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

NAME

Catalyst::Manual::Tutorial::MoreCatalystBasics - Catalyst Tutorial - Part 3: More Catalyst Application Development Basics

OVERVIEW

This is Part 3 of 10 for the Catalyst tutorial.

Tutorial Overview

  1. Introduction

  2. Catalyst Basics

  3. More Catalyst Basics

  4. Basic CRUD

  5. Authentication

  6. Authorization

  7. Debugging

  8. Testing

  9. Advanced CRUD

  10. Appendices

DESCRIPTION

This part of the tutorial builds on the work done in Part 2 to explore some features that are more typical of "real world" web applications. From this part of the tutorial onward, we will be building a simple book database application. Although the application will be too limited to be of use to anyone, it should provide a basic environment where we can explore a variety of features used in virtually all web applications.

You can checkout the source code for this example from the catalyst subversion repository as per the instructions in Catalyst::Manual::Tutorial::Intro.

CREATE A NEW APPLICATION

The remainder of the tutorial will build an application called MyApp. First use the Catalyst catalyst.pl script to initialize the framework for the MyApp application (make sure you aren't still inside the directory of the Hello application from the previous part of the tutorial or in a directory that already has a "MyApp" subdirectory):

    $ catalyst.pl MyApp
    created "MyApp"
    created "MyApp/script"
    created "MyApp/lib"
    created "MyApp/root"
    ...
    created "MyApp/script/myapp_create.pl"
    $ cd MyApp

This creates a similar skeletal structure to what we saw in Part 2 of the tutorial, except with MyApp and myapp substituted for Hello and hello.

EDIT THE LIST OF CATALYST PLUGINS

One of the greatest benefits of Catalyst is that it has such a large library of plugins and base classes available. Plugins are used to seamlessly integrate existing Perl modules into the overall Catalyst framework. In general, they do this by adding additional methods to the context object (generally written as $c) that Catalyst passes to every component throughout the framework.

By default, Catalyst enables three plugins/flags:

  • -Debug Flag

    Enables the Catalyst debug output you saw when we started the script/myapp_server.pl development server earlier. You can remove this item when you place your application into production.

    As you may have noticed, -Debug is not a plugin, but a flag. Although most of the items specified on the __PACKAGE__->setup line of your application class will be plugins, Catalyst supports a limited number of flag options (of these, -Debug is the most common). See the documentation for Catalyst.pm to get details on other flags (currently -Engine, -Home, and -Log).

    If you prefer, you can use the $c->debug method to enable debug messages.

    TIP: Depending on your needs, it can be helpful to permanently remove -Debug from lib/MyApp.pm and then use the -d option to script/myapp_server.pl to re-enable it just for the development server. We will not be using that approach in the tutorial, but feel free to make use of it in your own projects.

  • Catalyst::Plugin::ConfigLoader

    ConfigLoader provides an automatic way to load configurable parameters for your application from a central Config::General file (versus having the values hard-coded inside your Perl modules). Config::General uses syntax very similar to Apache configuration files. We will see how to use this feature of Catalyst during the authentication and authorization sections (Part 5 and Part 6).

    IMPORTANT NOTE: If you are using a version of Catalyst::Devel prior to version 1.06, be aware that Catalyst changed the default format from YAML to the more straightforward Config::General style. This tutorial uses the newer myapp.conf file for Config::General. However, Catalyst supports both formats and will automatically use either myapp.conf or myapp.yml (or any other format supported by Catalyst::Plugin::ConfigLoader and Config::Any). If you are using a version of Catalyst::Devel prior to 1.06, you can convert to the newer format by simply creating the myapp.conf file manually and deleting myapp.yml. The default contents of the myapp.conf you create should only consist of one line:

        name MyApp

    TIP: This script can be useful for converting between configuration formats:

        perl -Ilib -e 'use MyApp; use Config::General;
            Config::General->new->save_file("myapp.conf", MyApp->config);'
  • Catalyst::Plugin::Static::Simple

    Static::Simple provides an easy method of serving static content such as images and CSS files under the development server.

For our application, we want to add one new plugin into the mix. To do this, edit lib/MyApp.pm (this file is generally referred to as your application class) and delete the lines with:

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

Then replace it with:

    # Load plugins
    use Catalyst qw/-Debug
                ConfigLoader
                Static::Simple
                
                StackTrace
                /;

Note: Recent versions of Catalyst::Devel have used a variety of techniques to load these plugins/flags. For example, you might see the following:

    __PACKAGE__->setup(qw/-Debug ConfigLoader Static::Simple/);

Don't let these variations confuse you -- they all accomplish the same result.

This tells Catalyst to start using one new plugin, Catalyst::Plugin::StackTrace, to add a stack trace to the standard Catalyst "debug screen" (the screen Catalyst sends to your browser when an error occurs). Be aware that StackTrace output appears in your browser, not in the console window from which you're running your application, which is where logging output usually goes.

Notes:

  • __PACKAGE__ is just a shorthand way of referencing the name of the package where it is used. Therefore, in MyApp.pm, __PACKAGE__ is equivalent to MyApp.

  • You will want to disable StackTrace before you put your application into production, but it can be helpful during development.

  • When specifying plugins on the __PACKAGE__->setup line, you can omit Catalyst::Plugin:: from the name. Additionally, you can spread the plugin names across multiple lines as shown here, or place them all on one (or more) lines as with the default configuration.

CREATE A CATALYST CONTROLLER

As discussed earlier, controllers are where you write methods that interact with user input. Typically, controller methods respond to GET and POST messages from the user's web browser.

Use the Catalyst create script to add a controller for book-related actions:

    $ script/myapp_create.pl controller Books
     exists "/home/me/MyApp/script/../lib/MyApp/Controller"
     exists "/home/me/MyApp/script/../t"
    created "/home/me/MyApp/script/../lib/MyApp/Controller/Books.pm"
    created "/home/me/MyApp/script/../t/controller_Books.t"

Then edit lib/MyApp/Controller/Books.pm (as discussed in Part 2 of the Tutorial, Catalyst has a separate directory under lib/MyApp for each of the three parts of MVC: Model, View, and Controller) and add the following method to the controller:

    =head2 list
    
    Fetch all book objects and pass to books/list.tt2 in stash to be displayed
    
    =cut
    
    sub list : Local {
        # Retrieve the usual Perl OO '$self' for this object. $c is the Catalyst
        # 'Context' that's used to 'glue together' the various components
        # that make up the application
        my ($self, $c) = @_;
    
        # Retrieve all of the book records as book model objects and store in the
        # stash where they can be accessed by the TT template
        # $c->stash->{books} = [$c->model('DB::Books')->all];
        # But, for now, use this code until we create the model later
        $c->stash->{books} = '';
    
        # Set the TT template to use.  You will almost always want to do this
        # in your action methods (action methods respond to user input in
        # your controllers).
        $c->stash->{template} = 'books/list.tt2';
    }

TIP: See Appendix 1 for tips on removing the leading spaces when cutting and pasting example code from POD-based documents.

Programmers experienced with object-oriented Perl should recognize $self as a reference to the object where this method was called. On the other hand, $c will be new to many Perl programmers who have not used Catalyst before (it's sometimes written as $context). The Context object is automatically passed to all Catalyst components. It is used to pass information between components and provide access to Catalyst and plugin functionality.

Catalyst actions are regular Perl methods, but they make use of attributes (the ": Local" next to the "sub list" in the code above) to provide additional information to the Catalyst dispatcher logic (note that the space between the colon and the attribute name is optional... you will see attributes written both ways). Most Catalyst Controllers use one of five action types:

  • :Private -- Use :Private for methods that you want to make into an action, but you do not want Catalyst to directly expose the action to your users. Catalyst will not map :Private methods to a URI. Use them for various sorts of "special" methods (the begin, auto, etc. discussed below) or for methods you want to be able to forward or detach to. (If the method is a plain old "helper method" that you don't want to be an action at all, then just define the method without any attribute -- you can call it in your code, but the Catalyst dispatcher will ignore it.)

    There are five types of "special" build-in :Private actions: begin, end, default, index, and auto.

    • With begin, end, default, index private actions, only the most specific action of each type will be called. For example, if you define a begin action in your controller it will override a begin action in your application/root controller -- only the action in your controller will be called.

    • Unlike the other actions where only a single method is called for each request, every auto action along the chain of namespaces will be called. Each auto action will be called from the application/root controller down through the most specific class.

  • :Path -- :Path actions let you map a method to an explicit URI path. For example, ":Path('list')" in lib/MyApp/Controller/Books.pm would match on the URL http://localhost:3000/books/list but ":Path('/list')" would match on http://localhost:3000/list. You can use :Args() to specify how many arguments an action should except. See "Action_types" in Catalyst::Manual::Intro for more information and a few examples.

  • :Local -- :Local is merely a shorthand for ":Path('_name_of_method_')". For example, these are equivalent: "sub create_book :Local {...}" and "sub create_book :Path('create_book') {...}".

  • :Global -- :Global is merely a shorthand for ":Path('/_name_of_method_')". For example, these are equivalent: "sub create_book :Global {...}" and "sub create_book :Path('/create_book') {...}".

  • :Chained -- Newer Catalyst applications tend to use the Chained dispatch form of action types because of its power and flexibility. It allows a series of controller methods to automatically be dispatched to service a single user request. See Catalyst::Manual::Tutorial::BasicCRUD and Catalyst::DispatchType::Chained for more information on chained actions.

You should refer to "Action_types" in Catalyst::Manual::Intro for additional information and for coverage of some lesser-used action types not discussed here (Regex and LocalRegex).

CATALYST VIEWS

As mentioned in Part 2 of the tutorial, views are where you render output, typically for display in the user's web browser (but also possibly using other display output-generation systems). The code in lib/MyApp/View selects the type of view to use, with the actual rendering template found in the root directory. As with virtually every aspect of Catalyst, options abound when it comes to the specific view technology you adopt inside your application. However, most Catalyst applications use the Template Toolkit, known as TT (for more information on TT, see http://www.template-toolkit.org). Other somewhat popular view technologies include Mason (http://www.masonhq.com and http://www.masonbook.com) and HTML::Template (http://html-template.sourceforge.net).

Create a Catalyst View

When using TT for the Catalyst view, there are two main helper scripts:

Both helpers are similar. TT creates the lib/MyApp/View/TT.pm file and leaves the creation of any hierarchical template organization entirely up to you. (It also creates a t/view_TT.t file for testing; test cases will be discussed in Part 8.) TTSite, on the other hand, creates a modular and hierarchical view layout with separate Template Toolkit (TT) files for common header and footer information, configuration values, a CSS stylesheet, and more.

While TTSite was useful to bootstrap a project, its use is now deprecated and to be considered historical. For most Catalyst applications it adds redundant functionality and structure; many in the Catalyst community recommend that it's easier to learn both Catalyst and Template Toolkit if you use the more basic TT approach. Consequently, this tutorial will use "plain old TT."

Enter the following command to enable the TT style of view rendering for this tutorial:

    $ script/myapp_create.pl view TT TT
     exists "/home/me/MyApp/script/../lib/MyApp/View"
     exists "/home/me/MyApp/script/../t"
     created "/home/me/MyApp/script/../lib/MyApp/View/TT.pm"
     created "/home/me/MyApp/script/../t/view_TT.t"

This simply creates a view called TT (the second 'TT' argument) in a file called TT.pm (the first 'TT' argument). It is now up to you to decide how you want to structure your view layout. For the tutorial, we will start with a very simple TT template to initially demonstrate the concepts, but quickly migrate to a more typical "wrapper page" type of configuration (where the "wrapper" controls the overall "look and feel" of your site from a single file or set of files).

Edit lib/MyApp/View/TT.pm and you should see that the default contents contains something similar to the following:

    __PACKAGE__->config(TEMPLATE_EXTENSION => '.tt');

And update it to match:

    __PACKAGE__->config(
        # Change default TT extension
        TEMPLATE_EXTENSION => '.tt2',
        # Set the location for TT files
        INCLUDE_PATH => [
                MyApp->path_to( 'root', 'src' ),
            ],
    );

NOTE: Make sure to add a comma after '.tt2' outside the single quote.

This changes the default extension for Template Toolkit from '.tt' to '.tt2' and changes the base directory for your template files from root to root/src. These changes from the default are done mostly to facilitate the application we're developing in this tutorial; as with most things Perl, there's more than one way to do it...

Note: We will use root/src as the base directory for our template files, which a full naming convention of root/src/_controller_name_/_action_name_.tt2. Another popular option is to use root/ as the base (with a full filename pattern of root/_controller_name_/_action_name_.tt2).

Create a TT Template Page

First create a directory for book-related TT templates:

    $ mkdir -p root/src/books

Then create root/src/books/list.tt2 in your editor and enter:

    [% # This is a TT comment.  The '-' at the end "chomps" the newline.  You won't -%]
    [% # see this "chomping" in your browser because HTML ignores blank lines, but  -%]
    [% # it WILL eliminate a blank line if you view the HTML source.  It's purely   -%]
    [%- # optional, but both the beginning and the ending TT tags support chomping. -%]
    
    [% # Provide a title -%]
    [% META title = 'Book List' -%]
    
    <table>
    <tr><th>Title</th><th>Rating</th><th>Author(s)</th></tr>
    [% # Display each book in a table row %]
    [% FOREACH book IN books -%]
      <tr>
        <td>[% book.title %]</td>
        <td>[% book.rating %]</td>
      </tr>
    [% END -%]
    </table>

As indicated by the inline comments above, the META title line uses TT's META feature to provide a title to the "wrapper" that we will create later. Meanwhile, the FOREACH loop iterates through each book model object and prints the title and rating fields.

If you are new to TT, the [% and %] tags are used to delimit TT code. TT supports a wide variety of directives for "calling" other files, looping, conditional logic, etc. In general, TT simplifies the usual range of Perl operators down to the single dot (.) operator. This applies to operations as diverse as method calls, hash lookups, and list index values (see http://search.cpan.org/perldoc?Template::Manual::Variables for details and examples). In addition to the usual Template module Pod documentation, you can access the TT manual at http://search.cpan.org/perldoc?Template::Manual.

TIP: While you can build all sorts of complex logic into your TT templates, you should in general keep the "code" part of your templates as simple as possible. If you need more complex logic, create helper methods in your model that abstract out a set of code into a single call from your TT template. (Note that the same is true of your controller logic as well -- complex sections of code in your controllers should often be pulled out and placed into your model objects.)

Test Run The Application

To test your work so far, first start the development server:

    $ script/myapp_server.pl

Then point your browser to http://localhost:3000 and you should still get the Catalyst welcome page. Next, change the URL in your browser to http://localhost:3000/books/list. If you have everything working so far, you should see a web page that displays nothing other than our column headers for "Title", "Rating", and "Author(s)" -- we will not see any books until we get the database and model working below.

If you run into problems getting your application to run correctly, it might be helpful to refer to some of the debugging techniques covered in the Debugging part of the tutorial.

CREATE A SQLITE DATABASE

In this step, we make a text file with the required SQL commands to create a database table and load some sample data. We will use SQLite, a popular database that is lightweight and easy to use. Open myapp01.sql in your editor and enter:

    --
    -- Create a very simple database to hold book and author information
    --
    CREATE TABLE books (
            id          INTEGER PRIMARY KEY,
            title       TEXT ,
            rating      INTEGER
    );
    -- 'book_authors' is a many-to-many join table between books & authors
    CREATE TABLE book_authors (
            book_id     INTEGER,
            author_id   INTEGER,
            PRIMARY KEY (book_id, author_id)
    );
    CREATE TABLE authors (
            id          INTEGER PRIMARY KEY,
            first_name  TEXT,
            last_name   TEXT
    );
    ---
    --- Load some sample data
    ---
    INSERT INTO books VALUES (1, 'CCSP SNRS Exam Certification Guide', 5);
    INSERT INTO books VALUES (2, 'TCP/IP Illustrated, Volume 1', 5);
    INSERT INTO books VALUES (3, 'Internetworking with TCP/IP Vol.1', 4);
    INSERT INTO books VALUES (4, 'Perl Cookbook', 5);
    INSERT INTO books VALUES (5, 'Designing with Web Standards', 5);
    INSERT INTO authors VALUES (1, 'Greg', 'Bastien');
    INSERT INTO authors VALUES (2, 'Sara', 'Nasseh');
    INSERT INTO authors VALUES (3, 'Christian', 'Degu');
    INSERT INTO authors VALUES (4, 'Richard', 'Stevens');
    INSERT INTO authors VALUES (5, 'Douglas', 'Comer');
    INSERT INTO authors VALUES (6, 'Tom', 'Christiansen');
    INSERT INTO authors VALUES (7, 'Nathan', 'Torkington');
    INSERT INTO authors VALUES (8, 'Jeffrey', 'Zeldman');
    INSERT INTO book_authors VALUES (1, 1);
    INSERT INTO book_authors VALUES (1, 2);
    INSERT INTO book_authors VALUES (1, 3);
    INSERT INTO book_authors VALUES (2, 4);
    INSERT INTO book_authors VALUES (3, 5);
    INSERT INTO book_authors VALUES (4, 6);
    INSERT INTO book_authors VALUES (4, 7);
    INSERT INTO book_authors VALUES (5, 8);

Then use the following command to build a myapp.db SQLite database:

    $ sqlite3 myapp.db < myapp01.sql

If you need to create the database more than once, you probably want to issue the rm myapp.db command to delete the database before you use the sqlite3 myapp.db < myapp01.sql command.

Once the myapp.db database file has been created and initialized, you can use the SQLite command line environment to do a quick dump of the database contents:

    $ sqlite3 myapp.db
    SQLite version 3.5.9
    Enter ".help" for instructions
    sqlite> select * from books;
    1|CCSP SNRS Exam Certification Guide|5
    2|TCP/IP Illustrated, Volume 1|5
    3|Internetworking with TCP/IP Vol.1|4
    4|Perl Cookbook|5
    5|Designing with Web Standards|5
    sqlite> .q
    $

Or:

    $ sqlite3 myapp.db "select * from books"
    1|CCSP SNRS Exam Certification Guide|5
    2|TCP/IP Illustrated, Volume 1|5
    3|Internetworking with TCP/IP Vol.1|4
    4|Perl Cookbook|5
    5|Designing with Web Standards|5

As with most other SQL tools, if you are using the full "interactive" environment you need to terminate your SQL commands with a ";" (it's not required if you do a single SQL statement on the command line). Use ".q" to exit from SQLite from the SQLite interactive mode and return to your OS command prompt.

For using other databases, such as PostgreSQL or MySQL, see Appendix 2.

DATABASE ACCESS WITH DBIx::Class

Catalyst can be used with virtually any form of persistent datastore available via Perl. For example, Catalyst::Model::DBI can be used to easily access databases through the traditional Perl DBI interface. However, most Catalyst applications use some form of ORM technology to automatically create and save model objects as they are used. Although Class::DBI has been a popular choice in the past, Matt Trout's DBIx::Class (abbreviated as "DBIC") has rapidly emerged as the Perl-based ORM technology of choice. Most new Catalyst applications rely on DBIC, as will this tutorial.

Create a Dynamic DBIC Model

Use the create=dynamic model helper option to build a model that dynamically reads your database structure every time the application starts:

    $ script/myapp_create.pl model DB DBIC::Schema MyApp::Schema create=dynamic dbi:SQLite:myapp.db
     exists "/home/me/MyApp/script/../lib/MyApp/Model"
     exists "/home/me/MyApp/script/../t"
     exists "/home/me/MyApp/script/../lib/MyApp"
    created "/home/me/MyApp/script/../lib/MyApp/Schema.pm"
    created "/home/me/MyApp/script/../lib/MyApp/Model/DB.pm"
    created "/home/me/MyApp/script/../t/model_DB.t"

The script/myapp_create.pl command breaks down like this:

  • DB is the name of the model class to be created by the helper in lib/MyApp/Model.

  • DBIC::Schema is the type of the model to create.

  • MyApp::Schema is the name of the DBIC schema file written to lib/MyApp/Schema.pm.

  • Because we specified create=dynamic to the helper, it use DBIx::Class::Schema::Loader to dynamically load the schema information from the database every time the application starts.

  • And finally, dbi:SQLite:myapp.db is the standard DBI connect string for use with SQLite.

NOTE: Although the create=dynamic option to the DBIC helper makes for a nifty demonstration, is only really suitable for very small applications. After this demonstration, you should almost always use the create=static option that we switch to below.

ENABLE THE MODEL IN THE CONTROLLER

Open lib/MyApp/Controller/Books.pm and un-comment the model code we left disabled earlier so that your version matches the following (un- comment the line containing [$c->model('DB::Books')->all] and delete the next 2 lines):

    =head2 list
    
    Fetch all book objects and pass to books/list.tt2 in stash to be displayed
    
    =cut
    
    sub list : Local {
        # Retrieve the usual Perl OO '$self' for this object. $c is the Catalyst
        # 'Context' that's used to 'glue together' the various components
        # that make up the application
        my ($self, $c) = @_;
    
        # Retrieve all of the book records as book model objects and store in the
        # stash where they can be accessed by the TT template
        $c->stash->{books} = [$c->model('DB::Books')->all];
    
        # Set the TT template to use.  You will almost always want to do this
        # in your action methods (action methods respond to user input in
        # your controllers).
        $c->stash->{template} = 'books/list.tt2';
    }

TIP: You may see the $c->model('DB::Books') un-commented above written as $c->model('DB')->resultset('Books'). The two are equivalent. Either way, $c->model returns a DBIx::Class::ResultSet which handles queries against the database and iterating over the set of results that are returned.

We are using the ->all to fetch all of the books. DBIC supports a wide variety of more advanced operations to easily do things like filtering and sorting the results. For example, the following could be used to sort the results by descending title:

    $c->model('DB::Books')->search({}, {order_by => 'title DESC'});

Some other examples are provided in "Complex WHERE clauses" in DBIx::Class::Manual::Cookbook, with additional information found at "search" in DBIx::Class::ResultSet, "Searching" in DBIx::Class::Manual::FAQ, DBIx::Class::Manual::Intro and Catalyst::Model::DBIC::Schema.

Test Run The Application

First, let's enable an environment variable that causes DBIx::Class to dump the SQL statements used to access the database. This is a helpful trick when you are trying to debug your database-oriented code:

    $ export DBIC_TRACE=1

This assumes you are using BASH as your shell -- adjust accordingly if you are using a different shell (for example, under tcsh, use setenv DBIC_TRACE 1).

NOTE: You can also set this in your code using $class->storage->debug(1);. See DBIx::Class::Manual::Troubleshooting for details (including options to log to file instead of displaying to the Catalyst development server log).

Then launch the Catalyst development server. The log output should display something like:

    $ script/myapp_server.pl
    [debug] Debug messages enabled
    [debug] Statistics enabled
    [debug] Loaded plugins:
    .----------------------------------------------------------------------------.
    | Catalyst::Plugin::ConfigLoader  0.20                                       |
    | Catalyst::Plugin::StackTrace  0.08                                         |
    | Catalyst::Plugin::Static::Simple  0.20                                     |
    '----------------------------------------------------------------------------'
    
    [debug] Loaded dispatcher "Catalyst::Dispatcher"
    [debug] Loaded engine "Catalyst::Engine::HTTP"
    [debug] Found home "/home/me/MyApp"
    [debug] Loaded Config "/home/me/MyApp/myapp.conf"
    [debug] Loaded components:
    .-----------------------------------------------------------------+----------.
    | Class                                                           | Type     |
    +-----------------------------------------------------------------+----------+
    | MyApp::Controller::Books                                        | instance |
    | MyApp::Controller::Root                                         | instance |
    | MyApp::Model::DB                                                | instance |
    | MyApp::Model::DB::Authors                                       | class    |
    | MyApp::Model::DB::BookAuthors                                   | class    |
    | MyApp::Model::DB::Books                                         | class    |
    | MyApp::View::TT                                                 | instance |
    '-----------------------------------------------------------------+----------'
    
    [debug] Loaded Private actions:
    .----------------------+--------------------------------------+--------------.
    | Private              | Class                                | Method       |
    +----------------------+--------------------------------------+--------------+
    | /default             | MyApp::Controller::Root              | default      |
    | /end                 | MyApp::Controller::Root              | end          |
    | /index               | MyApp::Controller::Root              | index        |
    | /books/index         | MyApp::Controller::Books             | index        |
    | /books/list          | MyApp::Controller::Books             | list         |
    '----------------------+--------------------------------------+--------------'
    
    [debug] Loaded Path actions:
    .-------------------------------------+--------------------------------------.
    | Path                                | Private                              |
    +-------------------------------------+--------------------------------------+
    | /                                   | /default                             |
    | /                                   | /index                               |
    | /books                              | /books/index                         |
    | /books/list                         | /books/list                          |
    '-------------------------------------+--------------------------------------'
    
    [info] MyApp powered by Catalyst 5.71000
    You can connect to your server at http://debian:3000

NOTE: Be sure you run the script/myapp_server.pl command from the 'base' directory of your application, not inside the script directory itself or it will not be able to locate the myapp.db database file. You can use a fully qualified or a relative path to locate the database file, but we did not specify that when we ran the model helper earlier.

Some things you should note in the output above:

  • Catalyst::Model::DBIC::Schema dynamically created three model classes, one to represent each of the three tables in our database (MyApp::Model::DB::Authors, MyApp::Model::DB::BookAuthors, and MyApp::Model::DB::Books).

  • The "list" action in our Books controller showed up with a path of /books/list.

Point your browser to http://localhost:3000 and you should still get the Catalyst welcome page.

Next, to view the book list, change the URL in your browser to http://localhost:3000/books/list. You should get a list of the five books loaded by the myapp01.sql script above without any formatting. The rating for each book should appear on each row, but the "Author(s)" column will still be blank (we will fill that in later).

Also notice in the output of the script/myapp_server.pl that DBIC used the following SQL to retrieve the data:

    SELECT me.id, me.title, me.rating FROM books me

because we enabled DBIC_TRACE.

You now have the beginnings of a simple but workable web application. Continue on to future sections and we will develop the application more fully.

CREATE A WRAPPER FOR THE VIEW

When using TT, you can (and should) create a wrapper that will literally wrap content around each of your templates. This is certainly useful as you have one main source for changing things that will appear across your entire site/application instead of having to edit many individual files.

Configure TT.pm For The Wrapper

In order to create a wrapper, you must first edit your TT view and tell it where to find your wrapper file. Your TT view is located in lib/MyApp/View/TT.pm.

Edit lib/MyApp/View/TT.pm and change it to match the following:

    __PACKAGE__->config(
        # Change default TT extension
        TEMPLATE_EXTENSION => '.tt2',
        # Set the location for TT files
        INCLUDE_PATH => [
                MyApp->path_to( 'root', 'src' ),
            ],
        # Set to 1 for detailed timer stats in your HTML as comments
        TIMER              => 0,
        # This is your wrapper template located in the 'root/src'
        WRAPPER => 'wrapper.tt2',
    );

Create the Wrapper Template File and Stylesheet

Next you need to set up your wrapper template. Basically, you'll want to take the overall layout of your site and put it into this file. For the tutorial, open root/src/wrapper.tt2 and input the following:

    <?xml version="1.0" encoding="UTF-8"?>
    <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
    <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
    <head>
    <title>[% template.title or "My Catalyst App!" %]</title>
    <link rel="stylesheet" href="[% c.uri_for('/static/css/main.css') %]" />
    </head>
    
    <body>
    <div id="outer">
    <div id="header">
        [%# Your logo could go here -%]
        <img src="[% c.uri_for('/static/images/btn_88x31_powered.png') %]" />
        [%# Insert the page title -%]
        <h1>[% template.title or site.title %]</h1>
    </div>
    
    <div id="bodyblock">
    <div id="menu">
        Navigation:
        <ul>
            <li><a href="[% c.uri_for('/books/list') %]">Home</a></li>
            <li><a href="[% c.uri_for('/') %]" title="Catalyst Welcome Page">Welcome</a></li>
        </ul>
    </div><!-- end menu -->
    
    <div id="content">
        [%# Status and error messages %]
        <span class="message">[% status_msg %]</span>
        <span class="error">[% error_msg %]</span>
        [%# This is where TT will stick all of your template's contents. -%]
        [% content %]
    </div><!-- end content -->
    </div><!-- end bodyblock -->
    
    <div id="footer">Copyright (c) your name goes here</div>
    </div><!-- end outer -->
    
    </body>
    </html>

Notice the status and error message sections in the code above:

    <span class="status">[% status_msg %]</span>
    <span class="error">[% error_msg %]</span>

If we set either message in the Catalyst stash (e.g., $c->stash->{status_msg} = 'Request was successful!') it will be displayed whenever any view used by that request is rendered. The message and error CSS styles can be customized to suit your needs in the root/static/css/main.css file we create below.

Notes:

  • The Catalyst stash only lasts for a single HTTP request. If you need to retain information across requests you can use Catalyst::Plugin::Session (we will use Catalyst sessions in the Authentication part of the tutorial).

  • Although it is beyond the scope of this tutorial, you may wish to use a JavaScript or AJAX tool such as jQuery (http://www.jquery.com) or Dojo (http://www.dojotoolkit.org).

Create A Basic Stylesheet

First create a central location for stylesheets under the static directory:

    $ mkdir root/static/css

Then open the file root/static/css/main.css (the file referenced in the stylesheet href link of our wrapper above) and add the following content:

    #header {
        text-align: center;
    }
    #header h1 {
        margin: 0;
    }
    #header img {
        float: right;
    }
    #footer {
        text-align: center;
        font-style: italic;
        padding-top: 20px;
    }
    #menu {
        font-weight: bold;
        background-color: #ddd;
    }
    #menu ul {
        list-style: none;
        float: left;
        margin: 0;
        padding: 0 0 50% 5px;
        font-weight: normal;
        background-color: #ddd;
        width: 100px;
    }
    #content {
        margin-left: 120px;
    }
    .message {
        color: #390;
    }
    .error {
        color: #f00;
    }

You may wish to check out a "CSS Framework" like Emastic (http://code.google.com/p/emastic/) as a way to quickly provide lots of high-quality CSS functionality.

Test Run The Application

Restart the development server and hit "Reload" in your web browser and you should now see a formatted version of our basic book list. Although our wrapper and stylesheet are obviously very simple, you should see how it allows us to control the overall look of an entire website from two central files. To add new pages to the site, just provide a template that fills in the content section of our wrapper template -- the wrapper will provide the overall feel of the page.

A STATIC DATABASE MODEL WITH DBIx::Class

Create Static DBIC Schema Files

Unlike the previous DBIC section where we had create=dynamic automatically discover the structure of the database every time the application started, here we will use static schema files for more control. This is typical of most "real world" applications.

One option would be to manually create a separate schema file for each table in the database, however, lets use the same DBIx::Class::Schema::Loader used earlier with create=dynamic to build the static files for us. First, lets remove the schema file created earlier:

    $ rm lib/MyApp/Schema.pm

Now regenerate the schema using the create=static option:

    $ script/myapp_create.pl model DB DBIC::Schema MyApp::Schema create=static dbi:SQLite:myapp.db
     exists "/home/me/MyApp/script/../lib/MyApp/Model"
     exists "/home/me/MyApp/script/../t"
    Dumping manual schema for MyApp::Schema to directory /home/me/MyApp/script/../lib ...
    Schema dump completed.
     exists "/home/me/MyApp/script/../lib/MyApp/Model/DB.pm"

We could have also deleted lib/MyApp/Model/DB.pm, but it would have regenerated the same file (note the exists in the output above). If you take a look at lib/MyApp/Model/DB.pm, it simply contains a reference to the actual schema file in lib/MyApp/Schema.pm along with the database connect string.

If you look in the lib/MyApp/Schema.pm file, you will find that it is no longer using DBIx::Class::Schema::Loader as its base class (DBIx::Class::Schema::Loader is only being used by the helper to load the schema once and then create the static files for us) and Schema.pm only contains a call to the load_classes method. You will also find that lib/MyApp contains a Schema subdirectory, with files inside this directory named according to each of the tables in our simple database (Authors.pm, BookAuthors.pm, and Books.pm). These three files are called "Result Classes" in DBIC nomenclature. Although the Result Class files are named after tables in our database, the classes correspond to the row-level data that is returned by DBIC (more on this later, especially in "EXPLORING THE POWER OF DBIC" in Catalyst::Manual::Tutorial::BasicCRUD.

The idea with the Result Source files created under lib/MyApp/Schema by the create=static option is to only edit the files below the # DO NOT MODIFY THIS OR ANYTHING ABOVE! warning. If you place all of your changes below that point in the file, you can regenerate the automatically created information at the top of each file should your database structure get updated.

Also note the "flow" of the model information across the various files and directories. Catalyst will initially load the model from lib/MyApp/Model/DB.pm. This file contains a reference to lib/MyApp/Schema.pm, so that file is loaded next. Finally, the call to load_classes in Schema.pm will load each of the "result class" files from the lib/MyApp/Schema subdirectory. The end result is that Catalyst will dynamically create three table-specific Catalyst models every time the application starts (you can see these three model files listed in the debug output generated when you launch the application).

NOTE: The version of Catalyst::Model::DBIC::Schema in Debian 5 uses the older DBIC load_classes vs. the newer load_namspaces technique. For new applications, please try to use load_namespaces since it more easily supports a very useful DBIC technique called "ResultSet Classes." We will migrate to load_namespaces in Part 4 (BasicCRUD) of this tutorial.

Updating the Generated DBIC Schema Files

Let's manually add some relationship information to the auto-generated Result Class files. (Note: if you are using a database other than SQLite, such as PostgreSQL, then the relationship could have been automatically placed in the Result Class files. If so, you can skip this step.) First edit lib/MyApp/Schema/Books.pm and add the following text below the # You can replace this text... comment:

    #
    # Set relationships:
    #
    
    # has_many():
    #   args:
    #     1) Name of relationship, DBIC will create accessor with this name
    #     2) Name of the model class referenced by this relationship
    #     3) Column name in *foreign* table (aka, foreign key in peer table)
    __PACKAGE__->has_many(book_authors => 'MyApp::Schema::BookAuthors', 'book_id');
    
    # many_to_many():
    #   args:
    #     1) Name of relationship, DBIC will create accessor with this name
    #     2) Name of has_many() relationship this many_to_many() is shortcut for
    #     3) Name of belongs_to() relationship in model class of has_many() above
    #   You must already have the has_many() defined to use a many_to_many().
    __PACKAGE__->many_to_many(authors => 'book_authors', 'author');

Note: Be careful to put this code above the 1; at the end of the file. As with any Perl package, we need to end the last line with a statement that evaluates to true. This is customarily done with 1; on a line by itself.

Important Note: Although this tutorial uses plural names for both the names of the SQL tables and therefore the Result Classes (after all, Schema::Loader automatically named the Result Classes from the names of the SQL tables it found), DBIC users prefer singular names for these items. Please try to use singular table and DBIC model/Result Class names in your applications. This tutorial will migrate to singular names as soon as possible (patches welcomed). Note that while singular is preferred for the DBIC model, plural is perfectly acceptable for the names of the controller classes. After all, the Books.pm controller operates on multiple books.

This code defines both a has_many and a many_to_many relationship. The many_to_many relationship is optional, but it makes it easier to map a book to its collection of authors. Without it, we would have to "walk" though the book_authors table as in $book->book_authors->first->author->last_name (we will see examples on how to use DBIC objects in your code soon, but note that because $book->book_authors can return multiple authors, we have to use first to display a single author). many_to_many allows us to use the shorter $book->authors->first->last_name. Note that you cannot define a many_to_many relationship without also having the has_many relationship in place.

Then edit lib/MyApp/Schema/Authors.pm and add relationship information as follows (again, be careful to put in above the 1; but below the # DO NOT MODIFY THIS OR ANYTHING ABOVE! comment):

    #
    # Set relationships:
    #
    
    # has_many():
    #   args:
    #     1) Name of relationship, DBIC will create accessor with this name
    #     2) Name of the model class referenced by this relationship
    #     3) Column name in *foreign* table (aka, foreign key in peer table)
    __PACKAGE__->has_many(book_author => 'MyApp::Schema::BookAuthors', 'author_id');
    
    # many_to_many():
    #   args:
    #     1) Name of relationship, DBIC will create accessor with this name
    #     2) Name of has_many() relationship this many_to_many() is shortcut for
    #     3) Name of belongs_to() relationship in model class of has_many() above
    #   You must already have the has_many() defined to use a many_to_many().
    __PACKAGE__->many_to_many(books => 'book_author', 'book');

Finally, do the same for the "join table," lib/MyApp/Schema/BookAuthors.pm:

    #
    # Set relationships:
    #
    
    # belongs_to():
    #   args:
    #     1) Name of relationship, DBIC will create accessor with this name
    #     2) Name of the model class referenced by this relationship
    #     3) Column name in *this* table
    __PACKAGE__->belongs_to(book => 'MyApp::Schema::Books', 'book_id');
    
    # belongs_to():
    #   args:
    #     1) Name of relationship, DBIC will create accessor with this name
    #     2) Name of the model class referenced by this relationship
    #     3) Column name in *this* table
    __PACKAGE__->belongs_to(author => 'MyApp::Schema::Authors', 'author_id');

Run The Application

Run the Catalyst "demo server" script with the DBIC_TRACE option (it might still be enabled from earlier in the tutorial, but here is an alternate way to specify the option just in case):

    $ DBIC_TRACE=1 script/myapp_server.pl

Make sure that the application loads correctly and that you see the three dynamically created model class (one for each of the table-specific schema classes we created).

Then hit the URL http://localhost:3000/books/list with your browser and be sure that the book list is displayed via the relationships established above. You can leave the development server running for the next step if you wish.

Note: You will not see the authors yet because the view does not yet use the new relations. Read on to the next section where we update the template to do that.

UPDATING THE VIEW

Let's add a new column to our book list page that takes advantage of the relationship information we manually added to our schema files in the previous section. Edit root/src/books/list.tt2 add add the following code below the existing table cell that contains book.rating (IOW, add a new table cell below the existing two <td> tags but above the closing </tr> and </table> tags):

    ...
    <td>
      [% # First initialize a TT variable to hold a list.  Then use a TT FOREACH -%]
      [% # loop in 'side effect notation' to load just the last names of the     -%]
      [% # authors into the list. Note that the 'push' TT vmethod does not print -%]
      [% # a value, so nothing will be printed here.  But, if you have something -%]
      [% # in TT that does return a method and you don't want it printed, you    -%]
      [% # can: 1) assign it to a bogus value, or 2) use the CALL keyword to     -%]
      [% # call it and discard the return value.                                 -%]
      [% tt_authors = [ ];
         tt_authors.push(author.last_name) FOREACH author = book.authors %]
      [% # Now use a TT 'virtual method' to display the author count in parens   -%]
      [% # Note the use of the TT filter "| html" to escape dangerous characters -%]
      ([% tt_authors.size | html %])
      [% # Use another TT vmethod to join & print the names & comma separators   -%]
      [% tt_authors.join(', ') | html %]
    </td>
    ...

Then hit "Reload" in your browser (note that you don't need to reload the development server or use the -r option when updating TT templates) and you should now see the number of authors each book has along with a comma-separated list of the authors' last names. (If you didn't leave the development server running from the previous step, you will obviously need to start it before you can refresh your browser window.)

If you are still running the development server with DBIC_TRACE enabled, you should also now see five more SELECT statements in the debug output (one for each book as the authors are being retrieved by DBIC):

    SELECT me.id, me.title, me.rating FROM books me:
    SELECT author.id, author.first_name, author.last_name FROM book_authors me  
    JOIN authors author ON ( author.id = me.author_id ) WHERE ( me.book_id = ? ): '1'
    SELECT author.id, author.first_name, author.last_name FROM book_authors me  
    JOIN authors author ON ( author.id = me.author_id ) WHERE ( me.book_id = ? ): '2'
    SELECT author.id, author.first_name, author.last_name FROM book_authors me  
    JOIN authors author ON ( author.id = me.author_id ) WHERE ( me.book_id = ? ): '3'
    SELECT author.id, author.first_name, author.last_name FROM book_authors me  
    JOIN authors author ON ( author.id = me.author_id ) WHERE ( me.book_id = ? ): '4'
    SELECT author.id, author.first_name, author.last_name FROM book_authors me  
    JOIN authors author ON ( author.id = me.author_id ) WHERE ( me.book_id = ? ): '5'

Also note in root/src/books/list.tt2 that we are using "| html", a type of TT filter, to escape characters such as < and > to &lt; and &gt; and avoid various types of dangerous hacks against your application. In a real application, you would probably want to put "| html" at the end of every field where a user has control over the information that can appear in that field (and can therefore inject markup or code if you don't "neutralize" those fields). In addition to "| html", Template Toolkit has a variety of other useful filters that can found in the documentation for Template::Filters.

RUNNING THE APPLICATION FROM THE COMMAND LINE

In some situations, it can be useful to run your application and display a page without using a browser. Catalyst lets you do this using the scripts/myapp_test.pl script. Just supply the URL you wish to display and it will run that request through the normal controller dispatch logic and use the appropriate view to render the output (obviously, complex pages may dump a lot of text to your terminal window). For example, if you type:

    $ script/myapp_test.pl "/books/list"

You should get the same text as if you visited http://localhost:3000/books/list with the normal development server and asked your browser to view the page source.

OPTIONAL INFORMATION

NOTE: The rest of this part of the tutorial is optional. You can skip to Part 4, Basic CRUD, if you wish.

Using RenderView for the Default View

Once your controller logic has processed the request from a user, it forwards processing to your view in order to generate the appropriate response output. Catalyst uses Catalyst::Action::RenderView by default to automatically performs this operation. If you look in lib/MyApp/Controller/Root.pm, you should see the empty definition for the sub end method:

    sub end : ActionClass('RenderView') {}

The following bullet points provide a quick overview of the RenderView process:

  • Root.pm is designed to hold application-wide logic.

  • At the end of a given user request, Catalyst will call the most specific end method that's appropriate. For example, if the controller for a request has an end method defined, it will be called. However, if the controller does not define a controller-specific end method, the "global" end method in Root.pm will be called.

  • Because the definition includes an ActionClass attribute, the Catalyst::Action::RenderView logic will be executed after any code inside the definition of sub end is run. See Catalyst::Manual::Actions for more information on ActionClass.

  • Because sub end is empty, this effectively just runs the default logic in RenderView. However, you can easily extend the RenderView logic by adding your own code inside the empty method body ({}) created by the Catalyst Helpers when we first ran the catalyst.pl to initialize our application. See Catalyst::Action::RenderView for more detailed information on how to extended RenderView in sub end.

Using The Default Template Name

By default, Catalyst::View::TT will look for a template that uses the same name as your controller action, allowing you to save the step of manually specifying the template name in each action. For example, this would allow us to remove the $c->stash->{template} = 'books/list.tt2'; line of our list action in the Books controller. Open lib/MyApp/Controller/Books.pm in your editor and comment out this line to match the following (only the $c->stash->{template} line has changed):

    =head2 list
    
    Fetch all book objects and pass to books/list.tt2 in stash to be displayed
    
    =cut
    
    sub list : Local {
        # Retrieve the usual Perl OO '$self' for this object. $c is the Catalyst
        # 'Context' that's used to 'glue together' the various components
        # that make up the application
        my ($self, $c) = @_;
    
        # Retrieve all of the book records as book model objects and store in the
        # stash where they can be accessed by the TT template
        $c->stash->{books} = [$c->model('DB::Books')->all];
    
        # Set the TT template to use.  You will almost always want to do this
        # in your action methods (actions methods respond to user input in
        # your controllers).
        #$c->stash->{template} = 'books/list.tt2';
    }

You should now be able to restart the development server as per the previous section and access the http://localhost:3000/books/list as before.

NOTE: Please note that if you use the default template technique, you will not be able to use either the $c->forward or the $c->detach mechanisms (these are discussed in Part 2 and Part 9 of the Tutorial).

Return To A Manually-Specified Template

In order to be able to use $c->forward and $c->detach later in the tutorial, you should remove the comment from the statement in sub list in lib/MyApp/Controller/Books.pm:

    $c->stash->{template} = 'books/list.tt2';

Then delete the TEMPLATE_EXTENSION line in lib/MyApp/View/TT.pm.

You should then be able to restart the development server and access http://localhost:3000/books/list in the same manner as with earlier sections.

AUTHOR

Kennedy Clark, hkclark@gmail.com

Please report any errors, issues or suggestions to the author. The most recent version of the Catalyst Tutorial can be found at http://dev.catalyst.perl.org/repos/Catalyst/Catalyst-Manual/5.70/trunk/lib/Catalyst/Manual/Tutorial/.

Copyright 2006-2008, Kennedy Clark, under Creative Commons License (http://creativecommons.org/licenses/by-sa/3.0/us/).