1 =head1 Upgrading to Catalyst 5.80
3 Most applications and plugins should run unaltered on Catalyst 5.80.
5 However as a lot of refactoring work has taken place, and several changes have
6 been made which could cause incompatibilities. If your application or plugin
7 is using deprecated code, or relying on side-effects, then you could have
8 issues upgrading to this release.
10 Most issues found with pre-existing components have been easy to solve, and a
11 complete description of behaviour changes which may cause compatibility issues,
12 or warnings which are now emitted is included below to help if you have problems.
14 If you think you have found an upgrade related issue which is not covered in
15 this document, then please email the Catalyst list to discuss the problem.
17 =head1 Known backwards compatibility breakages.
19 =head2 Issues with Class::C3
21 Catalyst 5.80 uses L<Algorithm::C3> method dispatch order. This is built into
22 perl 5.10, and comes via L<Class::C3> for perl 5.8. This replaces L<NEXT>
23 with L<Class::C3::Adopt::NEXT>, forcing all components to resolve methods using
24 C3, rather than the unpredictable dispatch order of L<NEXT>.
26 To be able to do this, however, entails that the graph of superclasses for each
27 class must be linearizable using the C3 algorithm. Unfortunately, when
28 superclasses are being used as mixins, it is easy to get this wrong.
30 Most common is the case of:
32 package Component1; # Note, this is the common case
33 use base qw/Class::Accessor::Fast Class::Data::Inheritable/;
35 package Component2; # Accidentally saying it this way round causes fail.
36 use base qw/Class::Data::Inheritable Class::Accessor::Fast/;
39 use base qw/Component1 Component2/;
41 And the Catalyst plugin most often causing this, is
42 L<Catalyst::Plugin::Sesssion::Store::FastMMap> - if you are using this plugin
43 and see issues, then please upgrade!
45 This can, however, be found in your own application - the only solution is to
46 go through each base class of the class the error was reported against, until
47 you identify the ones in conflict, and resolve them.
49 =head2 Components which inherit from Moose::Object before Catalyst::Component
51 Moose components which say:
53 package TestApp::Controller::Example;
55 extends qw/Moose::Object Catalyst::Component/;
57 to use the constructor provided by Moose, whilst working (if you do some hacks
58 with the C< BUILDARGS > method), will not work with Catalyst 5.80 as
59 C<Catalyst::Component> inherits from C<Moose::Object>, and so C< @ISA > fails
62 The fix for this is to not inherit directly from C<Moose::Object>
63 yourself. Having components which do not inherit their constructor from
64 C<Catalyst::Component> is B<unsupported>, and has never been recommended,
65 therefore you're on your own if you're using this technique. You'll need
66 to detect the version of Catalyst your application is running with and deal
67 with it appropriately.
69 You will also see this issue if you do the following:
71 package TestApp::Controller::Example;
73 use base 'Catalyst::Controller';
75 as C< use base > appends to @ISA.
77 The correct way to use Moose in a component in a both forward and backwards
80 package TestApp::Controller::Root;
82 BEGIN { extends 'Catalyst::Component' }; # Or ::Controller, or whatever
84 Note that the C< extends > declaration needs to occur in a begin block for
85 L<attributes> to operate correctly.
87 You also don't get the L<Moose::Object> constructor, and therefore attribute
88 initialization will not work as normally expected. If you want to use Moose
89 attributes, then they need to be made lazy to correctly initialize.
91 Note that this only applies if your component needs to maintain component
92 backwards compatibility for Catalyst versions before 5.71001 - in 5.71001
93 attributes work as expected, and the BUILD method is called normally
94 (although BUILDARGS is not).
96 If you depend on Catalyst 5.8, then B<all> Moose features work as expected.
98 =head3 use Moose in MyApp
100 Similar to the above, this will also fail:
109 If you need to use Moose in your application class (e.g. for method modifiers
110 etc) then the correct technique is:
115 __PACKAGE__->setup(qw/
119 =head2 Anonymous closures installed directly into the symbol table
121 If you have any code which installs anonymous subroutine references directly
122 into the symbol table, you may encounter breakages. The simplest solution is
123 to use L<Sub::Name> to name the subroutine. Example:
125 # Original code, likely to break:
126 my $full_method_name = join('::', $package_name, $method_name);
127 *$full_method_name = sub { ... };
130 use Sub::Name 'subname';
131 my $full_method_name = join('::',$package_name, $method_name);
132 *$full_method_name = subname $full_method_name, sub { ... };
134 Additionally, you can take advantage of Catalysts use of L<Class::MOP> and
135 install the closure using the appropriate meta class. Example:
138 my $metaclass = Moose::Meta::Class->initialize($package_name);
139 $metaclass->add_method($method_name => sub { ... });
141 =head2 Hooking into application setup
143 To execute code during application start-up the following snippet in MyApp.pm
147 my ($class, @args) = @_;
148 $class->NEXT::setup(@args);
149 ... # things to do after the actual setup
152 With Catalyst 5.80 this won't work anymore. Due to the fact that Catalyst is
153 no longer using NEXT.pm for method resolution, this no longer works. The
154 functionality was only ever originally operational as L<NEXT> remembers what
155 methods have already been called, and will not call them again.
157 Using this now causes infinite recursion between MyApp::setup and
158 Catalyst::setup, due to other backwards compatibility issues related to how
159 plugin setup works. Moose method modifiers like C<< before|after|around 'setup
160 => sub { ... }; >> also will not operate correctly on the setup method.
162 The right way to do it is this:
164 after setup_finalize => sub {
165 ... # things to do after the actual setup
168 The setup_finalize hook was introduced as a way to avoid this issue.
170 =head2 Components with a new method which returns false
172 Previously, if you had a component which inherited from Catalyst::COMPONENT,
173 but overrode the new method to return false, then your class' configuration
174 would be blessed into a hash on your behalf, and this would be returned from
175 the COMPONENT method.
177 This behaviour makes no sense, and so has been removed. Implementing your own
178 C< new > method in components is B<highly> discouraged, instead, you should
179 inherit the new method from Catalyst::Component, and use Mooses BUILD
180 functionality and/or Moose attributes to perform any construction work
181 necessary for your class.
183 =head2 __PACKAGE__->mk_accessor('meta');
185 Won't work due to a limitation of L<Moose>. This is currently being fixed
188 =head2 Class::Data::Inheritable side effects
190 Previously, writing to a class data accessor would copy the accessor method
191 down into your package.
193 This behaviour has been removed. Whilst the class data is still stored
194 per-class, it is stored on the metaclass of the class defining the accessor.
196 Therefore anything relying on the side-effect of the accessor being copied down
199 The following test demonstrates the problem:
203 use base qw/Class::Data::Inheritable/;
204 __PACKAGE__->mk_classdata('foo');
209 use base qw/BaseClass/;
212 BaseClass->foo('base class');
213 Child->foo('sub class');
216 isnt(BaseClass->can('foo'), Child->can('foo'));
218 =head2 Extending Catalyst::Request or other classes in an ad-hoc manor using mk_accessors
220 Previously, it was possible to add additional accessors to Catalyst::Request
221 (or other classes) by calling the mk_accessors class method.
223 This is no longer supported - users should make a sub-class of the class whose
224 behaviour they would like to change, rather than globally polluting the
227 =head2 Confused multiple inheritance with Catalyst::Component::COMPONENT
229 Previously, Catalyst's COMPONENT method would delegate to the method on the
230 right hand side, which could then delegate back again with NEXT. This (as it
231 is insane AND makes no sense with C3 method dispatch order), and is therefore
234 If a COMPONENT method is detected in the inheritance hierarchy to the right
235 hand side of Catalyst::Component::COMPONENT, then the following warning
236 message will be emitted:
238 There is a COMPONENT method resolving after Catalyst::Component
241 The correct fix is to re-arrange your class' inheritance hierarchy so that the
242 COMPONENT method you would like to inherit is the first (left-hand most)
243 COMPONENT method in your @ISA.
247 =head2 Catalyst::Base
249 Any code using L<Catalyst::Base> will now warn, and this module will be removed
252 =head2 Methods in Catalyst::Dispatcher
254 The following methods in Catalyst::Dispatcher are both an implementation
255 detail, which may change in the 5.8X release series, and therefore their use
256 is highly deprecated.
264 =item registered_dispatch_types
266 =item method_action_class
274 The first time one of these methods is called, a warning will be emitted:
276 Class $class is calling the deprecated method Catalyst::Dispatcher::$public_method_name,
277 this will be removed in Catalyst 5.9X
279 You should B<NEVER> be calling any of these methods from application code.
281 Plugins authors and maintainers whose plugins currently call these methods
282 should change to using the public API, or, if you do not feel the public API
283 adequately supports your use-case, please email the development list to
284 discuss what API features you need so that you can be appropriately supported.
286 =head2 Class naming to packages defined does not correspond.
288 In this version of Catalyst, if a component is loaded from disk, but no
289 symbols are defined in that component's name space after it is loaded, this
290 warning will be issued:
292 require $class was successful but the package is not defined.
294 This is to protect against confusing bugs caused by mis-typing package names,
295 and will become a fatal error in a future version.
297 Please note that 'inner packages' (via L<Devel::InnerPackage>) are still fully
298 supported, this warning is only issued when component file naming does not map
299 to B<any> of the packages defined within that component.
301 =head2 $c->plugin method
303 Calling the plugin method is deprecated, and calling it at run time is B<highly
306 Instead you are recommended to use L< Catalyst::Model::Adaptor > or similar to
307 compose the functionality you need outside of the main application name space.
309 Calling the plugin method will not be supported past Catalyst 5.81.