dbix::class::helper::resultset::random(3pm) [debian man page]

DBIx::Class::Helper::ResultSet::Random(3pm)		User Contributed Perl Documentation	       DBIx::Class::Helper::ResultSet::Random(3pm)

NAME

       DBIx::Class::Helper::ResultSet::Random - Get random rows from a ResultSet

VERSION

       version 2.013002

SYNOPSIS

	# note that this is normally a component for a ResultSet
	package MySchema::ResultSet::Bar;

	use strict;
	use warnings;

	use parent 'DBIx::Class::ResultSet';

	__PACKAGE__->load_components('Helper::ResultSet::Random');

	# in code using resultset:
	my $random_row	= $schema->resultset('Bar')->rand->single;

DESCRIPTION

       This component allows convenient selection of random rows.  See "NOTE" in DBIx::Class::Helper::ResultSet for a nice way to apply it to your
       entire schema.

       Currently this works by doing something akin to

	SELECT TOP($x) from $table ORDER BY RANDOM()

       Lots of people think this is slow.  My own benchmarks show that doing the above, for 10 rows in a table with just over 8 million rows, is
       nearly instant.	Although that was with SQL Server, and different databases will handle that differently.

       So please, if you have performance issues and want this to work with your database, get in touch and I will do what I can to get it to work
       quickly enough to suite your needs.

METHODS

   rand
       This method takes a single argument, being the size of the random ResultSet to return.  It defaults to 1.  This Component will throw
       exceptions if the argument is not an integer or not greater than zero.

   _rand_order_by
       This module currently does an "ORDER BY" on some db specific function.  If for some reason it guesses incorrectly for your database the
       easiest way to fix that in the short-term (ie without patching upstream) is to override this method.  So for example, if your db uses
       "RAND()" instead of "RANDOM()" and it's not in the predefined list of dbs you could just do the following in your ResultSet class:

	sub _rand_order_by { 'RAND()' }

AUTHOR

       Arthur Axel "fREW" Schmidt <frioux+cpan@gmail.com>

COPYRIGHT AND LICENSE

       This software is copyright (c) 2012 by Arthur Axel "fREW" Schmidt.

       This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.

perl v5.14.2							    2012-06-18			       DBIx::Class::Helper::ResultSet::Random(3pm)

DBIx::Class::Helper::ResultSet::SearchOr(3pm)		User Contributed Perl Documentation	     DBIx::Class::Helper::ResultSet::SearchOr(3pm)

NAME

       DBIx::Class::Helper::ResultSet::SearchOr - Combine ResultSet searches with OR's

VERSION

       version 2.013002

SYNOPSIS

	package MyApp::Schema::ResultSet::Tests;

	use base 'DBIx::Class::ResultSet';

	__PACKAGE__->load_components(qw(Helper::ResultSet::IgnoreWantarray Helper::ResultSet::SearchOr));

	sub failed {
	  my $self = shift;

	  my $me = $self->current_source_alias;

	  $self->search({ "$me.passed" => '0' });
	}

	sub untested {
	  my $self = shift;

	  my $me = $self->current_source_alias;

	  $self->search({ "$me.passed" => undef });
	}

	sub not_passed {
	  my $self = shift;

	  my $me = $self->current_source_alias;

	  $self->search_or([$self->failed, $self->untested]);
	}

	1;

DESCRIPTION

       I would argue that the most important feature of DBIx::Class is the fact that you can "chain" ResultSet searches.  Unfortunately this can
       cause problems when you need to reuse multiple ResultSet methods as... well as or's.  In the past I got around this by doing:

	$rs->foo->union([ $rs->bar]);

       While this works, it can generate some hairy SQL pretty fast.  This Helper is supposed to basically be a lightweight union.  Note that it
       therefor has a number of "LIMITATIONS".	The thing that makes this module special is that the ResultSet that is doing the "search_or" ing
       still limits everything correctly.  To be clear, the following only returns $user's friends that match either of the following criteria:

	my $friend_rs = $schema->resultset('Friend');
	my @internet_friends = $user->friends->search_or([
	  $friend_rs->on_facebook,
	  $friend_rs->on_twitter,
	])->all;

       With a union, you'd have to implement it like this:

	$user->friends->on_facebook->union([ $user->friends->on_twitter ]);

       The union will work, but it will generate more complex SQL that may have lower performance on your database.

       See "NOTE" in DBIx::Class::Helper::ResultSet for a nice way to apply it to your entire schema.

METHODS

   search_or
	my $new_rs = $rs->search_or([ $rs->foo, $rs->bar ]);

       "search_or" takes a single arrayref of ResultSets.  The ResultSets must point to the same source or you will get an error message.
       Additionally, no check is made to ensure that more than one ResultSet is in the ArrayRef, but only passing one ResultSet would not make any
       sense.

LIMITATIONS

       Because this module us basically an expression union and not a true union, "JOIN"'s won't Just Work.  If you have a ResultSet method that
       uses a "JOIN" and you want to "OR" it with another method, you'll need to do something like this:

	my @authors = $authors->search(undef, { join => 'books' })->search_or([
	   $authors->wrote_good_books,
	   $authors->wrote_bestselling_books,
	])->all;

       Furthermore, if you want to "OR" two methods that "JOIN" in the same relationship via alternate paths you must use union.

AUTHOR

       Arthur Axel "fREW" Schmidt <frioux+cpan@gmail.com>

COPYRIGHT AND LICENSE

       This software is copyright (c) 2012 by Arthur Axel "fREW" Schmidt.

       This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.

perl v5.14.2							    2012-06-18			     DBIx::Class::Helper::ResultSet::SearchOr(3pm)