1 package DBIx::Class::ResultSet;
5 use base qw/DBIx::Class/;
7 use DBIx::Class::ResultSetColumn;
8 use Scalar::Util qw/blessed weaken/;
10 use Data::Compare (); # no imports!!! guard against insane architecture
12 # not importing first() as it will clash with our own method
16 # De-duplication in _merge_attr() is disabled, but left in for reference
17 # (the merger is used for other things that ought not to be de-duped)
18 *__HM_DEDUP = sub () { 0 };
28 __PACKAGE__->mk_group_accessors('simple' => qw/_result_class result_source/);
32 DBIx::Class::ResultSet - Represents a query used for fetching a set of results.
36 my $users_rs = $schema->resultset('User');
37 while( $user = $users_rs->next) {
38 print $user->username;
41 my $registered_users_rs = $schema->resultset('User')->search({ registered => 1 });
42 my @cds_in_2005 = $schema->resultset('CD')->search({ year => 2005 })->all();
46 A ResultSet is an object which stores a set of conditions representing
47 a query. It is the backbone of DBIx::Class (i.e. the really
48 important/useful bit).
50 No SQL is executed on the database when a ResultSet is created, it
51 just stores all the conditions needed to create the query.
53 A basic ResultSet representing the data of an entire table is returned
54 by calling C<resultset> on a L<DBIx::Class::Schema> and passing in a
55 L<Source|DBIx::Class::Manual::Glossary/Source> name.
57 my $users_rs = $schema->resultset('User');
59 A new ResultSet is returned from calling L</search> on an existing
60 ResultSet. The new one will contain all the conditions of the
61 original, plus any new conditions added in the C<search> call.
63 A ResultSet also incorporates an implicit iterator. L</next> and L</reset>
64 can be used to walk through all the L<DBIx::Class::Row>s the ResultSet
67 The query that the ResultSet represents is B<only> executed against
68 the database when these methods are called:
69 L</find>, L</next>, L</all>, L</first>, L</single>, L</count>.
71 If a resultset is used in a numeric context it returns the L</count>.
72 However, if it is used in a boolean context it is B<always> true. So if
73 you want to check if a resultset has any results, you must use C<if $rs
76 =head1 CUSTOM ResultSet CLASSES THAT USE Moose
78 If you want to make your custom ResultSet classes with L<Moose>, use a template
81 package MyApp::Schema::ResultSet::User;
84 use namespace::autoclean;
86 extends 'DBIx::Class::ResultSet';
88 sub BUILDARGS { $_[2] }
92 __PACKAGE__->meta->make_immutable;
96 The L<MooseX::NonMoose> is necessary so that the L<Moose> constructor does not
97 clash with the regular ResultSet constructor. Alternatively, you can use:
99 __PACKAGE__->meta->make_immutable(inline_constructor => 0);
101 The L<BUILDARGS|Moose::Manual::Construction/BUILDARGS> is necessary because the
102 signature of the ResultSet C<new> is C<< ->new($source, \%args) >>.
106 =head2 Chaining resultsets
108 Let's say you've got a query that needs to be run to return some data
109 to the user. But, you have an authorization system in place that
110 prevents certain users from seeing certain information. So, you want
111 to construct the basic query in one method, but add constraints to it in
116 my $request = $self->get_request; # Get a request object somehow.
117 my $schema = $self->result_source->schema;
119 my $cd_rs = $schema->resultset('CD')->search({
120 title => $request->param('title'),
121 year => $request->param('year'),
124 $cd_rs = $self->apply_security_policy( $cd_rs );
126 return $cd_rs->all();
129 sub apply_security_policy {
138 =head3 Resolving conditions and attributes
140 When a resultset is chained from another resultset, conditions and
141 attributes with the same keys need resolving.
143 L</join>, L</prefetch>, L</+select>, L</+as> attributes are merged
144 into the existing ones from the original resultset.
146 The L</where> and L</having> attributes, and any search conditions, are
147 merged with an SQL C<AND> to the existing condition from the original
150 All other attributes are overridden by any new ones supplied in the
153 =head2 Multiple queries
155 Since a resultset just defines a query, you can do all sorts of
156 things with it with the same object.
158 # Don't hit the DB yet.
159 my $cd_rs = $schema->resultset('CD')->search({
160 title => 'something',
164 # Each of these hits the DB individually.
165 my $count = $cd_rs->count;
166 my $most_recent = $cd_rs->get_column('date_released')->max();
167 my @records = $cd_rs->all;
169 And it's not just limited to SELECT statements.
175 $cd_rs->create({ artist => 'Fred' });
177 Which is the same as:
179 $schema->resultset('CD')->create({
180 title => 'something',
185 See: L</search>, L</count>, L</get_column>, L</all>, L</create>.
193 =item Arguments: L<$source|DBIx::Class::ResultSource>, L<\%attrs?|/ATTRIBUTES>
195 =item Return Value: L<$resultset|/search>
199 The resultset constructor. Takes a source object (usually a
200 L<DBIx::Class::ResultSourceProxy::Table>) and an attribute hash (see
201 L</ATTRIBUTES> below). Does not perform any queries -- these are
202 executed as needed by the other methods.
204 Generally you never construct a resultset manually. Instead you get one
206 C<< $schema->L<resultset|DBIx::Class::Schema/resultset>('$source_name') >>
207 or C<< $another_resultset->L<search|/search>(...) >> (the later called in
210 my $rs = $schema->resultset('CD')->search({ title => '100th Window' });
216 If called on an object, proxies to L</new_result> instead, so
218 my $cd = $schema->resultset('CD')->new({ title => 'Spoon' });
220 will return a CD object, not a ResultSet, and is equivalent to:
222 my $cd = $schema->resultset('CD')->new_result({ title => 'Spoon' });
224 Please also keep in mind that many internals call L</new_result> directly,
225 so overloading this method with the idea of intercepting new result object
226 creation B<will not work>. See also warning pertaining to L</create>.
234 return $class->new_result(@_) if ref $class;
236 my ($source, $attrs) = @_;
237 $source = $source->resolve
238 if $source->isa('DBIx::Class::ResultSourceHandle');
239 $attrs = { %{$attrs||{}} };
241 if ($attrs->{page}) {
242 $attrs->{rows} ||= 10;
245 $attrs->{alias} ||= 'me';
248 result_source => $source,
249 cond => $attrs->{where},
254 # if there is a dark selector, this means we are already in a
255 # chain and the cleanup/sanification was taken care of by
257 $self->_normalize_selection($attrs)
258 unless $attrs->{_dark_selector};
261 $attrs->{result_class} || $source->result_class
271 =item Arguments: L<$cond|DBIx::Class::SQLMaker> | undef, L<\%attrs?|/ATTRIBUTES>
273 =item Return Value: $resultset (scalar context) | L<@result_objs|DBIx::Class::Manual::ResultClass> (list context)
277 my @cds = $cd_rs->search({ year => 2001 }); # "... WHERE year = 2001"
278 my $new_rs = $cd_rs->search({ year => 2005 });
280 my $new_rs = $cd_rs->search([ { year => 2005 }, { year => 2004 } ]);
281 # year = 2005 OR year = 2004
283 In list context, C<< ->all() >> is called implicitly on the resultset, thus
284 returning a list of L<result|DBIx::Class::Manual::ResultClass> objects instead.
285 To avoid that, use L</search_rs>.
287 If you need to pass in additional attributes but no additional condition,
288 call it as C<search(undef, \%attrs)>.
290 # "SELECT name, artistid FROM $artist_table"
291 my @all_artists = $schema->resultset('Artist')->search(undef, {
292 columns => [qw/name artistid/],
295 For a list of attributes that can be passed to C<search>, see
296 L</ATTRIBUTES>. For more examples of using this function, see
297 L<Searching|DBIx::Class::Manual::Cookbook/Searching>. For a complete
298 documentation for the first argument, see L<SQL::Abstract>
299 and its extension L<DBIx::Class::SQLMaker>.
301 For more help on using joins with search, see L<DBIx::Class::Manual::Joining>.
305 Note that L</search> does not process/deflate any of the values passed in the
306 L<SQL::Abstract>-compatible search condition structure. This is unlike other
307 condition-bound methods L</new_result>, L</create> and L</find>. The user must ensure
308 manually that any value passed to this method will stringify to something the
309 RDBMS knows how to deal with. A notable example is the handling of L<DateTime>
310 objects, for more info see:
311 L<DBIx::Class::Manual::Cookbook/Formatting DateTime objects in queries>.
317 my $rs = $self->search_rs( @_ );
322 elsif (defined wantarray) {
326 # we can be called by a relationship helper, which in
327 # turn may be called in void context due to some braindead
328 # overload or whatever else the user decided to be clever
329 # at this particular day. Thus limit the exception to
330 # external code calls only
331 $self->throw_exception ('->search is *not* a mutator, calling it in void context makes no sense')
332 if (caller)[0] !~ /^\QDBIx::Class::/;
342 =item Arguments: L<$cond|DBIx::Class::SQLMaker>, L<\%attrs?|/ATTRIBUTES>
344 =item Return Value: L<$resultset|/search>
348 This method does the same exact thing as search() except it will
349 always return a resultset, even in list context.
356 my $rsrc = $self->result_source;
357 my ($call_cond, $call_attrs);
359 # Special-case handling for (undef, undef) or (undef)
360 # Note that (foo => undef) is valid deprecated syntax
361 @_ = () if not scalar grep { defined $_ } @_;
367 # fish out attrs in the ($condref, $attr) case
368 elsif (@_ == 2 and ( ! defined $_[0] or (ref $_[0]) ne '') ) {
369 ($call_cond, $call_attrs) = @_;
372 $self->throw_exception('Odd number of arguments to search')
376 carp_unique 'search( %condition ) is deprecated, use search( \%condition ) instead'
377 unless $rsrc->result_class->isa('DBIx::Class::CDBICompat');
379 for my $i (0 .. $#_) {
381 $self->throw_exception ('All keys in condition key/value pairs must be plain scalars')
382 if (! defined $_[$i] or ref $_[$i] ne '');
388 # see if we can keep the cache (no $rs changes)
390 my %safe = (alias => 1, cache => 1);
391 if ( ! List::Util::first { !$safe{$_} } keys %$call_attrs and (
394 ref $_[0] eq 'HASH' && ! keys %{$_[0]}
396 ref $_[0] eq 'ARRAY' && ! @{$_[0]}
398 $cache = $self->get_cache;
401 my $old_attrs = { %{$self->{attrs}} };
402 my $old_having = delete $old_attrs->{having};
403 my $old_where = delete $old_attrs->{where};
405 my $new_attrs = { %$old_attrs };
407 # take care of call attrs (only if anything is changing)
408 if ($call_attrs and keys %$call_attrs) {
410 # copy for _normalize_selection
411 $call_attrs = { %$call_attrs };
413 my @selector_attrs = qw/select as columns cols +select +as +columns include_columns/;
415 # reset the current selector list if new selectors are supplied
416 if (List::Util::first { exists $call_attrs->{$_} } qw/columns cols select as/) {
417 delete @{$old_attrs}{(@selector_attrs, '_dark_selector')};
420 # Normalize the new selector list (operates on the passed-in attr structure)
421 # Need to do it on every chain instead of only once on _resolved_attrs, in
422 # order to allow detection of empty vs partial 'as'
423 $call_attrs->{_dark_selector} = $old_attrs->{_dark_selector}
424 if $old_attrs->{_dark_selector};
425 $self->_normalize_selection ($call_attrs);
427 # start with blind overwriting merge, exclude selector attrs
428 $new_attrs = { %{$old_attrs}, %{$call_attrs} };
429 delete @{$new_attrs}{@selector_attrs};
431 for (@selector_attrs) {
432 $new_attrs->{$_} = $self->_merge_attr($old_attrs->{$_}, $call_attrs->{$_})
433 if ( exists $old_attrs->{$_} or exists $call_attrs->{$_} );
436 # older deprecated name, use only if {columns} is not there
437 if (my $c = delete $new_attrs->{cols}) {
438 if ($new_attrs->{columns}) {
439 carp "Resultset specifies both the 'columns' and the legacy 'cols' attributes - ignoring 'cols'";
442 $new_attrs->{columns} = $c;
447 # join/prefetch use their own crazy merging heuristics
448 foreach my $key (qw/join prefetch/) {
449 $new_attrs->{$key} = $self->_merge_joinpref_attr($old_attrs->{$key}, $call_attrs->{$key})
450 if exists $call_attrs->{$key};
453 # stack binds together
454 $new_attrs->{bind} = [ @{ $old_attrs->{bind} || [] }, @{ $call_attrs->{bind} || [] } ];
458 for ($old_where, $call_cond) {
460 $new_attrs->{where} = $self->_stack_cond (
461 $_, $new_attrs->{where}
466 if (defined $old_having) {
467 $new_attrs->{having} = $self->_stack_cond (
468 $old_having, $new_attrs->{having}
472 my $rs = (ref $self)->new($rsrc, $new_attrs);
474 $rs->set_cache($cache) if ($cache);
480 sub _normalize_selection {
481 my ($self, $attrs) = @_;
484 $attrs->{'+columns'} = $self->_merge_attr($attrs->{'+columns'}, delete $attrs->{include_columns})
485 if exists $attrs->{include_columns};
487 # columns are always placed first, however
489 # Keep the X vs +X separation until _resolved_attrs time - this allows to
490 # delay the decision on whether to use a default select list ($rsrc->columns)
491 # allowing stuff like the remove_columns helper to work
493 # select/as +select/+as pairs need special handling - the amount of select/as
494 # elements in each pair does *not* have to be equal (think multicolumn
495 # selectors like distinct(foo, bar) ). If the selector is bare (no 'as'
496 # supplied at all) - try to infer the alias, either from the -as parameter
497 # of the selector spec, or use the parameter whole if it looks like a column
498 # name (ugly legacy heuristic). If all fails - leave the selector bare (which
499 # is ok as well), but make sure no more additions to the 'as' chain take place
500 for my $pref ('', '+') {
502 my ($sel, $as) = map {
503 my $key = "${pref}${_}";
505 my $val = [ ref $attrs->{$key} eq 'ARRAY'
507 : $attrs->{$key} || ()
509 delete $attrs->{$key};
513 if (! @$as and ! @$sel ) {
516 elsif (@$as and ! @$sel) {
517 $self->throw_exception(
518 "Unable to handle ${pref}as specification (@$as) without a corresponding ${pref}select"
522 # no as part supplied at all - try to deduce (unless explicit end of named selection is declared)
523 # if any @$as has been supplied we assume the user knows what (s)he is doing
524 # and blindly keep stacking up pieces
525 unless ($attrs->{_dark_selector}) {
528 if ( ref $_ eq 'HASH' and exists $_->{-as} ) {
529 push @$as, $_->{-as};
531 # assume any plain no-space, no-parenthesis string to be a column spec
532 # FIXME - this is retarded but is necessary to support shit like 'count(foo)'
533 elsif ( ! ref $_ and $_ =~ /^ [^\s\(\)]+ $/x) {
536 # if all else fails - raise a flag that no more aliasing will be allowed
538 $attrs->{_dark_selector} = {
540 string => ($dark_sel_dumper ||= do {
541 require Data::Dumper::Concise;
542 Data::Dumper::Concise::DumperObject()->Indent(0);
543 })->Values([$_])->Dump
551 elsif (@$as < @$sel) {
552 $self->throw_exception(
553 "Unable to handle an ${pref}as specification (@$as) with less elements than the corresponding ${pref}select"
556 elsif ($pref and $attrs->{_dark_selector}) {
557 $self->throw_exception(
558 "Unable to process named '+select', resultset contains an unnamed selector $attrs->{_dark_selector}{string}"
564 $attrs->{"${pref}select"} = $self->_merge_attr($attrs->{"${pref}select"}, $sel);
565 $attrs->{"${pref}as"} = $self->_merge_attr($attrs->{"${pref}as"}, $as);
570 my ($self, $left, $right) = @_;
572 # collapse single element top-level conditions
573 # (single pass only, unlikely to need recursion)
574 for ($left, $right) {
575 if (ref $_ eq 'ARRAY') {
583 elsif (ref $_ eq 'HASH') {
584 my ($first, $more) = keys %$_;
587 if (! defined $first) {
591 elsif (! defined $more) {
592 if ($first eq '-and' and ref $_->{'-and'} eq 'HASH') {
595 elsif ($first eq '-or' and ref $_->{'-or'} eq 'ARRAY') {
602 # merge hashes with weeding out of duplicates (simple cases only)
603 if (ref $left eq 'HASH' and ref $right eq 'HASH') {
605 # shallow copy to destroy
606 $right = { %$right };
607 for (grep { exists $right->{$_} } keys %$left) {
608 # the use of eq_deeply here is justified - the rhs of an
609 # expression can contain a lot of twisted weird stuff
610 delete $right->{$_} if Data::Compare::Compare( $left->{$_}, $right->{$_} );
613 $right = undef unless keys %$right;
617 if (defined $left xor defined $right) {
618 return defined $left ? $left : $right;
620 elsif (! defined $left) {
624 return { -and => [ $left, $right ] };
628 =head2 search_literal
630 B<CAVEAT>: C<search_literal> is provided for Class::DBI compatibility and
631 should only be used in that context. C<search_literal> is a convenience
632 method. It is equivalent to calling C<< $schema->search(\[]) >>, but if you
633 want to ensure columns are bound correctly, use L</search>.
635 See L<DBIx::Class::Manual::Cookbook/Searching> and
636 L<DBIx::Class::Manual::FAQ/Searching> for searching techniques that do not
637 require C<search_literal>.
641 =item Arguments: $sql_fragment, @standalone_bind_values
643 =item Return Value: L<$resultset|/search> (scalar context) | L<@result_objs|DBIx::Class::Manual::ResultClass> (list context)
647 my @cds = $cd_rs->search_literal('year = ? AND title = ?', qw/2001 Reload/);
648 my $newrs = $artist_rs->search_literal('name = ?', 'Metallica');
650 Pass a literal chunk of SQL to be added to the conditional part of the
653 Example of how to use C<search> instead of C<search_literal>
655 my @cds = $cd_rs->search_literal('cdid = ? AND (artist = ? OR artist = ?)', (2, 1, 2));
656 my @cds = $cd_rs->search(\[ 'cdid = ? AND (artist = ? OR artist = ?)', [ 'cdid', 2 ], [ 'artist', 1 ], [ 'artist', 2 ] ]);
661 my ($self, $sql, @bind) = @_;
663 if ( @bind && ref($bind[-1]) eq 'HASH' ) {
666 return $self->search(\[ $sql, map [ {} => $_ ], @bind ], ($attr || () ));
673 =item Arguments: \%columns_values | @pk_values, { key => $unique_constraint, L<%attrs|/ATTRIBUTES> }?
675 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass> | undef
679 Finds and returns a single row based on supplied criteria. Takes either a
680 hashref with the same format as L</create> (including inference of foreign
681 keys from related objects), or a list of primary key values in the same
682 order as the L<primary columns|DBIx::Class::ResultSource/primary_columns>
683 declaration on the L</result_source>.
685 In either case an attempt is made to combine conditions already existing on
686 the resultset with the condition passed to this method.
688 To aid with preparing the correct query for the storage you may supply the
689 C<key> attribute, which is the name of a
690 L<unique constraint|DBIx::Class::ResultSource/add_unique_constraint> (the
691 unique constraint corresponding to the
692 L<primary columns|DBIx::Class::ResultSource/primary_columns> is always named
693 C<primary>). If the C<key> attribute has been supplied, and DBIC is unable
694 to construct a query that satisfies the named unique constraint fully (
695 non-NULL values for each column member of the constraint) an exception is
698 If no C<key> is specified, the search is carried over all unique constraints
699 which are fully defined by the available condition.
701 If no such constraint is found, C<find> currently defaults to a simple
702 C<< search->(\%column_values) >> which may or may not do what you expect.
703 Note that this fallback behavior may be deprecated in further versions. If
704 you need to search with arbitrary conditions - use L</search>. If the query
705 resulting from this fallback produces more than one row, a warning to the
706 effect is issued, though only the first row is constructed and returned as
709 In addition to C<key>, L</find> recognizes and applies standard
710 L<resultset attributes|/ATTRIBUTES> in the same way as L</search> does.
712 Note that if you have extra concerns about the correctness of the resulting
713 query you need to specify the C<key> attribute and supply the entire condition
714 as an argument to find (since it is not always possible to perform the
715 combination of the resultset condition with the supplied one, especially if
716 the resultset condition contains literal sql).
718 For example, to find a row by its primary key:
720 my $cd = $schema->resultset('CD')->find(5);
722 You can also find a row by a specific unique constraint:
724 my $cd = $schema->resultset('CD')->find(
726 artist => 'Massive Attack',
727 title => 'Mezzanine',
729 { key => 'cd_artist_title' }
732 See also L</find_or_create> and L</update_or_create>.
738 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
740 my $rsrc = $self->result_source;
743 if (exists $attrs->{key}) {
744 $constraint_name = defined $attrs->{key}
746 : $self->throw_exception("An undefined 'key' resultset attribute makes no sense")
750 # Parse out the condition from input
753 if (ref $_[0] eq 'HASH') {
754 $call_cond = { %{$_[0]} };
757 # if only values are supplied we need to default to 'primary'
758 $constraint_name = 'primary' unless defined $constraint_name;
760 my @c_cols = $rsrc->unique_constraint_columns($constraint_name);
762 $self->throw_exception(
763 "No constraint columns, maybe a malformed '$constraint_name' constraint?"
766 $self->throw_exception (
767 'find() expects either a column/value hashref, or a list of values '
768 . "corresponding to the columns of the specified unique constraint '$constraint_name'"
769 ) unless @c_cols == @_;
772 @{$call_cond}{@c_cols} = @_;
776 for my $key (keys %$call_cond) {
778 my $keyref = ref($call_cond->{$key})
780 my $relinfo = $rsrc->relationship_info($key)
782 my $val = delete $call_cond->{$key};
784 next if $keyref eq 'ARRAY'; # has_many for multi_create
786 my $rel_q = $rsrc->_resolve_condition(
787 $relinfo->{cond}, $val, $key, $key
789 die "Can't handle complex relationship conditions in find" if ref($rel_q) ne 'HASH';
790 @related{keys %$rel_q} = values %$rel_q;
794 # relationship conditions take precedence (?)
795 @{$call_cond}{keys %related} = values %related;
797 my $alias = exists $attrs->{alias} ? $attrs->{alias} : $self->{attrs}{alias};
799 if (defined $constraint_name) {
800 $final_cond = $self->_qualify_cond_columns (
802 $self->_build_unique_cond (
810 elsif ($self->{attrs}{accessor} and $self->{attrs}{accessor} eq 'single') {
811 # This means that we got here after a merger of relationship conditions
812 # in ::Relationship::Base::search_related (the row method), and furthermore
813 # the relationship is of the 'single' type. This means that the condition
814 # provided by the relationship (already attached to $self) is sufficient,
815 # as there can be only one row in the database that would satisfy the
819 # no key was specified - fall down to heuristics mode:
820 # run through all unique queries registered on the resultset, and
821 # 'OR' all qualifying queries together
822 my (@unique_queries, %seen_column_combinations);
823 for my $c_name ($rsrc->unique_constraint_names) {
824 next if $seen_column_combinations{
825 join "\x00", sort $rsrc->unique_constraint_columns($c_name)
828 push @unique_queries, try {
829 $self->_build_unique_cond ($c_name, $call_cond, 'croak_on_nulls')
833 $final_cond = @unique_queries
834 ? [ map { $self->_qualify_cond_columns($_, $alias) } @unique_queries ]
835 : $self->_non_unique_find_fallback ($call_cond, $attrs)
839 # Run the query, passing the result_class since it should propagate for find
840 my $rs = $self->search ($final_cond, {result_class => $self->result_class, %$attrs});
841 if ($rs->_resolved_attrs->{collapse}) {
843 carp "Query returned more than one row" if $rs->next;
851 # This is a stop-gap method as agreed during the discussion on find() cleanup:
852 # http://lists.scsys.co.uk/pipermail/dbix-class/2010-October/009535.html
854 # It is invoked when find() is called in legacy-mode with insufficiently-unique
855 # condition. It is provided for overrides until a saner way forward is devised
857 # *NOTE* This is not a public method, and it's *GUARANTEED* to disappear down
858 # the road. Please adjust your tests accordingly to catch this situation early
859 # DBIx::Class::ResultSet->can('_non_unique_find_fallback') is reasonable
861 # The method will not be removed without an adequately complete replacement
862 # for strict-mode enforcement
863 sub _non_unique_find_fallback {
864 my ($self, $cond, $attrs) = @_;
866 return $self->_qualify_cond_columns(
868 exists $attrs->{alias}
870 : $self->{attrs}{alias}
875 sub _qualify_cond_columns {
876 my ($self, $cond, $alias) = @_;
878 my %aliased = %$cond;
879 for (keys %aliased) {
880 $aliased{"$alias.$_"} = delete $aliased{$_}
887 sub _build_unique_cond {
888 my ($self, $constraint_name, $extra_cond, $croak_on_null) = @_;
890 my @c_cols = $self->result_source->unique_constraint_columns($constraint_name);
892 # combination may fail if $self->{cond} is non-trivial
893 my ($final_cond) = try {
894 $self->_merge_with_rscond ($extra_cond)
899 # trim out everything not in $columns
900 $final_cond = { map {
901 exists $final_cond->{$_}
902 ? ( $_ => $final_cond->{$_} )
906 if (my @missing = grep
907 { ! ($croak_on_null ? defined $final_cond->{$_} : exists $final_cond->{$_}) }
910 $self->throw_exception( sprintf ( "Unable to satisfy requested constraint '%s', no values for column(s): %s",
912 join (', ', map { "'$_'" } @missing),
919 !$ENV{DBIC_NULLABLE_KEY_NOWARN}
921 my @undefs = sort grep { ! defined $final_cond->{$_} } (keys %$final_cond)
923 carp_unique ( sprintf (
924 "NULL/undef values supplied for requested unique constraint '%s' (NULL "
925 . 'values in column(s): %s). This is almost certainly not what you wanted, '
926 . 'though you can set DBIC_NULLABLE_KEY_NOWARN to disable this warning.',
928 join (', ', map { "'$_'" } @undefs),
935 =head2 search_related
939 =item Arguments: $rel_name, $cond?, L<\%attrs?|/ATTRIBUTES>
941 =item Return Value: L<$resultset|/search> (scalar context) | L<@result_objs|DBIx::Class::Manual::ResultClass> (list context)
945 $new_rs = $cd_rs->search_related('artist', {
949 Searches the specified relationship, optionally specifying a condition and
950 attributes for matching records. See L</ATTRIBUTES> for more information.
952 In list context, C<< ->all() >> is called implicitly on the resultset, thus
953 returning a list of result objects instead. To avoid that, use L</search_related_rs>.
955 See also L</search_related_rs>.
960 return shift->related_resultset(shift)->search(@_);
963 =head2 search_related_rs
965 This method works exactly the same as search_related, except that
966 it guarantees a resultset, even in list context.
970 sub search_related_rs {
971 return shift->related_resultset(shift)->search_rs(@_);
978 =item Arguments: none
980 =item Return Value: L<$cursor|DBIx::Class::Cursor>
984 Returns a storage-driven cursor to the given resultset. See
985 L<DBIx::Class::Cursor> for more information.
992 return $self->{cursor} ||= do {
993 my $attrs = { %{$self->_resolved_attrs } };
994 $self->result_source->storage->select(
995 $attrs->{from}, $attrs->{select}, $attrs->{where}, $attrs
1004 =item Arguments: L<$cond?|DBIx::Class::SQLMaker>
1006 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass> | undef
1010 my $cd = $schema->resultset('CD')->single({ year => 2001 });
1012 Inflates the first result without creating a cursor if the resultset has
1013 any records in it; if not returns C<undef>. Used by L</find> as a lean version
1016 While this method can take an optional search condition (just like L</search>)
1017 being a fast-code-path it does not recognize search attributes. If you need to
1018 add extra joins or similar, call L</search> and then chain-call L</single> on the
1019 L<DBIx::Class::ResultSet> returned.
1025 As of 0.08100, this method enforces the assumption that the preceding
1026 query returns only one row. If more than one row is returned, you will receive
1029 Query returned more than one row
1031 In this case, you should be using L</next> or L</find> instead, or if you really
1032 know what you are doing, use the L</rows> attribute to explicitly limit the size
1035 This method will also throw an exception if it is called on a resultset prefetching
1036 has_many, as such a prefetch implies fetching multiple rows from the database in
1037 order to assemble the resulting object.
1044 my ($self, $where) = @_;
1046 $self->throw_exception('single() only takes search conditions, no attributes. You want ->search( $cond, $attrs )->single()');
1049 my $attrs = { %{$self->_resolved_attrs} };
1051 $self->throw_exception(
1052 'single() can not be used on resultsets prefetching has_many. Use find( \%cond ) or next() instead'
1053 ) if $attrs->{collapse};
1056 if (defined $attrs->{where}) {
1059 [ map { ref $_ eq 'ARRAY' ? [ -or => $_ ] : $_ }
1060 $where, delete $attrs->{where} ]
1063 $attrs->{where} = $where;
1067 my $data = [ $self->result_source->storage->select_single(
1068 $attrs->{from}, $attrs->{select},
1069 $attrs->{where}, $attrs
1071 return undef unless @$data;
1072 $self->{stashed_rows} = [ $data ];
1073 $self->_construct_objects->[0];
1079 # Recursively collapse the query, accumulating values for each column.
1081 sub _collapse_query {
1082 my ($self, $query, $collapsed) = @_;
1086 if (ref $query eq 'ARRAY') {
1087 foreach my $subquery (@$query) {
1088 next unless ref $subquery; # -or
1089 $collapsed = $self->_collapse_query($subquery, $collapsed);
1092 elsif (ref $query eq 'HASH') {
1093 if (keys %$query and (keys %$query)[0] eq '-and') {
1094 foreach my $subquery (@{$query->{-and}}) {
1095 $collapsed = $self->_collapse_query($subquery, $collapsed);
1099 foreach my $col (keys %$query) {
1100 my $value = $query->{$col};
1101 $collapsed->{$col}{$value}++;
1113 =item Arguments: L<$cond?|DBIx::Class::SQLMaker>
1115 =item Return Value: L<$resultsetcolumn|DBIx::Class::ResultSetColumn>
1119 my $max_length = $rs->get_column('length')->max;
1121 Returns a L<DBIx::Class::ResultSetColumn> instance for a column of the ResultSet.
1126 my ($self, $column) = @_;
1127 my $new = DBIx::Class::ResultSetColumn->new($self, $column);
1135 =item Arguments: L<$cond|DBIx::Class::SQLMaker>, L<\%attrs?|/ATTRIBUTES>
1137 =item Return Value: L<$resultset|/search> (scalar context) | L<@result_objs|DBIx::Class::Manual::ResultClass> (list context)
1141 # WHERE title LIKE '%blue%'
1142 $cd_rs = $rs->search_like({ title => '%blue%'});
1144 Performs a search, but uses C<LIKE> instead of C<=> as the condition. Note
1145 that this is simply a convenience method retained for ex Class::DBI users.
1146 You most likely want to use L</search> with specific operators.
1148 For more information, see L<DBIx::Class::Manual::Cookbook>.
1150 This method is deprecated and will be removed in 0.09. Use L</search()>
1151 instead. An example conversion is:
1153 ->search_like({ foo => 'bar' });
1157 ->search({ foo => { like => 'bar' } });
1164 'search_like() is deprecated and will be removed in DBIC version 0.09.'
1165 .' Instead use ->search({ x => { -like => "y%" } })'
1166 .' (note the outer pair of {}s - they are important!)'
1168 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
1169 my $query = ref $_[0] eq 'HASH' ? { %{shift()} }: {@_};
1170 $query->{$_} = { 'like' => $query->{$_} } for keys %$query;
1171 return $class->search($query, { %$attrs });
1178 =item Arguments: $first, $last
1180 =item Return Value: L<$resultset|/search> (scalar context) | L<@result_objs|DBIx::Class::Manual::ResultClass> (list context)
1184 Returns a resultset or object list representing a subset of elements from the
1185 resultset slice is called on. Indexes are from 0, i.e., to get the first
1186 three records, call:
1188 my ($one, $two, $three) = $rs->slice(0, 2);
1193 my ($self, $min, $max) = @_;
1194 my $attrs = {}; # = { %{ $self->{attrs} || {} } };
1195 $attrs->{offset} = $self->{attrs}{offset} || 0;
1196 $attrs->{offset} += $min;
1197 $attrs->{rows} = ($max ? ($max - $min + 1) : 1);
1198 return $self->search(undef, $attrs);
1199 #my $slice = (ref $self)->new($self->result_source, $attrs);
1200 #return (wantarray ? $slice->all : $slice);
1207 =item Arguments: none
1209 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass> | undef
1213 Returns the next element in the resultset (C<undef> is there is none).
1215 Can be used to efficiently iterate over records in the resultset:
1217 my $rs = $schema->resultset('CD')->search;
1218 while (my $cd = $rs->next) {
1222 Note that you need to store the resultset object, and call C<next> on it.
1223 Calling C<< resultset('Table')->next >> repeatedly will always return the
1224 first record from the resultset.
1231 if (my $cache = $self->get_cache) {
1232 $self->{all_cache_position} ||= 0;
1233 return $cache->[$self->{all_cache_position}++];
1236 if ($self->{attrs}{cache}) {
1237 delete $self->{pager};
1238 $self->{all_cache_position} = 1;
1239 return ($self->all)[0];
1242 return shift(@{$self->{stashed_objects}}) if @{ $self->{stashed_objects}||[] };
1244 $self->{stashed_objects} = $self->_construct_objects
1247 return shift @{$self->{stashed_objects}};
1250 # Constructs as many objects as it can in one pass while respecting
1251 # cursor laziness. Several modes of operation:
1253 # * Always builds everything present in @{$self->{stashed_rows}}
1254 # * If called with $fetch_all true - pulls everything off the cursor and
1255 # builds all objects in one pass
1256 # * If $self->_resolved_attrs->{collapse} is true, checks the order_by
1257 # and if the resultset is ordered properly by the left side:
1258 # * Fetches stuff off the cursor until the "master object" changes,
1259 # and saves the last extra row (if any) in @{$self->{stashed_rows}}
1261 # * Just fetches, and collapses/constructs everything as if $fetch_all
1262 # was requested (there is no other way to collapse except for an
1264 # * If no collapse is requested - just get the next row, construct and
1266 sub _construct_objects {
1267 my ($self, $fetch_all) = @_;
1269 my $rsrc = $self->result_source;
1270 my $attrs = $self->_resolved_attrs;
1271 my $cursor = $self->cursor;
1273 # this will be used as both initial raw-row collector AND as a RV of
1274 # _construct_objects. Not regrowing the array twice matters a lot...
1275 # a suprising amount actually
1276 my $rows = (delete $self->{stashed_rows}) || [];
1278 # FIXME SUBOPTIMAL - we can do better, cursor->next/all (well diff. methods) should return a ref
1279 $rows = [ @$rows, $cursor->all ];
1281 elsif (!$attrs->{collapse}) {
1282 # FIXME SUBOPTIMAL - we can do better, cursor->next/all (well diff. methods) should return a ref
1283 push @$rows, do { my @r = $cursor->next; @r ? \@r : () }
1287 $attrs->{_ordered_for_collapse} ||= (!$attrs->{order_by}) ? undef : do {
1288 my $st = $rsrc->schema->storage;
1291 ( $st->_extract_order_criteria($attrs->{order_by}) )
1294 my $colinfos = $st->_resolve_column_info($attrs->{from}, \@ord_cols);
1296 for (0 .. $#ord_cols) {
1298 ! $colinfos->{$ord_cols[$_]}
1300 $colinfos->{$ord_cols[$_]}{-result_source} != $rsrc
1302 splice @ord_cols, $_;
1307 # since all we check here are the start of the order_by belonging to the
1308 # top level $rsrc, a present identifying set will mean that the resultset
1309 # is ordered by its leftmost table in a tsable manner
1310 (@ord_cols and $rsrc->_identifying_column_set({ map
1311 { $colinfos->{$_}{-colname} => $colinfos->{$_} }
1316 if ($attrs->{_ordered_for_collapse}) {
1317 push @$rows, do { my @r = $cursor->next; @r ? \@r : () };
1319 # instead of looping over ->next, use ->all in stealth mode
1320 # *without* calling a ->reset afterwards
1321 # FIXME - encapsulation breach, got to be a better way
1322 elsif (! $cursor->{_done}) {
1323 push @$rows, $cursor->all;
1324 $cursor->{_done} = 1;
1329 return undef unless @$rows;
1331 my $res_class = $self->result_class;
1332 my $inflator = $res_class->can ('inflate_result')
1333 or $self->throw_exception("Inflator $res_class does not provide an inflate_result() method");
1335 my $infmap = $attrs->{as};
1337 if (!$attrs->{collapse} and $attrs->{_single_object_inflation}) {
1338 # construct a much simpler array->hash folder for the one-table cases right here
1340 # FIXME SUBOPTIMAL this is a very very very hot spot
1341 # while rather optimal we can *still* do much better, by
1342 # building a smarter [Row|HRI]::inflate_result(), and
1343 # switch to feeding it data via a much leaner interface
1345 # crude unscientific benchmarking indicated the shortcut eval is not worth it for
1346 # this particular resultset size
1348 my @as_idx = 0..$#$infmap;
1349 for my $r (@$rows) {
1350 $r = $inflator->($res_class, $rsrc, { map { $infmap->[$_] => $r->[$_] } @as_idx } );
1355 '$_ = $inflator->($res_class, $rsrc, { %s }) for @$rows',
1356 join (', ', map { "\$infmap->[$_] => \$_->[$_]" } 0..$#$infmap )
1361 $self->{_row_parser} ||= eval sprintf 'sub { %s }', $rsrc->_mk_row_parser({
1362 inflate_map => $infmap,
1363 selection => $attrs->{select},
1364 collapse => $attrs->{collapse},
1365 premultiplied => $attrs->{_main_source_premultiplied},
1368 # modify $rows in-place, shrinking/extending as necessary
1369 $self->{_row_parser}->($rows, $fetch_all ? () : (
1370 # FIXME SUBOPTIMAL - we can do better, cursor->next/all (well diff. methods) should return a ref
1371 sub { my @r = $cursor->next or return; \@r }, # how the collapser gets more rows
1372 ($self->{stashed_rows} = []), # where does it stuff excess
1375 $_ = $inflator->($res_class, $rsrc, @$_) for @$rows;
1379 if ($attrs->{record_filter}) {
1380 $_ = $attrs->{record_filter}->($_) for @$rows;
1386 =head2 result_source
1390 =item Arguments: L<$result_source?|DBIx::Class::ResultSource>
1392 =item Return Value: L<$result_source|DBIx::Class::ResultSource>
1396 An accessor for the primary ResultSource object from which this ResultSet
1403 =item Arguments: $result_class?
1405 =item Return Value: $result_class
1409 An accessor for the class to use when creating result objects. Defaults to
1410 C<< result_source->result_class >> - which in most cases is the name of the
1411 L<"table"|DBIx::Class::Manual::Glossary/"ResultSource"> class.
1413 Note that changing the result_class will also remove any components
1414 that were originally loaded in the source class via
1415 L<DBIx::Class::ResultSource/load_components>. Any overloaded methods
1416 in the original source class will not run.
1421 my ($self, $result_class) = @_;
1422 if ($result_class) {
1423 unless (ref $result_class) { # don't fire this for an object
1424 $self->ensure_class_loaded($result_class);
1426 $self->_result_class($result_class);
1427 # THIS LINE WOULD BE A BUG - this accessor specifically exists to
1428 # permit the user to set result class on one result set only; it only
1429 # chains if provided to search()
1430 #$self->{attrs}{result_class} = $result_class if ref $self;
1432 $self->_result_class;
1439 =item Arguments: L<$cond|DBIx::Class::SQLMaker>, L<\%attrs?|/ATTRIBUTES>
1441 =item Return Value: $count
1445 Performs an SQL C<COUNT> with the same query as the resultset was built
1446 with to find the number of elements. Passing arguments is equivalent to
1447 C<< $rs->search ($cond, \%attrs)->count >>
1453 return $self->search(@_)->count if @_ and defined $_[0];
1454 return scalar @{ $self->get_cache } if $self->get_cache;
1456 my $attrs = { %{ $self->_resolved_attrs } };
1458 # this is a little optimization - it is faster to do the limit
1459 # adjustments in software, instead of a subquery
1460 my ($rows, $offset) = delete @{$attrs}{qw/rows offset/};
1463 if ($self->_has_resolved_attr (qw/collapse group_by/)) {
1464 $crs = $self->_count_subq_rs ($attrs);
1467 $crs = $self->_count_rs ($attrs);
1469 my $count = $crs->next;
1471 $count -= $offset if $offset;
1472 $count = $rows if $rows and $rows < $count;
1473 $count = 0 if ($count < 0);
1482 =item Arguments: L<$cond|DBIx::Class::SQLMaker>, L<\%attrs?|/ATTRIBUTES>
1484 =item Return Value: L<$count_rs|DBIx::Class::ResultSetColumn>
1488 Same as L</count> but returns a L<DBIx::Class::ResultSetColumn> object.
1489 This can be very handy for subqueries:
1491 ->search( { amount => $some_rs->count_rs->as_query } )
1493 As with regular resultsets the SQL query will be executed only after
1494 the resultset is accessed via L</next> or L</all>. That would return
1495 the same single value obtainable via L</count>.
1501 return $self->search(@_)->count_rs if @_;
1503 # this may look like a lack of abstraction (count() does about the same)
1504 # but in fact an _rs *must* use a subquery for the limits, as the
1505 # software based limiting can not be ported if this $rs is to be used
1506 # in a subquery itself (i.e. ->as_query)
1507 if ($self->_has_resolved_attr (qw/collapse group_by offset rows/)) {
1508 return $self->_count_subq_rs;
1511 return $self->_count_rs;
1516 # returns a ResultSetColumn object tied to the count query
1519 my ($self, $attrs) = @_;
1521 my $rsrc = $self->result_source;
1522 $attrs ||= $self->_resolved_attrs;
1524 my $tmp_attrs = { %$attrs };
1525 # take off any limits, record_filter is cdbi, and no point of ordering nor locking a count
1526 delete @{$tmp_attrs}{qw/rows offset order_by record_filter for/};
1528 # overwrite the selector (supplied by the storage)
1529 $tmp_attrs->{select} = $rsrc->storage->_count_select ($rsrc, $attrs);
1530 $tmp_attrs->{as} = 'count';
1532 my $tmp_rs = $rsrc->resultset_class->new($rsrc, $tmp_attrs)->get_column ('count');
1538 # same as above but uses a subquery
1540 sub _count_subq_rs {
1541 my ($self, $attrs) = @_;
1543 my $rsrc = $self->result_source;
1544 $attrs ||= $self->_resolved_attrs;
1546 my $sub_attrs = { %$attrs };
1547 # extra selectors do not go in the subquery and there is no point of ordering it, nor locking it
1548 delete @{$sub_attrs}{qw/collapse columns as select _prefetch_selector_range order_by for/};
1550 # if we multi-prefetch we group_by something unique, as this is what we would
1551 # get out of the rs via ->next/->all. We *DO WANT* to clobber old group_by regardless
1552 if ( $attrs->{collapse} ) {
1553 $sub_attrs->{group_by} = [ map { "$attrs->{alias}.$_" } @{
1554 $rsrc->_identifying_column_set || $self->throw_exception(
1555 'Unable to construct a unique group_by criteria properly collapsing the '
1556 . 'has_many prefetch before count()'
1561 # Calculate subquery selector
1562 if (my $g = $sub_attrs->{group_by}) {
1564 my $sql_maker = $rsrc->storage->sql_maker;
1566 # necessary as the group_by may refer to aliased functions
1568 for my $sel (@{$attrs->{select}}) {
1569 $sel_index->{$sel->{-as}} = $sel
1570 if (ref $sel eq 'HASH' and $sel->{-as});
1573 # anything from the original select mentioned on the group-by needs to make it to the inner selector
1574 # also look for named aggregates referred in the having clause
1575 # having often contains scalarrefs - thus parse it out entirely
1577 if ($attrs->{having}) {
1578 local $sql_maker->{having_bind};
1579 local $sql_maker->{quote_char} = $sql_maker->{quote_char};
1580 local $sql_maker->{name_sep} = $sql_maker->{name_sep};
1581 unless (defined $sql_maker->{quote_char} and length $sql_maker->{quote_char}) {
1582 $sql_maker->{quote_char} = [ "\x00", "\xFF" ];
1583 # if we don't unset it we screw up retarded but unfortunately working
1584 # 'MAX(foo.bar)' => { '>', 3 }
1585 $sql_maker->{name_sep} = '';
1588 my ($lquote, $rquote, $sep) = map { quotemeta $_ } ($sql_maker->_quote_chars, $sql_maker->name_sep);
1590 my $having_sql = $sql_maker->_parse_rs_attrs ({ having => $attrs->{having} });
1593 # search for both a proper quoted qualified string, for a naive unquoted scalarref
1594 # and if all fails for an utterly naive quoted scalar-with-function
1595 while ($having_sql =~ /
1596 $rquote $sep $lquote (.+?) $rquote
1598 [\s,] \w+ \. (\w+) [\s,]
1600 [\s,] $lquote (.+?) $rquote [\s,]
1602 my $part = $1 || $2 || $3; # one of them matched if we got here
1603 unless ($seen_having{$part}++) {
1610 my $colpiece = $sel_index->{$_} || $_;
1612 # unqualify join-based group_by's. Arcane but possible query
1613 # also horrible horrible hack to alias a column (not a func.)
1614 # (probably need to introduce SQLA syntax)
1615 if ($colpiece =~ /\./ && $colpiece !~ /^$attrs->{alias}\./) {
1618 $colpiece = \ sprintf ('%s AS %s', map { $sql_maker->_quote ($_) } ($colpiece, $as) );
1620 push @{$sub_attrs->{select}}, $colpiece;
1624 my @pcols = map { "$attrs->{alias}.$_" } ($rsrc->primary_columns);
1625 $sub_attrs->{select} = @pcols ? \@pcols : [ 1 ];
1628 return $rsrc->resultset_class
1629 ->new ($rsrc, $sub_attrs)
1631 ->search ({}, { columns => { count => $rsrc->storage->_count_select ($rsrc, $attrs) } })
1632 ->get_column ('count');
1639 =head2 count_literal
1641 B<CAVEAT>: C<count_literal> is provided for Class::DBI compatibility and
1642 should only be used in that context. See L</search_literal> for further info.
1646 =item Arguments: $sql_fragment, @standalone_bind_values
1648 =item Return Value: $count
1652 Counts the results in a literal query. Equivalent to calling L</search_literal>
1653 with the passed arguments, then L</count>.
1657 sub count_literal { shift->search_literal(@_)->count; }
1663 =item Arguments: none
1665 =item Return Value: L<@result_objs|DBIx::Class::Manual::ResultClass>
1669 Returns all elements in the resultset.
1676 $self->throw_exception("all() doesn't take any arguments, you probably wanted ->search(...)->all()");
1679 delete @{$self}{qw/stashed_rows stashed_objects/};
1681 if (my $c = $self->get_cache) {
1685 $self->cursor->reset;
1687 my $objs = $self->_construct_objects('fetch_all') || [];
1689 $self->set_cache($objs) if $self->{attrs}{cache};
1698 =item Arguments: none
1700 =item Return Value: $self
1704 Resets the resultset's cursor, so you can iterate through the elements again.
1705 Implicitly resets the storage cursor, so a subsequent L</next> will trigger
1713 delete @{$self}{qw/stashed_rows stashed_objects/};
1714 $self->{all_cache_position} = 0;
1715 $self->cursor->reset;
1723 =item Arguments: none
1725 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass> | undef
1729 L<Resets|/reset> the resultset (causing a fresh query to storage) and returns
1730 an object for the first result (or C<undef> if the resultset is empty).
1735 return $_[0]->reset->next;
1741 # Determines whether and what type of subquery is required for the $rs operation.
1742 # If grouping is necessary either supplies its own, or verifies the current one
1743 # After all is done delegates to the proper storage method.
1745 sub _rs_update_delete {
1746 my ($self, $op, $values) = @_;
1748 my $rsrc = $self->result_source;
1749 my $storage = $rsrc->schema->storage;
1751 my $attrs = { %{$self->_resolved_attrs} };
1753 my $join_classifications;
1754 my $existing_group_by = delete $attrs->{group_by};
1756 # do we need a subquery for any reason?
1758 defined $existing_group_by
1760 # if {from} is unparseable wrap a subq
1761 ref($attrs->{from}) ne 'ARRAY'
1763 # limits call for a subq
1764 $self->_has_resolved_attr(qw/rows offset/)
1767 # simplify the joinmap, so we can further decide if a subq is necessary
1768 if (!$needs_subq and @{$attrs->{from}} > 1) {
1769 $attrs->{from} = $storage->_prune_unused_joins ($attrs->{from}, $attrs->{select}, $self->{cond}, $attrs);
1771 # check if there are any joins left after the prune
1772 if ( @{$attrs->{from}} > 1 ) {
1773 $join_classifications = $storage->_resolve_aliastypes_from_select_args (
1774 [ @{$attrs->{from}}[1 .. $#{$attrs->{from}}] ],
1780 # any non-pruneable joins imply subq
1781 $needs_subq = scalar keys %{ $join_classifications->{restricting} || {} };
1785 # check if the head is composite (by now all joins are thrown out unless $needs_subq)
1787 (ref $attrs->{from}[0]) ne 'HASH'
1789 ref $attrs->{from}[0]{ $attrs->{from}[0]{-alias} }
1793 # do we need anything like a subquery?
1794 if (! $needs_subq) {
1795 # Most databases do not allow aliasing of tables in UPDATE/DELETE. Thus
1796 # a condition containing 'me' or other table prefixes will not work
1797 # at all. Tell SQLMaker to dequalify idents via a gross hack.
1799 my $sqla = $rsrc->storage->sql_maker;
1800 local $sqla->{_dequalify_idents} = 1;
1801 \[ $sqla->_recurse_where($self->{cond}) ];
1805 # we got this far - means it is time to wrap a subquery
1806 my $idcols = $rsrc->_identifying_column_set || $self->throw_exception(
1808 "Unable to perform complex resultset %s() without an identifying set of columns on source '%s'",
1814 # make a new $rs selecting only the PKs (that's all we really need for the subq)
1815 delete $attrs->{$_} for qw/collapse select _prefetch_selector_range as/;
1816 $attrs->{columns} = [ map { "$attrs->{alias}.$_" } @$idcols ];
1817 $attrs->{group_by} = \ ''; # FIXME - this is an evil hack, it causes the optimiser to kick in and throw away the LEFT joins
1818 my $subrs = (ref $self)->new($rsrc, $attrs);
1820 if (@$idcols == 1) {
1821 $cond = { $idcols->[0] => { -in => $subrs->as_query } };
1823 elsif ($storage->_use_multicolumn_in) {
1824 # no syntax for calling this properly yet
1825 # !!! EXPERIMENTAL API !!! WILL CHANGE !!!
1826 $cond = $storage->sql_maker->_where_op_multicolumn_in (
1827 $idcols, # how do I convey a list of idents...? can binds reside on lhs?
1832 # if all else fails - get all primary keys and operate over a ORed set
1833 # wrap in a transaction for consistency
1834 # this is where the group_by/multiplication starts to matter
1838 keys %{ $join_classifications->{multiplying} || {} }
1840 # make sure if there is a supplied group_by it matches the columns compiled above
1841 # perfectly. Anything else can not be sanely executed on most databases so croak
1842 # right then and there
1843 if ($existing_group_by) {
1844 my @current_group_by = map
1845 { $_ =~ /\./ ? $_ : "$attrs->{alias}.$_" }
1850 join ("\x00", sort @current_group_by)
1852 join ("\x00", sort @{$attrs->{columns}} )
1854 $self->throw_exception (
1855 "You have just attempted a $op operation on a resultset which does group_by"
1856 . ' on columns other than the primary keys, while DBIC internally needs to retrieve'
1857 . ' the primary keys in a subselect. All sane RDBMS engines do not support this'
1858 . ' kind of queries. Please retry the operation with a modified group_by or'
1859 . ' without using one at all.'
1864 $subrs = $subrs->search({}, { group_by => $attrs->{columns} });
1867 $guard = $storage->txn_scope_guard;
1870 for my $row ($subrs->cursor->all) {
1872 { $idcols->[$_] => $row->[$_] }
1879 my $res = $storage->$op (
1881 $op eq 'update' ? $values : (),
1885 $guard->commit if $guard;
1894 =item Arguments: \%values
1896 =item Return Value: $underlying_storage_rv
1900 Sets the specified columns in the resultset to the supplied values in a
1901 single query. Note that this will not run any accessor/set_column/update
1902 triggers, nor will it update any result object instances derived from this
1903 resultset (this includes the contents of the L<resultset cache|/set_cache>
1904 if any). See L</update_all> if you need to execute any on-update
1905 triggers or cascades defined either by you or a
1906 L<result component|DBIx::Class::Manual::Component/WHAT IS A COMPONENT>.
1908 The return value is a pass through of what the underlying
1909 storage backend returned, and may vary. See L<DBI/execute> for the most
1914 Note that L</update> does not process/deflate any of the values passed in.
1915 This is unlike the corresponding L<DBIx::Class::Row/update>. The user must
1916 ensure manually that any value passed to this method will stringify to
1917 something the RDBMS knows how to deal with. A notable example is the
1918 handling of L<DateTime> objects, for more info see:
1919 L<DBIx::Class::Manual::Cookbook/Formatting DateTime objects in queries>.
1924 my ($self, $values) = @_;
1925 $self->throw_exception('Values for update must be a hash')
1926 unless ref $values eq 'HASH';
1928 return $self->_rs_update_delete ('update', $values);
1935 =item Arguments: \%values
1937 =item Return Value: 1
1941 Fetches all objects and updates them one at a time via
1942 L<DBIx::Class::Row/update>. Note that C<update_all> will run DBIC defined
1943 triggers, while L</update> will not.
1948 my ($self, $values) = @_;
1949 $self->throw_exception('Values for update_all must be a hash')
1950 unless ref $values eq 'HASH';
1952 my $guard = $self->result_source->schema->txn_scope_guard;
1953 $_->update({%$values}) for $self->all; # shallow copy - update will mangle it
1962 =item Arguments: none
1964 =item Return Value: $underlying_storage_rv
1968 Deletes the rows matching this resultset in a single query. Note that this
1969 will not run any delete triggers, nor will it alter the
1970 L<in_storage|DBIx::Class::Row/in_storage> status of any result object instances
1971 derived from this resultset (this includes the contents of the
1972 L<resultset cache|/set_cache> if any). See L</delete_all> if you need to
1973 execute any on-delete triggers or cascades defined either by you or a
1974 L<result component|DBIx::Class::Manual::Component/WHAT IS A COMPONENT>.
1976 The return value is a pass through of what the underlying storage backend
1977 returned, and may vary. See L<DBI/execute> for the most common case.
1983 $self->throw_exception('delete does not accept any arguments')
1986 return $self->_rs_update_delete ('delete');
1993 =item Arguments: none
1995 =item Return Value: 1
1999 Fetches all objects and deletes them one at a time via
2000 L<DBIx::Class::Row/delete>. Note that C<delete_all> will run DBIC defined
2001 triggers, while L</delete> will not.
2007 $self->throw_exception('delete_all does not accept any arguments')
2010 my $guard = $self->result_source->schema->txn_scope_guard;
2011 $_->delete for $self->all;
2020 =item Arguments: [ \@column_list, \@row_values+ ] | [ \%col_data+ ]
2022 =item Return Value: L<\@result_objects|DBIx::Class::Manual::ResultClass> (scalar context) | L<@result_objects|DBIx::Class::Manual::ResultClass> (list context)
2026 Accepts either an arrayref of hashrefs or alternatively an arrayref of
2033 The context of this method call has an important effect on what is
2034 submitted to storage. In void context data is fed directly to fastpath
2035 insertion routines provided by the underlying storage (most often
2036 L<DBI/execute_for_fetch>), bypassing the L<new|DBIx::Class::Row/new> and
2037 L<insert|DBIx::Class::Row/insert> calls on the
2038 L<Result|DBIx::Class::Manual::ResultClass> class, including any
2039 augmentation of these methods provided by components. For example if you
2040 are using something like L<DBIx::Class::UUIDColumns> to create primary
2041 keys for you, you will find that your PKs are empty. In this case you
2042 will have to explicitly force scalar or list context in order to create
2047 In non-void (scalar or list) context, this method is simply a wrapper
2048 for L</create>. Depending on list or scalar context either a list of
2049 L<Result|DBIx::Class::Manual::ResultClass> objects or an arrayref
2050 containing these objects is returned.
2052 When supplying data in "arrayref of arrayrefs" invocation style, the
2053 first element should be a list of column names and each subsequent
2054 element should be a data value in the earlier specified column order.
2057 $Arstist_rs->populate([
2058 [ qw( artistid name ) ],
2059 [ 100, 'A Formally Unknown Singer' ],
2060 [ 101, 'A singer that jumped the shark two albums ago' ],
2061 [ 102, 'An actually cool singer' ],
2064 For the arrayref of hashrefs style each hashref should be a structure
2065 suitable for passing to L</create>. Multi-create is also permitted with
2068 $schema->resultset("Artist")->populate([
2069 { artistid => 4, name => 'Manufactured Crap', cds => [
2070 { title => 'My First CD', year => 2006 },
2071 { title => 'Yet More Tweeny-Pop crap', year => 2007 },
2074 { artistid => 5, name => 'Angsty-Whiny Girl', cds => [
2075 { title => 'My parents sold me to a record company', year => 2005 },
2076 { title => 'Why Am I So Ugly?', year => 2006 },
2077 { title => 'I Got Surgery and am now Popular', year => 2007 }
2082 If you attempt a void-context multi-create as in the example above (each
2083 Artist also has the related list of CDs), and B<do not> supply the
2084 necessary autoinc foreign key information, this method will proxy to the
2085 less efficient L</create>, and then throw the Result objects away. In this
2086 case there are obviously no benefits to using this method over L</create>.
2093 # cruft placed in standalone method
2094 my $data = $self->_normalize_populate_args(@_);
2096 return unless @$data;
2098 if(defined wantarray) {
2099 my @created = map { $self->create($_) } @$data;
2100 return wantarray ? @created : \@created;
2103 my $first = $data->[0];
2105 # if a column is a registered relationship, and is a non-blessed hash/array, consider
2106 # it relationship data
2107 my (@rels, @columns);
2108 my $rsrc = $self->result_source;
2109 my $rels = { map { $_ => $rsrc->relationship_info($_) } $rsrc->relationships };
2110 for (keys %$first) {
2111 my $ref = ref $first->{$_};
2112 $rels->{$_} && ($ref eq 'ARRAY' or $ref eq 'HASH')
2118 my @pks = $rsrc->primary_columns;
2120 ## do the belongs_to relationships
2121 foreach my $index (0..$#$data) {
2123 # delegate to create() for any dataset without primary keys with specified relationships
2124 if (grep { !defined $data->[$index]->{$_} } @pks ) {
2126 if (grep { ref $data->[$index]{$r} eq $_ } qw/HASH ARRAY/) { # a related set must be a HASH or AoH
2127 my @ret = $self->populate($data);
2133 foreach my $rel (@rels) {
2134 next unless ref $data->[$index]->{$rel} eq "HASH";
2135 my $result = $self->related_resultset($rel)->create($data->[$index]->{$rel});
2136 my ($reverse_relname, $reverse_relinfo) = %{$rsrc->reverse_relationship_info($rel)};
2137 my $related = $result->result_source->_resolve_condition(
2138 $reverse_relinfo->{cond},
2144 delete $data->[$index]->{$rel};
2145 $data->[$index] = {%{$data->[$index]}, %$related};
2147 push @columns, keys %$related if $index == 0;
2151 ## inherit the data locked in the conditions of the resultset
2152 my ($rs_data) = $self->_merge_with_rscond({});
2153 delete @{$rs_data}{@columns};
2155 ## do bulk insert on current row
2156 $rsrc->storage->insert_bulk(
2158 [@columns, keys %$rs_data],
2159 [ map { [ @$_{@columns}, values %$rs_data ] } @$data ],
2162 ## do the has_many relationships
2163 foreach my $item (@$data) {
2167 foreach my $rel (@rels) {
2168 next unless ref $item->{$rel} eq "ARRAY" && @{ $item->{$rel} };
2170 $main_row ||= $self->new_result({map { $_ => $item->{$_} } @pks});
2172 my $child = $main_row->$rel;
2174 my $related = $child->result_source->_resolve_condition(
2175 $rels->{$rel}{cond},
2181 my @rows_to_add = ref $item->{$rel} eq 'ARRAY' ? @{$item->{$rel}} : ($item->{$rel});
2182 my @populate = map { {%$_, %$related} } @rows_to_add;
2184 $child->populate( \@populate );
2191 # populate() argumnets went over several incarnations
2192 # What we ultimately support is AoH
2193 sub _normalize_populate_args {
2194 my ($self, $arg) = @_;
2196 if (ref $arg eq 'ARRAY') {
2200 elsif (ref $arg->[0] eq 'HASH') {
2203 elsif (ref $arg->[0] eq 'ARRAY') {
2205 my @colnames = @{$arg->[0]};
2206 foreach my $values (@{$arg}[1 .. $#$arg]) {
2207 push @ret, { map { $colnames[$_] => $values->[$_] } (0 .. $#colnames) };
2213 $self->throw_exception('Populate expects an arrayref of hashrefs or arrayref of arrayrefs');
2220 =item Arguments: none
2222 =item Return Value: L<$pager|Data::Page>
2226 Returns a L<Data::Page> object for the current resultset. Only makes
2227 sense for queries with a C<page> attribute.
2229 To get the full count of entries for a paged resultset, call
2230 C<total_entries> on the L<Data::Page> object.
2237 return $self->{pager} if $self->{pager};
2239 my $attrs = $self->{attrs};
2240 if (!defined $attrs->{page}) {
2241 $self->throw_exception("Can't create pager for non-paged rs");
2243 elsif ($attrs->{page} <= 0) {
2244 $self->throw_exception('Invalid page number (page-numbers are 1-based)');
2246 $attrs->{rows} ||= 10;
2248 # throw away the paging flags and re-run the count (possibly
2249 # with a subselect) to get the real total count
2250 my $count_attrs = { %$attrs };
2251 delete @{$count_attrs}{qw/rows offset page pager/};
2253 my $total_rs = (ref $self)->new($self->result_source, $count_attrs);
2255 require DBIx::Class::ResultSet::Pager;
2256 return $self->{pager} = DBIx::Class::ResultSet::Pager->new(
2257 sub { $total_rs->count }, #lazy-get the total
2259 $self->{attrs}{page},
2267 =item Arguments: $page_number
2269 =item Return Value: L<$resultset|/search>
2273 Returns a resultset for the $page_number page of the resultset on which page
2274 is called, where each page contains a number of rows equal to the 'rows'
2275 attribute set on the resultset (10 by default).
2280 my ($self, $page) = @_;
2281 return (ref $self)->new($self->result_source, { %{$self->{attrs}}, page => $page });
2288 =item Arguments: \%col_data
2290 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass>
2294 Creates a new result object in the resultset's result class and returns
2295 it. The row is not inserted into the database at this point, call
2296 L<DBIx::Class::Row/insert> to do that. Calling L<DBIx::Class::Row/in_storage>
2297 will tell you whether the result object has been inserted or not.
2299 Passes the hashref of input on to L<DBIx::Class::Row/new>.
2304 my ($self, $values) = @_;
2306 $self->throw_exception( "new_result takes only one argument - a hashref of values" )
2309 $self->throw_exception( "new_result expects a hashref" )
2310 unless (ref $values eq 'HASH');
2312 my ($merged_cond, $cols_from_relations) = $self->_merge_with_rscond($values);
2316 @$cols_from_relations
2317 ? (-cols_from_relations => $cols_from_relations)
2319 -result_source => $self->result_source, # DO NOT REMOVE THIS, REQUIRED
2322 return $self->result_class->new(\%new);
2325 # _merge_with_rscond
2327 # Takes a simple hash of K/V data and returns its copy merged with the
2328 # condition already present on the resultset. Additionally returns an
2329 # arrayref of value/condition names, which were inferred from related
2330 # objects (this is needed for in-memory related objects)
2331 sub _merge_with_rscond {
2332 my ($self, $data) = @_;
2334 my (%new_data, @cols_from_relations);
2336 my $alias = $self->{attrs}{alias};
2338 if (! defined $self->{cond}) {
2339 # just massage $data below
2341 elsif ($self->{cond} eq $DBIx::Class::ResultSource::UNRESOLVABLE_CONDITION) {
2342 %new_data = %{ $self->{attrs}{related_objects} || {} }; # nothing might have been inserted yet
2343 @cols_from_relations = keys %new_data;
2345 elsif (ref $self->{cond} ne 'HASH') {
2346 $self->throw_exception(
2347 "Can't abstract implicit construct, resultset condition not a hash"
2351 # precendence must be given to passed values over values inherited from
2352 # the cond, so the order here is important.
2353 my $collapsed_cond = $self->_collapse_cond($self->{cond});
2354 my %implied = %{$self->_remove_alias($collapsed_cond, $alias)};
2356 while ( my($col, $value) = each %implied ) {
2357 my $vref = ref $value;
2363 (keys %$value)[0] eq '='
2365 $new_data{$col} = $value->{'='};
2367 elsif( !$vref or $vref eq 'SCALAR' or blessed($value) ) {
2368 $new_data{$col} = $value;
2375 %{ $self->_remove_alias($data, $alias) },
2378 return (\%new_data, \@cols_from_relations);
2381 # _has_resolved_attr
2383 # determines if the resultset defines at least one
2384 # of the attributes supplied
2386 # used to determine if a subquery is neccessary
2388 # supports some virtual attributes:
2390 # This will scan for any joins being present on the resultset.
2391 # It is not a mere key-search but a deep inspection of {from}
2394 sub _has_resolved_attr {
2395 my ($self, @attr_names) = @_;
2397 my $attrs = $self->_resolved_attrs;
2401 for my $n (@attr_names) {
2402 if (grep { $n eq $_ } (qw/-join/) ) {
2403 $extra_checks{$n}++;
2407 my $attr = $attrs->{$n};
2409 next if not defined $attr;
2411 if (ref $attr eq 'HASH') {
2412 return 1 if keys %$attr;
2414 elsif (ref $attr eq 'ARRAY') {
2422 # a resolved join is expressed as a multi-level from
2424 $extra_checks{-join}
2426 ref $attrs->{from} eq 'ARRAY'
2428 @{$attrs->{from}} > 1
2436 # Recursively collapse the condition.
2438 sub _collapse_cond {
2439 my ($self, $cond, $collapsed) = @_;
2443 if (ref $cond eq 'ARRAY') {
2444 foreach my $subcond (@$cond) {
2445 next unless ref $subcond; # -or
2446 $collapsed = $self->_collapse_cond($subcond, $collapsed);
2449 elsif (ref $cond eq 'HASH') {
2450 if (keys %$cond and (keys %$cond)[0] eq '-and') {
2451 foreach my $subcond (@{$cond->{-and}}) {
2452 $collapsed = $self->_collapse_cond($subcond, $collapsed);
2456 foreach my $col (keys %$cond) {
2457 my $value = $cond->{$col};
2458 $collapsed->{$col} = $value;
2468 # Remove the specified alias from the specified query hash. A copy is made so
2469 # the original query is not modified.
2472 my ($self, $query, $alias) = @_;
2474 my %orig = %{ $query || {} };
2477 foreach my $key (keys %orig) {
2479 $unaliased{$key} = $orig{$key};
2482 $unaliased{$1} = $orig{$key}
2483 if $key =~ m/^(?:\Q$alias\E\.)?([^.]+)$/;
2493 =item Arguments: none
2495 =item Return Value: \[ $sql, L<@bind_values|/DBIC BIND VALUES> ]
2499 Returns the SQL query and bind vars associated with the invocant.
2501 This is generally used as the RHS for a subquery.
2508 my $attrs = { %{ $self->_resolved_attrs } };
2513 # my ($sql, \@bind, \%dbi_bind_attrs) = _select_args_to_query (...)
2514 # $sql also has no wrapping parenthesis in list ctx
2516 my $sqlbind = $self->result_source->storage
2517 ->_select_args_to_query ($attrs->{from}, $attrs->{select}, $attrs->{where}, $attrs);
2526 =item Arguments: \%col_data, { key => $unique_constraint, L<%attrs|/ATTRIBUTES> }?
2528 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass>
2532 my $artist = $schema->resultset('Artist')->find_or_new(
2533 { artist => 'fred' }, { key => 'artists' });
2535 $cd->cd_to_producer->find_or_new({ producer => $producer },
2536 { key => 'primary });
2538 Find an existing record from this resultset using L</find>. if none exists,
2539 instantiate a new result object and return it. The object will not be saved
2540 into your storage until you call L<DBIx::Class::Row/insert> on it.
2542 You most likely want this method when looking for existing rows using a unique
2543 constraint that is not the primary key, or looking for related rows.
2545 If you want objects to be saved immediately, use L</find_or_create> instead.
2547 B<Note>: Make sure to read the documentation of L</find> and understand the
2548 significance of the C<key> attribute, as its lack may skew your search, and
2549 subsequently result in spurious new objects.
2551 B<Note>: Take care when using C<find_or_new> with a table having
2552 columns with default values that you intend to be automatically
2553 supplied by the database (e.g. an auto_increment primary key column).
2554 In normal usage, the value of such columns should NOT be included at
2555 all in the call to C<find_or_new>, even when set to C<undef>.
2561 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
2562 my $hash = ref $_[0] eq 'HASH' ? shift : {@_};
2563 if (keys %$hash and my $row = $self->find($hash, $attrs) ) {
2566 return $self->new_result($hash);
2573 =item Arguments: \%col_data
2575 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass>
2579 Attempt to create a single new row or a row with multiple related rows
2580 in the table represented by the resultset (and related tables). This
2581 will not check for duplicate rows before inserting, use
2582 L</find_or_create> to do that.
2584 To create one row for this resultset, pass a hashref of key/value
2585 pairs representing the columns of the table and the values you wish to
2586 store. If the appropriate relationships are set up, foreign key fields
2587 can also be passed an object representing the foreign row, and the
2588 value will be set to its primary key.
2590 To create related objects, pass a hashref of related-object column values
2591 B<keyed on the relationship name>. If the relationship is of type C<multi>
2592 (L<DBIx::Class::Relationship/has_many>) - pass an arrayref of hashrefs.
2593 The process will correctly identify columns holding foreign keys, and will
2594 transparently populate them from the keys of the corresponding relation.
2595 This can be applied recursively, and will work correctly for a structure
2596 with an arbitrary depth and width, as long as the relationships actually
2597 exists and the correct column data has been supplied.
2599 Instead of hashrefs of plain related data (key/value pairs), you may
2600 also pass new or inserted objects. New objects (not inserted yet, see
2601 L</new_result>), will be inserted into their appropriate tables.
2603 Effectively a shortcut for C<< ->new_result(\%col_data)->insert >>.
2605 Example of creating a new row.
2607 $person_rs->create({
2608 name=>"Some Person",
2609 email=>"somebody@someplace.com"
2612 Example of creating a new row and also creating rows in a related C<has_many>
2613 or C<has_one> resultset. Note Arrayref.
2616 { artistid => 4, name => 'Manufactured Crap', cds => [
2617 { title => 'My First CD', year => 2006 },
2618 { title => 'Yet More Tweeny-Pop crap', year => 2007 },
2623 Example of creating a new row and also creating a row in a related
2624 C<belongs_to> resultset. Note Hashref.
2627 title=>"Music for Silly Walks",
2630 name=>"Silly Musician",
2638 When subclassing ResultSet never attempt to override this method. Since
2639 it is a simple shortcut for C<< $self->new_result($attrs)->insert >>, a
2640 lot of the internals simply never call it, so your override will be
2641 bypassed more often than not. Override either L<DBIx::Class::Row/new>
2642 or L<DBIx::Class::Row/insert> depending on how early in the
2643 L</create> process you need to intervene. See also warning pertaining to
2651 my ($self, $attrs) = @_;
2652 $self->throw_exception( "create needs a hashref" )
2653 unless ref $attrs eq 'HASH';
2654 return $self->new_result($attrs)->insert;
2657 =head2 find_or_create
2661 =item Arguments: \%col_data, { key => $unique_constraint, L<%attrs|/ATTRIBUTES> }?
2663 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass>
2667 $cd->cd_to_producer->find_or_create({ producer => $producer },
2668 { key => 'primary' });
2670 Tries to find a record based on its primary key or unique constraints; if none
2671 is found, creates one and returns that instead.
2673 my $cd = $schema->resultset('CD')->find_or_create({
2675 artist => 'Massive Attack',
2676 title => 'Mezzanine',
2680 Also takes an optional C<key> attribute, to search by a specific key or unique
2681 constraint. For example:
2683 my $cd = $schema->resultset('CD')->find_or_create(
2685 artist => 'Massive Attack',
2686 title => 'Mezzanine',
2688 { key => 'cd_artist_title' }
2691 B<Note>: Make sure to read the documentation of L</find> and understand the
2692 significance of the C<key> attribute, as its lack may skew your search, and
2693 subsequently result in spurious row creation.
2695 B<Note>: Because find_or_create() reads from the database and then
2696 possibly inserts based on the result, this method is subject to a race
2697 condition. Another process could create a record in the table after
2698 the find has completed and before the create has started. To avoid
2699 this problem, use find_or_create() inside a transaction.
2701 B<Note>: Take care when using C<find_or_create> with a table having
2702 columns with default values that you intend to be automatically
2703 supplied by the database (e.g. an auto_increment primary key column).
2704 In normal usage, the value of such columns should NOT be included at
2705 all in the call to C<find_or_create>, even when set to C<undef>.
2707 See also L</find> and L</update_or_create>. For information on how to declare
2708 unique constraints, see L<DBIx::Class::ResultSource/add_unique_constraint>.
2710 If you need to know if an existing row was found or a new one created use
2711 L</find_or_new> and L<DBIx::Class::Row/in_storage> instead. Don't forget
2712 to call L<DBIx::Class::Row/insert> to save the newly created row to the
2715 my $cd = $schema->resultset('CD')->find_or_new({
2717 artist => 'Massive Attack',
2718 title => 'Mezzanine',
2722 if( !$cd->in_storage ) {
2729 sub find_or_create {
2731 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
2732 my $hash = ref $_[0] eq 'HASH' ? shift : {@_};
2733 if (keys %$hash and my $row = $self->find($hash, $attrs) ) {
2736 return $self->create($hash);
2739 =head2 update_or_create
2743 =item Arguments: \%col_data, { key => $unique_constraint, L<%attrs|/ATTRIBUTES> }?
2745 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass>
2749 $resultset->update_or_create({ col => $val, ... });
2751 Like L</find_or_create>, but if a row is found it is immediately updated via
2752 C<< $found_row->update (\%col_data) >>.
2755 Takes an optional C<key> attribute to search on a specific unique constraint.
2758 # In your application
2759 my $cd = $schema->resultset('CD')->update_or_create(
2761 artist => 'Massive Attack',
2762 title => 'Mezzanine',
2765 { key => 'cd_artist_title' }
2768 $cd->cd_to_producer->update_or_create({
2769 producer => $producer,
2775 B<Note>: Make sure to read the documentation of L</find> and understand the
2776 significance of the C<key> attribute, as its lack may skew your search, and
2777 subsequently result in spurious row creation.
2779 B<Note>: Take care when using C<update_or_create> with a table having
2780 columns with default values that you intend to be automatically
2781 supplied by the database (e.g. an auto_increment primary key column).
2782 In normal usage, the value of such columns should NOT be included at
2783 all in the call to C<update_or_create>, even when set to C<undef>.
2785 See also L</find> and L</find_or_create>. For information on how to declare
2786 unique constraints, see L<DBIx::Class::ResultSource/add_unique_constraint>.
2788 If you need to know if an existing row was updated or a new one created use
2789 L</update_or_new> and L<DBIx::Class::Row/in_storage> instead. Don't forget
2790 to call L<DBIx::Class::Row/insert> to save the newly created row to the
2795 sub update_or_create {
2797 my $attrs = (@_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {});
2798 my $cond = ref $_[0] eq 'HASH' ? shift : {@_};
2800 my $row = $self->find($cond, $attrs);
2802 $row->update($cond);
2806 return $self->create($cond);
2809 =head2 update_or_new
2813 =item Arguments: \%col_data, { key => $unique_constraint, L<%attrs|/ATTRIBUTES> }?
2815 =item Return Value: L<$result|DBIx::Class::Manual::ResultClass>
2819 $resultset->update_or_new({ col => $val, ... });
2821 Like L</find_or_new> but if a row is found it is immediately updated via
2822 C<< $found_row->update (\%col_data) >>.
2826 # In your application
2827 my $cd = $schema->resultset('CD')->update_or_new(
2829 artist => 'Massive Attack',
2830 title => 'Mezzanine',
2833 { key => 'cd_artist_title' }
2836 if ($cd->in_storage) {
2837 # the cd was updated
2840 # the cd is not yet in the database, let's insert it
2844 B<Note>: Make sure to read the documentation of L</find> and understand the
2845 significance of the C<key> attribute, as its lack may skew your search, and
2846 subsequently result in spurious new objects.
2848 B<Note>: Take care when using C<update_or_new> with a table having
2849 columns with default values that you intend to be automatically
2850 supplied by the database (e.g. an auto_increment primary key column).
2851 In normal usage, the value of such columns should NOT be included at
2852 all in the call to C<update_or_new>, even when set to C<undef>.
2854 See also L</find>, L</find_or_create> and L</find_or_new>.
2860 my $attrs = ( @_ > 1 && ref $_[$#_] eq 'HASH' ? pop(@_) : {} );
2861 my $cond = ref $_[0] eq 'HASH' ? shift : {@_};
2863 my $row = $self->find( $cond, $attrs );
2864 if ( defined $row ) {
2865 $row->update($cond);
2869 return $self->new_result($cond);
2876 =item Arguments: none
2878 =item Return Value: L<\@result_objs|DBIx::Class::Manual::ResultClass> | undef
2882 Gets the contents of the cache for the resultset, if the cache is set.
2884 The cache is populated either by using the L</prefetch> attribute to
2885 L</search> or by calling L</set_cache>.
2897 =item Arguments: L<\@result_objs|DBIx::Class::Manual::ResultClass>
2899 =item Return Value: L<\@result_objs|DBIx::Class::Manual::ResultClass>
2903 Sets the contents of the cache for the resultset. Expects an arrayref
2904 of objects of the same class as those produced by the resultset. Note that
2905 if the cache is set, the resultset will return the cached objects rather
2906 than re-querying the database even if the cache attr is not set.
2908 The contents of the cache can also be populated by using the
2909 L</prefetch> attribute to L</search>.
2914 my ( $self, $data ) = @_;
2915 $self->throw_exception("set_cache requires an arrayref")
2916 if defined($data) && (ref $data ne 'ARRAY');
2917 $self->{all_cache} = $data;
2924 =item Arguments: none
2926 =item Return Value: undef
2930 Clears the cache for the resultset.
2935 shift->set_cache(undef);
2942 =item Arguments: none
2944 =item Return Value: true, if the resultset has been paginated
2952 return !!$self->{attrs}{page};
2959 =item Arguments: none
2961 =item Return Value: true, if the resultset has been ordered with C<order_by>.
2969 return scalar $self->result_source->storage->_extract_order_criteria($self->{attrs}{order_by});
2972 =head2 related_resultset
2976 =item Arguments: $rel_name
2978 =item Return Value: L<$resultset|/search>
2982 Returns a related resultset for the supplied relationship name.
2984 $artist_rs = $schema->resultset('CD')->related_resultset('Artist');
2988 sub related_resultset {
2989 my ($self, $rel) = @_;
2991 $self->{related_resultsets} ||= {};
2992 return $self->{related_resultsets}{$rel} ||= do {
2993 my $rsrc = $self->result_source;
2994 my $rel_info = $rsrc->relationship_info($rel);
2996 $self->throw_exception(
2997 "search_related: result source '" . $rsrc->source_name .
2998 "' has no such relationship $rel")
3001 my $attrs = $self->_chain_relationship($rel);
3003 my $join_count = $attrs->{seen_join}{$rel};
3005 my $alias = $self->result_source->storage
3006 ->relname_to_table_alias($rel, $join_count);
3008 # since this is search_related, and we already slid the select window inwards
3009 # (the select/as attrs were deleted in the beginning), we need to flip all
3010 # left joins to inner, so we get the expected results
3011 # read the comment on top of the actual function to see what this does
3012 $attrs->{from} = $rsrc->schema->storage->_inner_join_to_node ($attrs->{from}, $alias);
3015 #XXX - temp fix for result_class bug. There likely is a more elegant fix -groditi
3016 delete @{$attrs}{qw(result_class alias)};
3020 if (my $cache = $self->get_cache) {
3021 if ($cache->[0] && $cache->[0]->related_resultset($rel)->get_cache) {
3022 $new_cache = [ map { @{$_->related_resultset($rel)->get_cache||[]} }
3027 my $rel_source = $rsrc->related_source($rel);
3031 # The reason we do this now instead of passing the alias to the
3032 # search_rs below is that if you wrap/overload resultset on the
3033 # source you need to know what alias it's -going- to have for things
3034 # to work sanely (e.g. RestrictWithObject wants to be able to add
3035 # extra query restrictions, and these may need to be $alias.)
3037 my $rel_attrs = $rel_source->resultset_attributes;
3038 local $rel_attrs->{alias} = $alias;
3040 $rel_source->resultset
3044 where => $attrs->{where},
3047 $new->set_cache($new_cache) if $new_cache;
3052 =head2 current_source_alias
3056 =item Arguments: none
3058 =item Return Value: $source_alias
3062 Returns the current table alias for the result source this resultset is built
3063 on, that will be used in the SQL query. Usually it is C<me>.
3065 Currently the source alias that refers to the result set returned by a
3066 L</search>/L</find> family method depends on how you got to the resultset: it's
3067 C<me> by default, but eg. L</search_related> aliases it to the related result
3068 source name (and keeps C<me> referring to the original result set). The long
3069 term goal is to make L<DBIx::Class> always alias the current resultset as C<me>
3070 (and make this method unnecessary).
3072 Thus it's currently necessary to use this method in predefined queries (see
3073 L<DBIx::Class::Manual::Cookbook/Predefined searches>) when referring to the
3074 source alias of the current result set:
3076 # in a result set class
3078 my ($self, $user) = @_;
3080 my $me = $self->current_source_alias;
3082 return $self->search({
3083 "$me.modified" => $user->id,
3089 sub current_source_alias {
3090 return (shift->{attrs} || {})->{alias} || 'me';
3093 =head2 as_subselect_rs
3097 =item Arguments: none
3099 =item Return Value: L<$resultset|/search>
3103 Act as a barrier to SQL symbols. The resultset provided will be made into a
3104 "virtual view" by including it as a subquery within the from clause. From this
3105 point on, any joined tables are inaccessible to ->search on the resultset (as if
3106 it were simply where-filtered without joins). For example:
3108 my $rs = $schema->resultset('Bar')->search({'x.name' => 'abc'},{ join => 'x' });
3110 # 'x' now pollutes the query namespace
3112 # So the following works as expected
3113 my $ok_rs = $rs->search({'x.other' => 1});
3115 # But this doesn't: instead of finding a 'Bar' related to two x rows (abc and
3116 # def) we look for one row with contradictory terms and join in another table
3117 # (aliased 'x_2') which we never use
3118 my $broken_rs = $rs->search({'x.name' => 'def'});
3120 my $rs2 = $rs->as_subselect_rs;
3122 # doesn't work - 'x' is no longer accessible in $rs2, having been sealed away
3123 my $not_joined_rs = $rs2->search({'x.other' => 1});
3125 # works as expected: finds a 'table' row related to two x rows (abc and def)
3126 my $correctly_joined_rs = $rs2->search({'x.name' => 'def'});
3128 Another example of when one might use this would be to select a subset of
3129 columns in a group by clause:
3131 my $rs = $schema->resultset('Bar')->search(undef, {
3132 group_by => [qw{ id foo_id baz_id }],
3133 })->as_subselect_rs->search(undef, {
3134 columns => [qw{ id foo_id }]
3137 In the above example normally columns would have to be equal to the group by,
3138 but because we isolated the group by into a subselect the above works.
3142 sub as_subselect_rs {
3145 my $attrs = $self->_resolved_attrs;
3147 my $fresh_rs = (ref $self)->new (
3148 $self->result_source
3151 # these pieces will be locked in the subquery
3152 delete $fresh_rs->{cond};
3153 delete @{$fresh_rs->{attrs}}{qw/where bind/};
3155 return $fresh_rs->search( {}, {
3157 $attrs->{alias} => $self->as_query,
3158 -alias => $attrs->{alias},
3159 -rsrc => $self->result_source,
3161 alias => $attrs->{alias},
3165 # This code is called by search_related, and makes sure there
3166 # is clear separation between the joins before, during, and
3167 # after the relationship. This information is needed later
3168 # in order to properly resolve prefetch aliases (any alias
3169 # with a relation_chain_depth less than the depth of the
3170 # current prefetch is not considered)
3172 # The increments happen twice per join. An even number means a
3173 # relationship specified via a search_related, whereas an odd
3174 # number indicates a join/prefetch added via attributes
3176 # Also this code will wrap the current resultset (the one we
3177 # chain to) in a subselect IFF it contains limiting attributes
3178 sub _chain_relationship {
3179 my ($self, $rel) = @_;
3180 my $source = $self->result_source;
3181 my $attrs = { %{$self->{attrs}||{}} };
3183 # we need to take the prefetch the attrs into account before we
3184 # ->_resolve_join as otherwise they get lost - captainL
3185 my $join = $self->_merge_joinpref_attr( $attrs->{join}, $attrs->{prefetch} );
3187 delete @{$attrs}{qw/join prefetch collapse group_by distinct select as columns +select +as +columns/};
3189 my $seen = { %{ (delete $attrs->{seen_join}) || {} } };
3192 my @force_subq_attrs = qw/offset rows group_by having/;
3195 ($attrs->{from} && ref $attrs->{from} ne 'ARRAY')
3197 $self->_has_resolved_attr (@force_subq_attrs)
3199 # Nuke the prefetch (if any) before the new $rs attrs
3200 # are resolved (prefetch is useless - we are wrapping
3201 # a subquery anyway).
3202 my $rs_copy = $self->search;
3203 $rs_copy->{attrs}{join} = $self->_merge_joinpref_attr (
3204 $rs_copy->{attrs}{join},
3205 delete $rs_copy->{attrs}{prefetch},
3210 -alias => $attrs->{alias},
3211 $attrs->{alias} => $rs_copy->as_query,
3213 delete @{$attrs}{@force_subq_attrs, qw/where bind/};
3214 $seen->{-relation_chain_depth} = 0;
3216 elsif ($attrs->{from}) { #shallow copy suffices
3217 $from = [ @{$attrs->{from}} ];
3222 -alias => $attrs->{alias},
3223 $attrs->{alias} => $source->from,
3227 my $jpath = ($seen->{-relation_chain_depth})
3228 ? $from->[-1][0]{-join_path}
3231 my @requested_joins = $source->_resolve_join(
3238 push @$from, @requested_joins;
3240 $seen->{-relation_chain_depth}++;
3242 # if $self already had a join/prefetch specified on it, the requested
3243 # $rel might very well be already included. What we do in this case
3244 # is effectively a no-op (except that we bump up the chain_depth on
3245 # the join in question so we could tell it *is* the search_related)
3248 # we consider the last one thus reverse
3249 for my $j (reverse @requested_joins) {
3250 my ($last_j) = keys %{$j->[0]{-join_path}[-1]};
3251 if ($rel eq $last_j) {
3252 $j->[0]{-relation_chain_depth}++;
3258 unless ($already_joined) {
3259 push @$from, $source->_resolve_join(
3267 $seen->{-relation_chain_depth}++;
3269 return {%$attrs, from => $from, seen_join => $seen};
3272 sub _resolved_attrs {
3274 return $self->{_attrs} if $self->{_attrs};
3276 my $attrs = { %{ $self->{attrs} || {} } };
3277 my $source = $self->result_source;
3278 my $alias = $attrs->{alias};
3280 # default selection list
3281 $attrs->{columns} = [ $source->columns ]
3282 unless List::Util::first { exists $attrs->{$_} } qw/columns cols select as/;
3284 # merge selectors together
3285 for (qw/columns select as/) {
3286 $attrs->{$_} = $self->_merge_attr($attrs->{$_}, delete $attrs->{"+$_"})
3287 if $attrs->{$_} or $attrs->{"+$_"};
3290 # disassemble columns
3292 if (my $cols = delete $attrs->{columns}) {
3293 for my $c (ref $cols eq 'ARRAY' ? @$cols : $cols) {
3294 if (ref $c eq 'HASH') {
3295 for my $as (sort keys %$c) {
3296 push @sel, $c->{$as};
3307 # when trying to weed off duplicates later do not go past this point -
3308 # everything added from here on is unbalanced "anyone's guess" stuff
3309 my $dedup_stop_idx = $#as;
3311 push @as, @{ ref $attrs->{as} eq 'ARRAY' ? $attrs->{as} : [ $attrs->{as} ] }
3313 push @sel, @{ ref $attrs->{select} eq 'ARRAY' ? $attrs->{select} : [ $attrs->{select} ] }
3314 if $attrs->{select};
3316 # assume all unqualified selectors to apply to the current alias (legacy stuff)
3317 $_ = (ref $_ or $_ =~ /\./) ? $_ : "$alias.$_" for @sel;
3319 # disqualify all $alias.col as-bits (inflate-map mandated)
3320 $_ = ($_ =~ /^\Q$alias.\E(.+)$/) ? $1 : $_ for @as;
3322 # de-duplicate the result (remove *identical* select/as pairs)
3323 # and also die on duplicate {as} pointing to different {select}s
3324 # not using a c-style for as the condition is prone to shrinkage
3327 while ($i <= $dedup_stop_idx) {
3328 if ($seen->{"$sel[$i] \x00\x00 $as[$i]"}++) {
3333 elsif ($seen->{$as[$i]}++) {
3334 $self->throw_exception(
3335 "inflate_result() alias '$as[$i]' specified twice with different SQL-side {select}-ors"
3343 $attrs->{select} = \@sel;
3344 $attrs->{as} = \@as;
3346 $attrs->{from} ||= [{
3348 -alias => $self->{attrs}{alias},
3349 $self->{attrs}{alias} => $source->from,
3352 if ( $attrs->{join} || $attrs->{prefetch} ) {
3354 $self->throw_exception ('join/prefetch can not be used with a custom {from}')
3355 if ref $attrs->{from} ne 'ARRAY';
3357 my $join = (delete $attrs->{join}) || {};
3359 if ( defined $attrs->{prefetch} ) {
3360 $join = $self->_merge_joinpref_attr( $join, $attrs->{prefetch} );
3363 $attrs->{from} = # have to copy here to avoid corrupting the original
3365 @{ $attrs->{from} },
3366 $source->_resolve_join(
3369 { %{ $attrs->{seen_join} || {} } },
3370 ( $attrs->{seen_join} && keys %{$attrs->{seen_join}})
3371 ? $attrs->{from}[-1][0]{-join_path}
3378 if ( defined $attrs->{order_by} ) {
3379 $attrs->{order_by} = (
3380 ref( $attrs->{order_by} ) eq 'ARRAY'
3381 ? [ @{ $attrs->{order_by} } ]
3382 : [ $attrs->{order_by} || () ]
3386 if ($attrs->{group_by} and ref $attrs->{group_by} ne 'ARRAY') {
3387 $attrs->{group_by} = [ $attrs->{group_by} ];
3390 # generate the distinct induced group_by early, as prefetch will be carried via a
3391 # subquery (since a group_by is present)
3392 if (delete $attrs->{distinct}) {
3393 if ($attrs->{group_by}) {
3394 carp_unique ("Useless use of distinct on a grouped resultset ('distinct' is ignored when a 'group_by' is present)");
3397 # distinct affects only the main selection part, not what prefetch may
3399 $attrs->{group_by} = $source->storage->_group_over_selection (
3407 # generate selections based on the prefetch helper
3409 $prefetch = $self->_merge_joinpref_attr( {}, delete $attrs->{prefetch} )
3410 if defined $attrs->{prefetch};
3414 $self->throw_exception("Unable to prefetch, resultset contains an unnamed selector $attrs->{_dark_selector}{string}")
3415 if $attrs->{_dark_selector};
3417 $attrs->{collapse} = 1;
3419 # this is a separate structure (we don't look in {from} directly)
3420 # as the resolver needs to shift things off the lists to work
3421 # properly (identical-prefetches on different branches)
3423 if (ref $attrs->{from} eq 'ARRAY') {
3425 my $start_depth = $attrs->{seen_join}{-relation_chain_depth} || 0;
3427 for my $j ( @{$attrs->{from}}[1 .. $#{$attrs->{from}} ] ) {
3428 next unless $j->[0]{-alias};
3429 next unless $j->[0]{-join_path};
3430 next if ($j->[0]{-relation_chain_depth} || 0) < $start_depth;
3432 my @jpath = map { keys %$_ } @{$j->[0]{-join_path}};
3435 $p = $p->{$_} ||= {} for @jpath[ ($start_depth/2) .. $#jpath]; #only even depths are actual jpath boundaries
3436 push @{$p->{-join_aliases} }, $j->[0]{-alias};
3440 my @prefetch = $source->_resolve_prefetch( $prefetch, $alias, $join_map );
3442 # we need to somehow mark which columns came from prefetch
3444 my $sel_end = $#{$attrs->{select}};
3445 $attrs->{_prefetch_selector_range} = [ $sel_end + 1, $sel_end + @prefetch ];
3448 push @{ $attrs->{select} }, (map { $_->[0] } @prefetch);
3449 push @{ $attrs->{as} }, (map { $_->[1] } @prefetch);
3452 if ( ! List::Util::first { $_ =~ /\./ } @{$attrs->{as}} ) {
3453 $attrs->{_single_object_inflation} = 1;
3454 $attrs->{collapse} = 0;
3457 # run through the resulting joinstructure (starting from our current slot)
3458 # and unset collapse if proven unnesessary
3460 # also while we are at it find out if the current root source has
3461 # been premultiplied by previous related_source chaining
3463 # this allows to predict whether a root object with all other relation
3464 # data set to NULL is in fact unique
3465 if ($attrs->{collapse}) {
3467 if (ref $attrs->{from} eq 'ARRAY') {
3469 if (@{$attrs->{from}} <= 1) {
3470 # no joins - no collapse
3471 $attrs->{collapse} = 0;
3474 # find where our table-spec starts
3475 my @fromlist = @{$attrs->{from}};
3477 my $t = shift @fromlist;
3480 # me vs join from-spec distinction - a ref means non-root
3481 if (ref $t eq 'ARRAY') {
3483 $is_multi ||= ! $t->{-is_single};
3485 last if ($t->{-alias} && $t->{-alias} eq $alias);
3486 $attrs->{_main_source_premultiplied} ||= $is_multi;
3489 # no non-singles remaining, nor any premultiplication - nothing to collapse
3491 ! $attrs->{_main_source_premultiplied}
3493 ! List::Util::first { ! $_->[0]{-is_single} } @fromlist
3495 $attrs->{collapse} = 0;
3501 # if we can not analyze the from - err on the side of safety
3502 $attrs->{_main_source_premultiplied} = 1;
3506 if (! $attrs->{order_by} and $attrs->{collapse}) {
3507 # default order for collapsing unless the user asked for something
3508 $attrs->{order_by} = [ map { "$alias.$_" } $source->primary_columns ];
3509 $attrs->{_ordered_for_collapse} = 1;
3510 $attrs->{_order_is_artificial} = 1;
3513 # if both page and offset are specified, produce a combined offset
3514 # even though it doesn't make much sense, this is what pre 081xx has
3516 if (my $page = delete $attrs->{page}) {
3518 ($attrs->{rows} * ($page - 1))
3520 ($attrs->{offset} || 0)
3524 return $self->{_attrs} = $attrs;
3528 my ($self, $attr) = @_;
3530 if (ref $attr eq 'HASH') {
3531 return $self->_rollout_hash($attr);
3532 } elsif (ref $attr eq 'ARRAY') {
3533 return $self->_rollout_array($attr);
3539 sub _rollout_array {
3540 my ($self, $attr) = @_;
3543 foreach my $element (@{$attr}) {
3544 if (ref $element eq 'HASH') {
3545 push( @rolled_array, @{ $self->_rollout_hash( $element ) } );
3546 } elsif (ref $element eq 'ARRAY') {
3547 # XXX - should probably recurse here
3548 push( @rolled_array, @{$self->_rollout_array($element)} );
3550 push( @rolled_array, $element );
3553 return \@rolled_array;
3557 my ($self, $attr) = @_;
3560 foreach my $key (keys %{$attr}) {
3561 push( @rolled_array, { $key => $attr->{$key} } );
3563 return \@rolled_array;
3566 sub _calculate_score {
3567 my ($self, $a, $b) = @_;
3569 if (defined $a xor defined $b) {
3572 elsif (not defined $a) {
3576 if (ref $b eq 'HASH') {
3577 my ($b_key) = keys %{$b};
3578 if (ref $a eq 'HASH') {
3579 my ($a_key) = keys %{$a};
3580 if ($a_key eq $b_key) {
3581 return (1 + $self->_calculate_score( $a->{$a_key}, $b->{$b_key} ));
3586 return ($a eq $b_key) ? 1 : 0;
3589 if (ref $a eq 'HASH') {
3590 my ($a_key) = keys %{$a};
3591 return ($b eq $a_key) ? 1 : 0;
3593 return ($b eq $a) ? 1 : 0;
3598 sub _merge_joinpref_attr {
3599 my ($self, $orig, $import) = @_;
3601 return $import unless defined($orig);
3602 return $orig unless defined($import);
3604 $orig = $self->_rollout_attr($orig);
3605 $import = $self->_rollout_attr($import);
3608 foreach my $import_element ( @{$import} ) {
3609 # find best candidate from $orig to merge $b_element into
3610 my $best_candidate = { position => undef, score => 0 }; my $position = 0;
3611 foreach my $orig_element ( @{$orig} ) {
3612 my $score = $self->_calculate_score( $orig_element, $import_element );
3613 if ($score > $best_candidate->{score}) {
3614 $best_candidate->{position} = $position;
3615 $best_candidate->{score} = $score;
3619 my ($import_key) = ( ref $import_element eq 'HASH' ) ? keys %{$import_element} : ($import_element);
3620 $import_key = '' if not defined $import_key;
3622 if ($best_candidate->{score} == 0 || exists $seen_keys->{$import_key}) {
3623 push( @{$orig}, $import_element );
3625 my $orig_best = $orig->[$best_candidate->{position}];
3626 # merge orig_best and b_element together and replace original with merged
3627 if (ref $orig_best ne 'HASH') {
3628 $orig->[$best_candidate->{position}] = $import_element;
3629 } elsif (ref $import_element eq 'HASH') {
3630 my ($key) = keys %{$orig_best};
3631 $orig->[$best_candidate->{position}] = { $key => $self->_merge_joinpref_attr($orig_best->{$key}, $import_element->{$key}) };
3634 $seen_keys->{$import_key} = 1; # don't merge the same key twice
3637 return @$orig ? $orig : ();
3645 require Hash::Merge;
3646 my $hm = Hash::Merge->new;
3648 $hm->specify_behavior({
3651 my ($defl, $defr) = map { defined $_ } (@_[0,1]);
3653 if ($defl xor $defr) {
3654 return [ $defl ? $_[0] : $_[1] ];
3659 elsif (__HM_DEDUP and $_[0] eq $_[1]) {
3663 return [$_[0], $_[1]];
3667 return $_[1] if !defined $_[0];
3668 return $_[1] if __HM_DEDUP and List::Util::first { $_ eq $_[0] } @{$_[1]};
3669 return [$_[0], @{$_[1]}]
3672 return [] if !defined $_[0] and !keys %{$_[1]};
3673 return [ $_[1] ] if !defined $_[0];
3674 return [ $_[0] ] if !keys %{$_[1]};
3675 return [$_[0], $_[1]]
3680 return $_[0] if !defined $_[1];
3681 return $_[0] if __HM_DEDUP and List::Util::first { $_ eq $_[1] } @{$_[0]};
3682 return [@{$_[0]}, $_[1]]
3685 my @ret = @{$_[0]} or return $_[1];
3686 return [ @ret, @{$_[1]} ] unless __HM_DEDUP;
3687 my %idx = map { $_ => 1 } @ret;
3688 push @ret, grep { ! defined $idx{$_} } (@{$_[1]});
3692 return [ $_[1] ] if ! @{$_[0]};
3693 return $_[0] if !keys %{$_[1]};
3694 return $_[0] if __HM_DEDUP and List::Util::first { $_ eq $_[1] } @{$_[0]};
3695 return [ @{$_[0]}, $_[1] ];
3700 return [] if !keys %{$_[0]} and !defined $_[1];
3701 return [ $_[0] ] if !defined $_[1];
3702 return [ $_[1] ] if !keys %{$_[0]};
3703 return [$_[0], $_[1]]
3706 return [] if !keys %{$_[0]} and !@{$_[1]};
3707 return [ $_[0] ] if !@{$_[1]};
3708 return $_[1] if !keys %{$_[0]};
3709 return $_[1] if __HM_DEDUP and List::Util::first { $_ eq $_[0] } @{$_[1]};
3710 return [ $_[0], @{$_[1]} ];
3713 return [] if !keys %{$_[0]} and !keys %{$_[1]};
3714 return [ $_[0] ] if !keys %{$_[1]};
3715 return [ $_[1] ] if !keys %{$_[0]};
3716 return [ $_[0] ] if $_[0] eq $_[1];
3717 return [ $_[0], $_[1] ];
3720 } => 'DBIC_RS_ATTR_MERGER');
3724 return $hm->merge ($_[1], $_[2]);
3728 sub STORABLE_freeze {
3729 my ($self, $cloning) = @_;
3730 my $to_serialize = { %$self };
3732 # A cursor in progress can't be serialized (and would make little sense anyway)
3733 # the parser can be regenerated (and can't be serialized)
3734 delete @{$to_serialize}{qw/cursor _row_parser/};
3736 # nor is it sensical to store a not-yet-fired-count pager
3737 if ($to_serialize->{pager} and ref $to_serialize->{pager}{total_entries} eq 'CODE') {
3738 delete $to_serialize->{pager};
3741 Storable::nfreeze($to_serialize);
3744 # need this hook for symmetry
3746 my ($self, $cloning, $serialized) = @_;
3748 %$self = %{ Storable::thaw($serialized) };
3754 =head2 throw_exception
3756 See L<DBIx::Class::Schema/throw_exception> for details.
3760 sub throw_exception {
3763 if (ref $self and my $rsrc = $self->result_source) {
3764 $rsrc->throw_exception(@_)
3767 DBIx::Class::Exception->throw(@_);
3771 # XXX: FIXME: Attributes docs need clearing up
3775 Attributes are used to refine a ResultSet in various ways when
3776 searching for data. They can be passed to any method which takes an
3777 C<\%attrs> argument. See L</search>, L</search_rs>, L</find>,
3780 Default attributes can be set on the result class using
3781 L<DBIx::Class::ResultSource/resultset_attributes>. (Please read
3782 the CAVEATS on that feature before using it!)
3784 These are in no particular order:
3790 =item Value: ( $order_by | \@order_by | \%order_by )
3794 Which column(s) to order the results by.
3796 [The full list of suitable values is documented in
3797 L<SQL::Abstract/"ORDER BY CLAUSES">; the following is a summary of
3800 If a single column name, or an arrayref of names is supplied, the
3801 argument is passed through directly to SQL. The hashref syntax allows
3802 for connection-agnostic specification of ordering direction:
3804 For descending order:
3806 order_by => { -desc => [qw/col1 col2 col3/] }
3808 For explicit ascending order:
3810 order_by => { -asc => 'col' }
3812 The old scalarref syntax (i.e. order_by => \'year DESC') is still
3813 supported, although you are strongly encouraged to use the hashref
3814 syntax as outlined above.
3820 =item Value: \@columns
3824 Shortcut to request a particular set of columns to be retrieved. Each
3825 column spec may be a string (a table column name), or a hash (in which
3826 case the key is the C<as> value, and the value is used as the C<select>
3827 expression). Adds C<me.> onto the start of any column without a C<.> in
3828 it and sets C<select> from that, then auto-populates C<as> from
3829 C<select> as normal. (You may also use the C<cols> attribute, as in
3830 earlier versions of DBIC.)
3832 Essentially C<columns> does the same as L</select> and L</as>.
3834 columns => [ 'foo', { bar => 'baz' } ]
3838 select => [qw/foo baz/],
3845 =item Value: \@columns
3849 Indicates additional columns to be selected from storage. Works the same
3850 as L</columns> but adds columns to the selection. (You may also use the
3851 C<include_columns> attribute, as in earlier versions of DBIC). For
3854 $schema->resultset('CD')->search(undef, {
3855 '+columns' => ['artist.name'],
3859 would return all CDs and include a 'name' column to the information
3860 passed to object inflation. Note that the 'artist' is the name of the
3861 column (or relationship) accessor, and 'name' is the name of the column
3862 accessor in the related table.
3864 B<NOTE:> You need to explicitly quote '+columns' when defining the attribute.
3865 Not doing so causes Perl to incorrectly interpret +columns as a bareword with a
3866 unary plus operator before it.
3868 =head2 include_columns
3872 =item Value: \@columns
3876 Deprecated. Acts as a synonym for L</+columns> for backward compatibility.
3882 =item Value: \@select_columns
3886 Indicates which columns should be selected from the storage. You can use
3887 column names, or in the case of RDBMS back ends, function or stored procedure
3890 $rs = $schema->resultset('Employee')->search(undef, {
3893 { count => 'employeeid' },
3894 { max => { length => 'name' }, -as => 'longest_name' }
3899 SELECT name, COUNT( employeeid ), MAX( LENGTH( name ) ) AS longest_name FROM employee
3901 B<NOTE:> You will almost always need a corresponding L</as> attribute when you
3902 use L</select>, to instruct DBIx::Class how to store the result of the column.
3903 Also note that the L</as> attribute has nothing to do with the SQL-side 'AS'
3904 identifier aliasing. You can however alias a function, so you can use it in
3905 e.g. an C<ORDER BY> clause. This is done via the C<-as> B<select function
3906 attribute> supplied as shown in the example above.
3908 B<NOTE:> You need to explicitly quote '+select'/'+as' when defining the attributes.
3909 Not doing so causes Perl to incorrectly interpret them as a bareword with a
3910 unary plus operator before it.
3916 Indicates additional columns to be selected from storage. Works the same as
3917 L</select> but adds columns to the default selection, instead of specifying
3926 Indicates additional column names for those added via L</+select>. See L</as>.
3934 =item Value: \@inflation_names
3938 Indicates column names for object inflation. That is L</as> indicates the
3939 slot name in which the column value will be stored within the
3940 L<Row|DBIx::Class::Row> object. The value will then be accessible via this
3941 identifier by the C<get_column> method (or via the object accessor B<if one
3942 with the same name already exists>) as shown below. The L</as> attribute has
3943 B<nothing to do> with the SQL-side C<AS>. See L</select> for details.
3945 $rs = $schema->resultset('Employee')->search(undef, {
3948 { count => 'employeeid' },
3949 { max => { length => 'name' }, -as => 'longest_name' }
3958 If the object against which the search is performed already has an accessor
3959 matching a column name specified in C<as>, the value can be retrieved using
3960 the accessor as normal:
3962 my $name = $employee->name();
3964 If on the other hand an accessor does not exist in the object, you need to
3965 use C<get_column> instead:
3967 my $employee_count = $employee->get_column('employee_count');
3969 You can create your own accessors if required - see
3970 L<DBIx::Class::Manual::Cookbook> for details.
3976 =item Value: ($rel_name | \@rel_names | \%rel_names)
3980 Contains a list of relationships that should be joined for this query. For
3983 # Get CDs by Nine Inch Nails
3984 my $rs = $schema->resultset('CD')->search(
3985 { 'artist.name' => 'Nine Inch Nails' },
3986 { join => 'artist' }
3989 Can also contain a hash reference to refer to the other relation's relations.
3992 package MyApp::Schema::Track;
3993 use base qw/DBIx::Class/;
3994 __PACKAGE__->table('track');
3995 __PACKAGE__->add_columns(qw/trackid cd position title/);
3996 __PACKAGE__->set_primary_key('trackid');
3997 __PACKAGE__->belongs_to(cd => 'MyApp::Schema::CD');
4000 # In your application
4001 my $rs = $schema->resultset('Artist')->search(
4002 { 'track.title' => 'Teardrop' },
4004 join => { cd => 'track' },
4005 order_by => 'artist.name',
4009 You need to use the relationship (not the table) name in conditions,
4010 because they are aliased as such. The current table is aliased as "me", so
4011 you need to use me.column_name in order to avoid ambiguity. For example:
4013 # Get CDs from 1984 with a 'Foo' track
4014 my $rs = $schema->resultset('CD')->search(
4017 'tracks.name' => 'Foo'
4019 { join => 'tracks' }
4022 If the same join is supplied twice, it will be aliased to <rel>_2 (and
4023 similarly for a third time). For e.g.
4025 my $rs = $schema->resultset('Artist')->search({
4026 'cds.title' => 'Down to Earth',
4027 'cds_2.title' => 'Popular',
4029 join => [ qw/cds cds/ ],
4032 will return a set of all artists that have both a cd with title 'Down
4033 to Earth' and a cd with title 'Popular'.
4035 If you want to fetch related objects from other tables as well, see C<prefetch>
4038 NOTE: An internal join-chain pruner will discard certain joins while
4039 constructing the actual SQL query, as long as the joins in question do not
4040 affect the retrieved result. This for example includes 1:1 left joins
4041 that are not part of the restriction specification (WHERE/HAVING) nor are
4042 a part of the query selection.
4044 For more help on using joins with search, see L<DBIx::Class::Manual::Joining>.
4050 =item Value: ($rel_name | \@rel_names | \%rel_names)
4054 Contains one or more relationships that should be fetched along with
4055 the main query (when they are accessed afterwards the data will
4056 already be available, without extra queries to the database). This is
4057 useful for when you know you will need the related objects, because it
4058 saves at least one query:
4060 my $rs = $schema->resultset('Tag')->search(
4069 The initial search results in SQL like the following:
4071 SELECT tag.*, cd.*, artist.* FROM tag
4072 JOIN cd ON tag.cd = cd.cdid
4073 JOIN artist ON cd.artist = artist.artistid
4075 L<DBIx::Class> has no need to go back to the database when we access the
4076 C<cd> or C<artist> relationships, which saves us two SQL statements in this
4079 Simple prefetches will be joined automatically, so there is no need
4080 for a C<join> attribute in the above search.
4082 L</prefetch> can be used with the any of the relationship types and
4083 multiple prefetches can be specified together. Below is a more complex
4084 example that prefetches a CD's artist, its liner notes (if present),
4085 the cover image, the tracks on that cd, and the guests on those
4089 My::Schema::CD->belongs_to( artist => 'My::Schema::Artist' );
4090 My::Schema::CD->might_have( liner_note => 'My::Schema::LinerNotes' );
4091 My::Schema::CD->has_one( cover_image => 'My::Schema::Artwork' );
4092 My::Schema::CD->has_many( tracks => 'My::Schema::Track' );
4094 My::Schema::Artist->belongs_to( record_label => 'My::Schema::RecordLabel' );
4096 My::Schema::Track->has_many( guests => 'My::Schema::Guest' );
4099 my $rs = $schema->resultset('CD')->search(
4103 { artist => 'record_label'}, # belongs_to => belongs_to
4104 'liner_note', # might_have
4105 'cover_image', # has_one
4106 { tracks => 'guests' }, # has_many => has_many
4111 This will produce SQL like the following:
4113 SELECT cd.*, artist.*, record_label.*, liner_note.*, cover_image.*,
4117 ON artist.artistid = me.artistid
4118 JOIN record_label record_label
4119 ON record_label.labelid = artist.labelid
4120 LEFT JOIN track tracks
4121 ON tracks.cdid = me.cdid
4122 LEFT JOIN guest guests
4123 ON guests.trackid = track.trackid
4124 LEFT JOIN liner_notes liner_note
4125 ON liner_note.cdid = me.cdid
4126 JOIN cd_artwork cover_image
4127 ON cover_image.cdid = me.cdid
4130 Now the C<artist>, C<record_label>, C<liner_note>, C<cover_image>,
4131 C<tracks>, and C<guests> of the CD will all be available through the
4132 relationship accessors without the need for additional queries to the
4135 However, there is one caveat to be observed: it can be dangerous to
4136 prefetch more than one L<has_many|DBIx::Class::Relationship/has_many>
4137 relationship on a given level. e.g.:
4139 my $rs = $schema->resultset('CD')->search(
4143 'tracks', # has_many
4144 { cd_to_producer => 'producer' }, # has_many => belongs_to (i.e. m2m)
4149 The collapser currently can't identify duplicate tuples for multiple
4150 L<has_many|DBIx::Class::Relationship/has_many> relationships and as a
4151 result the second L<has_many|DBIx::Class::Relationship/has_many>
4152 relation could contain redundant objects.
4154 =head3 Using L</prefetch> with L</join>
4156 L</prefetch> implies a L</join> with the equivalent argument, and is
4157 properly merged with any existing L</join> specification. So the
4160 my $rs = $schema->resultset('CD')->search(
4161 {'record_label.name' => 'Music Product Ltd.'},
4163 join => {artist => 'record_label'},
4164 prefetch => 'artist',
4168 ... will work, searching on the record label's name, but only
4169 prefetching the C<artist>.
4171 =head3 Using L</prefetch> with L</select> / L</+select> / L</as> / L</+as>
4173 L</prefetch> implies a L</+select>/L</+as> with the fields of the
4174 prefetched relations. So given:
4176 my $rs = $schema->resultset('CD')->search(
4179 select => ['cd.title'],
4181 prefetch => 'artist',
4185 The L</select> becomes: C<'cd.title', 'artist.*'> and the L</as>
4186 becomes: C<'cd_title', 'artist.*'>.
4190 Prefetch does a lot of deep magic. As such, it may not behave exactly
4191 as you might expect.
4197 Prefetch uses the L</cache> to populate the prefetched relationships. This
4198 may or may not be what you want.
4202 If you specify a condition on a prefetched relationship, ONLY those
4203 rows that match the prefetched condition will be fetched into that relationship.
4204 This means that adding prefetch to a search() B<may alter> what is returned by
4205 traversing a relationship. So, if you have C<< Artist->has_many(CDs) >> and you do
4207 my $artist_rs = $schema->resultset('Artist')->search({
4213 my $count = $artist_rs->first->cds->count;
4215 my $artist_rs_prefetch = $artist_rs->search( {}, { prefetch => 'cds' } );
4217 my $prefetch_count = $artist_rs_prefetch->first->cds->count;
4219 cmp_ok( $count, '==', $prefetch_count, "Counts should be the same" );
4221 that cmp_ok() may or may not pass depending on the datasets involved. This
4222 behavior may or may not survive the 0.09 transition.
4230 =item Value: $source_alias
4234 Sets the source alias for the query. Normally, this defaults to C<me>, but
4235 nested search queries (sub-SELECTs) might need specific aliases set to
4236 reference inner queries. For example:
4239 ->related_resultset('CDs')
4240 ->related_resultset('Tracks')
4242 'track.id' => { -ident => 'none_search.id' },
4246 my $ids = $self->search({
4249 alias => 'none_search',
4250 group_by => 'none_search.id',
4251 })->get_column('id')->as_query;
4253 $self->search({ id => { -in => $ids } })
4255 This attribute is directly tied to L</current_source_alias>.
4265 Makes the resultset paged and specifies the page to retrieve. Effectively
4266 identical to creating a non-pages resultset and then calling ->page($page)
4269 If L</rows> attribute is not specified it defaults to 10 rows per page.
4271 When you have a paged resultset, L</count> will only return the number
4272 of rows in the page. To get the total, use the L</pager> and call
4273 C<total_entries> on it.
4283 Specifies the maximum number of rows for direct retrieval or the number of
4284 rows per page if the page attribute or method is used.
4290 =item Value: $offset
4294 Specifies the (zero-based) row number for the first row to be returned, or the
4295 of the first row of the first page if paging is used.
4297 =head2 software_limit
4301 =item Value: (0 | 1)
4305 When combined with L</rows> and/or L</offset> the generated SQL will not
4306 include any limit dialect stanzas. Instead the entire result will be selected
4307 as if no limits were specified, and DBIC will perform the limit locally, by
4308 artificially advancing and finishing the resulting L</cursor>.
4310 This is the recommended way of performing resultset limiting when no sane RDBMS
4311 implementation is available (e.g.
4312 L<Sybase ASE|DBIx::Class::Storage::DBI::Sybase::ASE> using the
4313 L<Generic Sub Query|DBIx::Class::SQLMaker::LimitDialects/GenericSubQ> hack)
4319 =item Value: \@columns
4323 A arrayref of columns to group by. Can include columns of joined tables.
4325 group_by => [qw/ column1 column2 ... /]
4331 =item Value: $condition
4335 HAVING is a select statement attribute that is applied between GROUP BY and
4336 ORDER BY. It is applied to the after the grouping calculations have been
4339 having => { 'count_employee' => { '>=', 100 } }
4341 or with an in-place function in which case literal SQL is required:
4343 having => \[ 'count(employee) >= ?', [ count => 100 ] ]
4349 =item Value: (0 | 1)
4353 Set to 1 to group by all columns. If the resultset already has a group_by
4354 attribute, this setting is ignored and an appropriate warning is issued.
4360 Adds to the WHERE clause.
4362 # only return rows WHERE deleted IS NULL for all searches
4363 __PACKAGE__->resultset_attributes({ where => { deleted => undef } });
4365 Can be overridden by passing C<< { where => undef } >> as an attribute
4368 For more complicated where clauses see L<SQL::Abstract/WHERE CLAUSES>.
4374 Set to 1 to cache search results. This prevents extra SQL queries if you
4375 revisit rows in your ResultSet:
4377 my $resultset = $schema->resultset('Artist')->search( undef, { cache => 1 } );
4379 while( my $artist = $resultset->next ) {
4383 $rs->first; # without cache, this would issue a query
4385 By default, searches are not cached.
4387 For more examples of using these attributes, see
4388 L<DBIx::Class::Manual::Cookbook>.
4394 =item Value: ( 'update' | 'shared' | \$scalar )
4398 Set to 'update' for a SELECT ... FOR UPDATE or 'shared' for a SELECT
4399 ... FOR SHARED. If \$scalar is passed, this is taken directly and embedded in the
4402 =head1 DBIC BIND VALUES
4404 Because DBIC may need more information to bind values than just the column name
4405 and value itself, it uses a special format for both passing and receiving bind
4406 values. Each bind value should be composed of an arrayref of
4407 C<< [ \%args => $val ] >>. The format of C<< \%args >> is currently:
4413 If present (in any form), this is what is being passed directly to bind_param.
4414 Note that different DBD's expect different bind args. (e.g. DBD::SQLite takes
4415 a single numerical type, while DBD::Pg takes a hashref if bind options.)
4417 If this is specified, all other bind options described below are ignored.
4421 If present, this is used to infer the actual bind attribute by passing to
4422 C<< $resolved_storage->bind_attribute_by_data_type() >>. Defaults to the
4423 "data_type" from the L<add_columns column info|DBIx::Class::ResultSource/add_columns>.
4425 Note that the data type is somewhat freeform (hence the sqlt_ prefix);
4426 currently drivers are expected to "Do the Right Thing" when given a common
4427 datatype name. (Not ideal, but that's what we got at this point.)
4431 Currently used to correctly allocate buffers for bind_param_inout().
4432 Defaults to "size" from the L<add_columns column info|DBIx::Class::ResultSource/add_columns>,
4433 or to a sensible value based on the "data_type".
4437 Used to fill in missing sqlt_datatype and sqlt_size attributes (if they are
4438 explicitly specified they are never overriden). Also used by some weird DBDs,
4439 where the column name should be available at bind_param time (e.g. Oracle).
4443 For backwards compatibility and convenience, the following shortcuts are
4446 [ $name => $val ] === [ { dbic_colname => $name }, $val ]
4447 [ \$dt => $val ] === [ { sqlt_datatype => $dt }, $val ]
4448 [ undef, $val ] === [ {}, $val ]
4450 =head1 AUTHOR AND CONTRIBUTORS
4452 See L<AUTHOR|DBIx::Class/AUTHOR> and L<CONTRIBUTORS|DBIx::Class/CONTRIBUTORS> in DBIx::Class
4456 You may distribute this code under the same terms as Perl itself.