1 package DBIx::Class::ResultSet;
9 use Carp::Clan qw/^DBIx::Class/;
10 use DBIx::Class::Exception;
13 use DBIx::Class::ResultSetColumn;
14 use DBIx::Class::ResultSourceHandle;
17 use base qw/DBIx::Class/;
19 __PACKAGE__->mk_group_accessors('simple' => qw/_result_class _source_handle/);
23 DBIx::Class::ResultSet - Represents a query used for fetching a set of results.
27 my $users_rs = $schema->resultset('User');
28 my $registered_users_rs = $schema->resultset('User')->search({ registered => 1 });
29 my @cds_in_2005 = $schema->resultset('CD')->search({ year => 2005 })->all();
33 A ResultSet is an object which stores a set of conditions representing
34 a query. It is the backbone of DBIx::Class (i.e. the really
35 important/useful bit).
37 No SQL is executed on the database when a ResultSet is created, it
38 just stores all the conditions needed to create the query.
40 A basic ResultSet representing the data of an entire table is returned
41 by calling C<resultset> on a L<DBIx::Class::Schema> and passing in a
42 L<Source|DBIx::Class::Manual::Glossary/Source> name.
44 my $users_rs = $schema->resultset('User');
46 A new ResultSet is returned from calling L</search> on an existing
47 ResultSet. The new one will contain all the conditions of the
48 original, plus any new conditions added in the C<search> call.
50 A ResultSet also incorporates an implicit iterator. L</next> and L</reset>
51 can be used to walk through all the L<DBIx::Class::Row>s the ResultSet
54 The query that the ResultSet represents is B<only> executed against
55 the database when these methods are called:
56 L</find> L</next> L</all> L</first> L</single> L</count>
60 =head2 Chaining resultsets
62 Let's say you've got a query that needs to be run to return some data
63 to the user. But, you have an authorization system in place that
64 prevents certain users from seeing certain information. So, you want
65 to construct the basic query in one method, but add constraints to it in
70 my $request = $self->get_request; # Get a request object somehow.
71 my $schema = $self->get_schema; # Get the DBIC schema object somehow.
73 my $cd_rs = $schema->resultset('CD')->search({
74 title => $request->param('title'),
75 year => $request->param('year'),
78 $self->apply_security_policy( $cd_rs );
83 sub apply_security_policy {
92 =head3 Resolving conditions and attributes
94 When a resultset is chained from another resultset, conditions and
95 attributes with the same keys need resolving.
97 L</join>, L</prefetch>, L</+select>, L</+as> attributes are merged
98 into the existing ones from the original resultset.
100 The L</where>, L</having> attribute, and any search conditions are
101 merged with an SQL C<AND> to the existing condition from the original
104 All other attributes are overridden by any new ones supplied in the
107 =head2 Multiple queries
109 Since a resultset just defines a query, you can do all sorts of
110 things with it with the same object.
112 # Don't hit the DB yet.
113 my $cd_rs = $schema->resultset('CD')->search({
114 title => 'something',
118 # Each of these hits the DB individually.
119 my $count = $cd_rs->count;
120 my $most_recent = $cd_rs->get_column('date_released')->max();
121 my @records = $cd_rs->all;
123 And it's not just limited to SELECT statements.
129 $cd_rs->create({ artist => 'Fred' });
131 Which is the same as:
133 $schema->resultset('CD')->create({
134 title => 'something',
139 See: L</search>, L</count>, L</get_column>, L</all>, L</create>.
143 If a resultset is used in a numeric context it returns the L</count>.
144 However, if it is used in a booleand context it is always true. So if
145 you want to check if a resultset has any results use C<if $rs != 0>.
146 C<if $rs> will always be true.
154 =item Arguments: $source, \%$attrs
156 =item Return Value: $rs
160 The resultset constructor. Takes a source object (usually a
161 L<DBIx::Class::ResultSourceProxy::Table>) and an attribute hash (see
162 L</ATTRIBUTES> below). Does not perform any queries -- these are
163 executed as needed by the other methods.
165 Generally you won't need to construct a resultset manually. You'll
166 automatically get one from e.g. a L</search> called in scalar context:
168 my $rs = $schema->resultset('CD')->search({ title => '100th Window' });
170 IMPORTANT: If called on an object, proxies to new_result instead so
172 my $cd = $schema->resultset('CD')->new({ title => 'Spoon' });
174 will return a CD object, not a ResultSet.
180 return $class->new_result(@_) if ref $class;
182 my ($source, $attrs) = @_;
183 $source = $source->handle
184 unless $source->isa('DBIx::Class::ResultSourceHandle');
185 $attrs = { %{$attrs||{}} };
187 if ($attrs->{page}) {
188 $attrs->{rows} ||= 10;
191 $attrs->{alias} ||= 'me';
193 # Creation of {} and bless separated to mitigate RH perl bug
194 # see https://bugzilla.redhat.com/show_bug.cgi?id=196836
196 _source_handle => $source,
197 cond => $attrs->{where},
206 $attrs->{result_class} || $source->resolve->result_class
216 =item Arguments: $cond, \%attrs?
218 =item Return Value: $resultset (scalar context), @row_objs (list context)
222 my @cds = $cd_rs->search({ year => 2001 }); # "... WHERE year = 2001"
223 my $new_rs = $cd_rs->search({ year => 2005 });
225 my $new_rs = $cd_rs->search([ { year => 2005 }, { year => 2004 } ]);
226 # year = 2005 OR year = 2004
228 If you need to pass in additional attributes but no additional condition,
229 call it as C<search(undef, \%attrs)>.
231 # "SELECT name, artistid FROM $artist_table"
232 my @all_artists = $schema->resultset('Artist')->search(undef, {
233 columns => [qw/name artistid/],
236 For a list of attributes that can be passed to C<search>, see
237 L</ATTRIBUTES>. For more examples of using this function, see
238 L<Searching|DBIx::Class::Manual::Cookbook/Searching>. For a complete
239 documentation for the first argument, see L<SQL::Abstract>.
241 For more help on using joins with search, see L<DBIx::Class::Manual::Joining>.
247 my $rs = $self->search_rs( @_ );
248 return (wantarray ? $rs->all : $rs);
255 =item Arguments: $cond, \%attrs?
257 =item Return Value: $resultset
261 This method does the same exact thing as search() except it will
262 always return a resultset, even in list context.
269 # Special-case handling for (undef, undef).
270 if ( @_ == 2 && !defined $_[1] && !defined $_[0] ) {
275 $attrs = pop(@_) if @_ > 1 and ref $_[$#_] eq 'HASH';
276 my $our_attrs = { %{$self->{attrs}} };
277 my $having = delete $our_attrs->{having};
278 my $where = delete $our_attrs->{where};
282 my %safe = (alias => 1, cache => 1);
285 (@_ && defined($_[0])) # @_ == () or (undef)
287 (keys %$attrs # empty attrs or only 'safe' attrs
288 && List::Util::first { !$safe{$_} } keys %$attrs)
290 # no search, effectively just a clone
291 $rows = $self->get_cache;
294 my $new_attrs = { %{$our_attrs}, %{$attrs} };
296 # merge new attrs into inherited
297 foreach my $key (qw/join prefetch +select +as bind/) {
298 next unless exists $attrs->{$key};
299 $new_attrs->{$key} = $self->_merge_attr($our_attrs->{$key}, $attrs->{$key});
304 (@_ == 1 || ref $_[0] eq "HASH")
306 (ref $_[0] eq 'HASH')
308 (keys %{ $_[0] } > 0)
316 ? $self->throw_exception("Odd number of arguments to search")
323 if (defined $where) {
324 $new_attrs->{where} = (
325 defined $new_attrs->{where}
328 ref $_ eq 'ARRAY' ? [ -or => $_ ] : $_
329 } $where, $new_attrs->{where}
336 $new_attrs->{where} = (
337 defined $new_attrs->{where}
340 ref $_ eq 'ARRAY' ? [ -or => $_ ] : $_
341 } $cond, $new_attrs->{where}
347 if (defined $having) {
348 $new_attrs->{having} = (
349 defined $new_attrs->{having}
352 ref $_ eq 'ARRAY' ? [ -or => $_ ] : $_
353 } $having, $new_attrs->{having}
359 my $rs = (ref $self)->new($self->result_source, $new_attrs);
361 $rs->set_cache($rows);
366 =head2 search_literal
370 =item Arguments: $sql_fragment, @bind_values
372 =item Return Value: $resultset (scalar context), @row_objs (list context)
376 my @cds = $cd_rs->search_literal('year = ? AND title = ?', qw/2001 Reload/);
377 my $newrs = $artist_rs->search_literal('name = ?', 'Metallica');
379 Pass a literal chunk of SQL to be added to the conditional part of the
382 CAVEAT: C<search_literal> is provided for Class::DBI compatibility and should
383 only be used in that context. C<search_literal> is a convenience method.
384 It is equivalent to calling $schema->search(\[]), but if you want to ensure
385 columns are bound correctly, use C<search>.
387 Example of how to use C<search> instead of C<search_literal>
389 my @cds = $cd_rs->search_literal('cdid = ? AND (artist = ? OR artist = ?)', (2, 1, 2));
390 my @cds = $cd_rs->search(\[ 'cdid = ? AND (artist = ? OR artist = ?)', [ 'cdid', 2 ], [ 'artist', 1 ], [ 'artist', 2 ] ]);
393 See L<DBIx::Class::Manual::Cookbook/Searching> and
394 L<DBIx::Class::Manual::FAQ/Searching> for searching techniques that do not
395 require C<search_literal>.
400 my ($self, $sql, @bind) = @_;
402 if ( @bind && ref($bind[-1]) eq 'HASH' ) {
405 return $self->search(\[ $sql, map [ __DUMMY__ => $_ ], @bind ], ($attr || () ));
412 =item Arguments: @values | \%cols, \%attrs?
414 =item Return Value: $row_object | undef
418 Finds a row based on its primary key or unique constraint. For example, to find
419 a row by its primary key:
421 my $cd = $schema->resultset('CD')->find(5);
423 You can also find a row by a specific unique constraint using the C<key>
424 attribute. For example:
426 my $cd = $schema->resultset('CD')->find('Massive Attack', 'Mezzanine', {
427 key => 'cd_artist_title'
430 Additionally, you can specify the columns explicitly by name:
432 my $cd = $schema->resultset('CD')->find(
434 artist => 'Massive Attack',
435 title => 'Mezzanine',
437 { key => 'cd_artist_title' }
440 If the C<key> is specified as C<primary>, it searches only on the primary key.
442 If no C<key> is specified, it searches on all unique constraints defined on the
443 source for which column data is provided, including the primary key.
445 If your table does not have a primary key, you B<must> provide a value for the
446 C<key> attribute matching one of the unique constraints on the source.
448 In addition to C<key>, L</find> recognizes and applies standard
449 L<resultset attributes|/ATTRIBUTES> in the same way as L</search> does.
451 Note: If your query does not return only one row, a warning is generated:
453 Query returned more than one row
455 See also L</find_or_create> and L</update_or_create>. For information on how to
456 declare unique constraints, see
457 L<DBIx::Class::ResultSource/add_unique_constraint>.
463 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
465 # Default to the primary key, but allow a specific key
466 my @cols = exists $attrs->{key}
467 ? $self->result_source->unique_constraint_columns($attrs->{key})
468 : $self->result_source->primary_columns;
469 $self->throw_exception(
470 "Can't find unless a primary key is defined or unique constraint is specified"
473 # Parse out a hashref from input
475 if (ref $_[0] eq 'HASH') {
476 $input_query = { %{$_[0]} };
478 elsif (@_ == @cols) {
480 @{$input_query}{@cols} = @_;
483 # Compatibility: Allow e.g. find(id => $value)
484 carp "Find by key => value deprecated; please use a hashref instead";
488 my (%related, $info);
490 KEY: foreach my $key (keys %$input_query) {
491 if (ref($input_query->{$key})
492 && ($info = $self->result_source->relationship_info($key))) {
493 my $val = delete $input_query->{$key};
494 next KEY if (ref($val) eq 'ARRAY'); # has_many for multi_create
495 my $rel_q = $self->result_source->_resolve_condition(
496 $info->{cond}, $val, $key
498 die "Can't handle OR join condition in find" if ref($rel_q) eq 'ARRAY';
499 @related{keys %$rel_q} = values %$rel_q;
502 if (my @keys = keys %related) {
503 @{$input_query}{@keys} = values %related;
507 # Build the final query: Default to the disjunction of the unique queries,
508 # but allow the input query in case the ResultSet defines the query or the
509 # user is abusing find
510 my $alias = exists $attrs->{alias} ? $attrs->{alias} : $self->{attrs}{alias};
512 if (exists $attrs->{key}) {
513 my @unique_cols = $self->result_source->unique_constraint_columns($attrs->{key});
514 my $unique_query = $self->_build_unique_query($input_query, \@unique_cols);
515 $query = $self->_add_alias($unique_query, $alias);
517 elsif ($self->{attrs}{accessor} and $self->{attrs}{accessor} eq 'single') {
518 # This means that we got here after a merger of relationship conditions
519 # in ::Relationship::Base::search_related (the row method), and furthermore
520 # the relationship is of the 'single' type. This means that the condition
521 # provided by the relationship (already attached to $self) is sufficient,
522 # as there can be only one row in the databse that would satisfy the
526 my @unique_queries = $self->_unique_queries($input_query, $attrs);
527 $query = @unique_queries
528 ? [ map { $self->_add_alias($_, $alias) } @unique_queries ]
529 : $self->_add_alias($input_query, $alias);
533 my $rs = $self->search ($query, $attrs);
534 if (keys %{$rs->_resolved_attrs->{collapse}}) {
536 carp "Query returned more than one row" if $rs->next;
546 # Add the specified alias to the specified query hash. A copy is made so the
547 # original query is not modified.
550 my ($self, $query, $alias) = @_;
552 my %aliased = %$query;
553 foreach my $col (grep { ! m/\./ } keys %aliased) {
554 $aliased{"$alias.$col"} = delete $aliased{$col};
562 # Build a list of queries which satisfy unique constraints.
564 sub _unique_queries {
565 my ($self, $query, $attrs) = @_;
567 my @constraint_names = exists $attrs->{key}
569 : $self->result_source->unique_constraint_names;
571 my $where = $self->_collapse_cond($self->{attrs}{where} || {});
572 my $num_where = scalar keys %$where;
574 my (@unique_queries, %seen_column_combinations);
575 foreach my $name (@constraint_names) {
576 my @constraint_cols = $self->result_source->unique_constraint_columns($name);
578 my $constraint_sig = join "\x00", sort @constraint_cols;
579 next if $seen_column_combinations{$constraint_sig}++;
581 my $unique_query = $self->_build_unique_query($query, \@constraint_cols);
583 my $num_cols = scalar @constraint_cols;
584 my $num_query = scalar keys %$unique_query;
586 my $total = $num_query + $num_where;
587 if ($num_query && ($num_query == $num_cols || $total == $num_cols)) {
588 # The query is either unique on its own or is unique in combination with
589 # the existing where clause
590 push @unique_queries, $unique_query;
594 return @unique_queries;
597 # _build_unique_query
599 # Constrain the specified query hash based on the specified column names.
601 sub _build_unique_query {
602 my ($self, $query, $unique_cols) = @_;
605 map { $_ => $query->{$_} }
606 grep { exists $query->{$_} }
611 =head2 search_related
615 =item Arguments: $rel, $cond, \%attrs?
617 =item Return Value: $new_resultset
621 $new_rs = $cd_rs->search_related('artist', {
625 Searches the specified relationship, optionally specifying a condition and
626 attributes for matching records. See L</ATTRIBUTES> for more information.
631 return shift->related_resultset(shift)->search(@_);
634 =head2 search_related_rs
636 This method works exactly the same as search_related, except that
637 it guarantees a restultset, even in list context.
641 sub search_related_rs {
642 return shift->related_resultset(shift)->search_rs(@_);
649 =item Arguments: none
651 =item Return Value: $cursor
655 Returns a storage-driven cursor to the given resultset. See
656 L<DBIx::Class::Cursor> for more information.
663 my $attrs = $self->_resolved_attrs_copy;
665 return $self->{cursor}
666 ||= $self->result_source->storage->select($attrs->{from}, $attrs->{select},
667 $attrs->{where},$attrs);
674 =item Arguments: $cond?
676 =item Return Value: $row_object?
680 my $cd = $schema->resultset('CD')->single({ year => 2001 });
682 Inflates the first result without creating a cursor if the resultset has
683 any records in it; if not returns nothing. Used by L</find> as a lean version of
686 While this method can take an optional search condition (just like L</search>)
687 being a fast-code-path it does not recognize search attributes. If you need to
688 add extra joins or similar, call L</search> and then chain-call L</single> on the
689 L<DBIx::Class::ResultSet> returned.
695 As of 0.08100, this method enforces the assumption that the preceeding
696 query returns only one row. If more than one row is returned, you will receive
699 Query returned more than one row
701 In this case, you should be using L</next> or L</find> instead, or if you really
702 know what you are doing, use the L</rows> attribute to explicitly limit the size
705 This method will also throw an exception if it is called on a resultset prefetching
706 has_many, as such a prefetch implies fetching multiple rows from the database in
707 order to assemble the resulting object.
714 my ($self, $where) = @_;
716 $self->throw_exception('single() only takes search conditions, no attributes. You want ->search( $cond, $attrs )->single()');
719 my $attrs = $self->_resolved_attrs_copy;
721 if (keys %{$attrs->{collapse}}) {
722 $self->throw_exception(
723 'single() can not be used on resultsets prefetching has_many. Use find( \%cond ) or next() instead'
728 if (defined $attrs->{where}) {
731 [ map { ref $_ eq 'ARRAY' ? [ -or => $_ ] : $_ }
732 $where, delete $attrs->{where} ]
735 $attrs->{where} = $where;
739 # XXX: Disabled since it doesn't infer uniqueness in all cases
740 # unless ($self->_is_unique_query($attrs->{where})) {
741 # carp "Query not guaranteed to return a single row"
742 # . "; please declare your unique constraints or use search instead";
745 my @data = $self->result_source->storage->select_single(
746 $attrs->{from}, $attrs->{select},
747 $attrs->{where}, $attrs
750 return (@data ? ($self->_construct_object(@data))[0] : undef);
756 # Try to determine if the specified query is guaranteed to be unique, based on
757 # the declared unique constraints.
759 sub _is_unique_query {
760 my ($self, $query) = @_;
762 my $collapsed = $self->_collapse_query($query);
763 my $alias = $self->{attrs}{alias};
765 foreach my $name ($self->result_source->unique_constraint_names) {
766 my @unique_cols = map {
768 } $self->result_source->unique_constraint_columns($name);
770 # Count the values for each unique column
771 my %seen = map { $_ => 0 } @unique_cols;
773 foreach my $key (keys %$collapsed) {
774 my $aliased = $key =~ /\./ ? $key : "$alias.$key";
775 next unless exists $seen{$aliased}; # Additional constraints are okay
776 $seen{$aliased} = scalar keys %{ $collapsed->{$key} };
779 # If we get 0 or more than 1 value for a column, it's not necessarily unique
780 return 1 unless grep { $_ != 1 } values %seen;
788 # Recursively collapse the query, accumulating values for each column.
790 sub _collapse_query {
791 my ($self, $query, $collapsed) = @_;
795 if (ref $query eq 'ARRAY') {
796 foreach my $subquery (@$query) {
797 next unless ref $subquery; # -or
798 $collapsed = $self->_collapse_query($subquery, $collapsed);
801 elsif (ref $query eq 'HASH') {
802 if (keys %$query and (keys %$query)[0] eq '-and') {
803 foreach my $subquery (@{$query->{-and}}) {
804 $collapsed = $self->_collapse_query($subquery, $collapsed);
808 foreach my $col (keys %$query) {
809 my $value = $query->{$col};
810 $collapsed->{$col}{$value}++;
822 =item Arguments: $cond?
824 =item Return Value: $resultsetcolumn
828 my $max_length = $rs->get_column('length')->max;
830 Returns a L<DBIx::Class::ResultSetColumn> instance for a column of the ResultSet.
835 my ($self, $column) = @_;
836 my $new = DBIx::Class::ResultSetColumn->new($self, $column);
844 =item Arguments: $cond, \%attrs?
846 =item Return Value: $resultset (scalar context), @row_objs (list context)
850 # WHERE title LIKE '%blue%'
851 $cd_rs = $rs->search_like({ title => '%blue%'});
853 Performs a search, but uses C<LIKE> instead of C<=> as the condition. Note
854 that this is simply a convenience method retained for ex Class::DBI users.
855 You most likely want to use L</search> with specific operators.
857 For more information, see L<DBIx::Class::Manual::Cookbook>.
859 This method is deprecated and will be removed in 0.09. Use L</search()>
860 instead. An example conversion is:
862 ->search_like({ foo => 'bar' });
866 ->search({ foo => { like => 'bar' } });
873 'search_like() is deprecated and will be removed in DBIC version 0.09.'
874 .' Instead use ->search({ x => { -like => "y%" } })'
875 .' (note the outer pair of {}s - they are important!)'
877 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
878 my $query = ref $_[0] eq 'HASH' ? { %{shift()} }: {@_};
879 $query->{$_} = { 'like' => $query->{$_} } for keys %$query;
880 return $class->search($query, { %$attrs });
887 =item Arguments: $first, $last
889 =item Return Value: $resultset (scalar context), @row_objs (list context)
893 Returns a resultset or object list representing a subset of elements from the
894 resultset slice is called on. Indexes are from 0, i.e., to get the first
897 my ($one, $two, $three) = $rs->slice(0, 2);
902 my ($self, $min, $max) = @_;
903 my $attrs = {}; # = { %{ $self->{attrs} || {} } };
904 $attrs->{offset} = $self->{attrs}{offset} || 0;
905 $attrs->{offset} += $min;
906 $attrs->{rows} = ($max ? ($max - $min + 1) : 1);
907 return $self->search(undef(), $attrs);
908 #my $slice = (ref $self)->new($self->result_source, $attrs);
909 #return (wantarray ? $slice->all : $slice);
916 =item Arguments: none
918 =item Return Value: $result?
922 Returns the next element in the resultset (C<undef> is there is none).
924 Can be used to efficiently iterate over records in the resultset:
926 my $rs = $schema->resultset('CD')->search;
927 while (my $cd = $rs->next) {
931 Note that you need to store the resultset object, and call C<next> on it.
932 Calling C<< resultset('Table')->next >> repeatedly will always return the
933 first record from the resultset.
939 if (my $cache = $self->get_cache) {
940 $self->{all_cache_position} ||= 0;
941 return $cache->[$self->{all_cache_position}++];
943 if ($self->{attrs}{cache}) {
944 $self->{all_cache_position} = 1;
945 return ($self->all)[0];
947 if ($self->{stashed_objects}) {
948 my $obj = shift(@{$self->{stashed_objects}});
949 delete $self->{stashed_objects} unless @{$self->{stashed_objects}};
953 exists $self->{stashed_row}
954 ? @{delete $self->{stashed_row}}
955 : $self->cursor->next
957 return undef unless (@row);
958 my ($row, @more) = $self->_construct_object(@row);
959 $self->{stashed_objects} = \@more if @more;
963 sub _construct_object {
964 my ($self, @row) = @_;
966 my $info = $self->_collapse_result($self->{_attrs}{as}, \@row)
968 my @new = $self->result_class->inflate_result($self->result_source, @$info);
969 @new = $self->{_attrs}{record_filter}->(@new)
970 if exists $self->{_attrs}{record_filter};
974 sub _collapse_result {
975 my ($self, $as_proto, $row) = @_;
977 # if the first row that ever came in is totally empty - this means we got
978 # hit by a smooth^Wempty left-joined resultset. Just noop in that case
979 # instead of producing a {}
988 return undef unless $has_def;
992 # 'foo' => [ undef, 'foo' ]
993 # 'foo.bar' => [ 'foo', 'bar' ]
994 # 'foo.bar.baz' => [ 'foo.bar', 'baz' ]
996 my @construct_as = map { [ (/^(?:(.*)\.)?([^.]+)$/) ] } @$as_proto;
998 my %collapse = %{$self->{_attrs}{collapse}||{}};
1002 # if we're doing collapsing (has_many prefetch) we need to grab records
1003 # until the PK changes, so fill @pri_index. if not, we leave it empty so
1004 # we know we don't have to bother.
1006 # the reason for not using the collapse stuff directly is because if you
1007 # had for e.g. two artists in a row with no cds, the collapse info for
1008 # both would be NULL (undef) so you'd lose the second artist
1010 # store just the index so we can check the array positions from the row
1011 # without having to contruct the full hash
1013 if (keys %collapse) {
1014 my %pri = map { ($_ => 1) } $self->result_source->primary_columns;
1015 foreach my $i (0 .. $#construct_as) {
1016 next if defined($construct_as[$i][0]); # only self table
1017 if (delete $pri{$construct_as[$i][1]}) {
1018 push(@pri_index, $i);
1020 last unless keys %pri; # short circuit (Johnny Five Is Alive!)
1024 # no need to do an if, it'll be empty if @pri_index is empty anyway
1026 my %pri_vals = map { ($_ => $copy[$_]) } @pri_index;
1030 do { # no need to check anything at the front, we always want the first row
1034 foreach my $this_as (@construct_as) {
1035 $const{$this_as->[0]||''}{$this_as->[1]} = shift(@copy);
1038 push(@const_rows, \%const);
1040 } until ( # no pri_index => no collapse => drop straight out
1043 do { # get another row, stash it, drop out if different PK
1045 @copy = $self->cursor->next;
1046 $self->{stashed_row} = \@copy;
1048 # last thing in do block, counts as true if anything doesn't match
1050 # check xor defined first for NULL vs. NOT NULL then if one is
1051 # defined the other must be so check string equality
1054 (defined $pri_vals{$_} ^ defined $copy[$_])
1055 || (defined $pri_vals{$_} && ($pri_vals{$_} ne $copy[$_]))
1060 my $alias = $self->{attrs}{alias};
1067 foreach my $const (@const_rows) {
1068 scalar @const_keys or do {
1069 @const_keys = sort { length($a) <=> length($b) } keys %$const;
1071 foreach my $key (@const_keys) {
1074 my @parts = split(/\./, $key);
1076 my $data = $const->{$key};
1077 foreach my $p (@parts) {
1078 $target = $target->[1]->{$p} ||= [];
1080 if ($cur eq ".${key}" && (my @ckey = @{$collapse{$cur}||[]})) {
1081 # collapsing at this point and on final part
1082 my $pos = $collapse_pos{$cur};
1083 CK: foreach my $ck (@ckey) {
1084 if (!defined $pos->{$ck} || $pos->{$ck} ne $data->{$ck}) {
1085 $collapse_pos{$cur} = $data;
1086 delete @collapse_pos{ # clear all positioning for sub-entries
1087 grep { m/^\Q${cur}.\E/ } keys %collapse_pos
1094 if (exists $collapse{$cur}) {
1095 $target = $target->[-1];
1098 $target->[0] = $data;
1100 $info->[0] = $const->{$key};
1108 =head2 result_source
1112 =item Arguments: $result_source?
1114 =item Return Value: $result_source
1118 An accessor for the primary ResultSource object from which this ResultSet
1125 =item Arguments: $result_class?
1127 =item Return Value: $result_class
1131 An accessor for the class to use when creating row objects. Defaults to
1132 C<< result_source->result_class >> - which in most cases is the name of the
1133 L<"table"|DBIx::Class::Manual::Glossary/"ResultSource"> class.
1135 Note that changing the result_class will also remove any components
1136 that were originally loaded in the source class via
1137 L<DBIx::Class::ResultSource/load_components>. Any overloaded methods
1138 in the original source class will not run.
1143 my ($self, $result_class) = @_;
1144 if ($result_class) {
1145 $self->ensure_class_loaded($result_class);
1146 $self->_result_class($result_class);
1148 $self->_result_class;
1155 =item Arguments: $cond, \%attrs??
1157 =item Return Value: $count
1161 Performs an SQL C<COUNT> with the same query as the resultset was built
1162 with to find the number of elements. Passing arguments is equivalent to
1163 C<< $rs->search ($cond, \%attrs)->count >>
1169 return $self->search(@_)->count if @_ and defined $_[0];
1170 return scalar @{ $self->get_cache } if $self->get_cache;
1172 my $attrs = $self->_resolved_attrs_copy;
1174 # this is a little optimization - it is faster to do the limit
1175 # adjustments in software, instead of a subquery
1176 my $rows = delete $attrs->{rows};
1177 my $offset = delete $attrs->{offset};
1180 if ($self->_has_resolved_attr (qw/collapse group_by/)) {
1181 $crs = $self->_count_subq_rs ($attrs);
1184 $crs = $self->_count_rs ($attrs);
1186 my $count = $crs->next;
1188 $count -= $offset if $offset;
1189 $count = $rows if $rows and $rows < $count;
1190 $count = 0 if ($count < 0);
1199 =item Arguments: $cond, \%attrs??
1201 =item Return Value: $count_rs
1205 Same as L</count> but returns a L<DBIx::Class::ResultSetColumn> object.
1206 This can be very handy for subqueries:
1208 ->search( { amount => $some_rs->count_rs->as_query } )
1210 As with regular resultsets the SQL query will be executed only after
1211 the resultset is accessed via L</next> or L</all>. That would return
1212 the same single value obtainable via L</count>.
1218 return $self->search(@_)->count_rs if @_;
1220 # this may look like a lack of abstraction (count() does about the same)
1221 # but in fact an _rs *must* use a subquery for the limits, as the
1222 # software based limiting can not be ported if this $rs is to be used
1223 # in a subquery itself (i.e. ->as_query)
1224 if ($self->_has_resolved_attr (qw/collapse group_by offset rows/)) {
1225 return $self->_count_subq_rs;
1228 return $self->_count_rs;
1233 # returns a ResultSetColumn object tied to the count query
1236 my ($self, $attrs) = @_;
1238 my $rsrc = $self->result_source;
1239 $attrs ||= $self->_resolved_attrs;
1241 my $tmp_attrs = { %$attrs };
1243 # take off any limits, record_filter is cdbi, and no point of ordering a count
1244 delete $tmp_attrs->{$_} for (qw/select as rows offset order_by record_filter/);
1246 # overwrite the selector (supplied by the storage)
1247 $tmp_attrs->{select} = $rsrc->storage->_count_select ($rsrc, $tmp_attrs);
1248 $tmp_attrs->{as} = 'count';
1250 # read the comment on top of the actual function to see what this does
1251 $tmp_attrs->{from} = $self->_switch_to_inner_join_if_needed (
1252 $tmp_attrs->{from}, $tmp_attrs->{alias}
1255 my $tmp_rs = $rsrc->resultset_class->new($rsrc, $tmp_attrs)->get_column ('count');
1261 # same as above but uses a subquery
1263 sub _count_subq_rs {
1264 my ($self, $attrs) = @_;
1266 my $rsrc = $self->result_source;
1267 $attrs ||= $self->_resolved_attrs_copy;
1269 my $sub_attrs = { %$attrs };
1271 # extra selectors do not go in the subquery and there is no point of ordering it
1272 delete $sub_attrs->{$_} for qw/collapse select _prefetch_select as order_by/;
1274 # if we prefetch, we group_by primary keys only as this is what we would get out
1275 # of the rs via ->next/->all. We DO WANT to clobber old group_by regardless
1276 if ( keys %{$attrs->{collapse}} ) {
1277 $sub_attrs->{group_by} = [ map { "$attrs->{alias}.$_" } ($rsrc->primary_columns) ]
1280 $sub_attrs->{select} = $rsrc->storage->_subq_count_select ($rsrc, $sub_attrs);
1282 # read the comment on top of the actual function to see what this does
1283 $sub_attrs->{from} = $self->_switch_to_inner_join_if_needed (
1284 $sub_attrs->{from}, $sub_attrs->{alias}
1287 # this is so that ordering can be thrown away in things like Top limit
1288 $sub_attrs->{-for_count_only} = 1;
1290 my $sub_rs = $rsrc->resultset_class->new ($rsrc, $sub_attrs);
1293 -alias => 'count_subq',
1294 -source_handle => $rsrc->handle,
1295 count_subq => $sub_rs->as_query,
1298 # the subquery replaces this
1299 delete $attrs->{$_} for qw/where bind collapse group_by having having_bind rows offset/;
1301 return $self->_count_rs ($attrs);
1305 # The DBIC relationship chaining implementation is pretty simple - every
1306 # new related_relationship is pushed onto the {from} stack, and the {select}
1307 # window simply slides further in. This means that when we count somewhere
1308 # in the middle, we got to make sure that everything in the join chain is an
1309 # actual inner join, otherwise the count will come back with unpredictable
1310 # results (a resultset may be generated with _some_ rows regardless of if
1311 # the relation which the $rs currently selects has rows or not). E.g.
1312 # $artist_rs->cds->count - normally generates:
1313 # SELECT COUNT( * ) FROM artist me LEFT JOIN cd cds ON cds.artist = me.artistid
1314 # which actually returns the number of artists * (number of cds || 1)
1316 # So what we do here is crawl {from}, determine if the current alias is at
1317 # the top of the stack, and if not - make sure the chain is inner-joined down
1320 sub _switch_to_inner_join_if_needed {
1321 my ($self, $from, $alias) = @_;
1323 # subqueries and other oddness is naturally not supported
1325 ref $from ne 'ARRAY'
1329 ref $from->[0] ne 'HASH'
1331 ! $from->[0]{-alias}
1333 $from->[0]{-alias} eq $alias
1338 for my $j (@{$from}[1 .. $#$from]) {
1339 if ($j->[0]{-alias} eq $alias) {
1340 $switch_branch = $j->[0]{-join_path};
1345 # something else went wrong
1346 return $from unless $switch_branch;
1348 # So it looks like we will have to switch some stuff around.
1349 # local() is useless here as we will be leaving the scope
1350 # anyway, and deep cloning is just too fucking expensive
1351 # So replace the inner hashref manually
1352 my @new_from = ($from->[0]);
1353 my $sw_idx = { map { $_ => 1 } @$switch_branch };
1355 for my $j (@{$from}[1 .. $#$from]) {
1356 my $jalias = $j->[0]{-alias};
1358 if ($sw_idx->{$jalias}) {
1359 my %attrs = %{$j->[0]};
1360 delete $attrs{-join_type};
1379 =head2 count_literal
1383 =item Arguments: $sql_fragment, @bind_values
1385 =item Return Value: $count
1389 Counts the results in a literal query. Equivalent to calling L</search_literal>
1390 with the passed arguments, then L</count>.
1394 sub count_literal { shift->search_literal(@_)->count; }
1400 =item Arguments: none
1402 =item Return Value: @objects
1406 Returns all elements in the resultset. Called implicitly if the resultset
1407 is returned in list context.
1414 $self->throw_exception("all() doesn't take any arguments, you probably wanted ->search(...)->all()");
1417 return @{ $self->get_cache } if $self->get_cache;
1421 if (keys %{$self->_resolved_attrs->{collapse}}) {
1422 # Using $self->cursor->all is really just an optimisation.
1423 # If we're collapsing has_many prefetches it probably makes
1424 # very little difference, and this is cleaner than hacking
1425 # _construct_object to survive the approach
1426 $self->cursor->reset;
1427 my @row = $self->cursor->next;
1429 push(@obj, $self->_construct_object(@row));
1430 @row = (exists $self->{stashed_row}
1431 ? @{delete $self->{stashed_row}}
1432 : $self->cursor->next);
1435 @obj = map { $self->_construct_object(@$_) } $self->cursor->all;
1438 $self->set_cache(\@obj) if $self->{attrs}{cache};
1447 =item Arguments: none
1449 =item Return Value: $self
1453 Resets the resultset's cursor, so you can iterate through the elements again.
1454 Implicitly resets the storage cursor, so a subsequent L</next> will trigger
1461 delete $self->{_attrs} if exists $self->{_attrs};
1462 $self->{all_cache_position} = 0;
1463 $self->cursor->reset;
1471 =item Arguments: none
1473 =item Return Value: $object?
1477 Resets the resultset and returns an object for the first result (if the
1478 resultset returns anything).
1483 return $_[0]->reset->next;
1489 # Determines whether and what type of subquery is required for the $rs operation.
1490 # If grouping is necessary either supplies its own, or verifies the current one
1491 # After all is done delegates to the proper storage method.
1493 sub _rs_update_delete {
1494 my ($self, $op, $values) = @_;
1496 my $rsrc = $self->result_source;
1498 my $needs_group_by_subq = $self->_has_resolved_attr (qw/collapse group_by -join/);
1499 my $needs_subq = $self->_has_resolved_attr (qw/row offset/);
1501 if ($needs_group_by_subq or $needs_subq) {
1503 # make a new $rs selecting only the PKs (that's all we really need)
1504 my $attrs = $self->_resolved_attrs_copy;
1506 delete $attrs->{$_} for qw/collapse select as/;
1507 $attrs->{columns} = [ map { "$attrs->{alias}.$_" } ($self->result_source->primary_columns) ];
1509 if ($needs_group_by_subq) {
1510 # make sure no group_by was supplied, or if there is one - make sure it matches
1511 # the columns compiled above perfectly. Anything else can not be sanely executed
1512 # on most databases so croak right then and there
1514 if (my $g = $attrs->{group_by}) {
1515 my @current_group_by = map
1516 { $_ =~ /\./ ? $_ : "$attrs->{alias}.$_" }
1521 join ("\x00", sort @current_group_by)
1523 join ("\x00", sort @{$attrs->{columns}} )
1525 $self->throw_exception (
1526 "You have just attempted a $op operation on a resultset which does group_by"
1527 . ' on columns other than the primary keys, while DBIC internally needs to retrieve'
1528 . ' the primary keys in a subselect. All sane RDBMS engines do not support this'
1529 . ' kind of queries. Please retry the operation with a modified group_by or'
1530 . ' without using one at all.'
1535 $attrs->{group_by} = $attrs->{columns};
1539 my $subrs = (ref $self)->new($rsrc, $attrs);
1541 return $self->result_source->storage->_subq_update_delete($subrs, $op, $values);
1544 return $rsrc->storage->$op(
1546 $op eq 'update' ? $values : (),
1547 $self->_cond_for_update_delete,
1553 # _cond_for_update_delete
1555 # update/delete require the condition to be modified to handle
1556 # the differing SQL syntax available. This transforms the $self->{cond}
1557 # appropriately, returning the new condition.
1559 sub _cond_for_update_delete {
1560 my ($self, $full_cond) = @_;
1563 $full_cond ||= $self->{cond};
1564 # No-op. No condition, we're updating/deleting everything
1565 return $cond unless ref $full_cond;
1567 if (ref $full_cond eq 'ARRAY') {
1571 foreach my $key (keys %{$_}) {
1573 $hash{$1} = $_->{$key};
1579 elsif (ref $full_cond eq 'HASH') {
1580 if ((keys %{$full_cond})[0] eq '-and') {
1582 my @cond = @{$full_cond->{-and}};
1583 for (my $i = 0; $i < @cond; $i++) {
1584 my $entry = $cond[$i];
1586 if (ref $entry eq 'HASH') {
1587 $hash = $self->_cond_for_update_delete($entry);
1590 $entry =~ /([^.]+)$/;
1591 $hash->{$1} = $cond[++$i];
1593 push @{$cond->{-and}}, $hash;
1597 foreach my $key (keys %{$full_cond}) {
1599 $cond->{$1} = $full_cond->{$key};
1604 $self->throw_exception("Can't update/delete on resultset with condition unless hash or array");
1615 =item Arguments: \%values
1617 =item Return Value: $storage_rv
1621 Sets the specified columns in the resultset to the supplied values in a
1622 single query. Return value will be true if the update succeeded or false
1623 if no records were updated; exact type of success value is storage-dependent.
1628 my ($self, $values) = @_;
1629 $self->throw_exception('Values for update must be a hash')
1630 unless ref $values eq 'HASH';
1632 return $self->_rs_update_delete ('update', $values);
1639 =item Arguments: \%values
1641 =item Return Value: 1
1645 Fetches all objects and updates them one at a time. Note that C<update_all>
1646 will run DBIC cascade triggers, while L</update> will not.
1651 my ($self, $values) = @_;
1652 $self->throw_exception('Values for update_all must be a hash')
1653 unless ref $values eq 'HASH';
1654 foreach my $obj ($self->all) {
1655 $obj->set_columns($values)->update;
1664 =item Arguments: none
1666 =item Return Value: $storage_rv
1670 Deletes the contents of the resultset from its result source. Note that this
1671 will not run DBIC cascade triggers. See L</delete_all> if you need triggers
1672 to run. See also L<DBIx::Class::Row/delete>.
1674 Return value will be the amount of rows deleted; exact type of return value
1675 is storage-dependent.
1681 $self->throw_exception('delete does not accept any arguments')
1684 return $self->_rs_update_delete ('delete');
1691 =item Arguments: none
1693 =item Return Value: 1
1697 Fetches all objects and deletes them one at a time. Note that C<delete_all>
1698 will run DBIC cascade triggers, while L</delete> will not.
1704 $self->throw_exception('delete_all does not accept any arguments')
1707 $_->delete for $self->all;
1715 =item Arguments: \@data;
1719 Accepts either an arrayref of hashrefs or alternatively an arrayref of arrayrefs.
1720 For the arrayref of hashrefs style each hashref should be a structure suitable
1721 forsubmitting to a $resultset->create(...) method.
1723 In void context, C<insert_bulk> in L<DBIx::Class::Storage::DBI> is used
1724 to insert the data, as this is a faster method.
1726 Otherwise, each set of data is inserted into the database using
1727 L<DBIx::Class::ResultSet/create>, and the resulting objects are
1728 accumulated into an array. The array itself, or an array reference
1729 is returned depending on scalar or list context.
1731 Example: Assuming an Artist Class that has many CDs Classes relating:
1733 my $Artist_rs = $schema->resultset("Artist");
1735 ## Void Context Example
1736 $Artist_rs->populate([
1737 { artistid => 4, name => 'Manufactured Crap', cds => [
1738 { title => 'My First CD', year => 2006 },
1739 { title => 'Yet More Tweeny-Pop crap', year => 2007 },
1742 { artistid => 5, name => 'Angsty-Whiny Girl', cds => [
1743 { title => 'My parents sold me to a record company' ,year => 2005 },
1744 { title => 'Why Am I So Ugly?', year => 2006 },
1745 { title => 'I Got Surgery and am now Popular', year => 2007 }
1750 ## Array Context Example
1751 my ($ArtistOne, $ArtistTwo, $ArtistThree) = $Artist_rs->populate([
1752 { name => "Artist One"},
1753 { name => "Artist Two"},
1754 { name => "Artist Three", cds=> [
1755 { title => "First CD", year => 2007},
1756 { title => "Second CD", year => 2008},
1760 print $ArtistOne->name; ## response is 'Artist One'
1761 print $ArtistThree->cds->count ## reponse is '2'
1763 For the arrayref of arrayrefs style, the first element should be a list of the
1764 fieldsnames to which the remaining elements are rows being inserted. For
1767 $Arstist_rs->populate([
1768 [qw/artistid name/],
1769 [100, 'A Formally Unknown Singer'],
1770 [101, 'A singer that jumped the shark two albums ago'],
1771 [102, 'An actually cool singer.'],
1774 Please note an important effect on your data when choosing between void and
1775 wantarray context. Since void context goes straight to C<insert_bulk> in
1776 L<DBIx::Class::Storage::DBI> this will skip any component that is overriding
1777 C<insert>. So if you are using something like L<DBIx-Class-UUIDColumns> to
1778 create primary keys for you, you will find that your PKs are empty. In this
1779 case you will have to use the wantarray context in order to create those
1785 my $self = shift @_;
1786 my $data = ref $_[0][0] eq 'HASH'
1787 ? $_[0] : ref $_[0][0] eq 'ARRAY' ? $self->_normalize_populate_args($_[0]) :
1788 $self->throw_exception('Populate expects an arrayref of hashes or arrayref of arrayrefs');
1790 if(defined wantarray) {
1792 foreach my $item (@$data) {
1793 push(@created, $self->create($item));
1795 return wantarray ? @created : \@created;
1797 my ($first, @rest) = @$data;
1799 my @names = grep {!ref $first->{$_}} keys %$first;
1800 my @rels = grep { $self->result_source->has_relationship($_) } keys %$first;
1801 my @pks = $self->result_source->primary_columns;
1803 ## do the belongs_to relationships
1804 foreach my $index (0..$#$data) {
1806 # delegate to create() for any dataset without primary keys with specified relationships
1807 if (grep { !defined $data->[$index]->{$_} } @pks ) {
1809 if (grep { ref $data->[$index]{$r} eq $_ } qw/HASH ARRAY/) { # a related set must be a HASH or AoH
1810 my @ret = $self->populate($data);
1816 foreach my $rel (@rels) {
1817 next unless ref $data->[$index]->{$rel} eq "HASH";
1818 my $result = $self->related_resultset($rel)->create($data->[$index]->{$rel});
1819 my ($reverse) = keys %{$self->result_source->reverse_relationship_info($rel)};
1820 my $related = $result->result_source->_resolve_condition(
1821 $result->result_source->relationship_info($reverse)->{cond},
1826 delete $data->[$index]->{$rel};
1827 $data->[$index] = {%{$data->[$index]}, %$related};
1829 push @names, keys %$related if $index == 0;
1833 ## do bulk insert on current row
1834 my @values = map { [ @$_{@names} ] } @$data;
1836 $self->result_source->storage->insert_bulk(
1837 $self->result_source,
1842 ## do the has_many relationships
1843 foreach my $item (@$data) {
1845 foreach my $rel (@rels) {
1846 next unless $item->{$rel} && ref $item->{$rel} eq "ARRAY";
1848 my $parent = $self->find(map {{$_=>$item->{$_}} } @pks)
1849 || $self->throw_exception('Cannot find the relating object.');
1851 my $child = $parent->$rel;
1853 my $related = $child->result_source->_resolve_condition(
1854 $parent->result_source->relationship_info($rel)->{cond},
1859 my @rows_to_add = ref $item->{$rel} eq 'ARRAY' ? @{$item->{$rel}} : ($item->{$rel});
1860 my @populate = map { {%$_, %$related} } @rows_to_add;
1862 $child->populate( \@populate );
1868 =head2 _normalize_populate_args ($args)
1870 Private method used by L</populate> to normalize its incoming arguments. Factored
1871 out in case you want to subclass and accept new argument structures to the
1872 L</populate> method.
1876 sub _normalize_populate_args {
1877 my ($self, $data) = @_;
1878 my @names = @{shift(@$data)};
1879 my @results_to_create;
1880 foreach my $datum (@$data) {
1881 my %result_to_create;
1882 foreach my $index (0..$#names) {
1883 $result_to_create{$names[$index]} = $$datum[$index];
1885 push @results_to_create, \%result_to_create;
1887 return \@results_to_create;
1894 =item Arguments: none
1896 =item Return Value: $pager
1900 Return Value a L<Data::Page> object for the current resultset. Only makes
1901 sense for queries with a C<page> attribute.
1903 To get the full count of entries for a paged resultset, call
1904 C<total_entries> on the L<Data::Page> object.
1911 return $self->{pager} if $self->{pager};
1913 my $attrs = $self->{attrs};
1914 $self->throw_exception("Can't create pager for non-paged rs")
1915 unless $self->{attrs}{page};
1916 $attrs->{rows} ||= 10;
1918 # throw away the paging flags and re-run the count (possibly
1919 # with a subselect) to get the real total count
1920 my $count_attrs = { %$attrs };
1921 delete $count_attrs->{$_} for qw/rows offset page pager/;
1922 my $total_count = (ref $self)->new($self->result_source, $count_attrs)->count;
1924 return $self->{pager} = Data::Page->new(
1927 $self->{attrs}{page}
1935 =item Arguments: $page_number
1937 =item Return Value: $rs
1941 Returns a resultset for the $page_number page of the resultset on which page
1942 is called, where each page contains a number of rows equal to the 'rows'
1943 attribute set on the resultset (10 by default).
1948 my ($self, $page) = @_;
1949 return (ref $self)->new($self->result_source, { %{$self->{attrs}}, page => $page });
1956 =item Arguments: \%vals
1958 =item Return Value: $rowobject
1962 Creates a new row object in the resultset's result class and returns
1963 it. The row is not inserted into the database at this point, call
1964 L<DBIx::Class::Row/insert> to do that. Calling L<DBIx::Class::Row/in_storage>
1965 will tell you whether the row object has been inserted or not.
1967 Passes the hashref of input on to L<DBIx::Class::Row/new>.
1972 my ($self, $values) = @_;
1973 $self->throw_exception( "new_result needs a hash" )
1974 unless (ref $values eq 'HASH');
1977 my $alias = $self->{attrs}{alias};
1980 defined $self->{cond}
1981 && $self->{cond} eq $DBIx::Class::ResultSource::UNRESOLVABLE_CONDITION
1983 %new = %{ $self->{attrs}{related_objects} || {} }; # nothing might have been inserted yet
1984 $new{-from_resultset} = [ keys %new ] if keys %new;
1986 $self->throw_exception(
1987 "Can't abstract implicit construct, condition not a hash"
1988 ) if ($self->{cond} && !(ref $self->{cond} eq 'HASH'));
1990 my $collapsed_cond = (
1992 ? $self->_collapse_cond($self->{cond})
1996 # precendence must be given to passed values over values inherited from
1997 # the cond, so the order here is important.
1998 my %implied = %{$self->_remove_alias($collapsed_cond, $alias)};
1999 while( my($col,$value) = each %implied ){
2000 if(ref($value) eq 'HASH' && keys(%$value) && (keys %$value)[0] eq '='){
2001 $new{$col} = $value->{'='};
2004 $new{$col} = $value if $self->_is_deterministic_value($value);
2010 %{ $self->_remove_alias($values, $alias) },
2011 -source_handle => $self->_source_handle,
2012 -result_source => $self->result_source, # DO NOT REMOVE THIS, REQUIRED
2015 return $self->result_class->new(\%new);
2018 # _is_deterministic_value
2020 # Make an effor to strip non-deterministic values from the condition,
2021 # to make sure new_result chokes less
2023 sub _is_deterministic_value {
2026 my $ref_type = ref $value;
2027 return 1 if $ref_type eq '' || $ref_type eq 'SCALAR';
2028 return 1 if Scalar::Util::blessed($value);
2032 # _has_resolved_attr
2034 # determines if the resultset defines at least one
2035 # of the attributes supplied
2037 # used to determine if a subquery is neccessary
2039 # supports some virtual attributes:
2041 # This will scan for any joins being present on the resultset.
2042 # It is not a mere key-search but a deep inspection of {from}
2045 sub _has_resolved_attr {
2046 my ($self, @attr_names) = @_;
2048 my $attrs = $self->_resolved_attrs;
2052 for my $n (@attr_names) {
2053 if (grep { $n eq $_ } (qw/-join/) ) {
2054 $extra_checks{$n}++;
2058 my $attr = $attrs->{$n};
2060 next if not defined $attr;
2062 if (ref $attr eq 'HASH') {
2063 return 1 if keys %$attr;
2065 elsif (ref $attr eq 'ARRAY') {
2073 # a resolved join is expressed as a multi-level from
2075 $extra_checks{-join}
2077 ref $attrs->{from} eq 'ARRAY'
2079 @{$attrs->{from}} > 1
2087 # Recursively collapse the condition.
2089 sub _collapse_cond {
2090 my ($self, $cond, $collapsed) = @_;
2094 if (ref $cond eq 'ARRAY') {
2095 foreach my $subcond (@$cond) {
2096 next unless ref $subcond; # -or
2097 $collapsed = $self->_collapse_cond($subcond, $collapsed);
2100 elsif (ref $cond eq 'HASH') {
2101 if (keys %$cond and (keys %$cond)[0] eq '-and') {
2102 foreach my $subcond (@{$cond->{-and}}) {
2103 $collapsed = $self->_collapse_cond($subcond, $collapsed);
2107 foreach my $col (keys %$cond) {
2108 my $value = $cond->{$col};
2109 $collapsed->{$col} = $value;
2119 # Remove the specified alias from the specified query hash. A copy is made so
2120 # the original query is not modified.
2123 my ($self, $query, $alias) = @_;
2125 my %orig = %{ $query || {} };
2128 foreach my $key (keys %orig) {
2130 $unaliased{$key} = $orig{$key};
2133 $unaliased{$1} = $orig{$key}
2134 if $key =~ m/^(?:\Q$alias\E\.)?([^.]+)$/;
2140 =head2 as_query (EXPERIMENTAL)
2144 =item Arguments: none
2146 =item Return Value: \[ $sql, @bind ]
2150 Returns the SQL query and bind vars associated with the invocant.
2152 This is generally used as the RHS for a subquery.
2154 B<NOTE>: This feature is still experimental.
2161 my $attrs = $self->_resolved_attrs_copy;
2166 # my ($sql, \@bind, \%dbi_bind_attrs) = _select_args_to_query (...)
2167 # $sql also has no wrapping parenthesis in list ctx
2169 my $sqlbind = $self->result_source->storage
2170 ->_select_args_to_query ($attrs->{from}, $attrs->{select}, $attrs->{where}, $attrs);
2179 =item Arguments: \%vals, \%attrs?
2181 =item Return Value: $rowobject
2185 my $artist = $schema->resultset('Artist')->find_or_new(
2186 { artist => 'fred' }, { key => 'artists' });
2188 $cd->cd_to_producer->find_or_new({ producer => $producer },
2189 { key => 'primary });
2191 Find an existing record from this resultset, based on its primary
2192 key, or a unique constraint. If none exists, instantiate a new result
2193 object and return it. The object will not be saved into your storage
2194 until you call L<DBIx::Class::Row/insert> on it.
2196 You most likely want this method when looking for existing rows using
2197 a unique constraint that is not the primary key, or looking for
2200 If you want objects to be saved immediately, use L</find_or_create>
2203 B<Note>: Take care when using C<find_or_new> with a table having
2204 columns with default values that you intend to be automatically
2205 supplied by the database (e.g. an auto_increment primary key column).
2206 In normal usage, the value of such columns should NOT be included at
2207 all in the call to C<find_or_new>, even when set to C<undef>.
2213 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
2214 my $hash = ref $_[0] eq 'HASH' ? shift : {@_};
2215 if (keys %$hash and my $row = $self->find($hash, $attrs) ) {
2218 return $self->new_result($hash);
2225 =item Arguments: \%vals
2227 =item Return Value: a L<DBIx::Class::Row> $object
2231 Attempt to create a single new row or a row with multiple related rows
2232 in the table represented by the resultset (and related tables). This
2233 will not check for duplicate rows before inserting, use
2234 L</find_or_create> to do that.
2236 To create one row for this resultset, pass a hashref of key/value
2237 pairs representing the columns of the table and the values you wish to
2238 store. If the appropriate relationships are set up, foreign key fields
2239 can also be passed an object representing the foreign row, and the
2240 value will be set to its primary key.
2242 To create related objects, pass a hashref of related-object column values
2243 B<keyed on the relationship name>. If the relationship is of type C<multi>
2244 (L<DBIx::Class::Relationship/has_many>) - pass an arrayref of hashrefs.
2245 The process will correctly identify columns holding foreign keys, and will
2246 transparrently populate them from the keys of the corresponding relation.
2247 This can be applied recursively, and will work correctly for a structure
2248 with an arbitrary depth and width, as long as the relationships actually
2249 exists and the correct column data has been supplied.
2252 Instead of hashrefs of plain related data (key/value pairs), you may
2253 also pass new or inserted objects. New objects (not inserted yet, see
2254 L</new>), will be inserted into their appropriate tables.
2256 Effectively a shortcut for C<< ->new_result(\%vals)->insert >>.
2258 Example of creating a new row.
2260 $person_rs->create({
2261 name=>"Some Person",
2262 email=>"somebody@someplace.com"
2265 Example of creating a new row and also creating rows in a related C<has_many>
2266 or C<has_one> resultset. Note Arrayref.
2269 { artistid => 4, name => 'Manufactured Crap', cds => [
2270 { title => 'My First CD', year => 2006 },
2271 { title => 'Yet More Tweeny-Pop crap', year => 2007 },
2276 Example of creating a new row and also creating a row in a related
2277 C<belongs_to>resultset. Note Hashref.
2280 title=>"Music for Silly Walks",
2283 name=>"Silly Musician",
2291 When subclassing ResultSet never attempt to override this method. Since
2292 it is a simple shortcut for C<< $self->new_result($attrs)->insert >>, a
2293 lot of the internals simply never call it, so your override will be
2294 bypassed more often than not. Override either L<new|DBIx::Class::Row/new>
2295 or L<insert|DBIx::Class::Row/insert> depending on how early in the
2296 L</create> process you need to intervene.
2303 my ($self, $attrs) = @_;
2304 $self->throw_exception( "create needs a hashref" )
2305 unless ref $attrs eq 'HASH';
2306 return $self->new_result($attrs)->insert;
2309 =head2 find_or_create
2313 =item Arguments: \%vals, \%attrs?
2315 =item Return Value: $rowobject
2319 $cd->cd_to_producer->find_or_create({ producer => $producer },
2320 { key => 'primary' });
2322 Tries to find a record based on its primary key or unique constraints; if none
2323 is found, creates one and returns that instead.
2325 my $cd = $schema->resultset('CD')->find_or_create({
2327 artist => 'Massive Attack',
2328 title => 'Mezzanine',
2332 Also takes an optional C<key> attribute, to search by a specific key or unique
2333 constraint. For example:
2335 my $cd = $schema->resultset('CD')->find_or_create(
2337 artist => 'Massive Attack',
2338 title => 'Mezzanine',
2340 { key => 'cd_artist_title' }
2343 B<Note>: Because find_or_create() reads from the database and then
2344 possibly inserts based on the result, this method is subject to a race
2345 condition. Another process could create a record in the table after
2346 the find has completed and before the create has started. To avoid
2347 this problem, use find_or_create() inside a transaction.
2349 B<Note>: Take care when using C<find_or_create> with a table having
2350 columns with default values that you intend to be automatically
2351 supplied by the database (e.g. an auto_increment primary key column).
2352 In normal usage, the value of such columns should NOT be included at
2353 all in the call to C<find_or_create>, even when set to C<undef>.
2355 See also L</find> and L</update_or_create>. For information on how to declare
2356 unique constraints, see L<DBIx::Class::ResultSource/add_unique_constraint>.
2360 sub find_or_create {
2362 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
2363 my $hash = ref $_[0] eq 'HASH' ? shift : {@_};
2364 if (keys %$hash and my $row = $self->find($hash, $attrs) ) {
2367 return $self->create($hash);
2370 =head2 update_or_create
2374 =item Arguments: \%col_values, { key => $unique_constraint }?
2376 =item Return Value: $rowobject
2380 $resultset->update_or_create({ col => $val, ... });
2382 First, searches for an existing row matching one of the unique constraints
2383 (including the primary key) on the source of this resultset. If a row is
2384 found, updates it with the other given column values. Otherwise, creates a new
2387 Takes an optional C<key> attribute to search on a specific unique constraint.
2390 # In your application
2391 my $cd = $schema->resultset('CD')->update_or_create(
2393 artist => 'Massive Attack',
2394 title => 'Mezzanine',
2397 { key => 'cd_artist_title' }
2400 $cd->cd_to_producer->update_or_create({
2401 producer => $producer,
2408 If no C<key> is specified, it searches on all unique constraints defined on the
2409 source, including the primary key.
2411 If the C<key> is specified as C<primary>, it searches only on the primary key.
2413 See also L</find> and L</find_or_create>. For information on how to declare
2414 unique constraints, see L<DBIx::Class::ResultSource/add_unique_constraint>.
2416 B<Note>: Take care when using C<update_or_create> with a table having
2417 columns with default values that you intend to be automatically
2418 supplied by the database (e.g. an auto_increment primary key column).
2419 In normal usage, the value of such columns should NOT be included at
2420 all in the call to C<update_or_create>, even when set to C<undef>.
2424 sub update_or_create {
2426 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
2427 my $cond = ref $_[0] eq 'HASH' ? shift : {@_};
2429 my $row = $self->find($cond, $attrs);
2431 $row->update($cond);
2435 return $self->create($cond);
2438 =head2 update_or_new
2442 =item Arguments: \%col_values, { key => $unique_constraint }?
2444 =item Return Value: $rowobject
2448 $resultset->update_or_new({ col => $val, ... });
2450 First, searches for an existing row matching one of the unique constraints
2451 (including the primary key) on the source of this resultset. If a row is
2452 found, updates it with the other given column values. Otherwise, instantiate
2453 a new result object and return it. The object will not be saved into your storage
2454 until you call L<DBIx::Class::Row/insert> on it.
2456 Takes an optional C<key> attribute to search on a specific unique constraint.
2459 # In your application
2460 my $cd = $schema->resultset('CD')->update_or_new(
2462 artist => 'Massive Attack',
2463 title => 'Mezzanine',
2466 { key => 'cd_artist_title' }
2469 if ($cd->in_storage) {
2470 # the cd was updated
2473 # the cd is not yet in the database, let's insert it
2477 B<Note>: Take care when using C<update_or_new> with a table having
2478 columns with default values that you intend to be automatically
2479 supplied by the database (e.g. an auto_increment primary key column).
2480 In normal usage, the value of such columns should NOT be included at
2481 all in the call to C<update_or_new>, even when set to C<undef>.
2483 See also L</find>, L</find_or_create> and L</find_or_new>.
2489 my $attrs = ( @_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {} );
2490 my $cond = ref $_[0] eq 'HASH' ? shift : {@_};
2492 my $row = $self->find( $cond, $attrs );
2493 if ( defined $row ) {
2494 $row->update($cond);
2498 return $self->new_result($cond);
2505 =item Arguments: none
2507 =item Return Value: \@cache_objects?
2511 Gets the contents of the cache for the resultset, if the cache is set.
2513 The cache is populated either by using the L</prefetch> attribute to
2514 L</search> or by calling L</set_cache>.
2526 =item Arguments: \@cache_objects
2528 =item Return Value: \@cache_objects
2532 Sets the contents of the cache for the resultset. Expects an arrayref
2533 of objects of the same class as those produced by the resultset. Note that
2534 if the cache is set the resultset will return the cached objects rather
2535 than re-querying the database even if the cache attr is not set.
2537 The contents of the cache can also be populated by using the
2538 L</prefetch> attribute to L</search>.
2543 my ( $self, $data ) = @_;
2544 $self->throw_exception("set_cache requires an arrayref")
2545 if defined($data) && (ref $data ne 'ARRAY');
2546 $self->{all_cache} = $data;
2553 =item Arguments: none
2555 =item Return Value: []
2559 Clears the cache for the resultset.
2564 shift->set_cache(undef);
2567 =head2 related_resultset
2571 =item Arguments: $relationship_name
2573 =item Return Value: $resultset
2577 Returns a related resultset for the supplied relationship name.
2579 $artist_rs = $schema->resultset('CD')->related_resultset('Artist');
2583 sub related_resultset {
2584 my ($self, $rel) = @_;
2586 $self->{related_resultsets} ||= {};
2587 return $self->{related_resultsets}{$rel} ||= do {
2588 my $rel_info = $self->result_source->relationship_info($rel);
2590 $self->throw_exception(
2591 "search_related: result source '" . $self->result_source->source_name .
2592 "' has no such relationship $rel")
2595 my ($from,$seen) = $self->_chain_relationship($rel);
2597 my $join_count = $seen->{$rel};
2598 my $alias = ($join_count > 1 ? join('_', $rel, $join_count) : $rel);
2600 #XXX - temp fix for result_class bug. There likely is a more elegant fix -groditi
2601 my %attrs = %{$self->{attrs}||{}};
2602 delete @attrs{qw(result_class alias)};
2606 if (my $cache = $self->get_cache) {
2607 if ($cache->[0] && $cache->[0]->related_resultset($rel)->get_cache) {
2608 $new_cache = [ map { @{$_->related_resultset($rel)->get_cache} }
2613 my $rel_source = $self->result_source->related_source($rel);
2617 # The reason we do this now instead of passing the alias to the
2618 # search_rs below is that if you wrap/overload resultset on the
2619 # source you need to know what alias it's -going- to have for things
2620 # to work sanely (e.g. RestrictWithObject wants to be able to add
2621 # extra query restrictions, and these may need to be $alias.)
2623 my $attrs = $rel_source->resultset_attributes;
2624 local $attrs->{alias} = $alias;
2626 $rel_source->resultset
2634 where => $self->{cond},
2639 $new->set_cache($new_cache) if $new_cache;
2644 =head2 current_source_alias
2648 =item Arguments: none
2650 =item Return Value: $source_alias
2654 Returns the current table alias for the result source this resultset is built
2655 on, that will be used in the SQL query. Usually it is C<me>.
2657 Currently the source alias that refers to the result set returned by a
2658 L</search>/L</find> family method depends on how you got to the resultset: it's
2659 C<me> by default, but eg. L</search_related> aliases it to the related result
2660 source name (and keeps C<me> referring to the original result set). The long
2661 term goal is to make L<DBIx::Class> always alias the current resultset as C<me>
2662 (and make this method unnecessary).
2664 Thus it's currently necessary to use this method in predefined queries (see
2665 L<DBIx::Class::Manual::Cookbook/Predefined searches>) when referring to the
2666 source alias of the current result set:
2668 # in a result set class
2670 my ($self, $user) = @_;
2672 my $me = $self->current_source_alias;
2674 return $self->search(
2675 "$me.modified" => $user->id,
2681 sub current_source_alias {
2684 return ($self->{attrs} || {})->{alias} || 'me';
2687 # This code is called by search_related, and makes sure there
2688 # is clear separation between the joins before, during, and
2689 # after the relationship. This information is needed later
2690 # in order to properly resolve prefetch aliases (any alias
2691 # with a relation_chain_depth less than the depth of the
2692 # current prefetch is not considered)
2694 # The increments happen in 1/2s to make it easier to correlate the
2695 # join depth with the join path. An integer means a relationship
2696 # specified via a search_related, whereas a fraction means an added
2697 # join/prefetch via attributes
2698 sub _chain_relationship {
2699 my ($self, $rel) = @_;
2700 my $source = $self->result_source;
2701 my $attrs = $self->{attrs};
2707 -source_handle => $source->handle,
2708 -alias => $attrs->{alias},
2709 $attrs->{alias} => $source->from,
2713 my $seen = { %{$attrs->{seen_join} || {} } };
2714 my $jpath = ($attrs->{seen_join} && keys %{$attrs->{seen_join}})
2715 ? $from->[-1][0]{-join_path}
2719 # we need to take the prefetch the attrs into account before we
2720 # ->_resolve_join as otherwise they get lost - captainL
2721 my $merged = $self->_merge_attr( $attrs->{join}, $attrs->{prefetch} );
2723 my @requested_joins = $source->_resolve_join(
2730 push @$from, @requested_joins;
2732 $seen->{-relation_chain_depth} += 0.5;
2734 # if $self already had a join/prefetch specified on it, the requested
2735 # $rel might very well be already included. What we do in this case
2736 # is effectively a no-op (except that we bump up the chain_depth on
2737 # the join in question so we could tell it *is* the search_related)
2741 # we consider the last one thus reverse
2742 for my $j (reverse @requested_joins) {
2743 if ($rel eq $j->[0]{-join_path}[-1]) {
2744 $j->[0]{-relation_chain_depth} += 0.5;
2750 # alternative way to scan the entire chain - not backwards compatible
2751 # for my $j (reverse @$from) {
2752 # next unless ref $j eq 'ARRAY';
2753 # if ($j->[0]{-join_path} && $j->[0]{-join_path}[-1] eq $rel) {
2754 # $j->[0]{-relation_chain_depth} += 0.5;
2755 # $already_joined++;
2760 unless ($already_joined) {
2761 push @$from, $source->_resolve_join(
2769 $seen->{-relation_chain_depth} += 0.5;
2771 return ($from,$seen);
2774 # too many times we have to do $attrs = { %{$self->_resolved_attrs} }
2775 sub _resolved_attrs_copy {
2777 return { %{$self->_resolved_attrs (@_)} };
2780 sub _resolved_attrs {
2782 return $self->{_attrs} if $self->{_attrs};
2784 my $attrs = { %{ $self->{attrs} || {} } };
2785 my $source = $self->result_source;
2786 my $alias = $attrs->{alias};
2788 $attrs->{columns} ||= delete $attrs->{cols} if exists $attrs->{cols};
2791 # build columns (as long as select isn't set) into a set of as/select hashes
2792 unless ( $attrs->{select} ) {
2794 my @cols = ( ref($attrs->{columns}) eq 'ARRAY' )
2795 ? @{ delete $attrs->{columns}}
2797 ( delete $attrs->{columns} )
2799 $source->storage->_order_select_columns(
2801 [ $source->columns ],
2807 ( ref($_) eq 'HASH' )
2811 /^\Q${alias}.\E(.+)$/
2825 # add the additional columns on
2826 foreach ( 'include_columns', '+columns' ) {
2827 push @colbits, map {
2828 ( ref($_) eq 'HASH' )
2830 : { ( split( /\./, $_ ) )[-1] => ( /\./ ? $_ : "${alias}.$_" ) }
2831 } ( ref($attrs->{$_}) eq 'ARRAY' ) ? @{ delete $attrs->{$_} } : delete $attrs->{$_} if ( $attrs->{$_} );
2834 # start with initial select items
2835 if ( $attrs->{select} ) {
2837 ( ref $attrs->{select} eq 'ARRAY' )
2838 ? [ @{ $attrs->{select} } ]
2839 : [ $attrs->{select} ];
2843 ref $attrs->{as} eq 'ARRAY'
2844 ? [ @{ $attrs->{as} } ]
2847 : [ map { m/^\Q${alias}.\E(.+)$/ ? $1 : $_ } @{ $attrs->{select} } ]
2852 # otherwise we intialise select & as to empty
2853 $attrs->{select} = [];
2857 # now add colbits to select/as
2858 push( @{ $attrs->{select} }, map { values( %{$_} ) } @colbits );
2859 push( @{ $attrs->{as} }, map { keys( %{$_} ) } @colbits );
2862 if ( $adds = delete $attrs->{'+select'} ) {
2863 $adds = [$adds] unless ref $adds eq 'ARRAY';
2865 @{ $attrs->{select} },
2866 map { /\./ || ref $_ ? $_ : "${alias}.$_" } @$adds
2869 if ( $adds = delete $attrs->{'+as'} ) {
2870 $adds = [$adds] unless ref $adds eq 'ARRAY';
2871 push( @{ $attrs->{as} }, @$adds );
2874 $attrs->{from} ||= [ {
2875 -source_handle => $source->handle,
2876 -alias => $self->{attrs}{alias},
2877 $self->{attrs}{alias} => $source->from,
2880 if ( $attrs->{join} || $attrs->{prefetch} ) {
2882 $self->throw_exception ('join/prefetch can not be used with a custom {from}')
2883 if ref $attrs->{from} ne 'ARRAY';
2885 my $join = delete $attrs->{join} || {};
2887 if ( defined $attrs->{prefetch} ) {
2888 $join = $self->_merge_attr( $join, $attrs->{prefetch} );
2891 $attrs->{from} = # have to copy here to avoid corrupting the original
2893 @{ $attrs->{from} },
2894 $source->_resolve_join(
2897 { %{ $attrs->{seen_join} || {} } },
2898 ($attrs->{seen_join} && keys %{$attrs->{seen_join}})
2899 ? $attrs->{from}[-1][0]{-join_path}
2906 if ( defined $attrs->{order_by} ) {
2907 $attrs->{order_by} = (
2908 ref( $attrs->{order_by} ) eq 'ARRAY'
2909 ? [ @{ $attrs->{order_by} } ]
2910 : [ $attrs->{order_by} || () ]
2914 if ($attrs->{group_by} and ref $attrs->{group_by} ne 'ARRAY') {
2915 $attrs->{group_by} = [ $attrs->{group_by} ];
2918 # generate the distinct induced group_by early, as prefetch will be carried via a
2919 # subquery (since a group_by is present)
2920 if (delete $attrs->{distinct}) {
2921 if ($attrs->{group_by}) {
2922 carp ("Useless use of distinct on a grouped resultset ('distinct' is ignored when a 'group_by' is present)");
2925 $attrs->{group_by} = [ grep { !ref($_) || (ref($_) ne 'HASH') } @{$attrs->{select}} ];
2929 $attrs->{collapse} ||= {};
2930 if ( my $prefetch = delete $attrs->{prefetch} ) {
2931 $prefetch = $self->_merge_attr( {}, $prefetch );
2933 my $prefetch_ordering = [];
2935 my $join_map = $self->_joinpath_aliases ($attrs->{from}, $attrs->{seen_join});
2938 $source->_resolve_prefetch( $prefetch, $alias, $join_map, $prefetch_ordering, $attrs->{collapse} );
2940 # we need to somehow mark which columns came from prefetch
2941 $attrs->{_prefetch_select} = [ map { $_->[0] } @prefetch ];
2943 push @{ $attrs->{select} }, @{$attrs->{_prefetch_select}};
2944 push @{ $attrs->{as} }, (map { $_->[1] } @prefetch);
2946 push( @{$attrs->{order_by}}, @$prefetch_ordering );
2947 $attrs->{_collapse_order_by} = \@$prefetch_ordering;
2950 # if both page and offset are specified, produce a combined offset
2951 # even though it doesn't make much sense, this is what pre 081xx has
2953 if (my $page = delete $attrs->{page}) {
2955 ($attrs->{rows} * ($page - 1))
2957 ($attrs->{offset} || 0)
2961 return $self->{_attrs} = $attrs;
2964 sub _joinpath_aliases {
2965 my ($self, $fromspec, $seen) = @_;
2968 return $paths unless ref $fromspec eq 'ARRAY';
2970 my $cur_depth = $seen->{-relation_chain_depth} || 0;
2972 if (int ($cur_depth) != $cur_depth) {
2973 $self->throw_exception ("-relation_chain_depth is not an integer, something went horribly wrong ($cur_depth)");
2976 for my $j (@$fromspec) {
2978 next if ref $j ne 'ARRAY';
2979 next if ($j->[0]{-relation_chain_depth} || 0) < $cur_depth;
2981 my $jpath = $j->[0]{-join_path};
2984 $p = $p->{$_} ||= {} for @{$jpath}[$cur_depth .. $#$jpath];
2985 push @{$p->{-join_aliases} }, $j->[0]{-alias};
2992 my ($self, $attr) = @_;
2994 if (ref $attr eq 'HASH') {
2995 return $self->_rollout_hash($attr);
2996 } elsif (ref $attr eq 'ARRAY') {
2997 return $self->_rollout_array($attr);
3003 sub _rollout_array {
3004 my ($self, $attr) = @_;
3007 foreach my $element (@{$attr}) {
3008 if (ref $element eq 'HASH') {
3009 push( @rolled_array, @{ $self->_rollout_hash( $element ) } );
3010 } elsif (ref $element eq 'ARRAY') {
3011 # XXX - should probably recurse here
3012 push( @rolled_array, @{$self->_rollout_array($element)} );
3014 push( @rolled_array, $element );
3017 return \@rolled_array;
3021 my ($self, $attr) = @_;
3024 foreach my $key (keys %{$attr}) {
3025 push( @rolled_array, { $key => $attr->{$key} } );
3027 return \@rolled_array;
3030 sub _calculate_score {
3031 my ($self, $a, $b) = @_;
3033 if (defined $a xor defined $b) {
3036 elsif (not defined $a) {
3040 if (ref $b eq 'HASH') {
3041 my ($b_key) = keys %{$b};
3042 if (ref $a eq 'HASH') {
3043 my ($a_key) = keys %{$a};
3044 if ($a_key eq $b_key) {
3045 return (1 + $self->_calculate_score( $a->{$a_key}, $b->{$b_key} ));
3050 return ($a eq $b_key) ? 1 : 0;
3053 if (ref $a eq 'HASH') {
3054 my ($a_key) = keys %{$a};
3055 return ($b eq $a_key) ? 1 : 0;
3057 return ($b eq $a) ? 1 : 0;
3063 my ($self, $orig, $import) = @_;
3065 return $import unless defined($orig);
3066 return $orig unless defined($import);
3068 $orig = $self->_rollout_attr($orig);
3069 $import = $self->_rollout_attr($import);
3072 foreach my $import_element ( @{$import} ) {
3073 # find best candidate from $orig to merge $b_element into
3074 my $best_candidate = { position => undef, score => 0 }; my $position = 0;
3075 foreach my $orig_element ( @{$orig} ) {
3076 my $score = $self->_calculate_score( $orig_element, $import_element );
3077 if ($score > $best_candidate->{score}) {
3078 $best_candidate->{position} = $position;
3079 $best_candidate->{score} = $score;
3083 my ($import_key) = ( ref $import_element eq 'HASH' ) ? keys %{$import_element} : ($import_element);
3085 if ($best_candidate->{score} == 0 || exists $seen_keys->{$import_key}) {
3086 push( @{$orig}, $import_element );
3088 my $orig_best = $orig->[$best_candidate->{position}];
3089 # merge orig_best and b_element together and replace original with merged
3090 if (ref $orig_best ne 'HASH') {
3091 $orig->[$best_candidate->{position}] = $import_element;
3092 } elsif (ref $import_element eq 'HASH') {
3093 my ($key) = keys %{$orig_best};
3094 $orig->[$best_candidate->{position}] = { $key => $self->_merge_attr($orig_best->{$key}, $import_element->{$key}) };
3097 $seen_keys->{$import_key} = 1; # don't merge the same key twice
3107 $self->_source_handle($_[0]->handle);
3109 $self->_source_handle->resolve;
3113 =head2 throw_exception
3115 See L<DBIx::Class::Schema/throw_exception> for details.
3119 sub throw_exception {
3122 if (ref $self && $self->_source_handle->schema) {
3123 $self->_source_handle->schema->throw_exception(@_)
3126 DBIx::Class::Exception->throw(@_);
3130 # XXX: FIXME: Attributes docs need clearing up
3134 Attributes are used to refine a ResultSet in various ways when
3135 searching for data. They can be passed to any method which takes an
3136 C<\%attrs> argument. See L</search>, L</search_rs>, L</find>,
3139 These are in no particular order:
3145 =item Value: ( $order_by | \@order_by | \%order_by )
3149 Which column(s) to order the results by.
3151 [The full list of suitable values is documented in
3152 L<SQL::Abstract/"ORDER BY CLAUSES">; the following is a summary of
3155 If a single column name, or an arrayref of names is supplied, the
3156 argument is passed through directly to SQL. The hashref syntax allows
3157 for connection-agnostic specification of ordering direction:
3159 For descending order:
3161 order_by => { -desc => [qw/col1 col2 col3/] }
3163 For explicit ascending order:
3165 order_by => { -asc => 'col' }
3167 The old scalarref syntax (i.e. order_by => \'year DESC') is still
3168 supported, although you are strongly encouraged to use the hashref
3169 syntax as outlined above.
3175 =item Value: \@columns
3179 Shortcut to request a particular set of columns to be retrieved. Each
3180 column spec may be a string (a table column name), or a hash (in which
3181 case the key is the C<as> value, and the value is used as the C<select>
3182 expression). Adds C<me.> onto the start of any column without a C<.> in
3183 it and sets C<select> from that, then auto-populates C<as> from
3184 C<select> as normal. (You may also use the C<cols> attribute, as in
3185 earlier versions of DBIC.)
3191 =item Value: \@columns
3195 Indicates additional columns to be selected from storage. Works the same
3196 as L</columns> but adds columns to the selection. (You may also use the
3197 C<include_columns> attribute, as in earlier versions of DBIC). For
3200 $schema->resultset('CD')->search(undef, {
3201 '+columns' => ['artist.name'],
3205 would return all CDs and include a 'name' column to the information
3206 passed to object inflation. Note that the 'artist' is the name of the
3207 column (or relationship) accessor, and 'name' is the name of the column
3208 accessor in the related table.
3210 =head2 include_columns
3214 =item Value: \@columns
3218 Deprecated. Acts as a synonym for L</+columns> for backward compatibility.
3224 =item Value: \@select_columns
3228 Indicates which columns should be selected from the storage. You can use
3229 column names, or in the case of RDBMS back ends, function or stored procedure
3232 $rs = $schema->resultset('Employee')->search(undef, {
3235 { count => 'employeeid' },
3240 When you use function/stored procedure names and do not supply an C<as>
3241 attribute, the column names returned are storage-dependent. E.g. MySQL would
3242 return a column named C<count(employeeid)> in the above example.
3248 Indicates additional columns to be selected from storage. Works the same as
3249 L</select> but adds columns to the selection.
3257 Indicates additional column names for those added via L</+select>. See L</as>.
3265 =item Value: \@inflation_names
3269 Indicates column names for object inflation. That is, C<as>
3270 indicates the name that the column can be accessed as via the
3271 C<get_column> method (or via the object accessor, B<if one already
3272 exists>). It has nothing to do with the SQL code C<SELECT foo AS bar>.
3274 The C<as> attribute is used in conjunction with C<select>,
3275 usually when C<select> contains one or more function or stored
3278 $rs = $schema->resultset('Employee')->search(undef, {
3281 { count => 'employeeid' }
3283 as => ['name', 'employee_count'],
3286 my $employee = $rs->first(); # get the first Employee
3288 If the object against which the search is performed already has an accessor
3289 matching a column name specified in C<as>, the value can be retrieved using
3290 the accessor as normal:
3292 my $name = $employee->name();
3294 If on the other hand an accessor does not exist in the object, you need to
3295 use C<get_column> instead:
3297 my $employee_count = $employee->get_column('employee_count');
3299 You can create your own accessors if required - see
3300 L<DBIx::Class::Manual::Cookbook> for details.
3302 Please note: This will NOT insert an C<AS employee_count> into the SQL
3303 statement produced, it is used for internal access only. Thus
3304 attempting to use the accessor in an C<order_by> clause or similar
3305 will fail miserably.
3307 To get around this limitation, you can supply literal SQL to your
3308 C<select> attibute that contains the C<AS alias> text, eg:
3310 select => [\'myfield AS alias']
3316 =item Value: ($rel_name | \@rel_names | \%rel_names)
3320 Contains a list of relationships that should be joined for this query. For
3323 # Get CDs by Nine Inch Nails
3324 my $rs = $schema->resultset('CD')->search(
3325 { 'artist.name' => 'Nine Inch Nails' },
3326 { join => 'artist' }
3329 Can also contain a hash reference to refer to the other relation's relations.
3332 package MyApp::Schema::Track;
3333 use base qw/DBIx::Class/;
3334 __PACKAGE__->table('track');
3335 __PACKAGE__->add_columns(qw/trackid cd position title/);
3336 __PACKAGE__->set_primary_key('trackid');
3337 __PACKAGE__->belongs_to(cd => 'MyApp::Schema::CD');
3340 # In your application
3341 my $rs = $schema->resultset('Artist')->search(
3342 { 'track.title' => 'Teardrop' },
3344 join => { cd => 'track' },
3345 order_by => 'artist.name',
3349 You need to use the relationship (not the table) name in conditions,
3350 because they are aliased as such. The current table is aliased as "me", so
3351 you need to use me.column_name in order to avoid ambiguity. For example:
3353 # Get CDs from 1984 with a 'Foo' track
3354 my $rs = $schema->resultset('CD')->search(
3357 'tracks.name' => 'Foo'
3359 { join => 'tracks' }
3362 If the same join is supplied twice, it will be aliased to <rel>_2 (and
3363 similarly for a third time). For e.g.
3365 my $rs = $schema->resultset('Artist')->search({
3366 'cds.title' => 'Down to Earth',
3367 'cds_2.title' => 'Popular',
3369 join => [ qw/cds cds/ ],
3372 will return a set of all artists that have both a cd with title 'Down
3373 to Earth' and a cd with title 'Popular'.
3375 If you want to fetch related objects from other tables as well, see C<prefetch>
3378 For more help on using joins with search, see L<DBIx::Class::Manual::Joining>.
3384 =item Value: ($rel_name | \@rel_names | \%rel_names)
3388 Contains one or more relationships that should be fetched along with
3389 the main query (when they are accessed afterwards the data will
3390 already be available, without extra queries to the database). This is
3391 useful for when you know you will need the related objects, because it
3392 saves at least one query:
3394 my $rs = $schema->resultset('Tag')->search(
3403 The initial search results in SQL like the following:
3405 SELECT tag.*, cd.*, artist.* FROM tag
3406 JOIN cd ON tag.cd = cd.cdid
3407 JOIN artist ON cd.artist = artist.artistid
3409 L<DBIx::Class> has no need to go back to the database when we access the
3410 C<cd> or C<artist> relationships, which saves us two SQL statements in this
3413 Simple prefetches will be joined automatically, so there is no need
3414 for a C<join> attribute in the above search.
3416 C<prefetch> can be used with the following relationship types: C<belongs_to>,
3417 C<has_one> (or if you're using C<add_relationship>, any relationship declared
3418 with an accessor type of 'single' or 'filter'). A more complex example that
3419 prefetches an artists cds, the tracks on those cds, and the tags associted
3420 with that artist is given below (assuming many-to-many from artists to tags):
3422 my $rs = $schema->resultset('Artist')->search(
3426 { cds => 'tracks' },
3427 { artist_tags => 'tags' }
3433 B<NOTE:> If you specify a C<prefetch> attribute, the C<join> and C<select>
3434 attributes will be ignored.
3436 B<CAVEATs>: Prefetch does a lot of deep magic. As such, it may not behave
3437 exactly as you might expect.
3443 Prefetch uses the L</cache> to populate the prefetched relationships. This
3444 may or may not be what you want.
3448 If you specify a condition on a prefetched relationship, ONLY those
3449 rows that match the prefetched condition will be fetched into that relationship.
3450 This means that adding prefetch to a search() B<may alter> what is returned by
3451 traversing a relationship. So, if you have C<< Artist->has_many(CDs) >> and you do
3453 my $artist_rs = $schema->resultset('Artist')->search({
3459 my $count = $artist_rs->first->cds->count;
3461 my $artist_rs_prefetch = $artist_rs->search( {}, { prefetch => 'cds' } );
3463 my $prefetch_count = $artist_rs_prefetch->first->cds->count;
3465 cmp_ok( $count, '==', $prefetch_count, "Counts should be the same" );
3467 that cmp_ok() may or may not pass depending on the datasets involved. This
3468 behavior may or may not survive the 0.09 transition.
3480 Makes the resultset paged and specifies the page to retrieve. Effectively
3481 identical to creating a non-pages resultset and then calling ->page($page)
3484 If L<rows> attribute is not specified it defaults to 10 rows per page.
3486 When you have a paged resultset, L</count> will only return the number
3487 of rows in the page. To get the total, use the L</pager> and call
3488 C<total_entries> on it.
3498 Specifes the maximum number of rows for direct retrieval or the number of
3499 rows per page if the page attribute or method is used.
3505 =item Value: $offset
3509 Specifies the (zero-based) row number for the first row to be returned, or the
3510 of the first row of the first page if paging is used.
3516 =item Value: \@columns
3520 A arrayref of columns to group by. Can include columns of joined tables.
3522 group_by => [qw/ column1 column2 ... /]
3528 =item Value: $condition
3532 HAVING is a select statement attribute that is applied between GROUP BY and
3533 ORDER BY. It is applied to the after the grouping calculations have been
3536 having => { 'count(employee)' => { '>=', 100 } }
3542 =item Value: (0 | 1)
3546 Set to 1 to group by all columns. If the resultset already has a group_by
3547 attribute, this setting is ignored and an appropriate warning is issued.
3553 Adds to the WHERE clause.
3555 # only return rows WHERE deleted IS NULL for all searches
3556 __PACKAGE__->resultset_attributes({ where => { deleted => undef } }); )
3558 Can be overridden by passing C<{ where => undef }> as an attribute
3565 Set to 1 to cache search results. This prevents extra SQL queries if you
3566 revisit rows in your ResultSet:
3568 my $resultset = $schema->resultset('Artist')->search( undef, { cache => 1 } );
3570 while( my $artist = $resultset->next ) {
3574 $rs->first; # without cache, this would issue a query
3576 By default, searches are not cached.
3578 For more examples of using these attributes, see
3579 L<DBIx::Class::Manual::Cookbook>.
3585 =item Value: \@from_clause
3589 The C<from> attribute gives you manual control over the C<FROM> clause of SQL
3590 statements generated by L<DBIx::Class>, allowing you to express custom C<JOIN>
3593 NOTE: Use this on your own risk. This allows you to shoot off your foot!
3595 C<join> will usually do what you need and it is strongly recommended that you
3596 avoid using C<from> unless you cannot achieve the desired result using C<join>.
3597 And we really do mean "cannot", not just tried and failed. Attempting to use
3598 this because you're having problems with C<join> is like trying to use x86
3599 ASM because you've got a syntax error in your C. Trust us on this.
3601 Now, if you're still really, really sure you need to use this (and if you're
3602 not 100% sure, ask the mailing list first), here's an explanation of how this
3605 The syntax is as follows -
3608 { <alias1> => <table1> },
3610 { <alias2> => <table2>, -join_type => 'inner|left|right' },
3611 [], # nested JOIN (optional)
3612 { <table1.column1> => <table2.column2>, ... (more conditions) },
3614 # More of the above [ ] may follow for additional joins
3621 ON <table1.column1> = <table2.column2>
3622 <more joins may follow>
3624 An easy way to follow the examples below is to remember the following:
3626 Anything inside "[]" is a JOIN
3627 Anything inside "{}" is a condition for the enclosing JOIN
3629 The following examples utilize a "person" table in a family tree application.
3630 In order to express parent->child relationships, this table is self-joined:
3632 # Person->belongs_to('father' => 'Person');
3633 # Person->belongs_to('mother' => 'Person');
3635 C<from> can be used to nest joins. Here we return all children with a father,
3636 then search against all mothers of those children:
3638 $rs = $schema->resultset('Person')->search(
3641 alias => 'mother', # alias columns in accordance with "from"
3643 { mother => 'person' },
3646 { child => 'person' },
3648 { father => 'person' },
3649 { 'father.person_id' => 'child.father_id' }
3652 { 'mother.person_id' => 'child.mother_id' }
3659 # SELECT mother.* FROM person mother
3662 # JOIN person father
3663 # ON ( father.person_id = child.father_id )
3665 # ON ( mother.person_id = child.mother_id )
3667 The type of any join can be controlled manually. To search against only people
3668 with a father in the person table, we could explicitly use C<INNER JOIN>:
3670 $rs = $schema->resultset('Person')->search(
3673 alias => 'child', # alias columns in accordance with "from"
3675 { child => 'person' },
3677 { father => 'person', -join_type => 'inner' },
3678 { 'father.id' => 'child.father_id' }
3685 # SELECT child.* FROM person child
3686 # INNER JOIN person father ON child.father_id = father.id
3688 You can select from a subquery by passing a resultset to from as follows.
3690 $schema->resultset('Artist')->search(
3692 { alias => 'artist2',
3693 from => [ { artist2 => $artist_rs->as_query } ],
3696 # and you'll get sql like this..
3697 # SELECT artist2.artistid, artist2.name, artist2.rank, artist2.charfield FROM
3698 # ( SELECT me.artistid, me.name, me.rank, me.charfield FROM artists me ) artist2
3700 If you need to express really complex joins, you
3701 can supply literal SQL to C<from> via a scalar reference. In this case
3702 the contents of the scalar will replace the table name associated with the
3705 WARNING: This technique might very well not work as expected on chained
3706 searches - you have been warned.
3708 # Assuming the Event resultsource is defined as:
3710 MySchema::Event->add_columns (
3713 is_auto_increment => 1,
3722 MySchema::Event->set_primary_key ('sequence');
3724 # This will get back the latest event for every location. The column
3725 # selector is still provided by DBIC, all we do is add a JOIN/WHERE
3726 # combo to limit the resultset
3728 $rs = $schema->resultset('Event');
3729 $table = $rs->result_source->name;
3730 $latest = $rs->search (
3733 (SELECT e1.* FROM $table e1
3735 ON e1.location = e2.location
3736 AND e1.sequence < e2.sequence
3737 WHERE e2.sequence is NULL
3742 # Equivalent SQL (with the DBIC chunks added):
3744 SELECT me.sequence, me.location, me.type FROM
3745 (SELECT e1.* FROM events e1
3747 ON e1.location = e2.location
3748 AND e1.sequence < e2.sequence
3749 WHERE e2.sequence is NULL
3756 =item Value: ( 'update' | 'shared' )
3760 Set to 'update' for a SELECT ... FOR UPDATE or 'shared' for a SELECT