added documentation reminder to install FCGI.pm when deploying to fastcgi

[catagits/Catalyst-Manual.git] / lib / Catalyst / Manual / Cookbook.pod

=head1 NAME

Catalyst::Manual::Cookbook - Cooking with Catalyst

=head1 DESCRIPTION

Yummy code like your mum used to bake!

=head1 RECIPES

=head1 Basics

These recipes cover some basic stuff that is worth knowing for
Catalyst developers.

=head2 Delivering a Custom Error Page

By default, Catalyst will display its own error page whenever it
encounters an error in your application. When running under C<-Debug>
mode, the error page is a useful screen including the error message
and L<Data::Dump> output of the relevant parts of the C<$c> context
object.  When not in C<-Debug>, users see a simple "Please come back
later" screen.

To use a custom error page, use a special C<end> method to
short-circuit the error processing. The following is an example; you
might want to adjust it further depending on the needs of your
application (for example, any calls to C<fillform> will probably need
to go into this C<end> method; see L<Catalyst::Plugin::FillInForm>).

    sub end : Private {
        my ( $self, $c ) = @_;

        if ( scalar @{ $c->error } ) {
            $c->stash->{errors}   = $c->error;
            $c->stash->{template} = 'errors.tt';
            $c->forward('MyApp::View::TT');
            $c->error(0);
        }

        return 1 if $c->response->status =~ /^3\d\d$/;
        return 1 if $c->response->body;

        unless ( $c->response->content_type ) {
            $c->response->content_type('text/html; charset=utf-8');
        }

        $c->forward('MyApp::View::TT');
    }

You can manually set errors in your code to trigger this page by calling

    $c->error( 'You broke me!' );

=head2 Disable statistics

Just add this line to your application class if you don't want those
nifty statistics in your debug messages.

    sub Catalyst::Log::info { }

=head2 Enable debug status in the environment

Normally you enable the debugging info by adding the C<-Debug> flag to
your C<use Catalyst> statement . However, you can also enable it using
environment variable, so you can (for example) get debug info without
modifying your application scripts. Just set C<CATALYST_DEBUG> or
C<E<lt>MYAPPE<gt>_DEBUG> to a true value.

=head2 Sessions

When you have your users identified, you will want to somehow remember
that fact, to save them from having to identify themselves for every
single page. One way to do this is to send the username and password
parameters in every single page, but that's ugly, and won't work for
static pages.

Sessions are a method of saving data related to some transaction, and
giving the whole collection a single ID. This ID is then given to the
user to return to us on every page they visit while logged in. The
usual way to do this is using a browser cookie.

Catalyst uses two types of plugins to represent sessions:

=head3 State

A State module is used to keep track of the state of the session
between the users browser, and your application.

A common example is the Cookie state module, which sends the browser a
cookie containing the session ID. It will use default value for the
cookie name and domain, so will "just work" when used.

=head3 Store

A Store module is used to hold all the data relating to your session,
for example the users ID, or the items for their shopping cart. You
can store data in memory (FastMmap), in a file (File) or in a database
(DBI).

=head3 Authentication magic

If you have included the session modules in your application, the
Authentication modules will automagically use your session to save and
retrieve the user data for you.

=head3 Using a session

Once the session modules are loaded, the session is available as C<<
$c->session >>, and can be writen to and read from as a simple hash
reference.

=head3 EXAMPLE

  package MyApp;
  use Moose;
  use namespace::autoclean;

  use Catalyst  qw/
                         Session
                         Session::Store::FastMmap
                         Session::State::Cookie
                   /;
  extends 'Catalyst';
  __PACKAGE__->setup;

  package MyApp::Controller::Foo;
  use Moose;
  use namespace::autoclean;
  BEGIN { extends 'Catalyst::Controller' };
  ## Write data into the session

  sub add_item : Local {
     my ( $self, $c ) = @_;

     my $item_id = $c->req->param("item");

     push @{ $c->session->{items} }, $item_id;

  }

  ## A page later we retrieve the data from the session:

  sub get_items : Local {
     my ( $self, $c ) = @_;

     $c->stash->{items_to_display} = $c->session->{items};

  }


=head3 More information

L<http://search.cpan.org/dist/Catalyst-Plugin-Session>

L<http://search.cpan.org/dist/Catalyst-Plugin-Session-State-Cookie>

L<http://search.cpan.org/dist/Catalyst-Plugin-Session-State-URI>

L<http://search.cpan.org/dist/Catalyst-Plugin-Session-Store-FastMmap>

L<http://search.cpan.org/dist/Catalyst-Plugin-Session-Store-File>

L<http://search.cpan.org/dist/Catalyst-Plugin-Session-Store-DBI>

=head2 Configure your application

You configure your application with the C<config> method in your
application class. This can be hard-coded, or brought in from a
separate configuration file.

=head3 Using Config::General

L<Config::General|Config::General> is a method for creating flexible
and readable configuration files. It's a great way to keep your
Catalyst application configuration in one easy-to-understand location.

Now create C<myapp.conf> in your application home:

  name     MyApp

  # session; perldoc Catalyst::Plugin::Session::FastMmap
  <Session>
    expires 3600
    rewrite 0
    storage /tmp/myapp.session
  </Session>

  # emails; perldoc Catalyst::Plugin::Email
  # this passes options as an array :(
  Mail SMTP
  Mail localhost

This is equivalent to:

  # configure base package
  __PACKAGE__->config( name => MyApp );
  # configure authentication
  __PACKAGE__->config->{authentication} = {
    user_class => 'MyApp::Model::MyDB::Customer',
    ...
  };
  # configure sessions
  __PACKAGE__->config->{session} = {
    expires => 3600,
    ...
  };
  # configure email sending
  __PACKAGE__->config->{email} = [qw/SMTP localhost/];

See also L<Config::General|Config::General>.

=head1 Skipping your VCS's directories

Catalyst uses Module::Pluggable to load Models, Views, and Controllers.
Module::Pluggable will scan through all directories and load modules
it finds.  Sometimes you might want to skip some of these directories,
for example when your version control system makes a subdirectory with
meta-information in every version-controlled directory.  While
Catalyst skips subversion and CVS directories already, there are other
source control systems.  Here is the configuration you need to add
their directories to the list to skip.

You can make Catalyst skip these directories using the Catalyst config:

  # Configure the application
  __PACKAGE__->config(
      name => 'MyApp',
      setup_components => { except => qr/SCCS/ },
  );

See the Module::Pluggable manual page for more information on B<except>
and other options.

=head1 Users and Access Control

Most multiuser, and some single-user web applications require that
users identify themselves, and the application is often required to
define those roles.  The recipes below describe some ways of doing
this.

=head2 Authentication (logging in)

This is extensively covered in other documentation; see in particular
L<Catalyst::Plugin::Authentication> and the Authentication chapter
of the Tutorial at L<Catalyst::Manual::Tutorial::06_Authorization>.

=head2 Pass-through login (and other actions)

An easy way of having assorted actions that occur during the processing
of a request that are orthogonal to its actual purpose - logins, silent
commands etc. Provide actions for these, but when they're required for
something else fill e.g. a form variable __login and have a sub begin
like so:

    sub begin : Private {
      my ($self, $c) = @_;
      foreach my $action (qw/login docommand foo bar whatever/) {
        if ($c->req->params->{"__${action}"}) {
          $c->forward($action);
        }
      }
    }

=head2 Authentication/Authorization

This is done in several steps:

=over 4

=item Verification

Getting the user to identify themselves, by giving you some piece of
information known only to you and the user. Then you can assume that
the user is who they say they are. This is called B<credential
verification>.

=item Authorization

Making sure the user only accesses functions you want them to
access. This is done by checking the verified user's data against your
internal list of groups, or allowed persons for the current page.

=back

=head3 Modules

The Catalyst Authentication system is made up of many interacting
modules, to give you the most flexibility possible.

=head4 Credential verifiers

A Credential module tables the user input, and passes it to a Store,
or some other system, for verification. Typically, a user object is
created by either this module or the Store and made accessible by a
C<< $c->user >> call.

Examples:

 Password - Simple username/password checking.
 HTTPD    - Checks using basic HTTP auth.
 TypeKey  - Check using the typekey system.

=head3 Storage backends

A Storage backend contains the actual data representing the users. It
is queried by the credential verifiers. Updating the store is not done
within this system; you will need to do it yourself.

Examples:

 DBIC     - Storage using a database via DBIx::Class.
 Minimal  - Storage using a simple hash (for testing).

=head3 User objects

A User object is created by either the storage backend or the
credential verifier, and is filled with the retrieved user information.

Examples:

 Hash     - A simple hash of keys and values.

=head3 ACL authorization

ACL stands for Access Control List. The ACL plugin allows you to
regulate access on a path-by-path basis, by listing which users, or
roles, have access to which paths.

=head3 Roles authorization

Authorization by roles is for assigning users to groups, which can
then be assigned to ACLs, or just checked when needed.

=head3 Logging in

When you have chosen your modules, all you need to do is call the C<<
$c->authenticate >> method. If called with no parameters, it will try to find
suitable parameters, such as B<username> and B<password>, or you can
pass it these values.

=head3 Checking roles

Role checking is done by using the C<< $c->check_user_roles >> method.
This will check using the currently logged-in user (via C<< $c->user
>>). You pass it the name of a role to check, and it returns true if
the user is a member.

=head3 EXAMPLE

  package MyApp;
  use Moose;
  use namespace::autoclean;
  extends qw/Catalyst/;
  use Catalyst qw/
    Authentication
    Authorization::Roles
  /;

  __PACKAGE__->config(
     authentication => {
         default_realm => 'test',
         realms => {
             test => {
                 credential => {
                     class          => 'Password',
                     password_field => 'password',
                     password_type  => 'self_check',
                 },
                 store => {
                     class => 'Htpasswd',
                     file => 'htpasswd',
                 },
             },
         },
     },
  );

  package MyApp::Controller::Root;
  use Moose;
  use namespace::autoclean;

  BEGIN { extends 'Catalyst::Controller' }

  __PACKAGE__->config(namespace => '');

  sub login : Local {
     my ($self, $c) = @_;

     if (    my $user = $c->req->param("user")
         and my $password = $c->req->param("password") )
     {
         if ( $c->authenticate( username => $user, password => $password ) ) {
              $c->res->body( "hello " . $c->user->name );
         } else {
            # login incorrect
         }
     }
     else {
         # invalid form input
     }
  }

  sub restricted : Local {
     my ( $self, $c ) = @_;

     $c->detach("unauthorized")
       unless $c->check_user_roles( "admin" );

     # do something restricted here
  }

=head3 Using authentication in a testing environment

Ideally, to write tests for authentication/authorization code one would
first set up a test database with known data, then use
L<Test::WWW::Mechanize::Catalyst> to simulate a user logging
in. Unfortunately this can be rather awkward, which is why it's a good
thing that the authentication framework is so flexible.

Instead of using a test database, one can simply change the
authentication store to something a bit easier to deal with in a
testing environment. Additionally, this has the advantage of not
modifying one's database, which can be problematic if one forgets to
use the testing instead of production database.

Alternatively, if you want to authenticate real users, but not have to
worry about their passwords, you can use
L<Catalyst::Authentication::Credential::Testing> to force all users to
authenticate with a global password.

=head3 More information

L<Catalyst::Plugin::Authentication> has a longer explanation.

=head2 Authorization

=head3 Introduction

Authorization is the step that comes after
authentication. Authentication establishes that the user agent is really
representing the user we think it's representing, and then authorization
determines what this user is allowed to do.

=head3 Role Based Access Control

Under role based access control each user is allowed to perform any
number of roles. For example, at a zoo no one but specially trained
personnel can enter the moose cage (Mynd you, møøse bites kan be
pretty nasti!). For example:

    package Zoo::Controller::MooseCage;

    sub feed_moose : Local {
        my ( $self, $c ) = @_;

        $c->model( "Moose" )->eat( $c->req->param("food") );
    }

With this action, anyone can just come into the moose cage and feed
the moose, which is a very dangerous thing. We need to restrict this
action, so that only a qualified moose feeder can perform that action.

The Authorization::Roles plugin lets us perform role based access
control checks. Let's load it:

    use parent qw/Catalyst/;
    use Catalyst qw/
                    Authentication
                    Authorization::Roles
                  /;

And now our action should look like this:

    sub feed_moose : Local {
        my ( $self, $c ) = @_;

        if ( $c->check_roles( "moose_feeder" ) ) {
            $c->model( "Moose" )->eat( $c->req->param("food") );
        } else {
            $c->stash->{error} = "unauthorized";
        }
    }

This checks C<< $c->user >>, and only if the user has B<all> the roles
in the list, a true value is returned.

C<check_roles> has a sister method, C<assert_roles>, which throws an
exception if any roles are missing.

Some roles that might actually make sense in, say, a forum application:

=over 4

=item *

administrator

=item *

moderator

=back

each with a distinct task (system administration versus content
administration).

=head3 Access Control Lists

Checking for roles all the time can be tedious and error prone.

The Authorization::ACL plugin lets us declare where we'd like checks
to be done automatically for us.

For example, we may want to completely block out anyone who isn't a
C<moose_feeder> from the entire C<MooseCage> controller:

    Zoo->deny_access_unless( "/moose_cage", [qw/moose_feeder/] );

The role list behaves in the same way as C<check_roles>. However, the
ACL plugin isn't limited to just interacting with the Roles plugin. We
can use a code reference instead. For example, to allow either moose
trainers or moose feeders into the moose cage, we can create a more
complex check:

    Zoo->deny_access_unless( "/moose_cage", sub {
        my $c = shift;
        $c->check_roles( "moose_trainer" ) || $c->check_roles( "moose_feeder" );
    });

The more specific a role, the earlier it will be checked. Let's say
moose feeders are now restricted to only the C<feed_moose> action,
while moose trainers get access everywhere:

    Zoo->deny_access_unless( "/moose_cage", [qw/moose_trainer/] );
    Zoo->allow_access_if( "/moose_cage/feed_moose", [qw/moose_feeder/]);

When the C<feed_moose> action is accessed the second check will be
made. If the user is a C<moose_feeder>, then access will be
immediately granted. Otherwise, the next rule in line will be tested -
the one checking for a C<moose_trainer>.  If this rule is not
satisfied, access will be immediately denied.

Rules applied to the same path will be checked in the order they were
added.

Lastly, handling access denial events is done by creating an
C<access_denied> private action:

    sub access_denied : Private {
        my ( $self, $c, $action ) = @_;
    }

This action works much like auto, in that it is inherited across
namespaces (not like object oriented code). This means that the
C<access_denied> action which is B<nearest> to the action which was
blocked will be triggered.

If this action does not exist, an error will be thrown, which you can
clean up in your C<end> private action instead.

Also, it's important to note that if you restrict access to "/" then
C<end>, C<default>, etc. will also be restricted.

   MyApp->acl_allow_root_internals;

will create rules that permit access to C<end>, C<begin>, and C<auto> in the
root of your app (but not in any other controller).

=head1 Models

Models are where application data belongs.  Catalyst is exteremely
flexible with the kind of models that it can use.  The recipes here
are just the start.

=head2 Using existing DBIC (etc.) classes with Catalyst

Many people have existing Model classes that they would like to use
with Catalyst (or, conversely, they want to write Catalyst models that
can be used outside of Catalyst, e.g.  in a cron job). It's trivial to
write a simple component in Catalyst that slurps in an outside Model:

    package MyApp::Model::DB;

    use base qw/Catalyst::Model::DBIC::Schema/;

    __PACKAGE__->config(
        schema_class => 'Some::DBIC::Schema',
        connect_info => ['dbi:SQLite:foo.db', '', '', {AutoCommit=>1}];
    );

    1;

and that's it! Now C<Some::DBIC::Schema> is part of your
Cat app as C<MyApp::Model::DB>.

=head2 DBIx::Class as a Catalyst Model

See L<Catalyst::Model::DBIC::Schema>.

=head2 Create accessors to preload static data once per server instance

When you have data that you want to load just once from the model at
startup, instead of for each request, use mk_group_accessors to
create accessors and tie them to resultsets in your package that
inherits from DBIx::Class::Schema:

    package My::Schema;
    use base qw/DBIx::Class::Schema/;
    __PACKAGE__->register_class('RESULTSOURCEMONIKER',
                                'My::Schema::RESULTSOURCE');
    __PACKAGE__->mk_group_accessors('simple' =>
                                qw(ACCESSORNAME1 ACCESSORNAME2 ACCESSORNAMEn));

    sub connection {
        my ($self, @rest) = @_;
        $self->next::method(@rest);
        # $self is now a live My::Schema object, complete with DB connection

        $self->ACCESSORNAME1([ $self->resultset('RESULTSOURCEMONIKER')->all ]);
        $self->ACCESSORNAME2([ $self->resultset('RESULTSOURCEMONIKER')->search({ COLUMN => { '<' => '30' } })->all ]);
        $self->ACCESSORNAMEn([ $self->resultset('RESULTSOURCEMONIKER')->find(1) ]);
    }

    1;

and now in the controller, you can now access any of these without a
per-request fetch:

    $c->stash->{something} = $c->model('My::Schema')->schema->ACCESSORNAME;


=head2 XMLRPC

Unlike SOAP, XMLRPC is a very simple (and elegant) web-services
protocol, exchanging small XML messages like these:

Request:

    POST /api HTTP/1.1
    TE: deflate,gzip;q=0.3
    Connection: TE, close
    Accept: text/xml
    Accept: multipart/*
    Host: 127.0.0.1:3000
    User-Agent: SOAP::Lite/Perl/0.60
    Content-Length: 192
    Content-Type: text/xml

    <?xml version="1.0" encoding="UTF-8"?>
    <methodCall>
        <methodName>add</methodName>
        <params>
            <param><value><int>1</int></value></param>
            <param><value><int>2</int></value></param>
        </params>
    </methodCall>

Response:

    Connection: close
    Date: Tue, 20 Dec 2005 07:45:55 GMT
    Content-Length: 133
    Content-Type: text/xml
    Status: 200
    X-Catalyst: 5.70

    <?xml version="1.0" encoding="us-ascii"?>
    <methodResponse>
        <params>
            <param><value><int>3</int></value></param>
        </params>
    </methodResponse>

Now follow these few steps to implement the application:

1. Install Catalyst (5.61 or later), Catalyst::Plugin::XMLRPC (0.06 or
later) and SOAP::Lite (for XMLRPCsh.pl).

2. Create an application framework:

    % catalyst.pl MyApp
    ...
    % cd MyApp

3. Add the XMLRPC plugin to MyApp.pm

    use Catalyst qw/-Debug Static::Simple XMLRPC/;

4. Add an API controller

    % ./script/myapp_create.pl controller API

5. Add a XMLRPC redispatch method and an add method with Remote
attribute to lib/MyApp/Controller/API.pm

    sub default :Path {
        my ( $self, $c ) = @_;
        $c->xmlrpc;
    }

    sub add : Remote {
        my ( $self, $c, $a, $b ) = @_;
        return $a + $b;
    }

The default action is the entry point for each XMLRPC request. It will
redispatch every request to methods with Remote attribute in the same
class.

The C<add> method is not a traditional action; it has no private or
public path. Only the XMLRPC dispatcher knows it exists.

6. That's it! You have built your first web service. Let's test it with
XMLRPCsh.pl (part of SOAP::Lite):

    % ./script/myapp_server.pl
    ...
    % XMLRPCsh.pl http://127.0.0.1:3000/api
    Usage: method[(parameters)]
    > add( 1, 2 )
    --- XMLRPC RESULT ---
    '3'

=head3 Tip

Your return data type is usually auto-detected, but you can easily
enforce a specific one.

    sub add : Remote {
        my ( $self, $c, $a, $b ) = @_;
        return RPC::XML::int->new( $a + $b );
    }

=head1 Views

Views pertain to the display of your application.  As with models,
Catalyst is uncommonly flexible.  The recipes below are just a start.

=head2 Catalyst::View::TT

One of the first things you probably want to do when starting a new
Catalyst application is set up your View. Catalyst doesn't care how you
display your data; you can choose to generate HTML, PDF files, or plain
text if you wanted.

Most Catalyst applications use a template system to generate their HTML,
and though there are several template systems available,
L<Template Toolkit|Template> is probably the most popular.

Once again, the Catalyst developers have done all the hard work, and
made things easy for the rest of us. Catalyst::View::TT provides the
interface to Template Toolkit, and provides Helpers which let us set it
up that much more easily.

=head3 Creating your View

Catalyst::View::TT provides two different helpers for us to use: TT and
TTSite.

=head4 TT

Create a basic Template Toolkit View using the provided helper script:

    script/myapp_create.pl view TT TT

This will create lib/MyApp/View/MyView.pm, which is going to be pretty
empty to start. However, it sets everything up that you need to get
started. You can now define which template you want and forward to your
view. For instance:

    sub hello : Local {
        my ( $self, $c ) = @_;

        $c->stash->{template} = 'hello.tt';

        $c->forward( $c->view('TT') );
    }

In practice you wouldn't do the forwarding manually, but would
use L<Catalyst::Action::RenderView>.

=head4 TTSite

Although the TT helper does create a functional, working view, you may
find yourself having to create the same template files and changing the
same options every time you create a new application. The TTSite helper
saves us even more time by creating the basic templates and setting some
common options for us.

Once again, you can use the helper script:

    script/myapp_create.pl view TT TTSite

This time, the helper sets several options for us in the generated View.

    __PACKAGE__->config({
        CATALYST_VAR => 'Catalyst',
        INCLUDE_PATH => [
            MyApp->path_to( 'root', 'src' ),
            MyApp->path_to( 'root', 'lib' )
        ],
        PRE_PROCESS  => 'config/main',
        WRAPPER      => 'site/wrapper',
        ERROR        => 'error.tt2',
        TIMER        => 0
    });

=over

=item

INCLUDE_PATH defines the directories that Template Toolkit should search
for the template files.

=item

PRE_PROCESS is used to process configuration options which are common to
every template file.

=item

WRAPPER is a file which is processed with each template, usually used to
easily provide a common header and footer for every page.

=back

In addition to setting these options, the TTSite helper also created the
template and config files for us! In the 'root' directory, you'll notice
two new directories: src and lib.

Several configuration files in root/lib/config are called by PRE_PROCESS.

The files in root/lib/site are the site-wide templates, called by
WRAPPER, and display the html framework, control the layout, and provide
the templates for the header and footer of your page. Using the template
organization provided makes it much easier to standardize pages and make
changes when they are (inevitably) needed.

The template files that you will create for your application will go
into root/src, and you don't need to worry about putting the the <html>
or <head> sections; just put in the content. The WRAPPER will the rest
of the page around your template for you.


=head3 $c->stash

Of course, having the template system include the header and footer for
you isn't all that we want our templates to do. We need to be able to
put data into our templates, and have it appear where and how we want
it, right? That's where the stash comes in.

In our controllers, we can add data to the stash, and then access it
from the template. For instance:

    sub hello : Local {
        my ( $self, $c ) = @_;

        $c->stash->{name} = 'Adam';

        $c->stash->{template} = 'hello.tt';

        $c->forward( $c->view('TT') );
    }

Then, in hello.tt:

    <strong>Hello, [% name %]!</strong>

When you view this page, it will display "Hello, Adam!"

All of the information in your stash is available, by its name/key, in
your templates. And your data don't have to be plain, old, boring
scalars. You can pass array references and hash references, too.

In your controller:

    sub hello : Local {
        my ( $self, $c ) = @_;

        $c->stash->{names} = [ 'Adam', 'Dave', 'John' ];

        $c->stash->{template} = 'hello.tt';

        $c->forward( $c->view('TT') );
    }

In hello.tt:

    [% FOREACH name IN names %]
        <strong>Hello, [% name %]!</strong><br />
    [% END %]

This allowed us to loop through each item in the arrayref, and display a
line for each name that we have.

This is the most basic usage, but Template Toolkit is quite powerful,
and allows you to truly keep your presentation logic separate from the
rest of your application.

=head3 $c->uri_for()

One of my favorite things about Catalyst is the ability to move an
application around without having to worry that everything is going to
break. One of the areas that used to be a problem was with the http
links in your template files. For example, suppose you have an
application installed at http://www.domain.com/Calendar. The links point
to "/Calendar", "/Calendar/2005", "/Calendar/2005/10", etc.  If you move
the application to be at http://www.mydomain.com/Tools/Calendar, then
all of those links will suddenly break.

That's where $c->uri_for() comes in. This function will merge its
parameters with either the base location for the app, or its current
namespace. Let's take a look at a couple of examples.

In your template, you can use the following:

    <a href="[% c.uri_for('/login') %]">Login Here</a>

Although the parameter starts with a forward slash, this is relative
to the application root, not the webserver root. This is important to
remember. So, if your application is installed at
http://www.domain.com/Calendar, then the link would be
http://www.mydomain.com/Calendar/Login. If you move your application
to a different domain or path, then that link will still be correct.

Likewise,

    <a href="[% c.uri_for('2005','10', '24') %]">October, 24 2005</a>

The first parameter does NOT have a forward slash, and so it will be
relative to the current namespace. If the application is installed at
http://www.domain.com/Calendar. and if the template is called from
MyApp::Controller::Display, then the link would become
http://www.domain.com/Calendar/Display/2005/10/24.

If you want to link to a parent uri of your current namespace you can
prefix the arguments with multiple '../':

    <a href="[% c.uri_for('../../view', stashed_object.id) %]">User view</a>

Once again, this allows you to move your application around without
having to worry about broken links. But there's something else, as
well. Since the links are generated by uri_for, you can use the same
template file by several different controllers, and each controller
will get the links that its supposed to. Since we believe in Don't
Repeat Yourself, this is particularly helpful if you have common
elements in your site that you want to keep in one file.

Further Reading:

L<http://search.cpan.org/perldoc?Catalyst>

L<http://search.cpan.org/perldoc?Catalyst%3A%3AView%3A%3ATT>

L<http://search.cpan.org/perldoc?Template>

=head2 Adding RSS feeds

Adding RSS feeds to your Catalyst applications is simple. We'll see two
different aproaches here, but the basic premise is that you forward to
the normal view action first to get the objects, then handle the output
differently.

=head3 Using TT templates

This is the aproach used in Agave (L<http://dev.rawmode.org/>).

    sub rss : Local {
        my ($self,$c) = @_;
        $c->forward('view');
        $c->stash->{template}='rss.tt';
    }

Then you need a template. Here's the one from Agave:

    <?xml version="1.0" encoding="UTF-8"?>
    <rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/">
      <channel>
	<title>[ [% blog.name || c.config.name || "Agave" %] ] RSS Feed</title>
        <link>[% base %]</link>
        <description>Recent posts</description>
        <language>en-us</language>
        <ttl>40</ttl>
     [% WHILE (post = posts.next) %]
      <item>
        <title>[% post.title %]</title>
        <description>[% post.formatted_teaser|html%]</description>
        <pubDate>[% post.pub_date %]</pubDate>
        <guid>[% post.full_uri %]</guid>
        <link>[% post.full_uri %]</link>
        <dc:creator>[% post.author.screenname %]</dc:creator>
      </item>
    [% END %]
      </channel>
    </rss>

=head3 Using XML::Feed

A more robust solution is to use L<XML::Feed>, as was done in the Catalyst
Advent Calendar. Assuming we have a C<view> action that populates
'entries' with some DBIx::Class iterator, the code would look something
like this:

    sub rss : Local {
        my ($self,$c) = @_;
        $c->forward('view'); # get the entries

        my $feed = XML::Feed->new('RSS');
        $feed->title( $c->config->{name} . ' RSS Feed' );
        $feed->link( $c->req->base ); # link to the site.
        $feed->description('Catalyst advent calendar'); Some description

        # Process the entries
        while( my $entry = $c->stash->{entries}->next ) {
            my $feed_entry = XML::Feed::Entry->new('RSS');
            $feed_entry->title($entry->title);
            $feed_entry->link( $c->uri_for($entry->link) );
            $feed_entry->issued( DateTime->from_epoch(epoch => $entry->created) );
            $feed->add_entry($feed_entry);
        }
        $c->res->body( $feed->as_xml );
   }

A little more code in the controller, but with this approach you're
pretty sure to get something that validates.

Note that for both of the above aproaches, you'll need to set the
content type like this:

    $c->res->content_type('application/rss+xml');

=head3 Final words

You could generalize the second variant easily by replacing 'RSS' with a
variable, so you can generate Atom feeds with the same code.

Now, go ahead and make RSS feeds for all your stuff. The world *needs*
updates on your goldfish!

=head2 Forcing the browser to download content

Sometimes you need your application to send content for download. For
example, you can generate a comma-separated values (CSV) file for your
users to download and import into their spreadsheet program.

Let's say you have an C<Orders> controller which generates a CSV file
in the C<export> action (i.e., C<http://localhost:3000/orders/export>):

    sub export : Local Args(0) {
        my ( $self, $c ) = @_;

        # In a real application, you'd generate this from the database
        my $csv = "1,5.99\n2,29.99\n3,3.99\n";

        $c->res->content_type('text/comma-separated-values');
        $c->res->body($csv);
    }

Normally the browser uses the last part of the URI to generate a
filename for data it cannot display. In this case your browser would
likely ask you to save a file named C<export>.

Luckily you can have the browser download the content with a specific
filename by setting the C<Content-Disposition> header:

    my $filename = 'Important Orders.csv';
    $c->res->header('Content-Disposition', qq[attachment; filename="$filename"]);

Note the use of quotes around the filename; this ensures that any
spaces in the filename are handled by the browser.

Put this right before calling C<< $c->res->body >> and your browser
will download a file named C<Important Orders.csv> instead of
C<export>.

You can also use this to have the browser download content which it
normally displays, such as JPEG images or even HTML. Just be sure to
set the appropriate content type and disposition.


=head1 Controllers

Controllers are the main point of communication between the web server
and your application.  Here we explore some aspects of how they work.

=head2 Action Types

=head3 Introduction

A Catalyst application is driven by one or more Controller
modules. There are a number of ways that Catalyst can decide which of
the methods in your controller modules it should call. Controller
methods are also called actions, because they determine how your
catalyst application should (re-)act to any given URL. When the
application is started up, catalyst looks at all your actions, and
decides which URLs they map to.

=head3 Type attributes

Each action is a normal method in your controller, except that it has an
L<attribute|attributes>
attached. These can be one of several types.

Assume our Controller module starts with the following package declaration:

 package MyApp::Controller::Buckets;

and we are running our application on localhost, port 3000 (the test
server default).

=over 4

=item Path

A Path attribute also takes an argument, this can be either a relative
or an absolute path. A relative path will be relative to the
controller namespace, an absolute path will represent an exact
matching URL.

 sub my_handles : Path('handles') { .. }

becomes

 http://localhost:3000/buckets/handles

and

 sub my_handles : Path('/handles') { .. }

becomes

 http://localhost:3000/handles

See also: L<Catalyst::DispatchType::Path>

=item Local

When using a Local attribute, no parameters are needed, instead, the
name of the action is matched in the URL. The namespaces created by
the name of the controller package is always part of the URL.

 sub my_handles : Local { .. }

becomes

 http://localhost:3000/buckets/my_handles

=item Global

A Global attribute is similar to a Local attribute, except that the
namespace of the controller is ignored, and matching starts at root.

 sub my_handles : Global { .. }

becomes

 http://localhost:3000/my_handles

=item Regex

By now you should have figured that a Regex attribute is just what it
sounds like. This one takes a regular expression, and matches starting
from root. These differ from the rest as they can match multiple URLs.

 sub my_handles : Regex('^handles') { .. }

matches

 http://localhost:3000/handles

and

 http://localhost:3000/handles_and_other_parts

etc.

See also: L<Catalyst::DispatchType::Regex>

=item LocalRegex

A LocalRegex is similar to a Regex, except it only matches below the current
controller namespace.

 sub my_handles : LocalRegex(^handles') { .. }

matches

 http://localhost:3000/buckets/handles

and

 http://localhost:3000/buckets/handles_and_other_parts

etc.

=item Chained

See L<Catalyst::DispatchType::Chained> for a description of how the chained
dispatch type works.

=item Private

Last but not least, there is the Private attribute, which allows you
to create your own internal actions, which can be forwarded to, but
won't be matched as URLs.

 sub my_handles : Private { .. }

becomes nothing at all..

Catalyst also predefines some special Private actions, which you can
override, these are:

=over 4

=item default

The default action will be called, if no other matching action is
found. If you don't have one of these in your namespace, or any sub
part of your namespace, you'll get an error page instead. If you want
to find out where it was the user was trying to go, you can look in
the request object using C<< $c->req->path >>.

 sub default :Path { .. }

works for all unknown URLs, in this controller namespace, or every one
if put directly into MyApp.pm.

=item index

The index action is called when someone tries to visit the exact
namespace of your controller. If index, default and matching Path
actions are defined, then index will be used instead of default and
Path.

 sub index :Path :Args(0) { .. }

becomes

 http://localhost:3000/buckets

=item begin

The begin action is called at the beginning of every request involving
this namespace directly, before other matching actions are called. It
can be used to set up variables/data for this particular part of your
app. A single begin action is called, its always the one most relevant
to the current namespace.

 sub begin : Private { .. }

is called once when

 http://localhost:3000/bucket/(anything)?

is visited.

=item end

Like begin, this action is always called for the namespace it is in,
after every other action has finished. It is commonly used to forward
processing to the View component. A single end action is called, its
always the one most relevant to the current namespace.


 sub end : Private { .. }

is called once after any actions when

 http://localhost:3000/bucket/(anything)?

is visited.

=item auto

Lastly, the auto action is magic in that B<every> auto action in the
chain of paths up to and including the ending namespace, will be
called. (In contrast, only one of the begin/end/default actions will
be called, the relevant one).

 package MyApp::Controller::Root;
 sub auto : Private { .. }

and

 sub auto : Private { .. }

will both be called when visiting

 http://localhost:3000/bucket/(anything)?

=back

=back

=head3 A word of warning

You can put root actions in your main MyApp.pm file, but this is deprecated,
please put your actions into your Root controller.

=head3 Flowchart

A graphical flowchart of how the dispatcher works can be found on the wiki at
L<http://dev.catalyst.perl.org/attachment/wiki/WikiStart/catalyst-flow.png>.

=head2 DRY Controllers with Chained actions

Imagine that you would like the following paths in your application:

=over

=item B<< /cd/<ID>/track/<ID> >>

Displays info on a particular track.

In the case of a multi-volume CD, this is the track sequence.

=item B<< /cd/<ID>/volume/<ID>/track/<ID> >>

Displays info on a track on a specific volume.

=back

Here is some example code, showing how to do this with chained controllers:

    package CD::Controller;
    use base qw/Catalyst::Controller/;

    sub root : Chained('/') PathPart('/cd') CaptureArgs(1) {
        my ($self, $c, $cd_id) = @_;
        $c->stash->{cd_id} = $cd_id;
        $c->stash->{cd} = $self->model('CD')->find_by_id($cd_id);
    }

    sub trackinfo : Chained('track') PathPart('') Args(0) RenderView {
        my ($self, $c) = @_;
    }

    package CD::Controller::ByTrackSeq;
    use base qw/CD::Controller/;

    sub track : Chained('root') PathPart('track') CaptureArgs(1) {
        my ($self, $c, $track_seq) = @_;
        $c->stash->{track} = $self->stash->{cd}->find_track_by_seq($track_seq);
    }

    package CD::Controller::ByTrackVolNo;
    use base qw/CD::Controller/;

    sub volume : Chained('root') PathPart('volume') CaptureArgs(1) {
        my ($self, $c, $volume) = @_;
        $c->stash->{volume} = $volume;
    }

    sub track : Chained('volume') PathPart('track') CaptureArgs(1) {
        my ($self, $c, $track_no) = @_;
        $c->stash->{track} = $self->stash->{cd}->find_track_by_vol_and_track_no(
            $c->stash->{volume}, $track_no
        );
    }

Note that adding other actions (i.e. chain endpoints) which operate on a track
is simply a matter of adding a new sub to CD::Controller - no code is duplicated,
even though there are two different methods of looking up a track.

This technique can be expanded as needed to fulfil your requirements - for example,
if you inherit the first action of a chain from a base class, then mixing in a
different base class can be used to duplicate an entire URL hieratchy at a different
point within your application.

=head2 Component-based Subrequests

See L<Catalyst::Plugin::SubRequest>.

=head2 File uploads

=head3 Single file upload with Catalyst

To implement uploads in Catalyst, you need to have a HTML form similar to
this:

    <form action="/upload" method="post" enctype="multipart/form-data">
      <input type="hidden" name="form_submit" value="yes">
      <input type="file" name="my_file">
      <input type="submit" value="Send">
    </form>

It's very important not to forget C<enctype="multipart/form-data"> in
the form.

Catalyst Controller module 'upload' action:

    sub upload : Global {
        my ($self, $c) = @_;

        if ( $c->request->parameters->{form_submit} eq 'yes' ) {

            if ( my $upload = $c->request->upload('my_file') ) {

                my $filename = $upload->filename;
                my $target   = "/tmp/upload/$filename";

                unless ( $upload->link_to($target) || $upload->copy_to($target) ) {
                    die( "Failed to copy '$filename' to '$target': $!" );
                }
            }
        }

        $c->stash->{template} = 'file_upload.html';
    }

=head3 Multiple file upload with Catalyst

Code for uploading multiple files from one form needs a few changes:

The form should have this basic structure:

    <form action="/upload" method="post" enctype="multipart/form-data">
      <input type="hidden" name="form_submit" value="yes">
      <input type="file" name="file1" size="50"><br>
      <input type="file" name="file2" size="50"><br>
      <input type="file" name="file3" size="50"><br>
      <input type="submit" value="Send">
    </form>

And in the controller:

    sub upload : Local {
        my ($self, $c) = @_;

        if ( $c->request->parameters->{form_submit} eq 'yes' ) {

            for my $field ( $c->req->upload ) {

                my $upload   = $c->req->upload($field);
                my $filename = $upload->filename;
                my $target   = "/tmp/upload/$filename";

                unless ( $upload->link_to($target) || $upload->copy_to($target) ) {
                    die( "Failed to copy '$filename' to '$target': $!" );
                }
            }
        }

        $c->stash->{template} = 'file_upload.html';
    }

C<for my $field ($c-E<gt>req->upload)> loops automatically over all file
input fields and gets input names. After that is basic file saving code,
just like in single file upload.

Notice: C<die>ing might not be what you want to do, when an error
occurs, but it works as an example. A better idea would be to store
error C<$!> in $c->stash->{error} and show a custom error template
displaying this message.

For more information about uploads and usable methods look at
L<Catalyst::Request::Upload> and L<Catalyst::Request>.

=head2 Forwarding with arguments

Sometimes you want to pass along arguments when forwarding to another
action. As of version 5.30, arguments can be passed in the call to
C<forward>; in earlier versions, you can manually set the arguments in
the Catalyst Request object:

  # version 5.30 and later:
  $c->forward('/wherever', [qw/arg1 arg2 arg3/]);

  # pre-5.30
  $c->req->args([qw/arg1 arg2 arg3/]);
  $c->forward('/wherever');

(See the L<Catalyst::Manual::Intro> Flow_Control section for more
information on passing arguments via C<forward>.)

=head2 Chained dispatch using base classes, and inner packages.

  package MyApp::Controller::Base;
  use base qw/Catalyst::Controller/;

  sub key1 : Chained('/')

=head2 Extending RenderView (formerly DefaultEnd)

The recommended approach for an C<end> action is to use
L<Catalyst::Action::RenderView> (taking the place of
L<Catalyst::Plugin::DefaultEnd>), which does what you usually need.
However there are times when you need to add a bit to it, but don't want
to write your own C<end> action.

You can extend it like this:

To add something to an C<end> action that is called before rendering
(this is likely to be what you want), simply place it in the C<end>
method:

    sub end : ActionClass('RenderView') {
      my ( $self, $c ) = @_;
      # do stuff here; the RenderView action is called afterwards
    }

To add things to an C<end> action that are called I<after> rendering,
you can set it up like this:

    sub render : ActionClass('RenderView') { }

    sub end : Private {
      my ( $self, $c ) = @_;
      $c->forward('render');
      # do stuff here
    }



=head1 Deployment

The recipes below describe aspects of the deployment process,
including web server engines and tips to improve application efficiency.

=head2 mod_perl Deployment

mod_perl is the best solution for many applications, but we'll list some pros
and cons so you can decide for yourself.  The other production deployment
option is FastCGI, for which see below.

=head3 Pros

=head4 Speed

mod_perl is very fast and your app will benefit from being loaded in memory
within each Apache process.

=head4 Shared memory for multiple apps

If you need to run several Catalyst apps on the same server, mod_perl will
share the memory for common modules.

=head3 Cons

=head4 Memory usage

Since your application is fully loaded in memory, every Apache process will
be rather large.  This means a large Apache process will be tied up while
serving static files, large files, or dealing with slow clients.  For this
reason, it is best to run a two-tiered web architecture with a lightweight
frontend server passing dynamic requests to a large backend mod_perl
server.

=head4 Reloading

Any changes made to the core code of your app require a full Apache restart.
Catalyst does not support Apache::Reload or StatINC.  This is another good
reason to run a frontend web server where you can set up an
C<ErrorDocument 502> page to report that your app is down for maintenance.

=head4 Cannot run multiple versions of the same app

It is not possible to run two different versions of the same application in
the same Apache instance because the namespaces will collide.

=head4 Setup

Now that we have that out of the way, let's talk about setting up mod_perl
to run a Catalyst app.

=head4 1. Install Catalyst::Engine::Apache

You should install the latest versions of both Catalyst and
Catalyst::Engine::Apache.  The Apache engines were separated from the
Catalyst core in version 5.50 to allow for updates to the engine without
requiring a new Catalyst release.

=head4 2. Install Apache with mod_perl

Both Apache 1.3 and Apache 2 are supported, although Apache 2 is highly
recommended.  With Apache 2, make sure you are using the prefork MPM and not
the worker MPM.  The reason for this is that many Perl modules are not
thread-safe and may have problems running within the threaded worker
environment.  Catalyst is thread-safe however, so if you know what you're
doing, you may be able to run using worker.

In Debian, the following commands should get you going.

    apt-get install apache2-mpm-prefork
    apt-get install libapache2-mod-perl2

=head4 3. Configure your application

Every Catalyst application will automagically become a mod_perl handler
when run within mod_perl.  This makes the configuration extremely easy.
Here is a basic Apache 2 configuration.

    PerlSwitches -I/var/www/MyApp/lib
    PerlModule MyApp

    <Location />
        SetHandler          modperl
        PerlResponseHandler MyApp
    </Location>

The most important line here is C<PerlModule MyApp>.  This causes mod_perl
to preload your entire application into shared memory, including all of your
controller, model, and view classes and configuration.  If you have -Debug
mode enabled, you will see the startup output scroll by when you first
start Apache.

For an example Apache 1.3 configuration, please see the documentation for
L<Catalyst::Engine::Apache::MP13>.

=head3 Test It

That's it, your app is now a full-fledged mod_perl application!  Try it out
by going to http://your.server.com/.

=head3 Other Options

=head4 Non-root location

You may not always want to run your app at the root of your server or virtual
host.  In this case, it's a simple change to run at any non-root location
of your choice.

    <Location /myapp>
        SetHandler          modperl
        PerlResponseHandler MyApp
    </Location>

When running this way, it is best to make use of the C<uri_for> method in
Catalyst for constructing correct links.

=head4 Static file handling

Static files can be served directly by Apache for a performance boost.

    DocumentRoot /var/www/MyApp/root
    <Location /static>
        SetHandler default-handler
    </Location>

This will let all files within root/static be handled directly by Apache.  In
a two-tiered setup, the frontend server should handle static files.
The configuration to do this on the frontend will vary.

The same is accomplished in lighttpd with the following snippet:

   $HTTP["url"] !~ "^/(?:img/|static/|css/|favicon.ico$)" {
         fastcgi.server = (
             "" => (
                 "MyApp" => (
                     "socket"       => "/tmp/myapp.socket",
                     "check-local"  => "disable",
                 )
             )
         )
    }

Which serves everything in the img, static, css directories
statically, as well as the favicon file.

Note the path of the application needs to be stated explicitly in the
web server configuration for both these recipes.

=head2 Catalyst on shared hosting

So, you want to put your Catalyst app out there for the whole world to
see, but you don't want to break the bank. There is an answer - if you
can get shared hosting with FastCGI and a shell, you can install your
Catalyst app in a local directory on your shared host. First, run

    perl -MCPAN -e shell

and go through the standard CPAN configuration process. Then exit out
without installing anything. Next, open your .bashrc and add

    export PATH=$HOME/local/bin:$HOME/local/script:$PATH
    perlversion=`perl -v | grep 'built for' | awk '{print $4}' | sed -e 's/v//;'`
    export PERL5LIB=$HOME/local/share/perl/$perlversion:$HOME/local/lib/perl/$perlversion:$HOME/local/lib:$PERL5LIB

and log out, then back in again (or run C<". .bashrc"> if you
prefer). Finally, edit C<.cpan/CPAN/MyConfig.pm> and add

    'make_install_arg' => qq[SITEPREFIX=$ENV{HOME}/local],
    'makepl_arg' => qq[INSTALLDIRS=site install_base=$ENV{HOME}/local],

Now you can install the modules you need using CPAN as normal; they
will be installed into your local directory, and perl will pick them
up. Finally, change directory into the root of your virtual host and
symlink your application's script directory in:

    cd path/to/mydomain.com
    ln -s ~/lib/MyApp/script script

And add the following lines to your .htaccess file (assuming the server
is setup to handle .pl as fcgi - you may need to rename the script to
myapp_fastcgi.fcgi and/or use a SetHandler directive):

  RewriteEngine On
  RewriteCond %{REQUEST_URI} !^/?script/myapp_fastcgi.pl
  RewriteRule ^(.*)$ script/myapp_fastcgi.pl/$1 [PT,L]

Now C<http://mydomain.com/> should now Just Work. Congratulations, now
you can tell your friends about your new website (or in our case, tell
the client it's time to pay the invoice :) )

=head2 FastCGI Deployment

FastCGI is a high-performance extension to CGI. It is suitable
for production environments.

=head3 Pros

=head4 Speed

FastCGI performs equally as well as mod_perl.  Don't let the 'CGI' fool you;
your app runs as multiple persistent processes ready to receive connections
from the web server.

=head4 App Server

When using external FastCGI servers, your application runs as a standalone
application server.  It may be restarted independently from the web server.
This allows for a more robust environment and faster reload times when
pushing new app changes.  The frontend server can even be configured to
display a friendly "down for maintenance" page while the application is
restarting.

=head4 Load-balancing

You can launch your application on multiple backend servers and allow the
frontend web server to load-balance between all of them.  And of course, if
one goes down, your app continues to run fine.

=head4 Multiple versions of the same app

Each FastCGI application is a separate process, so you can run different
versions of the same app on a single server.

=head4 Can run with threaded Apache

Since your app is not running inside of Apache, the faster mpm_worker module
can be used without worrying about the thread safety of your application.

=head3 Cons

You may have to disable mod_deflate.  If you experience page hangs with
mod_fastcgi then remove deflate.load and deflate.conf from mods-enabled/

=head4 More complex environment

With FastCGI, there are more things to monitor and more processes running
than when using mod_perl.

=head3 Setup

=head4 1. Install Apache with mod_fastcgi

mod_fastcgi for Apache is a third party module, and can be found at
L<http://www.fastcgi.com/>.  It is also packaged in many distributions,
for example, libapache2-mod-fastcgi in Debian. You will also need to install
the L<FCGI> module from cpan.

Important Note! If you experience difficulty properly rendering pages,
try disabling Apache's mod_deflate (Deflate Module), e.g. 'a2dismod deflate'.

=head4 2. Configure your application

    # Serve static content directly
    DocumentRoot  /var/www/MyApp/root
    Alias /static /var/www/MyApp/root/static

    FastCgiServer /var/www/MyApp/script/myapp_fastcgi.pl -processes 3
    Alias /myapp/ /var/www/MyApp/script/myapp_fastcgi.pl/

    # Or, run at the root
    Alias / /var/www/MyApp/script/myapp_fastcgi.pl/

The above commands will launch 3 app processes and make the app available at
/myapp/

=head3 Standalone server mode

While not as easy as the previous method, running your app as an external
server gives you much more flexibility.

First, launch your app as a standalone server listening on a socket.

    script/myapp_fastcgi.pl -l /tmp/myapp.socket -n 5 -p /tmp/myapp.pid -d

You can also listen on a TCP port if your web server is not on the same
machine.

    script/myapp_fastcgi.pl -l :8080 -n 5 -p /tmp/myapp.pid -d

You will probably want to write an init script to handle starting/stopping
of the app using the pid file.

Now, we simply configure Apache to connect to the running server.

    # 502 is a Bad Gateway error, and will occur if the backend server is down
    # This allows us to display a friendly static page that says "down for
    # maintenance"
    Alias /_errors /var/www/MyApp/root/error-pages
    ErrorDocument 502 /_errors/502.html

    FastCgiExternalServer /tmp/myapp.fcgi -socket /tmp/myapp.socket
    Alias /myapp/ /tmp/myapp.fcgi/

    # Or, run at the root
    Alias / /tmp/myapp.fcgi/

=head3 More Info

L<Catalyst::Engine::FastCGI>.

=head2 Development server deployment

The development server is a mini web server written in perl.  If you
expect a low number of hits or you don't need mod_perl/FastCGI speed,
you could use the development server as the application server with a
lightweight proxy web server at the front.  However, consider using
L<Catalyst::Engine::HTTP::Prefork> for this kind of deployment instead, since
it can better handle multiple concurrent requests without forking, or can
prefork a set number of servers for improved performance.

=head3 Pros

As this is an application server setup, the pros are the same as
FastCGI (with the exception of speed).
It is also:

=head4 Simple

The development server is what you create your code on, so if it works
here, it should work in production!

=head3 Cons

=head4 Speed

Not as fast as mod_perl or FastCGI. Needs to fork for each request
that comes in - make sure static files are served by the web server to
save forking.

=head3 Setup

=head4 Start up the development server

   script/myapp_server.pl -p 8080 -k  -f -pidfile=/tmp/myapp.pid

You will probably want to write an init script to handle stop/starting
the app using the pid file.

=head4 Configuring Apache

Make sure mod_proxy is enabled and add:

    # Serve static content directly
    DocumentRoot /var/www/MyApp/root
    Alias /static /var/www/MyApp/root/static

    ProxyRequests Off
    <Proxy *>
        Order deny,allow
        Allow from all
    </Proxy>

    # Need to specifically stop these paths from being passed to proxy
    ProxyPass /static !
    ProxyPass /favicon.ico !

    ProxyPass / http://localhost:8080/
    ProxyPassReverse / http://localhost:8080/

    # This is optional if you'd like to show a custom error page
    # if the proxy is not available
    ErrorDocument 502 /static/error_pages/http502.html

You can wrap the above within a VirtualHost container if you want
different apps served on the same host.

=head2 Quick deployment: Building PAR Packages

You have an application running on your development box, but then you
have to quickly move it to another one for
demonstration/deployment/testing...

PAR packages can save you from a lot of trouble here. They are usual Zip
files that contain a blib tree; you can even include all prereqs and a
perl interpreter by setting a few flags!

=head3 Follow these few points to try it out!

1. Install Catalyst and PAR 0.89 (or later)

    % perl -MCPAN -e 'install Catalyst'
    ...
    % perl -MCPAN -e 'install PAR'
    ...

2. Create a application

    % catalyst.pl MyApp
    ...
    % cd MyApp

Recent versions of Catalyst (5.62 and up) include
L<Module::Install::Catalyst>, which simplifies the process greatly.  From the shell in your application directory:

    % perl Makefile.PL
    % make catalyst_par

You can customise the PAR creation process by special "catalyst_par_*" commands
available from L<Module::Install::Catalyst>. You can add these commands in your
Makefile.PL just before the line containing "catalyst;"

    #Makefile.PL example with extra PAR options
    use inc::Module::Install;

    name 'MyApp';
    all_from 'lib\MyApp.pm';

    requires 'Catalyst::Runtime' => '5.80005';
    <snip>
    ...
    <snip>

    catalyst_par_core(1);      # bundle perl core modules in the resulting PAR
    catalyst_par_multiarch(1); # build a multi-architecture PAR file
    catalyst_par_classes(qw/
      Some::Additional::Module
      Some::Other::Module
    /);          # specify additional modules you want to be included into PAR
    catalyst;

    install_script glob('script/*.pl');
    auto_install;
    WriteAll;

Congratulations! Your package "myapp.par" is ready, the following
steps are just optional.

3. Test your PAR package with "parl" (no typo)

    % parl myapp.par
    Usage:
        [parl] myapp[.par] [script] [arguments]

      Examples:
        parl myapp.par myapp_server.pl -r
        myapp myapp_cgi.pl

      Available scripts:
        myapp_cgi.pl
        myapp_create.pl
        myapp_fastcgi.pl
        myapp_server.pl
        myapp_test.pl

    % parl myapp.par myapp_server.pl
    You can connect to your server at http://localhost:3000

Yes, this nifty little starter application gets automatically included.
You can also use "catalyst_par_script('myapp_server.pl')" to set a
default script to execute.

6. Want to create a binary that includes the Perl interpreter?

    % pp -o myapp myapp.par
    % ./myapp myapp_server.pl
    You can connect to your server at http://localhost:3000

=head2 Serving static content

Serving static content in Catalyst used to be somewhat tricky; the use
of L<Catalyst::Plugin::Static::Simple> makes everything much easier.
This plugin will automatically serve your static content during development,
but allows you to easily switch to Apache (or other server) in a
production environment.

=head3 Introduction to Static::Simple

Static::Simple is a plugin that will help to serve static content for your
application. By default, it will serve most types of files, excluding some
standard Template Toolkit extensions, out of your B<root> file directory. All
files are served by path, so if B<images/me.jpg> is requested, then
B<root/images/me.jpg> is found and served.

=head3 Usage

Using the plugin is as simple as setting your use line in MyApp.pm to include:

 use Catalyst qw/Static::Simple/;

and already files will be served.

=head3 Configuring

Static content is best served from a single directory within your root
directory. Having many different directories such as C<root/css> and
C<root/images> requires more code to manage, because you must separately
identify each static directory--if you decide to add a C<root/js>
directory, you'll need to change your code to account for it. In
contrast, keeping all static directories as subdirectories of a main
C<root/static> directory makes things much easier to manage. Here's an
example of a typical root directory structure:

    root/
    root/content.tt
    root/controller/stuff.tt
    root/header.tt
    root/static/
    root/static/css/main.css
    root/static/images/logo.jpg
    root/static/js/code.js


All static content lives under C<root/static>, with everything else being
Template Toolkit files.

=over 4

=item Include Path

You may of course want to change the default locations, and make
Static::Simple look somewhere else, this is as easy as:

 MyApp->config->{static}->{include_path} = [
  MyApp->config->{root},
  '/path/to/my/files'
 ];

When you override include_path, it will not automatically append the
normal root path, so you need to add it yourself if you still want
it. These will be searched in order given, and the first matching file
served.

=item Static directories

If you want to force some directories to be only static, you can set
them using paths relative to the root dir, or regular expressions:

 MyApp->config->{static}->{dirs} = [
   'static',
   qr/^(images|css)/,
 ];

=item File extensions

By default, the following extensions are not served (that is, they will
be processed by Catalyst): B<tmpl, tt, tt2, html, xhtml>. This list can
be replaced easily:

 MyApp->config->{static}->{ignore_extensions} = [
    qw/tmpl tt tt2 html xhtml/
 ];

=item Ignoring directories

Entire directories can be ignored. If used with include_path,
directories relative to the include_path dirs will also be ignored:

 MyApp->config->{static}->{ignore_dirs} = [ qw/tmpl css/ ];

=back

=head3 More information

L<http://search.cpan.org/dist/Catalyst-Plugin-Static-Simple/>

=head3 Serving manually with the Static plugin with HTTP::Daemon (myapp_server.pl)

In some situations you might want to control things more directly,
using L<Catalyst::Plugin::Static>.

In your main application class (MyApp.pm), load the plugin:

    use Catalyst qw/-Debug FormValidator Static OtherPlugin/;

You will also need to make sure your end method does I<not> forward
static content to the view, perhaps like this:

    sub end : Private {
        my ( $self, $c ) = @_;

        $c->forward( 'MyApp::View::TT' )
          unless ( $c->res->body || !$c->stash->{template} );
    }

This code will only forward to the view if a template has been
previously defined by a controller and if there is not already data in
C<$c-E<gt>res-E<gt>body>.

Next, create a controller to handle requests for the /static path. Use
the Helper to save time. This command will create a stub controller as
C<lib/MyApp/Controller/Static.pm>.

    $ script/myapp_create.pl controller Static

Edit the file and add the following methods:

    # serve all files under /static as static files
    sub default : Path('/static') {
        my ( $self, $c ) = @_;

        # Optional, allow the browser to cache the content
        $c->res->headers->header( 'Cache-Control' => 'max-age=86400' );

        $c->serve_static; # from Catalyst::Plugin::Static
    }

    # also handle requests for /favicon.ico
    sub favicon : Path('/favicon.ico') {
        my ( $self, $c ) = @_;

        $c->serve_static;
    }

You can also define a different icon for the browser to use instead of
favicon.ico by using this in your HTML header:

    <link rel="icon" href="/static/myapp.ico" type="image/x-icon" />

=head3 Common problems with the Static plugin

The Static plugin makes use of the C<shared-mime-info> package to
automatically determine MIME types. This package is notoriously
difficult to install, especially on win32 and OS X. For OS X the easiest
path might be to install Fink, then use C<apt-get install
shared-mime-info>. Restart the server, and everything should be fine.

Make sure you are using the latest version (>= 0.16) for best
results. If you are having errors serving CSS files, or if they get
served as text/plain instead of text/css, you may have an outdated
shared-mime-info version. You may also wish to simply use the following
code in your Static controller:

    if ($c->req->path =~ /css$/i) {
        $c->serve_static( "text/css" );
    } else {
        $c->serve_static;
    }

=head3 Serving Static Files with Apache

When using Apache, you can bypass Catalyst and any Static
plugins/controllers controller by intercepting requests for the
C<root/static> path at the server level. All that is required is to
define a DocumentRoot and add a separate Location block for your static
content. Here is a complete config for this application under mod_perl
1.x:

    <Perl>
        use lib qw(/var/www/MyApp/lib);
    </Perl>
    PerlModule MyApp

    <VirtualHost *>
        ServerName myapp.example.com
        DocumentRoot /var/www/MyApp/root
        <Location />
            SetHandler perl-script
            PerlHandler MyApp
        </Location>
        <LocationMatch "/(static|favicon.ico)">
            SetHandler default-handler
        </LocationMatch>
    </VirtualHost>

And here's a simpler example that'll get you started:

    Alias /static/ "/my/static/files/"
    <Location "/static">
        SetHandler none
    </Location>

=head2 Caching

Catalyst makes it easy to employ several different types of caching to
speed up your applications.

=head3 Cache Plugins

There are three wrapper plugins around common CPAN cache modules:
Cache::FastMmap, Cache::FileCache, and Cache::Memcached.  These can be
used to cache the result of slow operations.

The Catalyst Advent Calendar uses the FileCache plugin to cache the
rendered XHTML version of the source POD document.  This is an ideal
application for a cache because the source document changes
infrequently but may be viewed many times.

    use Catalyst qw/Cache::FileCache/;

    ...

    use File::stat;
    sub render_pod : Local {
        my ( self, $c ) = @_;

        # the cache is keyed on the filename and the modification time
        # to check for updates to the file.
        my $file  = $c->path_to( 'root', '2005', '11.pod' );
        my $mtime = ( stat $file )->mtime;

        my $cached_pod = $c->cache->get("$file $mtime");
        if ( !$cached_pod ) {
            $cached_pod = do_slow_pod_rendering();
            # cache the result for 12 hours
            $c->cache->set( "$file $mtime", $cached_pod, '12h' );
        }
        $c->stash->{pod} = $cached_pod;
    }

We could actually cache the result forever, but using a value such as 12 hours
allows old entries to be automatically expired when they are no longer needed.

=head3 Page Caching

Another method of caching is to cache the entire HTML page.  While this is
traditionally handled by a front-end proxy server like Squid, the Catalyst
PageCache plugin makes it trivial to cache the entire output from
frequently-used or slow actions.

Many sites have a busy content-filled front page that might look something
like this.  It probably takes a while to process, and will do the exact same
thing for every single user who views the page.

    sub front_page : Path('/') {
        my ( $self, $c ) = @_;

        $c->forward( 'get_news_articles' );
        $c->forward( 'build_lots_of_boxes' );
        $c->forward( 'more_slow_stuff' );

        $c->stash->{template} = 'index.tt';
    }

We can add the PageCache plugin to speed things up.

    use Catalyst qw/Cache::FileCache PageCache/;

    sub front_page : Path ('/') {
        my ( $self, $c ) = @_;

        $c->cache_page( 300 );

        # same processing as above
    }

Now the entire output of the front page, from <html> to </html>, will be
cached for 5 minutes.  After 5 minutes, the next request will rebuild the
page and it will be re-cached.

Note that the page cache is keyed on the page URI plus all parameters, so
requests for / and /?foo=bar will result in different cache items.  Also,
only GET requests will be cached by the plugin.

You can even get that front-end Squid proxy to help out by enabling HTTP
headers for the cached page.

    MyApp->config->{page_cache}->{set_http_headers} = 1;

This would now set the following headers so proxies and browsers may cache
the content themselves.

    Cache-Control: max-age=($expire_time - time)
    Expires: $expire_time
    Last-Modified: $cache_created_time

=head3 Template Caching

Template Toolkit provides support for caching compiled versions of your
templates.  To enable this in Catalyst, use the following configuration.
TT will cache compiled templates keyed on the file mtime, so changes will
still be automatically detected.

    package MyApp::View::TT;

    use strict;
    use warnings;
    use base 'Catalyst::View::TT';

    __PACKAGE__->config(
        COMPILE_DIR => '/tmp/template_cache',
    );

    1;

=head3 More Info

See the documentation for each cache plugin for more details and other
available configuration options.

L<Catalyst::Plugin::Cache::FastMmap>
L<Catalyst::Plugin::Cache::FileCache>
L<Catalyst::Plugin::Cache::Memcached>
L<Catalyst::Plugin::PageCache>
L<http://search.cpan.org/dist/Template-Toolkit/lib/Template/Manual/Config.pod#Caching_and_Compiling_Options>

=head1 Testing

Testing is an integral part of the web application development
process.  Tests make multi developer teams easier to coordinate, and
they help ensure that there are no nasty surprises after upgrades or
alterations.

=head2 Testing

Catalyst provides a convenient way of testing your application during
development and before deployment in a real environment.

C<Catalyst::Test> makes it possible to run the same tests both locally
(without an external daemon) and against a remote server via HTTP.

=head3 Tests

Let's examine a skeleton application's C<t/> directory:

    mundus:~/MyApp chansen$ ls -l t/
    total 24
    -rw-r--r--  1 chansen  chansen   95 18 Dec 20:50 01app.t
    -rw-r--r--  1 chansen  chansen  190 18 Dec 20:50 02pod.t
    -rw-r--r--  1 chansen  chansen  213 18 Dec 20:50 03podcoverage.t

=over 4

=item C<01app.t>

Verifies that the application loads, compiles, and returns a successful
response.

=item C<02pod.t>

Verifies that all POD is free from errors. Only executed if the C<TEST_POD>
environment variable is true.

=item C<03podcoverage.t>

Verifies that all methods/functions have POD coverage. Only executed if the
C<TEST_POD> environment variable is true.

=back

=head3 Creating tests

    mundus:~/MyApp chansen$ cat t/01app.t | perl -ne 'printf( "%2d  %s", $., $_ )'
    1  use Test::More tests => 2;
    2  use_ok( Catalyst::Test, 'MyApp' );
    3
    4  ok( request('/')->is_success );

The first line declares how many tests we are going to run, in this case
two. The second line tests and loads our application in test mode. The
fourth line verifies that our application returns a successful response.

C<Catalyst::Test> exports two functions, C<request> and C<get>. Each can
take three different arguments:

=over 4

=item A string which is a relative or absolute URI.

    request('/my/path');
    request('http://www.host.com/my/path');

=item An instance of C<URI>.

    request( URI->new('http://www.host.com/my/path') );

=item An instance of C<HTTP::Request>.

    request( HTTP::Request->new( GET => 'http://www.host.com/my/path') );

=back

C<request> returns an instance of C<HTTP::Response> and C<get> returns the
content (body) of the response.

=head3 Running tests locally

    mundus:~/MyApp chansen$ CATALYST_DEBUG=0 TEST_POD=1 prove --lib lib/ t/
    t/01app............ok
    t/02pod............ok
    t/03podcoverage....ok
    All tests successful.
    Files=3, Tests=4,  2 wallclock secs ( 1.60 cusr +  0.36 csys =  1.96 CPU)

C<CATALYST_DEBUG=0> ensures that debugging is off; if it's enabled you
will see debug logs between tests.

C<TEST_POD=1> enables POD checking and coverage.

C<prove> A command-line tool that makes it easy to run tests. You can
find out more about it from the links below.

=head3 Running tests remotely

    mundus:~/MyApp chansen$ CATALYST_SERVER=http://localhost:3000/ prove --lib lib/ t/01app.t
    t/01app....ok
    All tests successful.
    Files=1, Tests=2,  0 wallclock secs ( 0.40 cusr +  0.01 csys =  0.41 CPU)

C<CATALYST_SERVER=http://localhost:3000/> is the absolute deployment URI of
your application. In C<CGI> or C<FastCGI> it should be the host and path
to the script.

=head3 C<Test::WWW::Mechanize> and Catalyst

Be sure to check out C<Test::WWW::Mechanize::Catalyst>. It makes it easy to
test HTML, forms and links. A short example of usage:

    use Test::More tests => 6;
    use_ok( Test::WWW::Mechanize::Catalyst, 'MyApp' );

    my $mech = Test::WWW::Mechanize::Catalyst->new;
    $mech->get_ok("http://localhost/", 'Got index page');
    $mech->title_like( qr/^MyApp on Catalyst/, 'Got right index title' );
    ok( $mech->find_link( text_regex => qr/^Wiki/i ), 'Found link to Wiki' );
    ok( $mech->find_link( text_regex => qr/^Mailing-List/i ), 'Found link to Mailing-List' );
    ok( $mech->find_link( text_regex => qr/^IRC channel/i ), 'Found link to IRC channel' );

=head3 Further Reading

=over 4

=item Catalyst::Test

L<http://search.cpan.org/dist/Catalyst/lib/Catalyst/Test.pm>

=item Test::WWW::Mechanize::Catalyst

L<http://search.cpan.org/dist/Test-WWW-Mechanize-Catalyst/lib/Test/WWW/Mechanize/Catalyst.pm>

=item Test::WWW::Mechanize

L<http://search.cpan.org/dist/Test-WWW-Mechanize/Mechanize.pm>

=item WWW::Mechanize

L<http://search.cpan.org/dist/WWW-Mechanize/lib/WWW/Mechanize.pm>

=item LWP::UserAgent

L<http://search.cpan.org/dist/libwww-perl/lib/LWP/UserAgent.pm>

=item HTML::Form

L<http://search.cpan.org/dist/libwww-perl/lib/HTML/Form.pm>

=item HTTP::Message

L<http://search.cpan.org/dist/libwww-perl/lib/HTTP/Message.pm>

=item HTTP::Request

L<http://search.cpan.org/dist/libwww-perl/lib/HTTP/Request.pm>

=item HTTP::Request::Common

L<http://search.cpan.org/dist/libwww-perl/lib/HTTP/Request/Common.pm>

=item HTTP::Response

L<http://search.cpan.org/dist/libwww-perl/lib/HTTP/Response.pm>

=item HTTP::Status

L<http://search.cpan.org/dist/libwww-perl/lib/HTTP/Status.pm>

=item URI

L<http://search.cpan.org/dist/URI/URI.pm>

=item Test::More

L<http://search.cpan.org/dist/Test-Simple/lib/Test/More.pm>

=item Test::Pod

L<http://search.cpan.org/dist/Test-Pod/Pod.pm>

=item Test::Pod::Coverage

L<http://search.cpan.org/dist/Test-Pod-Coverage/Coverage.pm>

=item prove (Test::Harness)

L<http://search.cpan.org/dist/Test-Harness/bin/prove>

=back

=head3 More Information

L<http://search.cpan.org/perldoc?Catalyst::Plugin::Authorization::Roles>
L<http://search.cpan.org/perldoc?Catalyst::Plugin::Authorization::ACL>

=head1 AUTHORS

Catalyst Contributors, see Catalyst.pm

=head1 COPYRIGHT

This library is free software. You can redistribute it and/or modify it under
the same terms as Perl itself.

=cut