=head1 DESCRIPTION
This class handles relationships between the tables in your database
-model. It allows your to set up relationships, and to perform joins
-on searches.
+model. It allows you to set up relationships and perform joins on them.
-This POD details only the convenience methods for setting up standard
-relationship types. For more information see ::Relationship::Base
+Only the helper methods for setting up standard relationship types
+are documented here. For the basic, lower-level methods, see
+L<DBIx::Class::Relationship::Base>.
=head1 METHODS
-All convenience methods take a signature of the following format -
+All helper methods take the following arguments:
- __PACKAGE__>method_name('relname', 'Foreign::Class', $join?, $attrs?);
+ __PACKAGE__>method_name('relname', 'Foreign::Class', $cond, $attrs);
+
+Both C<$cond> and C<$attrs> are optional. Pass C<undef> for C<$cond> if
+you want to use the default value for it, but still want to set C<$attrs>.
+The following attributes are recognize:
+=over 4
+=item join_type
-=over 4
+Explicitly specifies the type of join to use in the relationship. Any SQL
+join type is valid, e.g. C<LEFT> or C<RIGHT>. It will be placed in the SQL
+command immediately before C<JOIN>.
-=item has_one
+=item proxy
- my $f_obj = $obj->relname;
+An arrayref containing a list of accessors in the foreign class to proxy in
+the main class. If, for example, you do the following:
+
+ __PACKAGE__->might_have(bar => 'Bar', undef, { proxy => qw[/ margle /] });
+
+Then, assuming Bar has an accessor named margle, you can do:
-Creates a one-one relationship with another class; defaults to PK-PK for
-the join condition unless a condition is specified.
+ my $obj = Foo->find(1);
+ $obj->margle(10); # set margle; Bar object is created if it doesn't exist
+
+=back
-=item might_have
+=head2 belongs_to
my $f_obj = $obj->relname;
-Creates an optional one-one relationship with another class; defaults to PK-PK
-for the join condition unless a condition is specified.
+ $obj->relname($new_f_obj);
+
+Creates a relationship where we store the foreign class' PK; if $join is a
+column name instead of a condition that is assumed to be the FK, if not
+has_many assumes the FK is the relname is that is a column on the current
+class.
-=item has_many
+=head2 has_many
my @f_obj = $obj->relname($cond?, $attrs?);
my $f_result_set = $obj->relname($cond?, $attrs?);
Creates a one-many relationship with another class;
-=item belongs_to
+=head2 might_have
my $f_obj = $obj->relname;
- $obj->relname($new_f_obj);
+Creates an optional one-one relationship with another class; defaults to PK-PK
+for the join condition unless a condition is specified.
-Creates a relationship where we store the foreign class' PK; if $join is a
-column name instead of a condition that is assumed to be the FK, if not
-has_many assumes the FK is the relname is that is a column on the current
-class.
+=head2 has_one
+
+ my $f_obj = $obj->relname;
+
+Creates a one-one relationship with another class; defaults to PK-PK for
+the join condition unless a condition is specified.
=cut
1;
-=back
-
=head1 AUTHORS
Matt S. Trout <mst@shadowcatsystems.co.uk>
sub belongs_to {
my ($class, $rel, $f_class, $cond, $attrs) = @_;
eval "require $f_class";
+ if ($@) {
+ $class->throw($@) unless $@ =~ /Can't locate/;
+ }
+
my %f_primaries;
$f_primaries{$_} = 1 for eval { $f_class->primary_columns };
my $f_loaded = !$@;
+
# single key relationship
- if (not defined $cond) {
- $class->throw("Can't infer join condition for ${rel} on ${class}; unable to load ${f_class}") unless $f_loaded;
- my ($pri, $too_many) = keys %f_primaries;
- $class->throw("Can't infer join condition for ${rel} on ${class}; ${f_class} has multiple primary key") if $too_many;
+ if (!ref $cond) {
+ my ($pri,$too_many);
+ if (!defined $cond) {
+ $class->throw("Can't infer join condition for ${rel} on ${class}; unable to load ${f_class}") unless $f_loaded;
+ ($pri, $too_many) = keys %f_primaries;
+ $class->throw("Can't infer join condition for ${rel} on ${class}; ${f_class} has no primary keys") unless defined $pri;
+ $class->throw("Can't infer join condition for ${rel} on ${class}; ${f_class} has multiple primary key") if $too_many;
+ }
+ else {
+ $pri = $cond;
+ }
my $acc_type = ($class->has_column($rel)) ? 'filter' : 'single';
$class->add_relationship($rel, $f_class,
{ "foreign.${pri}" => "self.${rel}" },
);
}
# multiple key relationship
- else {
+ elsif (ref $cond eq 'HASH') {
my $cond_rel;
for (keys %$cond) {
if (m/\./) { # Explicit join condition
{ accessor => 'single', %{$attrs || {}} }
);
}
+ else {
+ $class->throw('third argument for belongs_to must be undef, a column name, or a join condition');
+ }
return 1;
}
=head1 NAME
-DBIX::Class::ResultSet - Responsible for fetching and creating resultset.
+DBIx::Class::ResultSet - Responsible for fetching and creating resultset.
-=head1 SYNOPSIS;
+=head1 SYNOPSIS
-$rs=MyApp::DB::Class->search(registered=>1);
+my $rs = MyApp::DB::Class->search(registered => 1);
+my @rows = MyApp::DB::Class->search(foo => 'bar');
=head1 DESCRIPTION
-The resultset is also known as an iterator.
+The resultset is also known as an iterator. It is responsible for handling
+queries that may return an arbitrary number of rows, e.g. via C<search>
+or a C<has_many> relationship.
=head1 METHODS
-=over 4
+=head2 new($db_class, \%$attrs)
-=item new <db_class> <attrs>
-
-The resultset constructor. Takes a db class and an
-attribute hash (see below for more info on attributes)
+The resultset constructor. Takes a table class and an attribute hash
+(see below for more information on attributes). Does not perform
+any queries -- these are executed as needed by the other methods.
=cut
return $new;
}
-=item search
+=head2 search
my @obj = $rs->search({ foo => 3 }); # "... WHERE foo = 3"
my $new_rs = $rs->search({ foo => 3 });
return (wantarray ? $rs->all : $rs);
}
-=item search_literal
+=head2 search_literal
my @obj = $rs->search_literal($literal_where_cond, @bind);
my $new_rs = $rs->search_literal($literal_where_cond, @bind);
Pass a literal chunk of SQL to be added to the conditional part of the
resultset
-=cut
+=cut
+
sub search_literal {
my ($self, $cond, @vals) = @_;
my $attrs = (ref $vals[$#vals] eq 'HASH' ? { %{ pop(@vals) } } : {});
return $self->search(\$cond, $attrs);
}
-=item cursor
+=head2 cursor
-Return a storage driven cursor to the given resultset.
+Returns a storage-driven cursor to the given resultset.
=cut
$attrs->{where},$attrs);
}
-=item search_like
+=head2 search_like
Identical to search except defaults to 'LIKE' instead of '=' in condition
return $class->search($query, { %$attrs });
}
-=item slice <first> <last>
+=head2 slice($first, $last)
-return a number of elements from the given resultset.
+Returns a subset of elements from the resultset.
=cut
return (wantarray ? $slice->all : $slice);
}
-=item next
+=head2 next
-Returns the next element in this resultset.
+Returns the next element in the resultset (undef is there is none).
=cut
return $new;
}
-=item count
+=head2 count
-Performs an SQL count with the same query as the resultset was built
+Performs an SQL C<COUNT> with the same query as the resultset was built
with to find the number of elements. If passed arguments, does a search
on the resultset and counts the results of that.
: $self->{count};
}
-=item count_literal
+=head2 count_literal
-Calls search_literal with the passed arguments, then count
+Calls search_literal with the passed arguments, then count.
=cut
sub count_literal { shift->search_literal(@_)->count; }
-=item all
+=head2 all
-Returns all elements in the resultset. Is called implictly if the search
-method is used in list context.
+Returns all elements in the resultset. Called implictly if the resultset
+is returned in list context.
=cut
$self->cursor->all;
}
-=item reset
+=head2 reset
-Reset this resultset's cursor, so you can iterate through the elements again.
+Resets the resultset's cursor, so you can iterate through the elements again.
=cut
return $self;
}
-=item first
+=head2 first
-resets the resultset and returns the first element.
+Resets the resultset and returns the first element.
=cut
return $_[0]->reset->next;
}
-=item delete
+=head2 delete
Deletes all elements in the resultset.
*delete_all = \&delete; # Yeah, yeah, yeah ...
-=item pager
+=head2 pager
Returns a L<Data::Page> object for the current resultset. Only makes
sense for queries with page turned on.
return $self->{pager};
}
-=item page <page>
+=head2 page($page_num)
-Returns a new resultset representing a given page.
+Returns a new resultset for the specified page.
=cut
return $self->new($self->{source}, $attrs);
}
-=back
-
=head1 Attributes
-The resultset is responsible for handling the various attributes that
-can be passed in with the search functions. Here's an overview of them:
+The resultset takes various attributes that modify its behavior.
+Here's an overview of them:
+
+=head2 order_by
-=over 4
+Which column(s) to order the results by. This is currently passed
+through directly to SQL, so you can give e.g. C<foo DESC> for a
+descending order.
-=item order_by
+=head2 cols
-Which column to order the results by.
+Which columns should be retrieved.
-=item cols
+=head2 join
-Which cols should be retrieved on the first search.
+Contains a list of relationships that should be joined for this query. Can also
+contain a hash reference to refer to that relation's relations. So, if one column
+in your class C<belongs_to> foo and another C<belongs_to> bar, you can do
+C<< join => [qw/ foo bar /] >> to join both (and e.g. use them for C<order_by>).
+If a foo contains many margles and you want to join those too, you can do
+C<< join => { foo => 'margle' } >>. If you want to fetch the columns from the
+related table as well, see C<prefetch> below.
-=item join
+=head2 prefetch
-Contains a list of relations that should be joined for this query. Can also
-contain a hash referece to refer to that relation's relations.
+Contains a list of relationships that should be fetched along with the main
+query (when they are accessed afterwards they will have already been
+"prefetched"). This is useful for when you know you will need the related
+object(s), because it saves a query. Currently limited to prefetching
+one relationship deep, so unlike C<join>, prefetch must be an arrayref.
-=item from
+=head2 from
-This attribute can contain a arrayref of elements. each element can be another
+This attribute can contain a arrayref of elements. Each element can be another
arrayref, to nest joins, or it can be a hash which represents the two sides
of the join.
-*NOTE* Use this on your own risk. This allows you to shoot your foot off!
+NOTE: Use this on your own risk. This allows you to shoot your foot off!
-=item page
+=head2 page
-Should the resultset be paged? This can also be enabled by using the
-'page' option.
+For a paged resultset, specifies which page to retrieve. Leave unset
+for an unpaged resultset.
-=item rows
+=head2 rows
-For paged resultsset, how many rows per page
+For a paged resultset, how many rows per page
-=item offset
-
-For paged resultsset, which page to start on.
-
-=back
+=cut
1;
projects / dbsrgits/DBIx-Class.git / commitdiff