Class::DBI::Plugin::PseudoColumns version 0.03
==============================================

NAME
   Class::DBI::Plugin::PseudoColumns - an interface to use some pseudo columns

SYNOPSIS
    package Music::CD;
    use base 'Music::DBI';

    Music::CD->table('cd');
    Music::CD->columns(All => qw/cdid artist title year reldate properties/);
    use Class::DBI::Plugin::PseudoColumns;
    Music::CD->pseudo_columns(properties => qw/asin tag/);

    use Music::CD;
    my $cds = Music::CD->search(artist => 'Steve Vai');
    while (my $cd = $cds->next) {
        if ($cd->title =~ /^Real\s+Illusions/i) {
            $cd->asin('B0007GADZO');
        }
        $cd->tag(['rock', 'guitar', 'tricky play']);
        $cd->update;
    }

    my $bought_cd = Music::CD->create({
        artist  => 'Rolling Stones',
        title   => 'A Bigger Bang',
        year    => 2005,
        reldate => '2005-11-22',
        asin    => 'B000BP86LE',
        tag     => ['rock', 'blues', 'favorite'],
    });
    $bought_cd->set(
        artist => 'The Rolling Stones',
        title  => 'A Bigger Bang - Special Edition',
    );
    $bought_cd->update;

DESCRIPTION
   This module provides an easy way to use pseudo column in your
   "Class::DBI" based table classes. The ``pseudo column'' means a kind of
   column that is including an optical hashref string (via "Data::Dumper",
   by default). You can get and set with using the pseudo column accessors
   (same as always). But, you can't use the columns' name into your SQL,
   SQL interfaced methods, "ORDER BY" clause and "GROUP BY" clause, etc.
   This module is useful when you'd like to add an unimportant column
   without issuing "ALTER TABLE", and when you'd like to have related
   multiple data without normalizing table.

HOW TO USE
 Create a column
   You should create a huge size column into your table. like this:

    CREATE TABLE cd (
      cdid int UNSIGNED auto_increment,
      artist varchar(255),
      title varchar(255),
      year int,
      reldate date,
      properties text, # create for using pseudo column
      PRIMARY KEY (cdid)
    );

 Create a table class
   Almost same as usual.

    package Music::CD;
    use base 'Music::DBI';

    Music::CD->table('cd');
    Music::CD->columns(All => qw/cdid artist title year reldate properties/);

 Use it
   You will be able to use pseudo column with only to "use" this module.

    use Class::DBI::Plugin::PseudoColumns;

 Set your pseudo columns as your like
    Music::CD->pseudo_columns(properties => qw/asin tag/);

METHOD
   This module provides following class methods.

   * create(\%data);
       "create" method works almost same as "Class::DBI::create()" if
       there's some pseudo column in %data.

   * set(column => value[, column2 => value2, ...]);
       "set" method works almost same as "Class::DBI::set()" if there's
       some pseudo column in argument.

   * pseudo_columns([parent_colname => ('pseudo_column1'[, 'pseudo_column2'
   ...])]);
       You can set a pseudo columns' name with parent column's name.
       ``pseudo_column1'' ... will provide to you each pseudo column's
       accessor.

       if you want to take a list of pseudo columns, you can pass one
       argument to this method when you want to specify grouped parent
       column name.

        my @p_columns = Music::CD->pseudo_columns('properties');

       And if you want to take all columns list of pseudo columns, you
       don't need to pass any argument to this method.

        my @all_p_columns = Music::CD->pseudo_columns();

   * serializer(parent_colname => \&serializer_sub);
       You can set a specific serializing function for your pseudo columns.

        use Storable ();
        __PACKAGE__->serializer(properties => \&Storable::nfreeze);

       The default serializer is "Data::Dumper::Dumper" (when you don't
       specify).

   * deserializer(parent_coluname => \&deserializer_sub);
       You can set a specific deserializing function for your pseudo
       columns.

        use Storable ();
        __PACKAGE__->deserializer(properties => \&Storable::thaw);

       The default deserializer calls "eval()" (when you don't specify) for
       the dumped optical hashref string.

       NOTE: The subref for serializer/deserializer must return a really
       ``storable'' string for your database. example of above works under
       a MySQL environment, but you have to change to use some another
       ``storable'' filter (like "MIME::Base64") under SQLite environment
       (see t/05_serializer.t)

AUTHOR
   Koichi Taniguchi <[email protected]>

COPYRIGHT
   Copyright (c) 2006 Koichi Taniguchi. Japan. All rights reserved.

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

SEE ALSO
   Class::DBI, Data::Dumper