Remove implementation notes

[gitmo/MooseX-Role-Parameterized.git] / lib / MooseX / Role / Parameterized / Tutorial.pod

=pod

=head1 NAME

MooseX::Role::Parameterized::Tutorial - why and how

=head1 MOTIVATION

Roles are composable units of behavior. They are useful for factoring out
functionality common to many classes from any part of your class hierarchy. See
L<Moose::Cookbook::Roles::Recipe1> for an introduction to L<Moose::Role>.

While combining roles affords you a great deal of flexibility, individual roles
have very little in the way of configurability. Core Moose provides C<alias>
for renaming methods and C<excludes> for ignoring methods. These options are
primarily (perhaps solely) for disambiguating role conflicts. See
L<Moose::Cookbook::Roles::Recipe2> for more about C<alias> and C<excludes>.

Because roles serve many different masters, they usually provide only the least
common denominator of functionality. To empower roles further, more
configurability than C<alias> and C<excludes> is required. Perhaps your role
needs to know which method to call when it is done. Or what default value to
use for its C<url> attribute.

Parameterized roles offer exactly this solution.

=head1 USAGE

=head3 C<with>

The syntax of a class consuming a parameterized role has not changed from the
standard C<with>. You pass in parameters just like you pass in C<alias> and
C<excludes> to ordinary roles:

    with 'MyRole::InstrumentMethod' => {
        method_name => 'dbh_do',
        log_to      => 'query.log',
    };

You can still combine parameterized roles. You just need to specify parameters
immediately after the role they belong to:

    with (
        'My::Parameterized::Role' => {
            needs_better_example => 1,
        },
        'My::Other::Role',
    );

=head3 C<parameter>

Inside your parameterized role, you specify a set of parameters. This is
exactly like specifying the attributes of a class. Instead of C<has> you use
the keyword C<parameter>, but your parameters can use any options to C<has>.

    parameter 'delegation' => (
        isa       => 'HashRef|ArrayRef|RegexpRef',
        predicate => 'has_delegation',
    );

You do have to declare what parameters you accept, just like you have to
declare what attributes you accept for regular Moose objects.

=head3 C<role>

C<role> takes a block of code that will be used to generate your role with its
parameters bound. Here is where you declare parameterized components: use
C<has>, method modifiers, and so on. The C<role> block receives an argument,
which contains the parameters specified by C<with>. You can access the
parameters just like regular attributes on that object.

Each time you compose this parameterized role, the role {} block will be
executed. It will receive a new parameter object and produce an entirely new
role. That's the whole point, after all.

Due to limitations inherent in Perl, you must declare methods with
C<< method name => sub { ... } >> instead of the usual C<sub name { ... }>.
Your methods may, of course, close over the parameter object. This means that
your methods may use parameters however they wish!

=head1 USES

Ideally these will become fully-explained examples in something resembling
L<Moose::Cookbook>. But for now, only a braindump.

=over 4

=item Configure a role's attributes

You can rename methods with core Moose, but now you can rename attributes. You
can now also choose type, default value, whether it's required, B<traits>, etc.

    parameter traits => (
        isa     => 'ArrayRef[Str]',
        default => sub { [] },
    );

    parameter type => (
        isa     => 'Str',
        default => 'Any',
    );

    role {
        my $p = shift;

        has action => (
            traits => $p->traits,
            isa    => $p->type,
            ...
        );
    }

=item Inform a role of your class' attributes and methods

Core roles can require only methods with specific names. Now your roles can
require that you specify a method name you wish the role to instrument, or
which attributes to dump to a file.

    parameter instrument_method => (
        isa      => 'Str',
        required => 1,
    );

    role {
        my $p = shift;
        around $p->instrument_method => sub { ... };
    }

=item Arbitrary execution choices

Your role may be able to provide configuration in how the role's methods
operate. For example, you can tell the role whether to save intermediate
states.

    parameter save_intermediate => (
        isa     => 'Bool',
        default => 0,
    );

    role {
        my $p = shift;
        method process => sub {
            ...
            if ($p->save_intermediate) { ... }
            ...
        };
    }

=item Deciding a backend

Your role may be able to freeze and thaw your instances using L<YAML>, L<JSON>,
L<Storable>. Which backend to use can be a parameter.

    parameter format => (
        isa     => (enum ['Storable', 'YAML', 'JSON']),
        default => 'Storable',
    );

    role {
        my $p = shift;
        if ($p->format eq 'Storable') {
            method freeze => \&Storable::freeze;
            method thaw   => \&Storable::thaw;
        }
        elsif ($p->format eq 'YAML') {
            method freeze => \&YAML::Dump;
            method thaw   => \&YAML::Load;
        }
        ...
    }

=back

=cut