# RDF::Trine::Store
# -----------------------------------------------------------------------------
=head1 NAME
RDF::Trine::Store - RDF triplestore base class
=head1 VERSION
This document describes RDF::Trine::Store version 1.014
=head1 DESCRIPTION
RDF::Trine::Store provides a base class and common API for implementations of
triple/quadstores for use with the RDF::Trine framework. In general, it should
be used only be people implementing new stores. For interacting with stores
(e.g. to read, insert, and delete triples) the RDF::Trine::Model interface
should be used (using the model as an intermediary between the client/user and
the underlying store).
To be used by the RDF::Trine framework, store implementations must implement a
set of required functionality:
=over 4
=item * C<< new >>
=item * C<< get_statements >>
=item * C<< get_contexts >>
=item * C<< add_statement >>
=item * C<< remove_statement >>
=item * C<< count_statements >>
=item * C<< supports >>
=back
Implementations may also provide the following methods if a native
implementation would be more efficient than the default provided by
RDF::Trine::Store:
=over 4
=item * C<< get_pattern >>
=item * C<< get_sparql >>
=item * C<< remove_statements >>
=item * C<< size >>
=item * C<< nuke >>
=item * C<< _begin_bulk_ops >>
=item * C<< _end_bulk_ops >>
=back
=cut
package RDF::Trine::Store;
use strict;
use warnings;
no warnings 'redefine';
use Data::Dumper;
use Log::Log4perl;
use Carp qw(carp croak confess);
use Scalar::Util qw(blessed reftype);
use Module::Load::Conditional qw[can_load];
use RDF::Trine::Store::Memory;
use RDF::Trine::Store::Hexastore;
use RDF::Trine::Store::SPARQL;
use RDF::Trine::Store::LanguagePreference;
######################################################################
our ($VERSION, $HAVE_REDLAND, %STORE_CLASSES);
BEGIN {
$VERSION = '1.014';
if ($RDF::Redland::VERSION) {
$HAVE_REDLAND = 1;
}
can_load( modules => {
'RDF::Trine::Store::DBI' => undef,
} );
}
######################################################################
=head1 METHODS
=over 4
=cut
=item C<< new ( $data ) >>
Returns a new RDF::Trine::Store object based on the supplied data value.
This constructor delegates to one of the following methods depending on the
value of C<< $data >>:
* C<< new_with_string >> if C<< $data >> is not a reference
* C<< new_with_config >> if C<< $data >> is a HASH reference
* C<< new_with_object >> if C<< $data >> is a blessed object
=cut
sub new {
my $class = shift;
my $data = shift;
if (blessed($data)) {
return $class->new_with_object($data);
} elsif (ref($data)) {
return $class->new_with_config($data);
} else {
return $class->new_with_string($data);
}
}
=item C<< new_with_string ( $config ) >>
Returns a new RDF::Trine::Store object based on the supplied configuration
string. The format of the string specifies the Store subclass to be
instantiated as well as any required constructor arguments. These are separated
by a semicolon. An example configuration string for the DBI store would be:
DBI;mymodel;DBI:mysql:database=rdf;user;password
The format of the constructor arguments (everything after the first ';') is
specific to the Store subclass.
=cut
sub new_with_string {
my $proto = shift;
my $string = shift;
if (defined($string)) {
my ($subclass, $config) = split(/;/, $string, 2);
my $class = join('::', 'RDF::Trine::Store', $subclass);
if (can_load(modules => { $class => 0 }) and $class->can('_new_with_string')) {
return $class->_new_with_string( $config );
} else {
throw RDF::Trine::Error::UnimplementedError -text => "The class $class doesn't support the use of new_with_string";
}
} else {
throw RDF::Trine::Error::MethodInvocationError;
}
}
=item C<< new_with_config ( \%config ) >>
Returns a new RDF::Trine::Store object based on the supplied
configuration hashref. This requires the the Store subclass to be
supplied with a C<storetype> key, while other keys are required by the
Store subclasses, please refer to each subclass for specific
documentation.
An example invocation for the DBI store may be:
my $store = RDF::Trine::Store->new_with_config({
storetype => 'DBI',
name => 'mymodel',
dsn => 'DBI:mysql:database=rdf',
username => 'dahut',
password => 'Str0ngPa55w0RD'
});
=cut
sub new_with_config {
my $proto = shift;
my $config = shift;
if (defined($config)) {
my $class = $config->{storeclass} || join('::', 'RDF::Trine::Store', $config->{storetype});
if ($class->can('_new_with_config')) {
return $class->_new_with_config( $config );
} else {
throw RDF::Trine::Error::UnimplementedError -text => "The class $class doesn't support the use of new_with_config";
}
} else {
throw RDF::Trine::Error::MethodInvocationError;
}
}
=item C<< new_with_object ( $object ) >>
Returns a new RDF::Trine::Store object based on the supplied opaque C<< $object >>.
If the C<< $object >> is recognized by an available backend as being sufficient
to construct a store object, the store object will be returned. Otherwise undef
will be returned.
=cut
sub new_with_object {
my $proto = shift;
my $obj = shift;
foreach my $class (keys %STORE_CLASSES) {
if ($class->can('_new_with_object')) {
my $s = $class->_new_with_object( $obj );
if ($s) {
return $s;
}
}
}
return;
}
=item C<< nuke >>
Permanently removes the store and its data.
=cut
sub nuke {}
=item C<< class_by_name ( $name ) >>
Returns the class of the storage implementation with the given name.
For example, C<< 'Memory' >> would return C<< 'RDF::Trine::Store::Memory' >>.
=cut
sub class_by_name {
my $proto = shift;
my $name = shift;
foreach my $class (keys %STORE_CLASSES) {
if (lc($class) =~ m/::${name}$/i) {
return $class;
}
}
return;
}
=item C<< temporary_store >>
Returns a new temporary triplestore (using appropriate default values).
=cut
sub temporary_store {
return RDF::Trine::Store::Memory->new();
}
# =item C<< get_pattern ( $bgp [, $context] ) >>
#
# Returns a stream object of all bindings matching the specified graph pattern.
#
# =cut
sub _get_pattern {
my $self = shift;
my $bgp = shift;
my $context = shift;
my @args = @_;
my %args = @args;
if ($bgp->isa('RDF::Trine::Statement')) {
$bgp = RDF::Trine::Pattern->new($bgp);
} else {
$bgp = $bgp->sort_for_join_variables();
}
my %iter_args;
my @triples = $bgp->triples;
my ($iter);
if (1 == scalar(@triples)) {
my $t = shift(@triples);
my @nodes = $t->nodes;
my $size = scalar(@nodes);
my %vars;
my @names = qw(subject predicate object context);
foreach my $n (0 .. $#nodes) {
if ($nodes[$n]->isa('RDF::Trine::Node::Variable')) {
$vars{ $names[ $n ] } = $nodes[$n]->name;
}
}
my $_iter = $self->get_statements( @nodes );
if ($_iter->finished) {
return RDF::Trine::Iterator::Bindings->new( [], [] );
}
my @vars = values %vars;
my $sub = sub {
my $row = $_iter->next;
return unless ($row);
my %data = map { $vars{ $_ } => $row->$_() } (keys %vars);
return RDF::Trine::VariableBindings->new( \%data );
};
$iter = RDF::Trine::Iterator::Bindings->new( $sub, \@vars );
} else {
my $t = pop(@triples);
my $rhs = $self->get_pattern( RDF::Trine::Pattern->new( $t ), $context, @args );
my $lhs = $self->get_pattern( RDF::Trine::Pattern->new( @triples ), $context, @args );
my @inner;
while (my $row = $rhs->next) {
push(@inner, $row);
}
my @results;
while (my $row = $lhs->next) {
RESULT: foreach my $irow (@inner) {
my %keysa;
my @keysa = keys %$irow;
@keysa{ @keysa } = (1) x scalar(@keysa);
my @shared = grep { exists $keysa{ $_ } } (keys %$row);
foreach my $key (@shared) {
my $val_a = $irow->{ $key };
my $val_b = $row->{ $key };
next unless (defined($val_a) and defined($val_b));
my $equal = $val_a->equal( $val_b );
unless ($equal) {
next RESULT;
}
}
my $jrow = { (map { $_ => $irow->{$_} } grep { defined($irow->{$_}) } keys %$irow), (map { $_ => $row->{$_} } grep { defined($row->{$_}) } keys %$row) };
push(@results, RDF::Trine::VariableBindings->new($jrow));
}
}
$iter = RDF::Trine::Iterator::Bindings->new( \@results, [ $bgp->referenced_variables ] );
}
if (my $o = $args{ 'orderby' }) {
unless (reftype($o) eq 'ARRAY') {
throw RDF::Trine::Error::MethodInvocationError -text => "The orderby argument to get_pattern must be an ARRAY reference";
}
my @order;
my %order;
my @o = @$o;
my @sorted_by;
my %vars = map { $_ => 1 } $bgp->referenced_variables;
if (scalar(@o) % 2 != 0) {
throw RDF::Trine::Error::MethodInvocationError -text => "The orderby argument ARRAY to get_pattern must contain an even number of elements";
}
while (@o) {
my ($k,$dir) = splice(@o, 0, 2, ());
next unless ($vars{ $k });
unless ($dir =~ m/^ASC|DESC$/i) {
throw RDF::Trine::Error::MethodInvocationError -text => "The sort direction for key $k must be either 'ASC' or 'DESC' in get_pattern call";
}
my $asc = ($dir eq 'ASC') ? 1 : 0;
push(@order, $k);
$order{ $k } = $asc;
push(@sorted_by, $k, $dir);
}
my @results = $iter->get_all;
@results = _sort_bindings( \@results, \@order, \%order );
$iter_args{ sorted_by } = \@sorted_by;
return RDF::Trine::Iterator::Bindings->new( \@results, [ $bgp->referenced_variables ], %iter_args );
} else {
return $iter;
}
}
sub _sort_bindings {
my $res = shift;
my $o = shift;
my $dir = shift;
my @sorted = map { $_->[0] } sort { _sort_mapped_data($a,$b,$o,$dir) } map { _map_sort_data( $_, $o ) } @$res;
return @sorted;
}
sub _sort_mapped_data {
my $a = shift;
my $b = shift;
my $o = shift;
my $dir = shift;
foreach my $i (1 .. $#{ $a }) {
my $av = $a->[ $i ];
my $bv = $b->[ $i ];
my $key = $o->[ $i-1 ];
next unless (defined($av) or defined($bv));
my $cmp = RDF::Trine::Node::compare( $av, $bv );
unless ($dir->{ $key }) {
$cmp *= -1;
}
return $cmp if ($cmp);
}
return 0;
}
sub _map_sort_data {
my $res = shift;
my $o = shift;
my @data = ($res, map { $res->{ $_ } } @$o);
return \@data;
}
=item C<< get_statements ($subject, $predicate, $object [, $context] ) >>
Returns a stream object of all statements matching the specified subject,
predicate and objects. Any of the arguments may be undef to match any value.
=cut
sub get_statements;
=item C<< get_contexts >>
Returns an RDF::Trine::Iterator over the RDF::Trine::Node objects comprising
the set of contexts of the stored quads.
=cut
sub get_contexts;
=item C<< add_statement ( $statement [, $context] ) >>
Adds the specified C<$statement> to the underlying model.
=cut
sub add_statement;
=item C<< remove_statement ( $statement [, $context]) >>
Removes the specified C<$statement> from the underlying model.
=cut
sub remove_statement;
=item C<< remove_statements ( $subject, $predicate, $object [, $context]) >>
Removes the specified C<$statement> from the underlying model.
=cut
sub remove_statements { # Fallback implementation
my $self = shift;
my $iterator = $self->get_statements(@_);
while (my $st = $iterator->next) {
$self->remove_statement($st);
}
}
=item C<< count_statements ($subject, $predicate, $object) >>
Returns a count of all the statements matching the specified subject,
predicate and objects. Any of the arguments may be undef to match any value.
=cut
sub count_statements;
=item C<< size >>
Returns the number of statements in the store.
=cut
sub size {
my $self = shift;
return $self->count_statements( undef, undef, undef, undef );
}
=item C<< etag >>
If the store has the capability and knowledge to support caching, returns a
persistent token that will remain consistent as long as the store's data doesn't
change. This token is acceptable for use as an HTTP ETag.
=cut
sub etag {
return;
}
=item C<< supports ( [ $feature ] ) >>
If C<< $feature >> is specified, returns true if the feature is supported by the
store, false otherwise. If C<< $feature >> is not specified, returns a list of
supported features.
=cut
sub supports {
return;
}
sub _begin_bulk_ops {}
sub _end_bulk_ops {}
1;
__END__
=back
=head1 BUGS
Please report any bugs or feature requests to through the GitHub web interface
at L<https://github.com/kasei/perlrdf/issues>.
=head1 AUTHOR
Gregory Todd Williams C<< <gwilliams@cpan.org> >>
=head1 COPYRIGHT
Copyright (c) 2006-2012 Gregory Todd Williams. This
program is free software; you can redistribute it and/or modify it under
the same terms as Perl itself.
=cut
get_statements() XXX maybe this should instead follow the quad semantics?
get_statements( s, p, o )
return (s,p,o,nil) for all distinct (s,p,o)
get_statements( s, p, o, g )
return all (s,p,o,g)
add_statement( TRIPLE )
add (s, p, o, nil)
add_statement( TRIPLE, CONTEXT )
add (s, p, o, context)
add_statement( QUAD )
add (s, p, o, g )
add_statement( QUAD, CONTEXT )
throw exception
remove_statement( TRIPLE )
remove (s, p, o, nil)
remove_statement( TRIPLE, CONTEXT )
remove (s, p, o, context)
remove_statement( QUAD )
remove (s, p, o, g)
remove_statement( QUAD, CONTEXT )
throw exception
count_statements() XXX maybe this should instead follow the quad semantics?
count_statements( s, p, o )
count distinct (s,p,o) for all statements (s,p,o,g)
count_statements( s, p, o, g )
count (s,p,o,g)