Dancer::Plugin::DBIC - DBIx::Class interface for Dancer applications

version 0.2100

use Dancer;
use Dancer::Plugin::DBIC qw(schema resultset rset);

get '/users/:user_id' => sub {
    my $user = schema('default')->resultset('User')->find(param 'user_id');

    # If you are accessing the 'default' schema, then all the following
    # are equivalent to the above:
    $user = schema->resultset('User')->find(param 'user_id');
    $user = resultset('User')->find(param 'user_id');
    $user = rset('User')->find(param 'user_id');

    template user_profile => {
        user => $user
    };
};

dance;

This plugin makes it very easy to create Dancer applications that interface with databases. It automatically exports the keyword schema which returns a DBIx::Class::Schema object. You just need to configure your database connection information. For performance, schema objects are cached in memory and are lazy loaded the first time they are accessed.

Configuration can be done in your Dancer config file.

Here is a simple example. It defines one database named default:

plugins:
  DBIC:
    default:
      dsn: dbi:SQLite:dbname=myapp.db
      schema_class: MyApp::Schema

In this example, there are 2 databases configured named default and foo:

plugins:
  DBIC:
    default:
      dsn: dbi:SQLite:dbname=myapp.db
      schema_class: MyApp::Schema
    foo:
      dsn: dbi:Pg:dbname=foo
      schema_class: Foo::Schema
      user: bob
      password: secret
      options:
        RaiseError: 1
        PrintError: 1

Each database configured must at least have a dsn option. The dsn option should be the DBI driver connection string. All other options are optional.

If you only have one schema configured, or one of them is named default, you can call schema without an argument to get the only or default schema, respectively.

If a schema_class option is not provided, then DBIx::Class::Schema::Loader will be used to dynamically load the schema by introspecting the database corresponding to the dsn value. You need DBIx::Class::Schema::Loader installed for this to work.

WARNING: Dynamic loading is not recommended for production environments. It is almost always better to provide a schema_class option.

The schema_class option should be the name of your DBIx::Class::Schema class. See ["SCHEMA GENERATION"](#SCHEMA GENERATION) Optionally, a database configuration may have user, password, and options parameters as described in the documentation for connect() in DBI.

Alternatively, you may also declare your connection information inside an array named connect_info:

plugins:
  DBIC:
    default:
      schema_class: MyApp::Schema
      connect_info:
        - dbi:Pg:dbname=foo
        - bob
        - secret
        -
          RaiseError: 1
          PrintError: 1

You can also add database read slaves to your configuration with the replicated config option. This will automatically make your read queries go to a slave and your write queries go to the master. Keep in mind that this will require additional dependencies: DBIx::Class::Optional::Dependencies#Storage::Replicated See DBIx::Class::Storage::DBI::Replicated for more details. Here is an example configuration that adds two read slaves:

plugins:
  DBIC:
    default:
      schema_class: MyApp::Schema
      dsn: dbi:Pg:dbname=master
      replicated:
        balancer_type: ::Random     # optional
        balancer_args:              # optional
            auto_validate_every: 5  # optional
            master_read_weight:1    # optional
        # pool_type and pool_args are also allowed and are also optional
        replicants:
          -
            - dbi:Pg:dbname=slave1
            - user1
            - password1
            -
              quote_names: 1
              pg_enable_utf8: 1
          -
            - dbi:Pg:dbname=slave2
            - user2
            - password2
            -
              quote_names: 1
              pg_enable_utf8: 1

Schema aliases allow you to reference the same underlying database by multiple names. For example:

plugins:
  DBIC:
    default:
      dsn: dbi:Pg:dbname=master
      schema_class: MyApp::Schema
    slave1:
      alias: default

Now you can access the default schema with schema(), schema('default'), or schema('slave1'). This can come in handy if, for example, you have master/slave replication in your production environment but only a single database in your development environment. You can continue to reference schema('slave1') in your code in both environments by simply creating a schema alias in your development.yml config file, as shown above.

my $user = schema->resultset('User')->find('bob');

The schema keyword returns a DBIx::Class::Schema object ready for you to use. If you have configured only one database, then you can simply call schema with no arguments. If you have configured multiple databases, you can still call schema with no arguments if there is a database named default in the configuration. With no argument, the default schema is returned. Otherwise, you must provide schema() with the name of the database:

my $user = schema('foo')->resultset('User')->find('bob');

This is a convenience method that will save you some typing. Use this only when accessing the default schema.

my $user = resultset('User')->find('bob');

is equivalent to:

my $user = schema->resultset('User')->find('bob');

my $user = rset('User')->find('bob');

This is simply an alias for resultset.

Setting the schema_class option and having proper DBIx::Class classes is the recommended approach for performance and stability. You can use the dbicdump command line tool provided by DBIx::Class::Schema::Loader to help you. For example, if your app were named Foo, then you could run the following from the root of your project directory:

dbicdump -o dump_directory=./lib Foo::Schema dbi:SQLite:/path/to/foo.db

For this example, your schema_class setting would be 'Foo::Schema'.

• Alexis Sukrieh [email protected]
• Dagfinn Ilmari Mannsåker <https://github.com/ilmari>
• David Precious [email protected]
• Fabrice Gabolde <https://github.com/fgabolde>
• Franck Cuny [email protected]
• Steven Humphrey <https://github.com/shumphrey>
• Yanick Champoux <https://github.com/yanick>

• Al Newkirk [email protected]
• Naveed Massjouni [email protected]

This software is copyright (c) 2010 by awncorp.

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