lockf(3), nfs(5),
1089 Digest::SHA256(3), Crypt::Blowfish(3), Compress::Zlib(3)
1093 Copyright (c) 2007 Rob Kinyon. All Rights Reserved.
1094 This is free software, you may use it and distribute it under the same terms