NAME

RDFStore::Model - An implementation of the Model RDF API using tied hashes and implementing free-text search on literals

SYNOPSIS

	use RDFStore::NodeFactory;
	my $factory= new RDFStore::NodeFactory();
	my $statement = $factory->createStatement(
                        	$factory->createResource('http://perl.org'),
                        	$factory->createResource('http://iscool.org/schema/1.0/','label'),
                        	$factory->createLiteral('Cool Web site')
                                );
	my $statement1 = $factory->createStatement(
				$factory->createResource("http://www.altavista.com"),
				$factory->createResource("http://pen.jrc.it/schema/1.0/','author'),
				$factory->createLiteral("Who? :-)")
				);

	my $statement2 = $factory->createStatement(
				$factory->createUniqueResource(),
				$factory->createUniqueResource(),
				$factory->createLiteral("")
				);

	use RDFStore::Model;
	my $model = new RDFStore::Model( Name => 'store', Split => 20, Compression => 1, FreeText => 1 );

	$model->add($statement);
	$model->add($statement1);
	$model->add($statement2);
	my $model1 = $model->duplicate();

	print $model1->getDigest->equals( $model1->getDigest );
	print $model1->getDigest->hashCode;

	my $found = $model->find($statement2->subject,undef,undef);
	my $found1 = $model->find(undef,undef,undef,undef,'Cool'); #free-text search on literals :)

	#get Statements
	foreach ( @{$found->elements} ) {
        	print $_->getLabel(),"\n";
	};

	#or faster
	my $fetch;
	foreach ( @{$found->elements} ) {
		my $fetch=$_;  #avoid too many fetches from RDFStore::Model::Statements
        	print $fetch->getLabel(),"\n";
	};

	#or
	my($statements)=$found1->elements;
	for ( 0..$#{$statements} ) {
                print $statements->[$_]->getLabel(),"\n";
        };

	#get RDFNodes
	foreach ( keys %{$found->elements}) {
        	print $found->elements->{$_}->getLabel(),"\n";
	};

DESCRIPTION

An RDFStore::Stanford::Model implementation using Data::MagicTie tied arrays and hashes to store triplets. The actual store could be tied either to an in-memory, a local or remote database - see Data::MagicTie(3).

This modules implements a storage and iterator by leveraging on perltie(3) and the Data::MagicTie(3) interface. A compact indexing model is used that allows to make free-text indexing up to literals.

CONSTRUCTORS

The following methods construct/tie RDFStore::Model storages and objects:

$model = new RDFStore::Model( %whateveryoulikeit );

Create an new RDFStore::Model object and tie up a serie of Data::MagicTie hash databases to read/write/query RDFStore::RDFNode. The %whateveryoulikeit hash contains a set of configuration options about how and where store actual data. Most of the options correspond to the Data::MagicTie ones - see Data::MagicTie(3) Possible additional options are the following:

Name

This is a label used to identify a Persistent storage by name. It might correspond to a physical file system directory containing the indexes DBs. By default if no Name option is given the storage is assumed to be in-memory (e.g. RDFStore::Storage::find method return result sets as in-memory models by default unless specified differently). For local persistent storages a directory named liek this option is created in the current working directory with mode 0666)

Sync

Sync the RDFStore::Model with the underling Data::MagciTie GDS after each add() or remove().

Compression

Switch on Run Length Encoding (RLE) on the internal index; this option allows to save a lot of space (or memory) if you are going to manage large models. For tiny models is faster to use no compression.

FreeText

Enable free text searching on literals over a model (see find)

SEE ALSO

Data::MagicTie(3) Digest(3) RDFStore::Stanford::Digest::Digestable(3) RDFStore::Stanford::Digest(3) RDFStore::RDFNode RDFStore::Resource RDFStore::FindIndex(3)

AUTHOR

Alberto Reggiori <areggiori@webweaving.org>