1Catalyst::Upgrading(3)User Contributed Perl DocumentationCatalyst::Upgrading(3)
2
6 Catalyst::Upgrading - Instructions for upgrading to the latest Catalyst
7
9 A new "log_stats" method has been added. This will only affect
10 subclasses that have a method with this name added.
11
13 We changed the way the middleware stash works so that it no longer
14 localizes the PSGI env hashref. This was done to fix bugs where people
15 set PSGI ENV hash keys and found them to disappear in certain cases.
16 It also means that now if a sub applications sets stash variables, that
17 stash will now bubble up to the parent application. This may be a
18 breaking change for you since previous versions of this code did not
19 allow that. A workaround is to explicitly delete stash keys in your
20 sub application before returning control to the parent application.
21
23 In older versions of Catalyst one could construct a URI with a fragment
24 (such as https://localhost/foo/bar#fragment) by using a '#' in the path
25 or final argument, for example:
26 $c->uri_for($action, 'foo#fragment');
28 This behavior was never documented and would break if using the Unicode
30 plugin, or when adding a query to the arguments:
31 $c->uri_for($action, 'foo#fragment', +{ a=>1, b=>2});
33 would define a fragment like "#fragment?a=1&b=2".
35 When we introduced UTF-8 encoding by default in Catalyst 5.9008x this
37 side effect behavior was broken since we started encoding the '#' when
38 it was part of the URI path.
39 In version 5.90095 and 5.90096 we attempted to fix this, but all we
41 managed to do was break people with URIs that included '#' as part of
42 the path data, when it was not expected to be a fragment delimiter.
43 In general Catalyst prefers an explicit specification rather than
45 relying on side effects or domain specific mini languages. As a result
46 we are now defining how to set a fragment for a URI via ->uri_for:
47 $c->uri_for($action_or_path, \@captures_or_args, @args, \$query, \$fragment);
49 If you are relying on the previous side effect behavior your URLs will
51 now encode the '#' delimiter, which is going to be a breaking change
52 for you. You need to alter your code to match the new specification or
53 modify uri_for for your local case. Patches to solve this are very
54 welcomed, as long as they don't break existing test cases.
55 NOTE If you are using the string form of the first argument:
57 $c->uri_for('/foo/bar#baz')
59 construction, we do not attempt to encode this and it will make a URL
61 with a fragment of 'baz'.
62
64 The method "last_error" in "Catalyst" was actually returning the first
65 error. This has been fixed but there is a small chance it could be a
66 breaking issue for you. If this gives you trouble changing to
67 "shift_errors" is the easiest workaround (although that does modify the
68 error stack so if you are relying on that not being changed you should
69 try something like @{$c->errors}[-1] instead. Since this method is
70 relatively new and the cases when the error stack actually has more
71 than one error in it, we feel the exposure is very low, but bug reports
72 are very welcomed.
73
75 Catalyst::Utils has a new method 'inject_component' which works the
76 same as the method of the same name in CatalystX::InjectComponent. You
77 should start converting any use of the non core method in your code as
78 future changes to Catalyst will be synchronized to the core method
79 first. We reserve the right to cease support of the non core version
80 should we reach a point in time where it cannot be properly supported
81 as an external module. Luckily this should be a trivial search and
82 replace. Change all occurrences of:
83 CatalystX::InjectComponent->inject(...)
85 Into
87 Catalyst::Utils::inject_component(...)
89 and we expect everything to work the same (we'd consider it not working
91 the same to be a bug, and please report it.)
92 We also cored features from CatalystX::RoleApplicator to compose a role
94 into the request, response and stats classes. The main difference is
95 that with CatalystX::RoleApplicator you did:
96 package MyApp;
98 use Catalyst;
100 use CatalystX::RoleApplicator;
101 __PACKAGE__->apply_request_class_roles(
103 qw/My::Request::Role Other::Request::Role/);
104 Whereas now we have three class attributes, 'request_class_traits',
106 'response_class_traits' and 'stats_class_traits', so you use like this
107 (note this value is an ArrayRef)
108 package MyApp;
110 use Catalyst;
112 __PACKAGE__->request_class_traits([qw/
114 My::Request::Role
115 Other::Request::Role/]);
116 (And the same for response_class_traits and stats_class_traits. We
118 left off the traits for Engine, since that class does a lot less
119 nowadays, and dispatcher. If you used those and can share a use case,
120 we'd be likely to support them.
121 Lastly, we have some of the feature from
123 CatalystX::ComponentsFromConfig in core. This should mostly work the
124 same way in core, except for now the core version does not create an
125 automatic base wrapper class for your configured components (it
126 requires these to be catalyst components and injects them directly. So
127 if you make heavy use of custom base classes in
128 CatalystX::ComponentsFromConfig you might need a bit of work to use the
129 core version (although there is no reason to stop using
130 CatalystX::ComponentsFromConfig since it should continue to work fine
131 and we'd consider issues with it to be bugs). Here's one way to map
132 from CatalystX::ComponentsFromConfig to core:
133 In CatalystX::ComponentsFromConfig:
135 MyApp->config(
137 'Model::MyClass' => {
138 class => 'MyClass',
139 args => { %args },
140 });
142 and now in core:
144 MyApp->config(
146 inject_components => {
147 'Model::MyClass' => { from_component => 'My::Class' },
148 },
149 'Model::MyClass' => {
150 %args
151 },
152 );
153 Although the core behavior requires more code, it better separates
155 concerns as well as plays more into core Catalyst expectations of how
156 configuration should look.
157 Also we added a new develop console mode only warning when you call a
159 component with arguments that don't expect or do anything meaningful
160 with those args. Its possible if you are logging debug mode in
161 production (please don't...) this could add verbosity to those logs if
162 you also happen to be calling for components and passing pointless
163 arguments. We added this warning to help people not make this error
164 and to better understand the component resolution flow.
165
167 In this version of Catalyst we made a small change to Chained
168 Dispatching so that when two or more actions all have the same path
169 specification AND they all have Args(0), we break the tie by choosing
170 the last action defined, and not the first one defined. This was done
171 to normalize Chaining to following the 'longest Path wins, and when
172 several actions match the same Path specification we choose the last
173 defined.' rule. Previously Args(0) was hard coded to be a special case
174 such that the first action defined would match (which is not the case
175 when Args is not zero.)
176 Its possible that this could be a breaking change for you, if you had
178 used action roles (custom or otherwise) to add additional matching
179 rules to differentiate between several Args(0) actions that share the
180 same root action chain. For example if you have code now like this:
181 sub check_default :Chained(/) CaptureArgs(0) { ... }
183 sub default_get :Chained('check_default') PathPart('') Args(0) GET {
185 pop->res->body('get3');
186 }
187 sub default_post :Chained('check_default') PathPart('') Args(0) POST {
189 pop->res->body('post3');
190 }
191 sub chain_default :Chained('check_default') PathPart('') Args(0) {
193 pop->res->body('chain_default');
194 }
195 The way that chaining will work previous is that when two or more equal
197 actions can match, the 'top' one wins. So if the request is "GET
198 .../check_default" BOTH actions 'default_get' AND 'chain_default' would
199 match. To break the tie in the case when Args is 0, we'd previous take
200 the 'top' (or first defined) action. Unfortunately this treatment of
201 Args(0) is special case. In all other cases we choose the 'last
202 defined' action to break a tie. So this version of Catalyst changed
203 the dispatcher to make Args(0) no longer a special case for breaking
204 ties. This means that the above code must now become:
205 sub check_default :Chained(/) CaptureArgs(0) { ... }
207 sub chain_default :Chained('check_default') PathPart('') Args(0) {
209 pop->res->body('chain_default');
210 }
211 sub default_get :Chained('check_default') PathPart('') Args(0) GET {
213 pop->res->body('get3');
214 }
215 sub default_post :Chained('check_default') PathPart('') Args(0) POST {
217 pop->res->body('post3');
218 }
219 If we want it to work as expected (for example we we GET to match
221 'default_get' and POST to match 'default_post' and any other http
222 Method to match 'chain_default').
223 In other words Arg(0) and chained actions must now follow the normal
225 rule where in a tie the last defined action wins and you should place
226 all your less defined or 'catch all' actions first.
227 If this causes you trouble and you can't fix your code to conform, you
229 may set the application configuration setting
230 "use_chained_args_0_special_case" to true and that will revert you code
231 to the previous behavior.
232 More backwards compatibility options with UTF-8 changes
234 In order to give better backwards compatibility with the 5.90080+ UTF-8
235 changes we've added several configuration options around control of how
236 we try to decode your URL keywords / query parameters.
237 "do_not_decode_query"
239 If true, then do not try to character decode any wide characters in
241 your request URL query or keywords. Most readings of the relevant
242 specifications suggest these should be UTF-* encoded, which is the
243 default that Catalyst will use, however if you are creating a lot of
244 URLs manually or have external evil clients, this might cause you
245 trouble. If you find the changes introduced in Catalyst version
246 5.90080+ break some of your query code, you may disable the UTF-8
247 decoding globally using this configuration.
248 This setting takes precedence over "default_query_encoding" and
250 "decode_query_using_global_encoding"
251 "default_query_encoding"
253 By default we decode query and keywords in your request URL using
255 UTF-8, which is our reading of the relevant specifications. This
256 setting allows one to specify a fixed value for how to decode your
257 query. You might need this if you are doing a lot of custom encoding
258 of your URLs and not using UTF-8.
259 This setting take precedence over "decode_query_using_global_encoding".
261 "decode_query_using_global_encoding"
263 Setting this to true will default your query decoding to whatever your
265 general global encoding is (the default is UTF-8).
266
268 UTF8 encoding is now default. For temporary backwards compatibility,
269 if this change is causing you trouble, you can disable it by setting
270 the application configuration option to undef:
271 MyApp->config(encoding => undef);
273 But please consider this a temporary measure since it is the intention
275 that UTF8 is enabled going forwards and the expectation is that other
276 ecosystem projects will assume this as well. At some point you
277 application will not correctly function without this setting.
278 As of 5.90084 we've added two additional configuration flags for more
280 selective control over some encoding changes:
281 'skip_body_param_unicode_decoding' and
282 'skip_complex_post_part_handling'. You may use these to more
283 selectively disable new features while you are seeking a long term fix.
284 Please review CONFIGURATION in Catalyst.
285 For further information, please see Catalyst::UTF8
287 A number of projects in the wider ecosystem required minor updates to
289 be able to work correctly. Here's the known list:
290 Catalyst::View::TT, Catalyst::View::Mason, Catalyst::View::HTML::Mason,
292 Catalyst::View::Xslate, Test::WWW::Mechanize::Catalyst
293 You will need to update to modern versions in most cases, although
295 quite a few of these only needed minor test case and documentation
296 changes so you will need to review the changelog of each one that is
297 relevant to you to determine your true upgrade needs.
298
300 Starting in the v5.90059_001 development release, the regexp dispatch
301 type is no longer automatically included as a dependency. If you are
302 still using this dispatch type, you need to add
303 Catalyst::DispatchType::Regex into your build system.
304 The standalone distribution of Regexp will be supported for the time
306 being, but should we find that supporting it prevents us from moving
307 Catalyst forward in necessary ways, we reserve the right to drop that
308 support. It is highly recommended that you use this last stage of
309 deprecation to change your code.
310
312 Catalyst::Plugin::Unicode::Encoding is now core
313 The previously stand alone Unicode support module
314 Catalyst::Plugin::Unicode::Encoding has been brought into core as a
315 default plugin. Going forward, all you need is to add a configuration
316 setting for the encoding type. For example:
317 package Myapp::Web;
319 use Catalyst;
321 __PACKAGE__->config( encoding => 'UTF-8' );
323 Please note that this is different from the old stand alone plugin
325 which applied "UTF-8" encoding by default (that is, if you did not set
326 an explicit "encoding" configuration value, it assumed you wanted
327 UTF-8). In order to preserve backwards compatibility you will need to
328 explicitly turn it on via the configuration setting. THIS MIGHT CHANGE
329 IN THE FUTURE, so please consider starting to test your application
330 with proper UTF-8 support and remove all those crappy hacks you munged
331 into the code because you didn't know the Plugin existed :)
332 For people that are using the Plugin, you will note a startup warning
334 suggesting that you can remove it from the plugin list. When you do
335 so, please remember to add the configuration setting, since you can no
336 longer rely on the default being UTF-8. We'll add it for you if you
337 continue to use the stand alone plugin and we detect this, but this
338 backwards compatibility shim will likely be removed in a few releases
339 (trying to clean up the codebase after all).
340 If you have trouble with any of this, please bring it to the attention
342 of the Catalyst maintainer group.
343 basic async and event loop support
345 This version of Catalyst offers some support for using AnyEvent and
346 IO::Async event loops in your application. These changes should work
347 fine for most applications however if you are already trying to perform
348 some streaming, minor changes in this area of the code might affect
349 your functionality. Please see Catalyst::Response\write_fh for more
350 and for a basic example.
351 We consider this feature experimental. We will try not to break it,
353 but we reserve the right to make necessary changes to fix major issues
354 that people run into when the use this functionality in the wild.
355
357 Regex dispatch type is deprecated.
358 The Regex dispatchtype (Catalyst::DispatchType::Regex) has been
359 deprecated.
360 You are encouraged to move your application to Chained dispatch
362 (Catalyst::DispatchType::Chained).
363 If you cannot do so, please add a dependency to
365 Catalyst::DispatchType::Regex to your application's Makefile.PL
366
368 The major change is that Plack, a toolkit for using the PSGI
369 specification, now replaces most of the subclasses of Catalyst::Engine.
370 If you are using one of the standard subclasses of Catalyst::Engine
371 this should be a straightforward upgrade for you. It was a design goal
372 for this release to preserve as much backwards compatibility as
373 possible. However, since Plack is different from Catalyst::Engine, it
374 is possible that differences exist for edge cases. Therefore, we
375 recommend that care be taken with this upgrade and that testing should
376 be greater than would be the case with a minor point update. Please
377 inform the Catalyst developers of any problems so that we can fix them
378 and incorporate tests.
379 It is highly recommended that you become familiar with the Plack
381 ecosystem and documentation. Being able to take advantage of Plack
382 development and middleware is a major bonus to this upgrade.
383 Documentation about how to take advantage of Plack::Middleware by
384 writing your own ".psgi" file is contained in Catalyst::PSGI.
385 If you have created a custom subclass of <Catalyst:Engine>, you will
387 need to convert it to be a subclass of Plack::Handler.
388 If you are using the Plack engine, Catalyst::Engine::PSGI, this new
390 release supersedes that code.
391 If you are using a subclass of Catalyst::Engine that is aimed at
393 nonstandard or internal/testing uses, such as
394 Catalyst::Engine::Embeddable, you should still be able to continue
395 using that engine.
396 Advice for specific subclasses of Catalyst::Engine follows:
398 Upgrading the FastCGI Engine
400 No upgrade is needed if your myapp_fastcgi.pl script is already
401 upgraded to use Catalyst::Script::FastCGI.
402 Upgrading the mod_perl / Apache Engines
404 The engines that are built upon the various iterations of mod_perl,
405 Catalyst::Engine::Apache::MP13 (for mod_perl 1, and Apache 1.x) and
406 Catalyst::Engine::Apache2::MP20 (for mod_perl 2, and Apache 2.x),
407 should be seamless upgrades and will work using Plack::Handler::Apache1
408 or Plack::Handler::Apache2 as required.
409 Catalyst::Engine::Apache2::MP19, however, is no longer supported, as
411 Plack does not support mod_perl version 1.99. This is unlikely to be a
412 problem for anyone, as 1.99 was a brief beta-test release for mod_perl
413 2, and all users of mod_perl 1.99 are encouraged to upgrade to a
414 supported release of Apache 2 and mod_perl 2.
415 Upgrading the HTTP Engine
417 The default development server that comes with the Catalyst
418 distribution should continue to work as expected with no changes as
419 long as your "myapp_server" script is upgraded to use
420 Catalyst::Script::HTTP.
421 Upgrading the CGI Engine
423 If you were using Catalyst::Engine::CGI there is no upgrade needed if
424 your myapp_cgi.pl script is already upgraded to use
425 Catalyst::Script::CGI.
426 Upgrading Catalyst::Engine::HTTP::Prefork
428 If you were using Catalyst::Engine::HTTP::Prefork then Starman is
429 automatically loaded. You should (at least) change your "Makefile.PL"
430 to depend on Starman.
431 You can regenerate your "myapp_server.pl" script with "catalyst.pl" and
433 implement a "MyApp::Script::Server" class that looks like this:
434 package MyApp::Script::Server;
436 use Moose;
437 use namespace::autoclean;
438 extends 'CatalystX::Script::Server::Starman';
440 1;
442 This takes advantage of the new script system, and will add a number of
444 options to the standard server script as extra options are added by
445 Starman.
446 More information about these options can be seen at "SYNOPSIS" in
448 CatalystX::Script::Server::Starman.
449 An alternate route to implement this functionality is to write a simple
451 .psgi file for your application, and then use the plackup utility to
452 start the server.
453 Upgrading the PSGI Engine
455 If you were using Catalyst::Engine::PSGI, this new release supersedes
456 this engine in supporting Plack. By default the Engine is now always
457 Plack. As a result, you can remove the dependency on
458 Catalyst::Engine::PSGI in your "Makefile.PL".
459 Applications that were using Catalyst::Engine::PSGI previously should
461 entirely continue to work in this release with no changes.
462 However, if you have an "app.psgi" script, then you no longer need to
464 specify the PSGI engine. Instead, the Catalyst application class now
465 has a new method "psgi_app" which returns a PSGI compatible coderef
466 which you can wrap in the middleware of your choice.
467 Catalyst will use the .psgi for your application if it is located in
469 the "home" directory of the application.
470 For example, if you were using Catalyst::Engine::PSGI in the past, you
472 will have written (or generated) a "script/myapp.psgi" file similar to
473 this one:
474 use Plack::Builder;
476 use MyCatalytApp;
477 MyCatalystApp->setup_engine('PSGI');
479 builder {
481 enable ... # enable your desired middleware
482 sub { MyCatalystApp->run(@_) };
483 };
484 Instead, you now say:
486 use Plack::Builder;
488 use MyCatalystApp;
489 builder {
491 enable ... #enable your desired middleware
492 MyCatalystApp->psgi_app;
493 };
494 In the simplest case:
496 MyCatalystApp->setup_engine('PSGI');
498 my $app = sub { MyCatalystApp->run(@_) }
499 becomes
501 my $app = MyCatalystApp->psgi_app(@_);
503 NOT:
505 my $app = sub { MyCatalystApp->psgi_app(@_) };
507 # If you make ^^ this mistake, your app won't work, and will confuse the hell out of you!
508 You can now move "script/myapp.psgi" to "myapp.psgi", and the built-in
510 Catalyst scripts and your test suite will start using your .psgi file.
511 NOTE: If you rename your .psgi file without these modifications, then
513 any tests run via Catalyst::Test will not be compatible with the new
514 release, and will result in the development server starting, rather
515 than the expected test running.
516 NOTE: If you are directly accessing "$c->req->env" to get the PSGI
518 environment then this accessor is moved to "$c->engine->env", you will
519 need to update your code.
520 Engines which are known to be broken
522 The following engines DO NOT work as of Catalyst version 5.9. The core
523 team will be happy to work with the developers and/or users of these
524 engines to help them port to the new Plack/Engine system, but for now,
525 applications which are currently using these engines WILL NOT run
526 without modification to the engine code.
527 Catalyst::Engine::Wx
529 Catalyst::Engine::Zeus
530 Catalyst::Engine::JobQueue::POE
531 Catalyst::Engine::XMPP2
532 Catalyst::Engine::SCGI
533 Engines with unknown status
535 The following engines are untested or have unknown compatibility.
536 Reports are highly encouraged:
537 Catalyst::Engine::Mojo
539 Catalyst::Engine::Server (marked as Deprecated)
540 Catalyst::Engine::HTTP::POE (marked as Deprecated)
541 Plack functionality
543 See Catalyst::PSGI.
544 Tests in 5.9
546 Tests should generally work the same in Catalyst 5.9, but there are
547 some differences.
548 Previously, if using Catalyst::Test and doing local requests (against a
550 local server), if the application threw an exception then this
551 exception propagated into the test.
552 This behavior has been removed, and now a 500 response will be returned
554 to the test. This change standardizes behavior, so that local test
555 requests behave similarly to remote requests.
556
558 Most applications and plugins should run unaltered on Catalyst 5.80.
559 However, a lot of refactoring work has taken place, and several changes
561 have been made which could cause incompatibilities. If your application
562 or plugin is using deprecated code, or relying on side effects, then
563 you could have issues upgrading to this release.
564 Most issues found with existing components have been easy to solve.
566 This document provides a complete description of behavior changes which
567 may cause compatibility issues, and of new Catalyst warnings which
568 might be unclear.
569 If you think you have found an upgrade-related issue which is not
571 covered in this document, please email the Catalyst list to discuss the
572 problem.
573
575 Application class roles
576 You can only apply method modifiers after the application's "->setup"
577 method has been called. This means that modifiers will not work with
578 methods run during the call to "->setup".
579 See Catalyst::Manual::ExtendingCatalyst for more information about
581 using Moose in your applications.
582 Controller actions in Moose roles
584 You can use MooseX::MethodAttributes::Role if you want to declare
585 actions inside Moose roles.
586 Using Moose in Components
588 The correct way to use Moose in a component in a both forward and
589 backwards compatible way is:
590 package TestApp::Controller::Root;
592 use Moose;
593 BEGIN { extends 'Catalyst::Component' }; # Or ::Controller, or whatever
594 See "Components which inherit from Moose::Object before
596 Catalyst::Component".
597
599 Applications in a single file
600 Applications must be in their own file, and loaded at compile time.
601 This issue generally only affects the tests of CPAN distributions. Your
602 application will fail if you try to define an application inline in a
603 block, and use plugins which supply a " new " method, then use that
604 application latter in tests within the same file.
605 This is due to the fact that Catalyst is inlining a new method on your
607 application class allowing it to be compatible with Moose. The method
608 used to do this changed in 5.80004 to avoid the possibility of
609 reporting an 'Unknown Error' if your application failed to compile.
610 Issues with Class::C3
612 Catalyst 5.80 uses the Algorithm::C3 method dispatch order. This is
613 built into Perl 5.10, and comes via Class::C3 for Perl 5.8. This
614 replaces NEXT with Class::C3::Adopt::NEXT, forcing all components to
615 resolve methods using C3, rather than the unpredictable dispatch order
616 of NEXT.
617 This issue manifests itself by your application failing to start due to
619 an error message about having a non-linear @ISA.
620 The Catalyst plugin most often causing this is
622 Catalyst::Plugin::Session::Store::FastMmap - if you are using this
623 plugin and see issues, then please upgrade your plugins, as it has been
624 fixed. Note that Makefile.PL in the distribution will warn about known
625 incompatible components.
626 This issue can, however, be found in your own application - the only
628 solution is to go through each base class of the class the error was
629 reported against, until you identify the ones in conflict, and resolve
630 them.
631 To be able to generate a linear @ISA, the list of superclasses for each
633 class must be resolvable using the C3 algorithm. Unfortunately, when
634 superclasses are being used as mixins (to add functionality used in
635 your class), and with multiple inheritance, it is easy to get this
636 wrong.
637 Most common is the case of:
639 package Component1; # Note, this is the common case
641 use base qw/Class::Accessor::Fast Class::Data::Inheritable/;
642 package Component2; # Accidentally saying it this way causes a failure
644 use base qw/Class::Data::Inheritable Class::Accessor::Fast/;
645 package GoesBang;
647 use base qw/Component1 Component2/;
648 Any situation like this will cause your application to fail to start.
650 For additional documentation about this issue, and how to resolve it,
652 see Class::C3::Adopt::NEXT.
653 Components which inherit from Moose::Object before Catalyst::Component
655 Moose components which say:
656 package TestApp::Controller::Example;
658 use Moose;
659 extends qw/Moose::Object Catalyst::Component/;
660 to use the constructor provided by Moose, while working (if you do some
662 hacks with the " BUILDARGS " method), will not work with Catalyst 5.80
663 as "Catalyst::Component" inherits from "Moose::Object", and so @ISA
664 fails to linearize.
665 The correct way to use Moose in a component in a both forward and
667 backwards compatible way is:
668 package TestApp::Controller::Root;
670 use Moose;
671 BEGIN { extends 'Catalyst::Component' }; # Or ::Controller, or whatever
672 Note that the " extends " declaration needs to occur in a begin block
674 for attributes to operate correctly.
675 This way you do not inherit directly from "Moose::Object" yourself.
677 Having components which do not inherit their constructor from
678 "Catalyst::Component" is unsupported, and has never been recommended,
679 therefore you're on your own if you're using this technique. You'll
680 need to detect the version of Catalyst your application is running, and
681 deal with it appropriately.
682 You also don't get the Moose::Object constructor, and therefore
684 attribute initialization will not work as normally expected. If you
685 want to use Moose attributes, then they need to be made lazy to
686 correctly initialize.
687 Note that this only applies if your component needs to maintain
689 component backwards compatibility for Catalyst versions before 5.71001
690 - in 5.71001 attributes work as expected, and the BUILD method is
691 called normally (although BUILDARGS is not).
692 If you depend on Catalyst 5.8, then all Moose features work as
694 expected.
695 You will also see this issue if you do the following:
697 package TestApp::Controller::Example;
699 use Moose;
700 use base 'Catalyst::Controller';
701 as " use base " appends to @ISA.
703 use Moose in MyApp
705 Similar to the above, this will also fail:
707 package MyApp;
709 use Moose;
710 use Catalyst qw/
711 ConfigLoader
712 /;
713 __PACKAGE__->setup;
714 If you need to use Moose in your application class (e.g. for method
716 modifiers etc.) then the correct technique is:
717 package MyApp;
719 use Moose;
720 use Catalyst;
721 extends 'Catalyst';
723 __PACKAGE__->config( name => 'MyApp' );
725 __PACKAGE__->setup(qw/
726 ConfigLoader
727 /);
728 Anonymous closures installed directly into the symbol table
730 If you have any code which installs anonymous subroutine references
731 directly into the symbol table, you may encounter breakages. The
732 simplest solution is to use Sub::Name to name the subroutine. Example:
733 # Original code, likely to break:
735 my $full_method_name = join('::', $package_name, $method_name);
736 *$full_method_name = sub { ... };
737 # Fixed Code
739 use Sub::Name 'subname';
740 my $full_method_name = join('::',$package_name, $method_name);
741 *$full_method_name = subname $full_method_name, sub { ... };
742 Additionally, you can take advantage of Catalyst's use of Class::MOP
744 and install the closure using the appropriate metaclass. Example:
745 use Class::MOP;
747 my $metaclass = Moose::Meta::Class->initialize($package_name);
748 $metaclass->add_method($method_name => sub { ... });
749 Hooking into application setup
751 To execute code during application start-up, the following snippet in
752 MyApp.pm used to work:
753 sub setup {
755 my ($class, @args) = @_;
756 $class->NEXT::setup(@args);
757 ... # things to do after the actual setup
758 }
759 With Catalyst 5.80 this won't work anymore, because Catalyst no longer
761 uses NEXT.pm for method resolution. The functionality was only ever
762 originally operational as NEXT remembers what methods have already been
763 called, and will not call them again.
764 Using this now causes infinite recursion between MyApp::setup and
766 Catalyst::setup, due to other backwards compatibility issues related to
767 how plugin setup works. Moose method modifiers like
768 "before|after|around setup => sub { ... };" also will not operate
769 correctly on the setup method.
770 The right way to do it is this:
772 after setup_finalize => sub {
774 ... # things to do after the actual setup
775 };
776 The setup_finalize hook was introduced as a way to avoid this issue.
778 Components with a new method which returns false
780 Previously, if you had a component which inherited from
781 Catalyst::COMPONENT, but overrode the new method to return false, then
782 your class's configuration would be blessed into a hash on your behalf,
783 and this would be returned from the COMPONENT method.
784 This behavior makes no sense, and so has been removed. Implementing
786 your own " new " method in components is highly discouraged. Instead,
787 you should inherit the new method from Catalyst::Component, and use
788 Moose's BUILD functionality and/or Moose attributes to perform any
789 construction work necessary for your class.
790 __PACKAGE__->mk_accessor('meta');
792 Won't work due to a limitation of Moose. This is currently being fixed
793 inside Moose.
794 Class::Data::Inheritable side effects
796 Previously, writing to a class data accessor would copy the accessor
797 method down into your package.
798 This behavior has been removed. While the class data is still stored
800 per-class, it is stored on the metaclass of the class defining the
801 accessor.
802 Therefore anything relying on the side effect of the accessor being
804 copied down will be broken.
805 The following test demonstrates the problem:
807 {
809 package BaseClass;
810 use base qw/Class::Data::Inheritable/;
811 __PACKAGE__->mk_classdata('foo');
812 }
813 {
815 package Child;
816 use base qw/BaseClass/;
817 }
818 BaseClass->foo('base class');
820 Child->foo('sub class');
821 use Test::More;
823 isnt(BaseClass->can('foo'), Child->can('foo'));
824 Extending Catalyst::Request or other classes in an ad hoc manner using
826 mk_accessors
827 Previously, it was possible to add additional accessors to
828 Catalyst::Request (or other classes) by calling the mk_accessors class
829 method.
830 This is no longer supported - users should make a subclass of the class
832 whose behavior they would like to change, rather than globally
833 polluting the Catalyst objects.
834 Confused multiple inheritance with Catalyst::Component::COMPONENT
836 Previously, Catalyst's COMPONENT method would delegate to the method on
837 the right hand side, which could then delegate back again with NEXT.
838 This is poor practice, and in addition, makes no sense with C3 method
839 dispatch order, and is therefore no longer supported.
840 If a COMPONENT method is detected in the inheritance hierarchy to the
842 right hand side of Catalyst::Component::COMPONENT, then the following
843 warning message will be emitted:
844 There is a COMPONENT method resolving after Catalyst::Component
846 in ${next_package}.
847 The correct fix is to re-arrange your class's inheritance hierarchy so
849 that the COMPONENT method you would like to inherit is the first (left-
850 hand most) COMPONENT method in your @ISA.
851 Development server relying on environment variables
853 Previously, the development server would allow propagation of system
854 environment variables into the request environment, this has changed
855 with the adoption of Plack. You can use Plack::Middleware::ForceEnv to
856 achieve the same effect.
857
859 Actions in your application class
860 Having actions in your application class will now emit a warning at
861 application startup as this is deprecated. It is highly recommended
862 that these actions are moved into a MyApp::Controller::Root (as
863 demonstrated by the scaffold application generated by catalyst.pl).
864 This warning, also affects tests. You should move actions in your test,
866 creating a myTest::Controller::Root, like the following example:
867 package MyTest::Controller::Root;
869 use strict;
871 use warnings;
872 use parent 'Catalyst::Controller';
874 __PACKAGE__->config(namespace => '');
876 sub action : Local {
878 my ( $self, $c ) = @_;
879 $c->do_something;
880 }
881 1;
883 ::[MVC]:: naming scheme
885 Having packages called MyApp::[MVC]::XX is deprecated and can no longer
886 be generated by catalyst.pl
887 This is still supported, but it is recommended that you rename your
889 application components to Model/View/Controller.
890 A warning will be issued at application startup if the ::[MVC]:: naming
892 scheme is in use.
893 Catalyst::Base
895 Any code using Catalyst::Base will now emit a warning; this module will
896 be removed in a future release.
897 Methods in Catalyst::Dispatcher
899 The following methods in Catalyst::Dispatcher are implementation
900 details, which may change in the 5.8X release series, and therefore
901 their use is highly deprecated.
902 tree
904 dispatch_types
905 registered_dispatch_types
906 method_action_class
907 action_hash
908 container_hash
909 The first time one of these methods is called, a warning will be
911 emitted:
912 Class $class is calling the deprecated method Catalyst::Dispatcher::$public_method_name,
914 this will be removed in Catalyst 5.9
915 You should NEVER be calling any of these methods from application code.
917 Plugin authors and maintainers whose plugins currently call these
919 methods should change to using the public API, or, if you do not feel
920 the public API adequately supports your use case, please email the
921 development list to discuss what API features you need so that you can
922 be appropriately supported.
923 Class files with names that don't correspond to the packages they define
925 In this version of Catalyst, if a component is loaded from disk, but no
926 symbols are defined in that component's name space after it is loaded,
927 this warning will be issued:
928 require $class was successful but the package is not defined.
930 This is to protect against confusing bugs caused by mistyping package
932 names, and will become a fatal error in a future version.
933 Please note that 'inner packages' (via Devel::InnerPackage) are still
935 fully supported; this warning is only issued when component file naming
936 does not map to any of the packages defined within that component.
937 $c->plugin method
939 Calling the plugin method is deprecated, and calling it at run time is
940 highly deprecated.
941 Instead you are recommended to use Catalyst::Model::Adaptor or similar
943 to compose the functionality you need outside of the main application
944 name space.
945 Calling the plugin method will not be supported past Catalyst 5.81.
947perl v5.32.0 2020-07-28 Catalyst::Upgrading(3)