💯 Doug Bell


Beam::Runner - Configure, list, document, and execute runnable task objects


version 0.012


    beam run <container> <task> [<args...>]
    beam list
    beam list <container>
    beam help <container> <task>
    beam help


This distribution is an execution and organization system for runnable objects (tasks). This allows you to prepare a list of runnable tasks in configuration files and then execute them. This also allows easy discovery of configuration files and objects, and allows you to document your objects for your users.


A task is an object that consumes the Beam::Runnable role. This role requires only a run() method be implemented in the class. This run() method should accept all the arguments given on the command line. It can parse GNU-style options out of this array using "GetOptionsFromArray" in Getopt::Long.

Task modules can compose additional roles to easily add more features, like adding a timeout with Beam::Runnable::Timeout::Alarm.

Task modules are expected to have documentation that will be displayed by the beam list and beam help commands. The beam list command will display the NAME section of the documentation, and the beam help command will display the NAME, SYNOPSIS, DESCRIPTION, ARGUMENTS, OPTIONS, ENVIRONMENT, and SEE ALSO sections of the documentation.

Configuration Files

The configuration file is a Beam::Wire container file that describes objects. Some of these objects are marked as executable tasks by consuming the Beam::Runnable role.

The container file can have a special entry called $summary which has a short summary that will be displayed when using the beam list command.

Here's an example container file that has a summary, configures a DBIx::Class schema (using the schema class for CPAN Testers: CPAN::Testers::Schema), and configures a runnable task called to_metabase located in the class CPAN::Testers::Backend::Migrate::ToMetabase:

    # migrate.yml
    $summary: Migrate data between databases

        $class: CPAN::Testers::Schema
        $method: connect_from_config

        $class: CPAN::Testers::Backend::Migrate::ToMetabase
            $ref: _schema

For more information about container files, see the Beam::Wire documentation.


Here's a short tutorial for getting started with Beam::Runner. If you want to try it yourself, start with an empty directory.

Create a Task

To create a task, make a Perl module that uses the Beam::Runnable role and implements a run method. For an example, let's create a task that prints Hello, World! to the screen.

    package My::Runnable::Greeting;
    use Moo;
    with 'Beam::Runnable';
    sub run {
        my ( $self, @args ) = @_;
        print "Hello, World!\n";

If you're following along, save this in the lib/My/Runnable/Greeting.pm file.

Create a Configuration File

Now that we have a task to run, we need to create a configuration file (or a "container"). The configuration file is a YAML file that describes all the tasks we can run. Let's create an etc directory and name our container file etc/greet.yml.

Inside this file, we define our task. We have to give our task a simple name, like hello. Then we have to say what task class to run (in our case, My::Runnable::Greeting).

        $class: My::Runnable::Greeting

Run the Task

Now we can run our task. Before we do, we need to tell Beam::Runner where to find our code and our configuration by setting some environment variables:

    $ export PERL5LIB=lib:$PERL5LIB
    $ export BEAM_PATH=etc

The PERL5LIB environment variable adds directories for perl to search for modules (like our task module). The BEAM_PATH environment variable adds directories to search for configuration files (like ours).

To validate that our environment variables are set correctly, we can list the tasks:

    $ beam list
    - hello -- My::Runnable::Greeting

The beam list command looks through our BEAM_PATH directory, opens all the configuration files it finds, and lists all the Beam::Runnable objects inside (helpfully giving us the module name for us to find documentation).

Then, to run the command, we use beam run and give it the configuration file (greet) and the task (hello):

    $ beam run greet hello
    Hello, World!

Adding Documentation

Part of the additional benefits of defining tasks in Beam::Runnable modules is that the beam help command will show the documentation for the task. To do this, we must add documentation to our module.

This documentation is done as POD, Perl's system of documentation. Certain sections of the documentation will be shown: NAME, SYNOPSIS, DESCRIPTION, ARGUMENTS, OPTIONS, and SEE ALSO.

    =head1 NAME

    My::Runnable::Greeting - Greet the user

    =head1 SYNOPSIS

        beam run greet hello

    =head1 DESCRIPTION

    This task greets the user warmly and then exits.

    =head1 ARGUMENTS

    No arguments are allowed during a greeting.

    =head1 OPTIONS

    Greeting warmly is the only option.

    =head1 SEE ALSO


If we add this documentation to our lib/My/Runnable/Greeting.pm file, we can then run beam help to see the documentation:

    $ beam help greet hello
        My::Runnable::Greeting - Greet the user

            beam run greet hello

        This task greets the user warmly and then exits.

        No arguments are allowed during a greeting.

        Greeting warmly is the only option.


The beam list command will also use our new documentation to show the NAME section:

    $ beam list
    - hello -- My::Runnable::Greeting - Greet the user

Going Further

For more information on how to use the configuration file to create more complex objects like database connections, see Beam::Wire::Help::Config.

To learn how to run your tasks using a distributed job queue to parallelize and improve performance, see Beam::Minion.


beam, Beam::Runnable, Beam::Wire


Doug Bell <preaction@cpan.org>


This software is copyright (c) 2016 by Doug Bell.

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