5 Moose::Manual::Delta - Important Changes in Moose
9 This documents any important or noteworthy changes in Moose, with a
10 focus on backwards. This does duplicate data from the F<Changes> file,
11 but aims to provide more details and when possible workarounds.
13 Besides helping keep up with changes, you can also use this document
14 for finding the lowest version of Moose that supported a given
15 feature. If you encounter a problem and have a solution but don't see
16 it documented here, or think we missed an important feature, please
23 =item Calling $object->new() is no longer deprecated
25 We decided to undeprecate this. Now it just works.
33 =item Added Native delegation for Code refs
35 See L<Moose::Meta::Attribute::Native::Trait::Code> for details.
37 =item Calling $object->new() is deprecated
39 Moose has long supported this, but it's never really been documented, and we
40 don't think this is a good practice. If you want to construct an object from
41 an existing object, you should provide some sort of alternate constructor like
42 C<< $object->clone >>.
44 Calling C<< $object->new >> now issues a warning, and will be an error in a
47 =item Moose no longer warns if you call C<make_immutable> for a class with mutable ancestors
49 While in theory this is a good thing to warn about, we found so many
50 exceptions to this that doing this properly became quite problematic.
54 =head1 Version 0.89_02
58 =item New Native delegation methods from L<List::Util> and L<List::MoreUtils>
60 In particular, we now have C<reduce>, C<shuffle>, C<uniq>, and C<natatime>.
62 =item The Moose::Exporter with_caller feature is now deprecated
64 Use C<with_meta> instead. The C<with_caller> option will start warning in a
67 =item Moose now warns if you call C<make_immutable> for a class with mutable ancestors
69 This is dangerous because modifying a class after a subclass has been
70 immutabilized will lead to incorrect results in the subclass, due to inlining,
71 caching, etc. This occasionally happens accidentally, when a class loads one
72 of its subclasses in the middle of its class definition, so pointing out that
73 this may cause issues should be helpful. Metaclasses (classes that inherit
74 from L<Class::MOP::Object>) are currently exempt from this check, since at the
75 moment we aren't very consistent about which metaclasses we immutabilize.
77 =item C<enum> and C<duck_type> now take arrayrefs for all forms
79 Previously, calling these functions with a list would take the first element of
80 the list as the type constraint name, and use the remainder as the enum values
81 or method names. This makes the interface inconsistent with the anon-type forms
82 of these functions (which must take an arrayref), and a free-form list where
83 the first value is sometimes special is hard to validate (and harder to give
84 reasonable error messages for). These functions have been changed to take
85 arrayrefs in all their forms - so, C<< enum 'My::Type' => [qw(foo bar)] >> is
86 now the preferred way to create an enum type constraint. The old syntax still
87 works for now, but it will hopefully be deprecated and removed in a future
92 =head1 Version 0.89_01
94 L<Moose::Meta::Attribute::Native> has been moved into the Moose core from
95 L<MooseX::AttributeHelpers>. Major changes include:
99 =item C<traits>, not C<metaclass>
101 Method providers are only available via traits.
103 =item C<handles>, not C<provides> or C<curries>
105 The C<provides> syntax was like core Moose C<< handles => HASHREF >>
106 syntax, but with the keys and values reversed. This was confusing,
107 and AttributeHelpers now uses C<< handles => HASHREF >> in a way that
108 should be intuitive to anyone already familiar with how it is used for
111 The C<curries> functionality provided by AttributeHelpers has been
112 generalized to apply to all cases of C<< handles => HASHREF >>, though
113 not every piece of functionality has been ported (currying with a
114 CODEREF is not supported).
116 =item C<empty> is now C<is_empty>, and means empty, not non-empty
118 Previously, the C<empty> method provided by Arrays and Hashes returned true if
119 the attribute was B<not> empty (no elements). Now it returns true if the
120 attribute B<is> empty. It was also renamed to C<is_empty>, to reflect this.
122 =item C<find> was renamed to C<first>, and C<first> and C<last> were removed
124 L<List::Util> refers to the functionality that we used to provide under C<find>
125 as L<first|List::Util/first>, so that will likely be more familiar (and will
126 fit in better if we decide to add more List::Util functions). C<first> and
127 C<last> were removed, since their functionality is easily duplicated with
130 =item Helpers that take a coderef of one argument now use C<$_>
132 Subroutines passed as the first argument to C<first>, C<map>, and C<grep> now
133 receive their argument in C<$_> rather than as a parameter to the subroutine.
134 Helpers that take a coderef of two or more arguments remain using the argument
135 list (there are technical limitations to using C<$a> and C<$b> like C<sort>
138 See L<Moose::Meta::Attribute::Native> for the new documentation.
142 The C<alias> and C<excludes> role parameters have been renamed to C<-alias>
143 and C<-excludes>. The old names still work, but new code should use the new
144 names, and eventually the old ones will be deprecated and removed.
148 C<< use Moose -metaclass => 'Foo' >> now does alias resolution, just like
149 C<-traits> (and the C<metaclass> and C<traits> options to C<has>).
151 Added two functions C<meta_class_alias> and C<meta_attribute_alias> to
152 L<Moose::Util>, to simplify aliasing metaclasses and metatraits. This is
153 a wrapper around the old
155 package Moose::Meta::Class::Custom::Trait::FooTrait;
156 sub register_implementation { 'My::Meta::Trait' }
162 When an attribute generates I<no> accessors, we now warn. This is to help
163 users who forget the C<is> option. If you really do not want any accessors,
164 you can use C<< is => 'bare' >>. You can maintain back compat with older
165 versions of Moose by using something like:
167 ($Moose::VERSION >= 0.84 ? is => 'bare' : ())
169 When an accessor overwrites an existing method, we now warn. To work around
170 this warning (if you really must have this behavior), you can explicitly
171 remove the method before creating it as an accessor:
175 __PACKAGE__->meta->remove_method('foo');
181 When an unknown option is passed to C<has>, we now warn. You can silence
182 the warning by fixing your code. :)
184 The C<Role> type has been deprecated. On its own, it was useless,
185 since it just checked C<< $object->can('does') >>. If you were using
186 it as a parent type, just call C<role_type('Role::Name')> to create an
187 appropriate type instead.
191 C<use Moose::Exporter;> now imports C<strict> and C<warnings> into packages
196 C<DEMOLISHALL> and C<DEMOLISH> now receive an argument indicating whether or
197 not we are in global destruction.
201 Type constraints no longer run coercions for a value that already matches the
202 constraint. This may affect some (arguably buggy) edge case coercions that
203 rely on side effects in the C<via> clause.
207 L<Moose::Exporter> now accepts the C<-metaclass> option for easily
208 overriding the metaclass (without L<metaclass>). This works for classes
213 Added a C<duck_type> sugar function to L<Moose::Util::TypeConstraints>
214 to make integration with non-Moose classes easier. It simply checks if
215 C<< $obj->can() >> a list of methods.
217 A number of methods (mostly inherited from L<Class::MOP>) have been
218 renamed with a leading underscore to indicate their internal-ness. The
219 old method names will still work for a while, but will warn that the
220 method has been renamed. In a few cases, the method will be removed
221 entirely in the future. This may affect MooseX authors who were using
226 Calling C<subtype> with a name as the only argument now throws an
227 exception. If you want an anonymous subtype do:
229 my $subtype = subtype as 'Foo';
231 This is related to the changes in version 0.71_01.
233 The C<is_needed> method in L<Moose::Meta::Method::Destructor> is now
234 only usable as a class method. Previously, it worked as a class or
235 object method, with a different internal implementation for each
238 The internals of making a class immutable changed a lot in Class::MOP
239 0.78_02, and Moose's internals have changed along with it. The
240 external C<< $metaclass->make_immutable >> method still works the same
245 A mutable class accepted C<< Foo->new(undef) >> without complaint,
246 while an immutable class would blow up with an unhelpful error. Now,
247 in both cases we throw a helpful error instead.
249 This "feature" was originally added to allow for cases such as this:
257 return My::Class->new($args);
259 But we decided this is a bad idea and a little too magical, because it
260 can easily mask real errors.
262 =head1 Version 0.71_01
264 Calling C<type> or C<subtype> without the sugar helpers (C<as>,
265 C<where>, C<message>) is now deprecated.
267 As a side effect, this meant we ended up using Perl prototypes on
268 C<as>, and code like this will no longer work:
270 use Moose::Util::TypeConstraints;
271 use Declare::Constraints::Simple -All;
273 subtype 'ArrayOfInts'
275 => IsArrayRef(IsInt);
277 Instead it must be changed to this:
282 where => IsArrayRef(IsInt)
286 If you want to maintain backwards compat with older versions of Moose,
287 you must explicitly test Moose's C<VERSION>:
289 if ( Moose->VERSION < 0.71_01 ) {
290 subtype 'ArrayOfInts'
292 => IsArrayRef(IsInt);
298 where => IsArrayRef(IsInt)
305 We no longer pass the meta-attribute object as a final argument to
306 triggers. This actually changed for inlined code a while back, but the
307 non-inlined version and the docs were still out of date.
309 If by some chance you actually used this feature, the workaround is
310 simple. You fetch the attribute object from out of the C<$self>
311 that is passed as the first argument to trigger, like so:
317 my ( $self, $value ) = @_;
318 my $attr = $self->meta->find_attribute_by_name('foo');
326 If you created a subtype and passed a parent that Moose didn't know
327 about, it simply ignored the parent. Now it automatically creates the
328 parent as a class type. This may not be what you want, but is less
331 You could declare a name with subtype such as "Foo!Bar". Moose would
332 accept this allowed, but if you used it in a parameterized type such
333 as "ArrayRef[Foo!Bar]" it wouldn't work. We now do some vetting on
334 names created via the sugar functions, so that they can only contain
335 alphanumerics, ":", and ".".
339 Methods created via an attribute can now fulfill a C<requires>
340 declaration for a role. Honestly we don't know why Stevan didn't make
341 this work originally, he was just insane or something.
343 Stack traces from inlined code will now report the line and file as
344 being in your class, as opposed to in Moose guts.
346 =head1 Version 0.62_02
348 When a class does not provide all of a role's required methods, the
349 error thrown now mentions all of the missing methods, as opposed to
350 just the first missing method.
352 Moose will no longer inline a constructor for your class unless it
353 inherits its constructor from Moose::Object, and will warn when it
354 doesn't inline. If you want to force inlining anyway, pass
355 C<< replace_constructor => 1 >> to C<make_immutable>.
357 If you want to get rid of the warning, pass C<< inline_constructor =>
362 Removed the (deprecated) C<make_immutable> keyword.
364 Removing an attribute from a class now also removes delegation
365 (C<handles>) methods installed for that attribute. This is correct
366 behavior, but if you were wrongly relying on it you might get bit.
370 Roles now add methods by calling C<add_method>, not
371 C<alias_method>. They make sure to always provide a method object,
372 which will be cloned internally. This means that it is now possible to
373 track the source of a method provided by a role, and even follow its
374 history through intermediate roles. This means that methods added by
375 a role now show up when looking at a class's method list/map.
377 Parameter and Union args are now sorted, this makes Int|Str the same
378 constraint as Str|Int. Also, incoming type constraint strings are
379 normalized to remove all whitespace differences. This is mostly for
380 internals and should not affect outside code.
382 L<Moose::Exporter> will no longer remove a subroutine that the
383 exporting package re-exports. Moose re-exports the Carp::confess
384 function, among others. The reasoning is that we cannot know whether
385 you have also explicitly imported those functions for your own use, so
386 we err on the safe side and always keep them.
390 C<Moose::init_meta> should now be called as a method.
392 New modules for extension writers, L<Moose::Exporter> and
393 L<Moose::Util::MetaRole>.
395 =head1 Version 0.55_01
397 Implemented metaclass traits (and wrote a recipe for it):
399 use Moose -traits => 'Foo'
401 This should make writing small Moose extensions a little
406 Fixed C<coerce> to accept anon types just like C<subtype> can.
409 coerce $some_anon_type => from 'Str' => via { ... };
413 Added C<BUILDARGS>, a new step in C<< Moose::Object->new() >>.
417 Fixed how the C<< is => (ro|rw) >> works with custom defined
418 C<reader>, C<writer> and C<accessor> options. See the below table for
421 is => ro, writer => _foo # turns into (reader => foo, writer => _foo)
422 is => rw, writer => _foo # turns into (reader => foo, writer => _foo)
423 is => rw, accessor => _foo # turns into (accessor => _foo)
424 is => ro, accessor => _foo # error, accesor is rw
428 The C<before/around/after> method modifiers now support regexp
429 matching of method names. NOTE: this only works for classes, it is
430 currently not supported in roles, but, ... patches welcome.
432 The C<has> keyword for roles now accepts the same array ref form that
433 L<Moose>.pm does for classes.
435 A trigger on a read-only attribute is no longer an error, as it's
436 useful to trigger off of the constructor.
438 Subtypes of parameterizable types now are parameterizable types
443 Fixed issue where C<DEMOLISHALL> was eating the value in C<$@>, and so
444 not working correctly. It still kind of eats them, but so does vanilla
449 Inherited attributes may now be extended without restriction on the
450 type ('isa', 'does').
452 The entire set of Moose::Meta::TypeConstraint::* classes were
453 refactored in this release. If you were relying on their internals you
454 should test your code carefully.
458 Documenting the use of '+name' with attributes that come from recently
459 composed roles. It makes sense, people are using it, and so why not
460 just officially support it.
462 The C<< Moose::Meta::Class->create >> method now supports roles.
464 It is now possible to make anonymous enum types by passing C<enum> an
465 array reference instead of the C<< enum $name => @values >>.
469 Added the C<make_immutable> keyword as a shortcut to calling
470 C<make_immutable> on the meta object. This eventually got removed!
472 Made C<< init_arg => undef >> work in Moose. This means "do not accept
473 a constructor parameter for this attribute".
475 Type errors now use the provided message. Prior to this release they
480 Moose is now a postmodern object system :)
482 The Role system was completely refactored. It is 100% backwards
483 compat, but the internals were totally changed. If you relied on the
484 internals then you are advised to test carefully.
486 Added method exclusion and aliasing for Roles in this release.
488 Added the L<Moose::Util::TypeConstraints::OptimizedConstraints>
491 Passing a list of values to an accessor (which is only expecting one
492 value) used to be silently ignored, now it throws an error.
496 Added parameterized types and did a pretty heavy refactoring of the
497 type constraint system.
499 Better framework extendability and better support for "making your own
502 =head1 Version 0.25 or before
504 Honestly, you shouldn't be using versions of Moose that are this old,
505 so many bug fixes and speed improvements have been made you would be
506 crazy to not upgrade.
508 Also, I am tired of going through the Changelog so I am stopping here,
509 if anyone would like to continue this please feel free.
513 Stevan Little E<lt>stevan@iinteractive.comE<gt>
515 =head1 COPYRIGHT AND LICENSE
517 Copyright 2009 by Infinity Interactive, Inc.
519 L<http://www.iinteractive.com>
521 This library is free software; you can redistribute it and/or modify
522 it under the same terms as Perl itself.