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