3 Catalyst::Upgrading - Instructions for upgrading to the latest Catalyst
5 =head1 Upgrading to Catalyst 5.90040
7 This version of L<Catalyst> offers some support for using L<AnyEvent> and
8 L<IO::Async> event loops in your application. In order to achieve this goal
9 we needed to make some changes to the way the we finalize the HTTP response
10 such that sloppy code that closed over $c and leaked memory will no longer
11 work in some manner. For example you might accidently have:
13 $c->stash(my_model => sub { $c->model->find(shift) });
15 If you have old code that leaks memory in this way but otherwise seemed to
16 work, it will no longer complete the response properly.
18 If you don't want to fix your code, you can force the old behavior with the
19 global configuration key C<aggressively_close_writer_on_finalize_body>. This
20 of course will still leave you with a leaky application and you lose the new
21 event loop support, but your application will go back to completing its
22 response output. For example:
31 enable_catalyst_header => 1,
32 disable_component_resolution_regex_fallback => 1,
33 aggressively_close_writer_on_finalize_body => 1,
38 See L<Catalyst::Component::ContextClosure> for help on how to close over the
39 context safely, should you need to do this. See L<CatalystX::LeakChecker>
40 and L<Catalyst::Controller::LeakTracker> for help if you want to solve your
43 =head1 Upgrading to Catalyst 5.9
45 The major change is that L<Plack>, a toolkit for using the L<PSGI>
46 specification, now replaces most of the subclasses of L<Catalyst::Engine>. If
47 you are using one of the standard subclasses of L<Catalyst::Engine> this
48 should be a straightforward upgrade for you. It was a design goal for
49 this release to preserve as much backwards compatibility as possible.
50 However, since L<Plack> is different from L<Catalyst::Engine>, it is
51 possible that differences exist for edge cases. Therefore, we recommend
52 that care be taken with this upgrade and that testing should be greater
53 than would be the case with a minor point update. Please inform the
54 Catalyst developers of any problems so that we can fix them and
57 It is highly recommended that you become familiar with the L<Plack> ecosystem
58 and documentation. Being able to take advantage of L<Plack> development and
59 middleware is a major bonus to this upgrade. Documentation about how to
60 take advantage of L<Plack::Middleware> by writing your own C<< .psgi >> file
61 is contained in L<Catalyst::PSGI>.
63 If you have created a custom subclass of L<Catalyst:Engine>, you will
64 need to convert it to be a subclass of L<Plack::Handler>.
66 If you are using the L<Plack> engine, L<Catalyst::Engine::PSGI>, this new
67 release supersedes that code.
69 If you are using a subclass of L<Catalyst::Engine> that is aimed at
70 nonstandard or internal/testing uses, such as
71 L<Catalyst::Engine::Embeddable>, you should still be able to continue
74 Advice for specific subclasses of L<Catalyst::Engine> follows:
76 =head2 Upgrading the FastCGI Engine
78 No upgrade is needed if your myapp_fastcgi.pl script is already upgraded
79 to use L<Catalyst::Script::FastCGI>.
81 =head2 Upgrading the mod_perl / Apache Engines
83 The engines that are built upon the various iterations of mod_perl,
84 L<Catalyst::Engine::Apache::MP13> (for mod_perl 1, and Apache 1.x) and
85 L<Catalyst::Engine::Apache2::MP20> (for mod_perl 2, and Apache 2.x),
86 should be seamless upgrades and will work using using L<Plack::Handler::Apache1>
87 or L<Plack::Handler::Apache2> as required.
89 L<Catalyst::Engine::Apache2::MP19>, however, is no longer supported, as
90 Plack does not support mod_perl version 1.99. This is unlikely to be a
91 problem for anyone, as 1.99 was a brief beta-test release for mod_perl
92 2, and all users of mod_perl 1.99 are encouraged to upgrade to a
93 supported release of Apache 2 and mod_perl 2.
95 =head2 Upgrading the HTTP Engine
97 The default development server that comes with the L<Catalyst> distribution
98 should continue to work as expected with no changes as long as your C<myapp_server>
99 script is upgraded to use L<Catalyst::Script::HTTP>.
101 =head2 Upgrading the CGI Engine
103 If you were using L<Catalyst::Engine::CGI> there is no upgrade needed if your
104 myapp_cgi.pl script is already upgraded to use L<Catalyst::Script::CGI>.
106 =head2 Upgrading Catalyst::Engine::HTTP::Prefork
108 If you were using L<Catalyst::Engine::HTTP::Prefork> then L<Starman>
109 is automatically loaded. You should (at least) change your C<Makefile.PL>
110 to depend on Starman.
112 You can regenerate your C<myapp_server.pl> script with C<catalyst.pl>
113 and implement a C<MyApp::Script::Server> class that looks like this:
115 package MyApp::Script::Server;
117 use namespace::autoclean;
119 extends 'CatalystX::Script::Server::Starman';
123 This takes advantage of the new script system, and will add a number of
124 options to the standard server script as extra options are added by
127 More information about these options can be seen at
128 L<CatalystX::Script::Server::Starman/SYNOPSIS>.
130 An alternate route to implement this functionality is to write a simple .psgi
131 file for your application, and then use the L<plackup> utility to start the
134 =head2 Upgrading the PSGI Engine
136 If you were using L<Catalyst::Engine::PSGI>, this new release supersedes
137 this engine in supporting L<Plack>. By default the Engine is now always
138 L<Plack>. As a result, you can remove the dependency on
139 L<Catalyst::Engine::PSGI> in your C<Makefile.PL>.
141 Applications that were using L<Catalyst::Engine::PSGI>
142 previously should entirely continue to work in this release with no changes.
144 However, if you have an C<app.psgi> script, then you no longer need to
145 specify the PSGI engine. Instead, the L<Catalyst> application class now
146 has a new method C<psgi_app> which returns a L<PSGI> compatible coderef
147 which you can wrap in the middleware of your choice.
149 Catalyst will use the .psgi for your application if it is located in the C<home>
150 directory of the application.
152 For example, if you were using L<Catalyst::Engine::PSGI> in the past, you will
153 have written (or generated) a C<script/myapp.psgi> file similar to this one:
158 MyCatalystApp->setup_engine('PSGI');
161 enable ... # enable your desired middleware
162 sub { MyCatalystApp->run(@_) };
165 Instead, you now say:
171 enable ... #enable your desired middleware
172 MyCatalystApp->psgi_app;
175 In the simplest case:
177 MyCatalystApp->setup_engine('PSGI');
178 my $app = sub { MyCatalystApp->run(@_) }
182 my $app = MyCatalystApp->psgi_app(@_);
186 my $app = sub { MyCatalystApp->psgi_app(@_) };
187 # If you make ^^ this mistake, your app won't work, and will confuse the hell out of you!
189 You can now move C<< script/myapp.psgi >> to C<< myapp.psgi >>, and the built-in
190 Catalyst scripts and your test suite will start using your .psgi file.
192 B<NOTE:> If you rename your .psgi file without these modifications, then
193 any tests run via L<Catalyst::Test> will not be compatible with the new
194 release, and will result in the development server starting, rather than
195 the expected test running.
197 B<NOTE:> If you are directly accessing C<< $c->req->env >> to get the PSGI
198 environment then this accessor is moved to C<< $c->engine->env >>,
199 you will need to update your code.
201 =head2 Engines which are known to be broken
203 The following engines B<DO NOT> work as of Catalyst version 5.9. The
204 core team will be happy to work with the developers and/or users of
205 these engines to help them port to the new Plack/Engine system, but for
206 now, applications which are currently using these engines B<WILL NOT>
207 run without modification to the engine code.
211 =item Catalyst::Engine::Wx
213 =item Catalyst::Engine::Zeus
215 =item Catalyst::Engine::JobQueue::POE
217 =item Catalyst::Engine::XMPP2
219 =item Catalyst::Engine::SCGI
223 =head2 Engines with unknown status
225 The following engines are untested or have unknown compatibility.
226 Reports are highly encouraged:
230 =item Catalyst::Engine::Mojo
232 =item Catalyst::Engine::Server (marked as Deprecated)
234 =item Catalyst::Engine::HTTP::POE (marked as Deprecated)
238 =head2 Plack functionality
240 See L<Catalyst::PSGI>.
244 Tests should generally work the same in Catalyst 5.9, but there are
247 Previously, if using L<Catalyst::Test> and doing local requests (against
248 a local server), if the application threw an exception then this
249 exception propagated into the test.
251 This behavior has been removed, and now a 500 response will be returned
252 to the test. This change standardizes behavior, so that local test
253 requests behave similarly to remote requests.
255 =head2 Regex dispatch type is deprecated.
257 The Regex dispatchtype (L<Catalyst::DispatchType::Regex>) has been deprecated.
259 You are encouraged to move your application to Chained dispatch (L<Catalyst::DispatchType::Chained>).
261 If you cannot do so, please add a dependency to Catalyst::DispatchType::Regex to your application's
264 =head1 Upgrading to Catalyst 5.80
266 Most applications and plugins should run unaltered on Catalyst 5.80.
268 However, a lot of refactoring work has taken place, and several changes have
269 been made which could cause incompatibilities. If your application or plugin
270 is using deprecated code, or relying on side effects, then you could have
271 issues upgrading to this release.
273 Most issues found with existing components have been easy to
274 solve. This document provides a complete description of behavior changes
275 which may cause compatibility issues, and of new Catalyst warnings which
278 If you think you have found an upgrade-related issue which is not covered in
279 this document, please email the Catalyst list to discuss the problem.
281 =head1 Moose features
283 =head2 Application class roles
285 You can only apply method modifiers after the application's C<< ->setup >>
286 method has been called. This means that modifiers will not work with methods
287 run during the call to C<< ->setup >>.
289 See L<Catalyst::Manual::ExtendingCatalyst> for more information about using
290 L<Moose> in your applications.
292 =head2 Controller actions in Moose roles
294 You can use L<MooseX::MethodAttributes::Role> if you want to declare actions
297 =head2 Using Moose in Components
299 The correct way to use Moose in a component in a both forward and backwards
302 package TestApp::Controller::Root;
304 BEGIN { extends 'Catalyst::Component' }; # Or ::Controller, or whatever
306 See L<Components which inherit from Moose::Object before Catalyst::Component>.
308 =head1 Known backwards compatibility breakages
310 =head2 Applications in a single file
312 Applications must be in their own file, and loaded at compile time. This
313 issue generally only affects the tests of CPAN distributions. Your
314 application will fail if you try to define an application inline in a
315 block, and use plugins which supply a C< new > method, then use that
316 application latter in tests within the same file.
318 This is due to the fact that Catalyst is inlining a new method on your
319 application class allowing it to be compatible with Moose. The method
320 used to do this changed in 5.80004 to avoid the possibility of reporting
321 an 'Unknown Error' if your application failed to compile.
323 =head2 Issues with Class::C3
325 Catalyst 5.80 uses the L<Algorithm::C3> method dispatch order. This is
326 built into Perl 5.10, and comes via L<Class::C3> for Perl 5.8. This
327 replaces L<NEXT> with L<Class::C3::Adopt::NEXT>, forcing all components
328 to resolve methods using C3, rather than the unpredictable dispatch
331 This issue manifests itself by your application failing to start due to an
332 error message about having a non-linear @ISA.
334 The Catalyst plugin most often causing this is
335 L<Catalyst::Plugin::Session::Store::FastMmap> - if you are using this
336 plugin and see issues, then please upgrade your plugins, as it has been
337 fixed. Note that Makefile.PL in the distribution will warn about known
338 incompatible components.
340 This issue can, however, be found in your own application - the only solution is
341 to go through each base class of the class the error was reported against, until
342 you identify the ones in conflict, and resolve them.
344 To be able to generate a linear @ISA, the list of superclasses for each
345 class must be resolvable using the C3 algorithm. Unfortunately, when
346 superclasses are being used as mixins (to add functionality used in your class),
347 and with multiple inheritance, it is easy to get this wrong.
349 Most common is the case of:
351 package Component1; # Note, this is the common case
352 use base qw/Class::Accessor::Fast Class::Data::Inheritable/;
354 package Component2; # Accidentally saying it this way causes a failure
355 use base qw/Class::Data::Inheritable Class::Accessor::Fast/;
358 use base qw/Component1 Component2/;
360 Any situation like this will cause your application to fail to start.
362 For additional documentation about this issue, and how to resolve it, see
363 L<Class::C3::Adopt::NEXT>.
365 =head2 Components which inherit from Moose::Object before Catalyst::Component
367 Moose components which say:
369 package TestApp::Controller::Example;
371 extends qw/Moose::Object Catalyst::Component/;
373 to use the constructor provided by Moose, while working (if you do some hacks
374 with the C< BUILDARGS > method), will not work with Catalyst 5.80 as
375 C<Catalyst::Component> inherits from C<Moose::Object>, and so C< @ISA > fails
378 The correct way to use Moose in a component in a both forward and backwards
381 package TestApp::Controller::Root;
383 BEGIN { extends 'Catalyst::Component' }; # Or ::Controller, or whatever
385 Note that the C< extends > declaration needs to occur in a begin block for
386 L<attributes> to operate correctly.
388 This way you do not inherit directly from C<Moose::Object>
389 yourself. Having components which do not inherit their constructor from
390 C<Catalyst::Component> is B<unsupported>, and has never been recommended,
391 therefore you're on your own if you're using this technique. You'll need
392 to detect the version of Catalyst your application is running, and deal
393 with it appropriately.
395 You also don't get the L<Moose::Object> constructor, and therefore attribute
396 initialization will not work as normally expected. If you want to use Moose
397 attributes, then they need to be made lazy to correctly initialize.
399 Note that this only applies if your component needs to maintain component
400 backwards compatibility for Catalyst versions before 5.71001 - in 5.71001
401 attributes work as expected, and the BUILD method is called normally
402 (although BUILDARGS is not).
404 If you depend on Catalyst 5.8, then B<all> Moose features work as expected.
406 You will also see this issue if you do the following:
408 package TestApp::Controller::Example;
410 use base 'Catalyst::Controller';
412 as C< use base > appends to @ISA.
414 =head3 use Moose in MyApp
416 Similar to the above, this will also fail:
425 If you need to use Moose in your application class (e.g. for method modifiers
426 etc.) then the correct technique is:
434 __PACKAGE__->config( name => 'MyApp' );
435 __PACKAGE__->setup(qw/
439 =head2 Anonymous closures installed directly into the symbol table
441 If you have any code which installs anonymous subroutine references directly
442 into the symbol table, you may encounter breakages. The simplest solution is
443 to use L<Sub::Name> to name the subroutine. Example:
445 # Original code, likely to break:
446 my $full_method_name = join('::', $package_name, $method_name);
447 *$full_method_name = sub { ... };
450 use Sub::Name 'subname';
451 my $full_method_name = join('::',$package_name, $method_name);
452 *$full_method_name = subname $full_method_name, sub { ... };
454 Additionally, you can take advantage of Catalyst's use of L<Class::MOP> and
455 install the closure using the appropriate metaclass. Example:
458 my $metaclass = Moose::Meta::Class->initialize($package_name);
459 $metaclass->add_method($method_name => sub { ... });
461 =head2 Hooking into application setup
463 To execute code during application start-up, the following snippet in MyApp.pm
467 my ($class, @args) = @_;
468 $class->NEXT::setup(@args);
469 ... # things to do after the actual setup
472 With Catalyst 5.80 this won't work anymore, because Catalyst no longer
473 uses NEXT.pm for method resolution. The functionality was only ever
474 originally operational as L<NEXT> remembers what methods have already
475 been called, and will not call them again.
477 Using this now causes infinite recursion between MyApp::setup and
478 Catalyst::setup, due to other backwards compatibility issues related to how
479 plugin setup works. Moose method modifiers like C<< before|after|around setup
480 => sub { ... }; >> also will not operate correctly on the setup method.
482 The right way to do it is this:
484 after setup_finalize => sub {
485 ... # things to do after the actual setup
488 The setup_finalize hook was introduced as a way to avoid this issue.
490 =head2 Components with a new method which returns false
492 Previously, if you had a component which inherited from Catalyst::COMPONENT,
493 but overrode the new method to return false, then your class's configuration
494 would be blessed into a hash on your behalf, and this would be returned from
495 the COMPONENT method.
497 This behavior makes no sense, and so has been removed. Implementing your own
498 C< new > method in components is B<highly> discouraged. Instead, you should
499 inherit the new method from Catalyst::Component, and use Moose's BUILD
500 functionality and/or Moose attributes to perform any construction work
501 necessary for your class.
503 =head2 __PACKAGE__->mk_accessor('meta');
505 Won't work due to a limitation of L<Moose>. This is currently being fixed
508 =head2 Class::Data::Inheritable side effects
510 Previously, writing to a class data accessor would copy the accessor method
511 down into your package.
513 This behavior has been removed. While the class data is still stored
514 per-class, it is stored on the metaclass of the class defining the accessor.
516 Therefore anything relying on the side effect of the accessor being copied down
519 The following test demonstrates the problem:
523 use base qw/Class::Data::Inheritable/;
524 __PACKAGE__->mk_classdata('foo');
529 use base qw/BaseClass/;
532 BaseClass->foo('base class');
533 Child->foo('sub class');
536 isnt(BaseClass->can('foo'), Child->can('foo'));
538 =head2 Extending Catalyst::Request or other classes in an ad hoc manner using mk_accessors
540 Previously, it was possible to add additional accessors to Catalyst::Request
541 (or other classes) by calling the mk_accessors class method.
543 This is no longer supported - users should make a subclass of the class whose
544 behavior they would like to change, rather than globally polluting the
547 =head2 Confused multiple inheritance with Catalyst::Component::COMPONENT
549 Previously, Catalyst's COMPONENT method would delegate to the method on
550 the right hand side, which could then delegate back again with
551 NEXT. This is poor practice, and in addition, makes no sense with C3
552 method dispatch order, and is therefore no longer supported.
554 If a COMPONENT method is detected in the inheritance hierarchy to the right
555 hand side of Catalyst::Component::COMPONENT, then the following warning
556 message will be emitted:
558 There is a COMPONENT method resolving after Catalyst::Component
561 The correct fix is to re-arrange your class's inheritance hierarchy so that the
562 COMPONENT method you would like to inherit is the first (left-hand most)
563 COMPONENT method in your @ISA.
565 =head2 Development server relying on environment variables
567 Previously, the development server would allow propagation of system
568 environment variables into the request environment, this has changed with the
569 adoption of Plack. You can use L<Plack::Middleware::ForceEnv> to achieve the
574 =head2 Actions in your application class
576 Having actions in your application class will now emit a warning at application
577 startup as this is deprecated. It is highly recommended that these actions are moved
578 into a MyApp::Controller::Root (as demonstrated by the scaffold application
579 generated by catalyst.pl).
581 This warning, also affects tests. You should move actions in your test,
582 creating a myTest::Controller::Root, like the following example:
584 package MyTest::Controller::Root;
589 use parent 'Catalyst::Controller';
591 __PACKAGE__->config(namespace => '');
594 my ( $self, $c ) = @_;
600 =head2 ::[MVC]:: naming scheme
602 Having packages called MyApp::[MVC]::XX is deprecated and can no longer be generated
605 This is still supported, but it is recommended that you rename your application
606 components to Model/View/Controller.
608 A warning will be issued at application startup if the ::[MVC]:: naming scheme is
611 =head2 Catalyst::Base
613 Any code using L<Catalyst::Base> will now emit a warning; this
614 module will be removed in a future release.
616 =head2 Methods in Catalyst::Dispatcher
618 The following methods in Catalyst::Dispatcher are implementation
619 details, which may change in the 5.8X release series, and therefore their use
620 is highly deprecated.
628 =item registered_dispatch_types
630 =item method_action_class
638 The first time one of these methods is called, a warning will be emitted:
640 Class $class is calling the deprecated method Catalyst::Dispatcher::$public_method_name,
641 this will be removed in Catalyst 5.9
643 You should B<NEVER> be calling any of these methods from application code.
645 Plugin authors and maintainers whose plugins currently call these methods
646 should change to using the public API, or, if you do not feel the public API
647 adequately supports your use case, please email the development list to
648 discuss what API features you need so that you can be appropriately supported.
650 =head2 Class files with names that don't correspond to the packages they define
652 In this version of Catalyst, if a component is loaded from disk, but no
653 symbols are defined in that component's name space after it is loaded, this
654 warning will be issued:
656 require $class was successful but the package is not defined.
658 This is to protect against confusing bugs caused by mistyping package names,
659 and will become a fatal error in a future version.
661 Please note that 'inner packages' (via L<Devel::InnerPackage>) are still fully
662 supported; this warning is only issued when component file naming does not map
663 to B<any> of the packages defined within that component.
665 =head2 $c->plugin method
667 Calling the plugin method is deprecated, and calling it at run time is B<highly
670 Instead you are recommended to use L<Catalyst::Model::Adaptor> or similar to
671 compose the functionality you need outside of the main application name space.
673 Calling the plugin method will not be supported past Catalyst 5.81.