Class::MOP::load_class change

[gitmo/Moose.git] / lib / Moose / Cookbook / FAQ.pod

=pod

=head1 NAME

Moose::Cookbook::FAQ - Frequenty asked questions about Moose

=head1 FREQUENTLY ASKED QUESTIONS

=head2 Module Stability

=head3 Is Moose "production ready"?

Yes and No. Currently I have one web application in production 
using Moose, and at $work we are re-writing our core offering to 
use Moose. Several other people on #moose either have sites in 
production which use Moose, or are in the process of deploying 
sites which use Moose. 

The biggest barrier to widespread use of Moose in production 
right now is speed of development and speed of execution. 

Since development is still happening, regular upgrades are a 
fact of life. This can be hairy in production, so if this makes 
you quake with fear, you might want to wait a few months.

Then comes speed of execution. In some ways, Moose is actually 
pretty fast and makes great effort to stay out of your way when 
you don't want it there. However, certain parts of Moose are 
slow, such as compile time setup, introspection and object 
construction (only because it uses introspection). See 
L<Is Moose slow?> below for a deeper discussion on the subject.

=head3 Is Moose's API stable?

Yes and No. The external API, the one 90% of users will interact
with, is B<very stable> and any changes B<will be 100% backwards 
compatible>. The introspection API is I<mostly> stable, I still 
reserve the right to tweak that if needed, but I will do my 
absolute best to maintain backwards comptability here as well.

=head3 Is Moose slow?

Again, this one is tricky, so Yes I<and> No.

First let me say that I<nothing> in life is free, and that some 
Moose features do cost more than others. It is also the 
policy of Moose to B<only charge you for the features you use>, 
and to do our absolute best to not place any extra burdens on 
the execution of your code for features you are not using. 

Next, I will point out again that we are still in the "early 
adopter" phase, so speed it not that important yet. We are 
actually optimizing for "theoretical correctness" first, and 
we will optimize for speed later. It has been our experience 
that taking this approach allows for greater optimization 
capacity. 

And lastly, I want to reassure the speed junkies out there that 
we B<are> working on it. 

We have the immutable classes in Class::MOP, but which are not 
yet integrated with Moose. These allow you to "close" a class 
and then for many of it's more expensive methods to me memoized. 
Our tests indicated a performance comparable (and in some 
instances exceeding) that of hand-coded Perl.

We are also discussing and experimenting with L<Module::Compile>, 
and the idea of compiling highly optimized C<.pmc> files. And we 
have also mapped out some core methods as canidates for conversion 
to XS. 

=head3 When will Moose be 1.0 ready?

I expect (barring unforseen circumstances) that Moose will be 
at 1.0 by the end of this year (2006). Which means that it will be 
completely stable and provide a number of optimization options to 
suit the need for speed. 

Will I have addressed all your concerns by then? Will all the 
features you want be included? I don't know unless you tell me, 
so come over to #moose and we can talk.

=head2 Constructors

=head3 How do I write custom constructors with Moose?

Ideally, you should never write your own C<new> method, and should
use Moose's other features to handle your specific object construction
needs. Here are a few scenarios, and the Moose way to solve them;

If you need to call initializtion code post instance construction, 
then use the C<BUILD> method. This feature is taken directly from 
Perl 6. Every C<BUILD> method in your inheritence chain is called 
(in the correct order) immediately after the instance is constructed. 
This allows you to ensure that all your superclasses are initialized 
properly as well. This is the best approach to take (when possible)
because it makes subclassing your class much easier.

If you need to affect the constructor's parameters prior to the 
instance actually being constructed, you have a number of options.

First, there are I<coercions> (See the L<Moose::Cookbook::Recipe5> 
for a complete example and explaination of coercions). With 
coercions it is possible to morph argument values into the correct 
expected types. This approach is the most flexible and robust, but 
does have a slightly higher learning curve.

Second, using an C<around> method modifier on C<new> can be an 
effective way to affect the contents of C<@_> prior to letting 
Moose deal with it. This carries with it the extra burden for 
your subclasses, in that they have to be sure to explicitly 
call your C<new> and/or work around your C<new> to get to the 
version from L<Moose::Object>. 

The last approach is to use the standard Perl technique of calling 
the C<SUPER::new> within your own custom version of C<new>. This 
of course brings with it all the issues of the C<around> solution 
along with any issues C<SUPER::> might add as well.

In short, try to use C<BUILD> and coercions, they are your best 
bets.

=head3 How do I make non-Moose constuctors work with Moose? 

Moose provides it's own constructor, but it does it by making all 
Moose-based classes inherit from L<Moose::Object>. When inheriting
from a non-Moose class, the inheritence chain to L<Moose::Object> 
is broken. The simplest way to fix this is to simply explicitly 
inherit from L<Moose::Object> yourself. However, this does not 
always fix the issue of a constructor. Here is a basic example of 
how this can be worked around:

  package My::HTML::Template;
  use Moose;
  
  # explict inheritence 
  extends 'HTML::Template', 'Moose::Object';
  
  # explicit constructor
  sub new {
      my $class = shift;
      # call HTML::Template's constructor
      my $obj = $class->SUPER::new(@_);
      return $class->meta->new_object(
          # pass in the constructed object
          # using the special key __INSTANCE__
          __INSTANCE__ => $obj, @_
      );
  }

Of course this only works if both your Moose class, and the 
inherited non-Moose class use the same instance type (typically 
HASH refs). 

Other techniques can be used as well, such as creating the object 
using C<Moose::Object::new>, but calling the inherited non-Moose 
class's initializtion methods (if available). 

It is also entirely possible to just rely on HASH autovivification 
to create the slot's needed for Moose based attributes. Although 
this does somewhat restrict use of construction time attribute 
features.

In short, there are several ways to go about this, it is best to 
evaluate each case based on the class you wish to extend, and the 
features you wish to employ. As always, both IRC and the mailing 
list are great ways to get help finding the best approach.

=head2 Accessors

=head3 How do I tell Moose to use get/set accessors?

The easiest way to accomplish this is to use the C<reader> and 
C<writer> attribute options. Here is some example code:

  has 'bar' => (
      isa    => 'Baz',
      reader => 'get_bar', 
      writer => 'set_bar',
  );

Moose will still take advantage of type constraints, triggers, etc. 
when creating these methods. 

If you do not like this much typing, and wish it to be a default for 
your class. Please see L<Moose::Policy>, and more specifically the 
L<Moose::Policy::FollowPBP>. This will allow you to write this:

  has 'bar' => (
      isa => 'Baz',
      is  => 'rw',
  );

And have Moose create C<get_bar> and C<set_bar> instead of the usual 
C<bar>.

NOTE: This B<cannot> be set globally in Moose, as this would break 
other classes which are built with Moose.

=head3 How can I get Moose to inflate/deflate values in the accessor?

Well, the first question to ask is if you actually need both inflate 
and deflate.

If you only need to inflate, then I suggest using coercions. Here is 
some basic sample code for inflating a L<DateTime> object. 

  subtype 'DateTime'
      => as 'Object'
      => where { $_->isa('DateTime') };
      
  coerce 'DateTime'
      => from 'Str'
        => via { DateTime::Format::MySQL->parse_datetime($_) };
        
  has 'timestamp' => (is => 'rw', isa => 'DateTime', coerce => 1);

This creates a custom subtype for L<DateTime> objects, then attaches 
a coercion to that subtype. The C<timestamp> attribute is then told 
to expect a C<DateTime> type, and to try and coerce it. When a C<Str>
type is given to the C<timestamp> accessor, it will attempt to 
coerce the value into a C<DateTime> object using the code in found 
in the C<via> block. 

For a more detailed and complete example of coercions, see the 
L<Moose::Cookbook::Recipe5>.

If you need to deflate your attribute, the current best practice is to 
add an C<around> modifier to your accessor. Here is some example code:

  # a timestamp which stores as 
  # seconds from the epoch
  has 'timestamp' => (is => 'rw', isa => 'Int');
  
  around 'timestamp' => sub {
      my $next = shift;
      my ($self, $timestamp) = @_;
      # assume we get a DateTime object ...
      $next->($self, $timestamp->epoch);
  };

It is also possible to do deflation using coercion, but this tends 
to get quite complex and require many subtypes. An example of this 
is outside the scope of this document, ask on #moose or send a mail 
to the list.

Still another option is to write a custom attribute metaclass, which 
is also outside the scope of this document, but I would be happy to 
explain it on #moose or the mailing list.

=head2 Method Modfiers

=head3 How can I affect the values in C<@_> using C<before>?

You can't actually, C<before> only runs before the main method, 
and it cannot easily affect the execution of it. What you want is 
an C<around> method. 

=head3 Can I use C<before> to stop execution of a method?

Yes, but only if you throw an exception. If this is too drastic a 
measure then I suggest using C<around> instead. The C<around> method 
modifier is the only modifier which can actually stop the execution 
of the main method. Here is an example:

  around 'baz' => sub {
      my $next = shift;
      my ($self, %options) = @_;
      if ($options{bar} eq 'foo') {
          $next->($self, %options);
      }
      else {
          return 'bar';
      }
  };

By choosing not to call the C<$next> method, you can stop the 
execution of the main method.

=head2 Type Constraints

=head3 How can I have a custom error message for a type constraint?

Use the C<message> option when building the subtype. Like so:

  subtype 'NaturalLessThanTen' 
      => as 'Natural'
      => where { $_ < 10 }
      => message { "This number ($_) is not less than ten!" };

This will be called when a value fails to pass the C<NaturalLessThanTen>
constraint check. 

=head1 AUTHOR

Stevan Little E<lt>stevan@iinteractive.comE<gt>

=head1 COPYRIGHT AND LICENSE

Copyright 2006, 2007 by Infinity Interactive, Inc.

L<http://www.iinteractive.com>

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

=cut