3 Catalyst::Upgrading - Instructions for upgrading to the latest Catalyst
5 =head1 Upgrading to Catalyst 5.90
7 The major change is that L<Plack> now replaces most of the subclasses of
8 L<Catalyst::Engine>. If you are using one of the standard subclasses of
9 L<Catalyst::Engine> this should be a straightforward upgrade for you. It was
10 a design goal for this release to be as backwardly compatible as possible.
11 However since L<Plack> is different from L<Catalyst::Engine> it would be
12 possible that edge case differences would exist. Therefore we recommend care
13 be taken with this upgrade and that testing should be greater than would be
14 the case with a minor point update.
16 It is highly recommended that you become familar with the L<Plack> ecosystem
17 and documentation. Being able to take advantage of L<Plack> development and
18 middleware is a major bonus to this upgrade.
20 If you have created a custom subclass of L<Catalyst:Engine> you will need to
21 convert it to be a subclass of L<Plack::Handler>.
23 If you are using the L<Plack> engine, L<Catalyst::Engine::PSGI>, this new
24 release supercedes that code.
26 If you are using a subclass of L<Catalyst::Engine> that is aimed at nonstandard
27 or internal / testing uses, such as L<Catalyst::Engine::Embeddable> you should
28 still be able to continue using that engine.
30 Advice for specific subclasses of L<Catalyst::Engine> follows:
32 =head2 Upgrading the FastCGI Engine
34 No upgrade needed if your myapp_fastcgi.pl script is already upgraded
35 enough to use L<Catalyst::Script::FastCGI>.
37 =head2 Upgrading the mod_perl / Apache Engines
39 The engines that are build upon the various iterations of mod_perl,
40 L<Catalyst::Engine::Apache::MP13> and
41 L<Catalyst::Engine::Apache2::MP20> should be seemless upgrades and will
42 work using using L<Plack::Handler::Apache1> or L<Plack::Handler::Apache2>
45 L<Catalyst::Engine::Apache2::MP19>, is however no longer supported, as Plack
46 does not support mod_perl version 1.99??? FIXME - is this true?
48 =head2 Upgrading the HTTP Engine
50 The default development server that comes with the L<Catalyst> distribution
51 should continue to work as expected with no changes as long as your C<myapp_server>
52 script is upgraded to use L<Catalyst::Script::HTTP>.
54 =head2 Upgrading the CGI Engine
56 If you were using L<Catalyst::Engine::CGI> there is no upgrade needed if your
57 myapp_cgi.pl script is already upgraded enough to use L<Catalyst::Script::CGI>.
59 =head2 Upgrading the Preforking Engine
61 If you were using L<Catalyst::Engine::HTTP::Prefork> then L<Starman>
62 is automatically loaded.
64 XXX FIXME - note how to run Starman with different options.
66 =head2 Upgrading the PSGI Engine
68 If you were using L<Catalyst::Engine::PSGI> this new release supercedes this
69 engine in supporting L<Plack>. By default the Engine is now always L<Plack>.
70 As a result, you can stop depending on L<Catalyst::Engine::PSGI> in your
71 C<Makefile.PL>. Additionally, if you have an C<app.psgi> script you no longer
72 need to specify the PSGI engine. Instead, the L<Catalyst> application class
73 now has a new method C<psgi_app> which returns a L<Plack> compatible coderef.
75 For example, if you were using L<Catalyst::Engine::PSGI> in the past, you may
76 have written an C<app.psgi> file similar to this one:
81 MyCatalystApp->setup_engine('PSGI');
84 enable ... # enable your desired middleware
85 sub { MyCatalystApp->run(@_) };
88 Instead, you now just do
94 enable ... #enable your desired middleware
95 MyCatalystApp->psgi_app;
98 =head2 Engines with unknown status
100 The following engines have untested or unknown compatibility. Reports are
103 Catalyst::Engine::Embeddable - needs testing, should work?
104 Catalyst::Engine::XMPP2
105 Catalyst::Engine::SCGI
106 Catalyst::Engine::Mojo
107 Catalyst::Engine::Zeus - broken for ages
108 Catalyst::Engine::JobQueue::POE - broken for ages
110 Catalyst::Engine::Stomp - fixed
111 Catalyst::Engine::Server (Marked as Deprecated)
112 Catalyst::Engine::HTTP::POE (Marked as Deprecated)
114 =head2 Using middleware
116 XXX Should this be here or elsewhere?
118 =head2 Making an app.psgi file
120 =head2 Running with plackup?
122 =head1 Upgrading to Catalyst 5.80
124 Most applications and plugins should run unaltered on Catalyst 5.80.
126 However, a lot of refactoring work has taken place, and several changes have
127 been made which could cause incompatibilities. If your application or plugin
128 is using deprecated code, or relying on side effects, then you could have
129 issues upgrading to this release.
131 Most issues found with pre-existing components have been easy to
132 solve. This document provides a complete description of behavior changes
133 which may cause compatibility issues, and of new Catalyst warnings which
136 If you think you have found an upgrade-related issue which is not covered in
137 this document, please email the Catalyst list to discuss the problem.
139 =head1 Moose features
141 =head2 Application class roles
143 You can only apply method modifiers after the application's C<< ->setup >>
144 method has been called. This means that modifiers will not work with methods
145 which run during the call to C<< ->setup >>.
147 See L<Catalyst::Manual::ExtendingCatalyst> for more information about using
148 L<Moose> in your applications.
150 =head2 Controller actions in Moose roles
152 You can use L<MooseX::MethodAttributes::Role> if you want to declare actions
155 =head2 Using Moose in Components
157 The correct way to use Moose in a component in a both forward and backwards
160 package TestApp::Controller::Root;
162 BEGIN { extends 'Catalyst::Component' }; # Or ::Controller, or whatever
164 See L<Components which inherit from Moose::Object before Catalyst::Component>.
166 =head1 Known backwards compatibility breakages
168 =head2 Applications in a single file
170 Applications must be in their own file, and loaded at compile time. This
171 issue generally only affects the tests of CPAN distributions. Your
172 application will fail if you try to define an application inline in a
173 block, and use plugins which supply a C< new > method, then use that
174 application latter in tests within the same file.
176 This is due to the fact that Catalyst is inlining a new method on your
177 application class allowing it to be compatible with Moose. The method
178 used to do this changed in 5.80004 to avoid the possibility of reporting
179 an 'Unknown Error' if your application failed to compile.
181 =head2 Issues with Class::C3
183 Catalyst 5.80 uses the L<Algorithm::C3> method dispatch order. This is
184 built into Perl 5.10, and comes via L<Class::C3> for Perl 5.8. This
185 replaces L<NEXT> with L<Class::C3::Adopt::NEXT>, forcing all components
186 to resolve methods using C3, rather than the unpredictable dispatch
189 This issue is characterised by your application failing to start due to an
190 error message about having a non-linear @ISA.
192 The Catalyst plugin most often causing this is
193 L<Catalyst::Plugin::Session::Store::FastMmap> - if you are using this
194 plugin and see issues, then please upgrade your plugins, as it has been
195 fixed. Note that Makefile.PL in the distribution will warn about known
196 incompatible components.
198 This issue can, however, be found in your own application - the only solution is
199 to go through each base class of the class the error was reported against, until
200 you identify the ones in conflict, and resolve them.
202 To be able to generate a linear @ISA, the list of superclasses for each
203 class must be resolvable using the C3 algorithm. Unfortunately, when
204 superclasses are being used as mixins (to add functionality used in your class),
205 and with multiple inheritence, it is easy to get this wrong.
207 Most common is the case of:
209 package Component1; # Note, this is the common case
210 use base qw/Class::Accessor::Fast Class::Data::Inheritable/;
212 package Component2; # Accidentally saying it this way causes a failure
213 use base qw/Class::Data::Inheritable Class::Accessor::Fast/;
216 use base qw/Component1 Component2/;
218 Any situation like this will cause your application to fail to start.
220 For additional documentation about this issue, and how to resolve it, see
221 L<Class::C3::Adopt::NEXT>.
223 =head2 Components which inherit from Moose::Object before Catalyst::Component
225 Moose components which say:
227 package TestApp::Controller::Example;
229 extends qw/Moose::Object Catalyst::Component/;
231 to use the constructor provided by Moose, while working (if you do some hacks
232 with the C< BUILDARGS > method), will not work with Catalyst 5.80 as
233 C<Catalyst::Component> inherits from C<Moose::Object>, and so C< @ISA > fails
236 The correct way to use Moose in a component in a both forward and backwards
239 package TestApp::Controller::Root;
241 BEGIN { extends 'Catalyst::Component' }; # Or ::Controller, or whatever
243 Note that the C< extends > declaration needs to occur in a begin block for
244 L<attributes> to operate correctly.
246 This way you do not inherit directly from C<Moose::Object>
247 yourself. Having components which do not inherit their constructor from
248 C<Catalyst::Component> is B<unsupported>, and has never been recommended,
249 therefore you're on your own if you're using this technique. You'll need
250 to detect the version of Catalyst your application is running, and deal
251 with it appropriately.
253 You also don't get the L<Moose::Object> constructor, and therefore attribute
254 initialization will not work as normally expected. If you want to use Moose
255 attributes, then they need to be made lazy to correctly initialize.
257 Note that this only applies if your component needs to maintain component
258 backwards compatibility for Catalyst versions before 5.71001 - in 5.71001
259 attributes work as expected, and the BUILD method is called normally
260 (although BUILDARGS is not).
262 If you depend on Catalyst 5.8, then B<all> Moose features work as expected.
264 You will also see this issue if you do the following:
266 package TestApp::Controller::Example;
268 use base 'Catalyst::Controller';
270 as C< use base > appends to @ISA.
272 =head3 use Moose in MyApp
274 Similar to the above, this will also fail:
283 If you need to use Moose in your application class (e.g. for method modifiers
284 etc.) then the correct technique is:
292 __PACKAGE__->config( name => 'MyApp' );
293 __PACKAGE__->setup(qw/
297 =head2 Anonymous closures installed directly into the symbol table
299 If you have any code which installs anonymous subroutine references directly
300 into the symbol table, you may encounter breakages. The simplest solution is
301 to use L<Sub::Name> to name the subroutine. Example:
303 # Original code, likely to break:
304 my $full_method_name = join('::', $package_name, $method_name);
305 *$full_method_name = sub { ... };
308 use Sub::Name 'subname';
309 my $full_method_name = join('::',$package_name, $method_name);
310 *$full_method_name = subname $full_method_name, sub { ... };
312 Additionally, you can take advantage of Catalyst's use of L<Class::MOP> and
313 install the closure using the appropriate metaclass. Example:
316 my $metaclass = Moose::Meta::Class->initialize($package_name);
317 $metaclass->add_method($method_name => sub { ... });
319 =head2 Hooking into application setup
321 To execute code during application start-up, the following snippet in MyApp.pm
325 my ($class, @args) = @_;
326 $class->NEXT::setup(@args);
327 ... # things to do after the actual setup
330 With Catalyst 5.80 this won't work anymore, because Catalyst no longer
331 uses NEXT.pm for method resolution. The functionality was only ever
332 originally operational as L<NEXT> remembers what methods have already
333 been called, and will not call them again.
335 Using this now causes infinite recursion between MyApp::setup and
336 Catalyst::setup, due to other backwards compatibility issues related to how
337 plugin setup works. Moose method modifiers like C<< before|after|around setup
338 => sub { ... }; >> also will not operate correctly on the setup method.
340 The right way to do it is this:
342 after setup_finalize => sub {
343 ... # things to do after the actual setup
346 The setup_finalize hook was introduced as a way to avoid this issue.
348 =head2 Components with a new method which returns false
350 Previously, if you had a component which inherited from Catalyst::COMPONENT,
351 but overrode the new method to return false, then your class's configuration
352 would be blessed into a hash on your behalf, and this would be returned from
353 the COMPONENT method.
355 This behavior makes no sense, and so has been removed. Implementing your own
356 C< new > method in components is B<highly> discouraged. Instead, you should
357 inherit the new method from Catalyst::Component, and use Moose's BUILD
358 functionality and/or Moose attributes to perform any construction work
359 necessary for your class.
361 =head2 __PACKAGE__->mk_accessor('meta');
363 Won't work due to a limitation of L<Moose>. This is currently being fixed
366 =head2 Class::Data::Inheritable side effects
368 Previously, writing to a class data accessor would copy the accessor method
369 down into your package.
371 This behavior has been removed. While the class data is still stored
372 per-class, it is stored on the metaclass of the class defining the accessor.
374 Therefore anything relying on the side effect of the accessor being copied down
377 The following test demonstrates the problem:
381 use base qw/Class::Data::Inheritable/;
382 __PACKAGE__->mk_classdata('foo');
387 use base qw/BaseClass/;
390 BaseClass->foo('base class');
391 Child->foo('sub class');
394 isnt(BaseClass->can('foo'), Child->can('foo'));
396 =head2 Extending Catalyst::Request or other classes in an ad-hoc manner using mk_accessors
398 Previously, it was possible to add additional accessors to Catalyst::Request
399 (or other classes) by calling the mk_accessors class method.
401 This is no longer supported - users should make a subclass of the class whose
402 behavior they would like to change, rather than globally polluting the
405 =head2 Confused multiple inheritance with Catalyst::Component::COMPONENT
407 Previously, Catalyst's COMPONENT method would delegate to the method on
408 the right hand side, which could then delegate back again with
409 NEXT. This is poor practice, and in addition, makes no sense with C3
410 method dispatch order, and is therefore no longer supported.
412 If a COMPONENT method is detected in the inheritance hierarchy to the right
413 hand side of Catalyst::Component::COMPONENT, then the following warning
414 message will be emitted:
416 There is a COMPONENT method resolving after Catalyst::Component
419 The correct fix is to re-arrange your class's inheritance hierarchy so that the
420 COMPONENT method you would like to inherit is the first (left-hand most)
421 COMPONENT method in your @ISA.
425 =head2 Actions in your application class
427 Having actions in your application class will now emit a warning at application
428 startup as this is deprecated. It is highly recommended that these actions are moved
429 into a MyApp::Controller::Root (as demonstrated by the scaffold application
430 generated by catalyst.pl).
432 This warning, also affects tests. You should move actions in your test,
433 creating a myTest::Controller::Root, like the following example:
435 package MyTest::Controller::Root;
440 use parent 'Catalyst::Controller';
442 __PACKAGE__->config(namespace => '');
445 my ( $self, $c ) = @_;
451 =head2 ::[MVC]:: naming scheme
453 Having packages called MyApp::[MVC]::XX is deprecated and can no longer be generated
456 This is still supported, but it is recommended that you rename your application
457 components to Model/View/Controller.
459 A warning will be issued at application startup if the ::[MVC]:: naming scheme is
462 =head2 Catalyst::Base
464 Any code using L<Catalyst::Base> will now emit a warning; this
465 module will be removed in a future release.
467 =head2 Methods in Catalyst::Dispatcher
469 The following methods in Catalyst::Dispatcher are implementation
470 details, which may change in the 5.8X release series, and therefore their use
471 is highly deprecated.
479 =item registered_dispatch_types
481 =item method_action_class
489 The first time one of these methods is called, a warning will be emitted:
491 Class $class is calling the deprecated method Catalyst::Dispatcher::$public_method_name,
492 this will be removed in Catalyst 5.9X
494 You should B<NEVER> be calling any of these methods from application code.
496 Plugin authors and maintainers whose plugins currently call these methods
497 should change to using the public API, or, if you do not feel the public API
498 adequately supports your use case, please email the development list to
499 discuss what API features you need so that you can be appropriately supported.
501 =head2 Class files with names that don't correspond to the packages they define
503 In this version of Catalyst, if a component is loaded from disk, but no
504 symbols are defined in that component's name space after it is loaded, this
505 warning will be issued:
507 require $class was successful but the package is not defined.
509 This is to protect against confusing bugs caused by mistyping package names,
510 and will become a fatal error in a future version.
512 Please note that 'inner packages' (via L<Devel::InnerPackage>) are still fully
513 supported; this warning is only issued when component file naming does not map
514 to B<any> of the packages defined within that component.
516 =head2 $c->plugin method
518 Calling the plugin method is deprecated, and calling it at run time is B<highly
521 Instead you are recommended to use L<Catalyst::Model::Adaptor> or similar to
522 compose the functionality you need outside of the main application name space.
524 Calling the plugin method will not be supported past Catalyst 5.81.