The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.

=encoding utf8

=head1 NAME

Mojolicious::Guides::Tutorial - Get started with Mojolicious

=head1 TUTORIAL

A quick example-driven introduction to the wonders of L<Mojolicious::Lite>.
Almost everything you'll learn here also applies to full L<Mojolicious>
applications.

This is only the first of the L<Mojolicious::Guides>. Other guides delve deeper
into topics like L<growing|Mojolicious::Guides::Growing> a L<Mojolicious::Lite>
prototype into a well-structured L<Mojolicious> application,
L<routing|Mojolicious::Guides::Routing>,
L<rendering|Mojolicious::Guides::Rendering> and more. It is highly encouraged
that readers continue on to the remaining guides after reading this one.

=head2 Hello World

A simple Hello World application can look like this, L<strict>, L<warnings>,
L<utf8> and Perl 5.10 L<features|feature> are automatically enabled and a few
L<functions|Mojolicious::Lite/"FUNCTIONS"> imported, when you use
L<Mojolicious::Lite>, turning your script into a full featured web application.

  #!/usr/bin/env perl
  use Mojolicious::Lite;

  get '/' => sub {
    my $c = shift;
    $c->render(text => 'Hello World!');
  };

  app->start;

With L<Mojolicious::Command::generate::lite_app> there is also a helper command
to generate a small example application.

  $ mojo generate lite_app myapp.pl

=head2 Commands

Many different L<commands|Mojolicious::Commands/"COMMANDS"> are automatically
available from the command line. CGI and L<PSGI> environments can even be
detected and will usually just work without commands.

  $ ./myapp.pl daemon
  Server available at http://127.0.0.1:3000

  $ ./myapp.pl daemon -l http://*:8080
  Server available at http://127.0.0.1:8080

  $ ./myapp.pl cgi
  ...CGI output...

  $ ./myapp.pl get /
  Hello World!

  $ ./myapp.pl
  ...List of available commands (or automatically detected environment)...

A call to L<Mojolicious/"start"> (C<app-E<gt>start>), which starts the command
system, should be the last expression in your application, because its return
value can be significant.

  # Use @ARGV to pick a command
  app->start;

  # Start the "daemon" command
  app->start('daemon', '-l', 'http://*:8080');

=head2 Reloading

Your application will automatically reload itself if you start it with the
L<morbo> development web server, so you don't have to restart the server after
every change.

  $ morbo ./myapp.pl
  Server available at http://127.0.0.1:3000

For more information about how to deploy your application see also
L<Mojolicious::Guides::Cookbook/"DEPLOYMENT">.

=head2 Routes

Routes are basically just fancy paths that can contain different kinds of
placeholders and usually lead to an action, if they match the path part of the
request URL. The first argument passed to all actions (C<$c>) is a
L<Mojolicious::Controller> object, containing both the HTTP request and
response.

  use Mojolicious::Lite;

  # Route leading to an action that renders some text
  get '/foo' => sub {
    my $c = shift;
    $c->render(text => 'Hello World!');
  };

  app->start;

Response content is often generated by actions with
L<Mojolicious::Controller/"render">, but more about that later.

=head2 GET/POST parameters

All C<GET> and C<POST> parameters sent with the request are accessible via
L<Mojolicious::Controller/"param">.

  use Mojolicious::Lite;

  # /foo?user=sri
  get '/foo' => sub {
    my $c    = shift;
    my $user = $c->param('user');
    $c->render(text => "Hello $user.");
  };

  app->start;

=head2 Stash and templates

The L<Mojolicious::Controller/"stash"> is used to pass data to templates, which
can be inlined in the C<DATA> section. A few stash values like C<template>,
C<text> and C<data> are reserved and will be used by
L<Mojolicious::Controller/"render"> to decide how a response should be
generated.

  use Mojolicious::Lite;

  # Route leading to an action that renders a template
  get '/foo' => sub {
    my $c = shift;
    $c->stash(one => 23);
    $c->render(template => 'magic', two => 24);
  };

  app->start;
  __DATA__

  @@ magic.html.ep
  The magic numbers are <%= $one %> and <%= $two %>.

For more information about templates see also
L<Mojolicious::Guides::Rendering/"Embedded Perl">.

=head2 HTTP

L<Mojolicious::Controller/"req"> and L<Mojolicious::Controller/"res"> give you
full access to all HTTP features and information.

  use Mojolicious::Lite;

  # Access request information
  get '/agent' => sub {
    my $c    = shift;
    my $host = $c->req->url->to_abs->host;
    my $ua   = $c->req->headers->user_agent;
    $c->render(text => "Request by $ua reached $host.");
  };

  # Echo the request body and send custom header with response
  post '/echo' => sub {
    my $c = shift;
    $c->res->headers->header('X-Bender' => 'Bite my shiny metal ass!');
    $c->render(data => $c->req->body);
  };

  app->start;

You can test the more advanced examples right from the command line with
L<Mojolicious::Command::get>.

  $ ./myapp.pl get -v -M POST -c 'test' /echo

=head2 JSON

JSON is the most commonly used data-interchange format for web services.
L<Mojolicious> loves JSON and comes with the possibly fastest pure-Perl
implementation L<Mojo::JSON> built right in, which is accessible through
L<Mojo::Message/"json"> as well as the reserved stash value C<json>.

  use Mojolicious::Lite;

  # Modify the received JSON document and return it
  put '/reverse' => sub {
    my $c    = shift;
    my $hash = $c->req->json;
    $hash->{message} = reverse $hash->{message};
    $c->render(json => $hash);
  };

  app->start;

You can send JSON documents from the command line with
L<Mojolicious::Command::get>.

  $ ./myapp.pl get -M PUT -c '{"message":"Hello Mojo!"}' /reverse

=head2 Built-in C<exception> and C<not_found> pages

During development you will encounter these pages whenever you make a mistake,
they are gorgeous and contain a lot of valuable information that will aid you
in debugging your application.

  use Mojolicious::Lite;

  # Not found (404)
  get '/missing' => sub { shift->render(template => 'does_not_exist') };

  # Exception (500)
  get '/dies' => sub { die 'Intentional error' };

  app->start;

You can even use CSS selectors with L<Mojolicious::Command::get> to extract only
the information you're actually interested in.

  $ ./myapp.pl get /dies '#error'

And don't worry about revealing too much information on these pages, they are
only available during development, and will be replaced automatically with pages
that don't reveal any sensitive information in a production environment.

=head2 Route names

All routes can have a name associated with them, this allows automatic template
detection and backreferencing with L<Mojolicious::Controller/"url_for">, on
which many methods and helpers like
L<Mojolicious::Plugin::TagHelpers/"link_to"> rely.

  use Mojolicious::Lite;

  # Render the template "index.html.ep"
  get '/' => sub {
    my $c = shift;
    $c->render;
  } => 'index';

  # Render the template "hello.html.ep"
  get '/hello';

  app->start;
  __DATA__

  @@ index.html.ep
  <%= link_to Hello  => 'hello' %>.
  <%= link_to Reload => 'index' %>.

  @@ hello.html.ep
  Hello World!

Nameless routes get an automatically generated one assigned that is simply
equal to the route itself without non-word characters.

=head2 Layouts

Templates can have layouts too, you just select one with the helper
L<Mojolicious::Plugin::DefaultHelpers/"layout"> and place the result of the
current template with the helper
L<Mojolicious::Plugin::DefaultHelpers/"content">.

  use Mojolicious::Lite;

  get '/with_layout';

  app->start;
  __DATA__

  @@ with_layout.html.ep
  % title 'Green';
  % layout 'green';
  Hello World!

  @@ layouts/green.html.ep
  <!DOCTYPE html>
  <html>
    <head><title><%= title %></title></head>
    <body><%= content %></body>
  </html>

The stash or helpers like L<Mojolicious::Plugin::DefaultHelpers/"title"> can be
used to pass additional data to the layout.

=head2 Blocks

Template blocks can be used like normal Perl functions and are always delimited
by the C<begin> and C<end> keywords, they are the foundation for many helpers.

  use Mojolicious::Lite;

  get '/with_block' => 'block';

  app->start;
  __DATA__

  @@ block.html.ep
  % my $link = begin
    % my ($url, $name) = @_;
    Try <%= link_to $url => begin %><%= $name %><% end %>.
  % end
  <!DOCTYPE html>
  <html>
    <head><title>Sebastians frameworks</title></head>
    <body>
      %= $link->('http://mojolicious.org', 'Mojolicious')
      %= $link->('http://catalystframework.org', 'Catalyst')
    </body>
  </html>

=head2 Helpers

Helpers are little functions you can create with the keyword
L<Mojolicious::Lite/"helper"> and reuse throughout your whole application, from
actions to templates.

  use Mojolicious::Lite;

  # A helper to identify visitors
  helper whois => sub {
    my $c     = shift;
    my $agent = $c->req->headers->user_agent || 'Anonymous';
    my $ip    = $c->tx->remote_address;
    return "$agent ($ip)";
  };

  # Use helper in action and template
  get '/secret' => sub {
    my $c    = shift;
    my $user = $c->whois;
    $c->app->log->debug("Request from $user");
  };

  app->start;
  __DATA__

  @@ secret.html.ep
  We know who you are <%= whois %>.

A list of all built-in ones can be found in
L<Mojolicious::Plugin::DefaultHelpers> and L<Mojolicious::Plugin::TagHelpers>.

=head2 Placeholders

Route placeholders allow capturing parts of a request path until a C</> or C<.>
separator occurs, similar to the regular expression C<([^/.]+)>. Results are
accessible via L<Mojolicious::Controller/"stash"> and
L<Mojolicious::Controller/"param">.

  use Mojolicious::Lite;

  # /foo/test
  # /foo/test123
  get '/foo/:bar' => sub {
    my $c   = shift;
    my $bar = $c->stash('bar');
    $c->render(text => "Our :bar placeholder matched $bar");
  };

  # /testsomething/foo
  # /test123something/foo
  get '/(:bar)something/foo' => sub {
    my $c   = shift;
    my $bar = $c->param('bar');
    $c->render(text => "Our :bar placeholder matched $bar");
  };

  app->start;

To separate them from the surrounding text, you can surround your placeholders
with parentheses, which also makes the colon prefix optional.

=head2 Relaxed Placeholders

Relaxed placeholders allow matching of everything until a C</> occurs, similar
to the regular expression C<([^/]+)>.

  use Mojolicious::Lite;

  # /hello/test
  # /hello/test.html
  get '/hello/#you' => 'groovy';

  app->start;
  __DATA__

  @@ groovy.html.ep
  Your name is <%= $you %>.

=head2 Wildcard placeholders

Wildcard placeholders allow matching absolutely everything, including C</> and
C<.>, similar to the regular expression C<(.+)>.

  use Mojolicious::Lite;

  # /hello/test
  # /hello/test123
  # /hello/test.123/test/123
  get '/hello/*you' => 'groovy';

  app->start;
  __DATA__

  @@ groovy.html.ep
  Your name is <%= $you %>.

=head2 HTTP methods

Routes can be restricted to specific request methods with different keywords
like L<Mojolicious::Lite/"get"> and L<Mojolicious::Lite/"any">.

  use Mojolicious::Lite;

  # GET /hello
  get '/hello' => sub {
    my $c = shift;
    $c->render(text => 'Hello World!');
  };

  # PUT /hello
  put '/hello' => sub {
    my $c    = shift;
    my $size = length $c->req->body;
    $c->render(text => "You uploaded $size bytes to /hello.");
  };

  # GET|POST|PATCH /bye
  any ['GET', 'POST', 'PATCH'] => '/bye' => sub {
    my $c = shift;
    $c->render(text => 'Bye World!');
  };

  # * /whatever
  any '/whatever' => sub {
    my $c      = shift;
    my $method = $c->req->method;
    $c->render(text => "You called /whatever with $method.");
  };

  app->start;

=head2 Optional placeholders

All placeholders require a value, but by assigning them default values you can
make capturing optional.

  use Mojolicious::Lite;

  # /hello
  # /hello/Sara
  get '/hello/:name' => {name => 'Sebastian', day => 'Monday'} => sub {
    my $c = shift;
    $c->render(template => 'groovy', format => 'txt');
  };

  app->start;
  __DATA__

  @@ groovy.txt.ep
  My name is <%= $name %> and it is <%= $day %>.

Default values that don't belong to a placeholder simply get merged into the
stash all the time.

=head2 Restrictive placeholders

A very easy way to make placeholders more restrictive are alternatives, you just
make a list of possible values.

  use Mojolicious::Lite;

  # /test
  # /123
  any '/:foo' => [foo => ['test', '123']] => sub {
    my $c   = shift;
    my $foo = $c->param('foo');
    $c->render(text => "Our :foo placeholder matched $foo");
  };

  app->start;

All placeholders get compiled to a regular expression internally, this process
can also be customized. Just make sure not to use C<^> and C<$>, or capturing
groups C<(...)>, non-capturing groups C<(?:...)> are fine though.

  use Mojolicious::Lite;

  # /1
  # /123
  any '/:bar' => [bar => qr/\d+/] => sub {
    my $c   = shift;
    my $bar = $c->param('bar');
    $c->render(text => "Our :bar placeholder matched $bar");
  };

  app->start;

You can take a closer look at all the generated regular expressions with the
command L<Mojolicious::Command::routes>.

  $ ./myapp.pl routes -v

=head2 Under

Authentication and code shared between multiple routes can be realized easily
with routes generated by L<Mojolicious::Lite/"under">. All following routes are
only evaluated if the callback returned a true value.

  use Mojolicious::Lite;

  # Authenticate based on name parameter
  under sub {
    my $c = shift;

    # Authenticated
    my $name = $c->param('name') || '';
    return 1 if $name eq 'Bender';

    # Not authenticated
    $c->render(template => 'denied');
    return undef;
  };

  # Only reached when authenticated
  get '/' => 'index';

  app->start;
  __DATA__

  @@ denied.html.ep
  You are not Bender, permission denied.

  @@ index.html.ep
  Hi Bender.

Prefixing multiple routes is another good use for it.

  use Mojolicious::Lite;

  # /foo
  under '/foo';

  # /foo/bar
  get '/bar' => {text => 'foo bar'};

  # /foo/baz
  get '/baz' => {text => 'foo baz'};

  # / (reset)
  under '/' => {msg => 'whatever'};

  # /bar
  get '/bar' => {inline => '<%= $msg %> works'};

  app->start;

You can also group related routes with L<Mojolicious::Lite/"group">, which
allows nesting of routes generated with L<Mojolicious::Lite/"under">.

  use Mojolicious::Lite;

  # Global logic shared by all routes
  under sub {
    my $c = shift;
    return 1 if $c->req->headers->header('X-Bender');
    $c->render(text => "You're not Bender.");
    return undef;
  };

  # Admin section
  group {

    # Local logic shared only by routes in this group
    under '/admin' => sub {
      my $c = shift;
      return 1 if $c->req->headers->header('X-Awesome');
      $c->render(text => "You're not awesome enough.");
      return undef;
    };

    # GET /admin/dashboard
    get '/dashboard' => {text => 'Nothing to see here yet.'};
  };

  # GET /welcome
  get '/welcome' => {text => 'Hi Bender.'};

  app->start;

=head2 Formats

Formats can be automatically detected from file extensions like C<.html>, they
are used to find the right template and generate the correct C<Content-Type>
header.

  use Mojolicious::Lite;

  # /detection
  # /detection.html
  # /detection.txt
  get '/detection' => sub {
    my $c = shift;
    $c->render(template => 'detected');
  };

  app->start;
  __DATA__

  @@ detected.html.ep
  <!DOCTYPE html>
  <html>
    <head><title>Detected</title></head>
    <body>HTML was detected.</body>
  </html>

  @@ detected.txt.ep
  TXT was detected.

The default format is C<html>, and restrictive placeholders can be used to
limit possible values.

  use Mojolicious::Lite;

  # /hello.json
  # /hello.txt
  get '/hello' => [format => ['json', 'txt']] => sub {
    my $c = shift;
    return $c->render(json => {hello => 'world'})
      if $c->stash('format') eq 'json';
    $c->render(text => 'hello world');
  };

  app->start;

Or you can just disable format detection with a special type of restrictive
placeholder.

  use Mojolicious::Lite;

  # /hello
  get '/hello' => [format => 0] => {text => 'No format detection.'};

  # Disable detection and allow the following routes to re-enable it on demand
  under [format => 0];

  # /foo
  get '/foo' => {text => 'No format detection again.'};

  # /bar.txt
  get '/bar' => [format => 'txt'] => {text => ' Just one format.'};

  app->start;

=head2 Content negotiation

For resources with different representations and that require truly RESTful
content negotiation you can also use L<Mojolicious::Controller/"respond_to">.

  use Mojolicious::Lite;

  # /hello (Accept: application/json)
  # /hello (Accept: application/xml)
  # /hello.json
  # /hello.xml
  # /hello?format=json
  # /hello?format=xml
  get '/hello' => sub {
    my $c = shift;
    $c->respond_to(
      json => {json => {hello => 'world'}},
      xml  => {text => '<hello>world</hello>'},
      any  => {data => '', status => 204}
    );
  };

  app->start;

MIME type mappings can be extended or changed easily with
L<Mojolicious/"types">.

  app->types->type(rdf => 'application/rdf+xml');

=head2 Static files

Similar to templates, but with only a single file extension and optional Base64
encoding, static files can be inlined in the C<DATA> section and are served
automatically.

  use Mojolicious::Lite;

  app->start;
  __DATA__

  @@ something.js
  alert('hello!');

  @@ test.txt (base64)
  dGVzdCAxMjMKbGFsYWxh

External static files are not limited to a single file extension and will be
served automatically from a C<public> directory if it exists.

  $ mkdir public
  $ mv something.js public/something.js
  $ mv mojolicious.tar.gz public/mojolicious.tar.gz

Both have a higher precedence than routes for C<GET> and C<HEAD> requests.
Content negotiation with C<Range>, C<If-None-Match> and C<If-Modified-Since>
headers is supported as well and can be tested very easily with
L<Mojolicious::Command::get>.

  $ ./myapp.pl get /something.js -v -H 'Range: bytes=2-4'

=head2 External templates

External templates will be searched by the renderer in a C<templates> directory
if it exists.

  $ mkdir -p templates/foo
  $ echo 'Hello World!' > templates/foo/bar.html.ep

They have a higher precedence than templates in the C<DATA> section.

  use Mojolicious::Lite;

  # Render template "templates/foo/bar.html.ep"
  any '/external' => sub {
    my $c = shift;
    $c->render(template => 'foo/bar');
  };

  app->start;

=head2 Conditions

Conditions such as C<agent> and C<host> from
L<Mojolicious::Plugin::HeaderCondition> allow even more powerful route
constructs.

  use Mojolicious::Lite;

  # Firefox
  get '/foo' => (agent => qr/Firefox/) => sub {
    my $c = shift;
    $c->render(text => 'Congratulations, you are using a cool browser.');
  };

  # Internet Explorer
  get '/foo' => (agent => qr/Internet Explorer/) => sub {
    my $c = shift;
    $c->render(text => 'Dude, you really need to upgrade to Firefox.');
  };

  # http://mojolicious.org/bar
  get '/bar' => (host => 'mojolicious.org') => sub {
    my $c = shift;
    $c->render(text => 'Hello Mojolicious.');
  };

  app->start;

=head2 Sessions

Cookie-based sessions just work out of the box, as soon as you start using them
through the helper L<Mojolicious::Plugin::DefaultHelpers/"session">. Just be
aware that all session data gets serialized with L<Mojo::JSON> and stored
client-side, with a cryptographic signature to prevent tampering.

  use Mojolicious::Lite;

  # Access session data in action and template
  get '/counter' => sub {
    my $c = shift;
    $c->session->{counter}++;
  };

  app->start;
  __DATA__

  @@ counter.html.ep
  Counter: <%= session 'counter' %>

Note that you should use custom L<Mojolicious/"secrets"> to make signed cookies
really tamper resistant.

  app->secrets(['My secret passphrase here']);

=head2 File uploads

All files uploaded via C<multipart/form-data> request are automatically
available as L<Mojo::Upload> objects from L<Mojolicious::Controller/"param">.
And you don't have to worry about memory usage, because all files above 250KiB
will be automatically streamed into a temporary file. To build HTML forms more
efficiently, you can also use tag helpers like
L<Mojolicious::Plugin::TagHelpers/"form_for">.

  use Mojolicious::Lite;

  # Upload form in DATA section
  get '/' => 'form';

  # Multipart upload handler
  post '/upload' => sub {
    my $c = shift;

    # Check file size
    return $c->render(text => 'File is too big.', status => 200)
      if $c->req->is_limit_exceeded;

    # Process uploaded file
    return $c->redirect_to('form') unless my $example = $c->param('example');
    my $size = $example->size;
    my $name = $example->filename;
    $c->render(text => "Thanks for uploading $size byte file $name.");
  };

  app->start;
  __DATA__

  @@ form.html.ep
  <!DOCTYPE html>
  <html>
    <head><title>Upload</title></head>
    <body>
      %= form_for upload => (enctype => 'multipart/form-data') => begin
        %= file_field 'example'
        %= submit_button 'Upload'
      % end
    </body>
  </html>

To protect you from excessively large files there is also a limit of 16MiB by
default, which you can tweak with the attribute
L<Mojolicious/"max_request_size">.

  # Increase limit to 1GiB
  app->max_request_size(1073741824);

=head2 User agent

With L<Mojo::UserAgent>, which is available through the helper
L<Mojolicious::Plugin::DefaultHelpers/"ua">, there's a full featured HTTP and
WebSocket user agent built right in. Especially in combination with
L<Mojo::JSON> and L<Mojo::DOM> this can be a very powerful tool.

  use Mojolicious::Lite;

  # Blocking
  get '/headers' => sub {
    my $c   = shift;
    my $url = $c->param('url') || 'http://mojolicious.org';
    my $dom = $c->ua->get($url)->result->dom;
    $c->render(json => $dom->find('h1, h2, h3')->map('text')->to_array);
  };

  # Non-blocking
  get '/title' => sub {
    my $c = shift;
    $c->ua->get('mojolicious.org' => sub {
      my ($ua, $tx) = @_;
      $c->render(data => $tx->result->dom->at('title')->text);
    });
  };

  # Concurrent non-blocking
  get '/titles' => sub {
    my $c = shift;
    my $mojo = $ua->get_p('http://mojolicious.org');
    my $cpan = $ua->get_p('https://metacpan.org');
    Mojo::Promise->all($mojo, $cpan)->then(sub {
      my ($mojo, $cpan) = @_;
      $c->render(json => {
        mojo => $mojo->[0]->result->dom->at('title')->text,
        cpan => $cpan->[0]->result->dom->at('title')->text
      });
    })->wait;
  };

  app->start;

For more information about the user agent see also
L<Mojolicious::Guides::Cookbook/"USER AGENT">.

=head2 WebSockets

WebSocket applications have never been this simple before. Just receive
messages by subscribing to events such as
L<Mojo::Transaction::WebSocket/"json"> with L<Mojolicious::Controller/"on"> and
return them with L<Mojolicious::Controller/"send">.

  use Mojolicious::Lite;

  websocket '/echo' => sub {
    my $c = shift;
    $c->on(json => sub {
      my ($c, $hash) = @_;
      $hash->{msg} = "echo: $hash->{msg}";
      $c->send({json => $hash});
    });
  };

  get '/' => 'index';

  app->start;
  __DATA__

  @@ index.html.ep
  <!DOCTYPE html>
  <html>
    <head>
      <title>Echo</title>
      <script>
        var ws = new WebSocket('<%= url_for('echo')->to_abs %>');
        ws.onmessage = function (event) {
          document.body.innerHTML += JSON.parse(event.data).msg;
        };
        ws.onopen = function (event) {
          ws.send(JSON.stringify({msg: 'I ♥ Mojolicious!'}));
        };
      </script>
    </head>
  </html>

For more information about real-time web features see also
L<Mojolicious::Guides::Cookbook/"REAL-TIME WEB">.

=head2 Mode

You can use the L<Mojo::Log> object from L<Mojo/"log"> to portably collect
debug messages and automatically disable them later in a production setup by
changing the L<Mojolicious> operating mode, which can also be retrieved from
the attribute L<Mojolicious/"mode">.

  use Mojolicious::Lite;

  # Prepare mode specific message during startup
  my $msg = app->mode eq 'development' ? 'Development!' : 'Something else!';

  get '/' => sub {
    my $c = shift;
    $c->app->log->debug('Rendering mode specific message');
    $c->render(text => $msg);
  };

  app->log->debug('Starting application');
  app->start;

The default operating mode will usually be C<development> and can be changed
with command line options or the C<MOJO_MODE> and C<PLACK_ENV> environment
variables. A mode other than C<development> will raise the log level from
C<debug> to C<info>.

  $ ./myapp.pl daemon -m production

All messages will be written to C<STDERR> or a C<log/$mode.log> file if a
C<log> directory exists.

  $ mkdir log

Mode changes also affect a few other aspects of the framework, such as the
built-in C<exception> and C<not_found> pages. Once you switch modes from
C<development> to C<production>, no sensitive information will be revealed on
those pages anymore.

=head2 Testing

Testing your application is as easy as creating a C<t> directory and filling it
with normal Perl tests like C<t/basic.t>, which can be a lot of fun thanks to
L<Test::Mojo>.

  use Test::More;
  use Test::Mojo;

  use FindBin;
  require "$FindBin::Bin/../myapp.pl";

  my $t = Test::Mojo->new;
  $t->get_ok('/')->status_is(200)->content_like(qr/Funky/);

  done_testing();

Just run your tests with the command L<Mojolicious::Command::test> or L<prove>.

  $ ./myapp.pl test
  $ ./myapp.pl test -v t/basic.t
  $ prove -l -v t/basic.t

=head1 MORE

You can continue with L<Mojolicious::Guides> now or take a look at the
L<Mojolicious wiki|http://github.com/kraih/mojo/wiki>, which contains a lot more
documentation and examples by many different authors.

=head1 SUPPORT

If you have any questions the documentation might not yet answer, don't
hesitate to ask on the
L<mailing list|http://groups.google.com/group/mojolicious> or the official IRC
channel C<#mojo> on C<irc.perl.org>
(L<chat now!|https://chat.mibbit.com/?channel=%23mojo&server=irc.perl.org>).

=cut