Catalyst::Manual::Tutorial::03_MoreCatalystBasics - Catalyst Tutorial - Chapter 3: More Catalyst Application Development Basics
This is Chapter 3 of 10 for the Catalyst tutorial.
03_More Catalyst Basics
This chapter of the tutorial builds on the work done in Chapter 2 to explore some features that are more typical of "real world" web applications. From this chapter 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.
Source code for the tutorial in included in the /home/catalyst/Final directory of the Tutorial Virtual machine (one subdirectory per chapter). There are also instructions for downloading the code in Catalyst::Manual::Tutorial::01_Intro.
Please take a look at "STARTING WITH THE TUTORIAL VIRTUAL MACHINE" in Catalyst::Manual::Tutorial::01_Intro before doing the rest of this tutorial. Although the tutorial should work correctly under most any recent version of Perl running on any operating system, the tutorial has been written using the virtual machine that is available for download. The entire tutorial has been tested to be sure it runs correctly in this environment, so it is the most trouble-free way to get started with Catalyst.
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 chapter 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"
Change to application directory and Run "perl Makefile.PL" to make sure your install is complete
And change the "MyApp" directory the helper created:
$ cd MyApp
This creates a similar skeletal structure to what we saw in Chapter 2 of the tutorial, except with MyApp
and myapp
substituted for Hello
and hello
. (As noted in Chapter 2, omit the ".pl" from the command if you are using Strawberry Perl.)
One of the greatest benefits of Catalyst is that it has such a large library of base classes and plugins available that you can use to easily add functionality to your application. 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.
Take a look at the file lib/MyApp.pm that the helper created above. 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.
To be technically correct, it turns out that -Debug
is not a plugin, but a flag. Although most of the items specified on the use Catalyst
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
, -Log
, and -Stats
).
If you prefer, there are several other ways to enable debug output:
the $c->debug
method on the $c
Catalyst context object
the -d
option on the script/myapp_server.pl script
the CATALYST_DEBUG=1
environment variable (or CATALYST_DEBUG=0
to temporarily disable debug output)
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 when needed. 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 (Chapter 5 and Chapter 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 way to serve static content, such as images and CSS files, from the development server.
For our application, we want to add one new plugin to 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 additional plugin, Catalyst::Plugin::StackTrace, to add a stack trace near the top of 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.
Make sure when adding new plugins you also include them as a new dependency within the Makefile.PL file. For example, after adding the StackTrace plugin the Makefile.PL should include the following line:
requires 'Catalyst::Plugin::StackTrace';
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, 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 line.
If you want to see what the StackTrace error screen looks like, edit lib/MyApp/Controller/Root.pm and put a die "Oops";
command in the sub index :Path :Args(0)
method. Then start the development server and open http://localhost:3000/
in your browser. You should get a screen that starts with "Caught exception in MyApp::Controller::Root->index" with sections showing a stacktrace, information about the Request and Response objects, the stash (something we will learn about soon), and the applications configuration. Just don't forget to remove the die before you continue the tutorial! :-)
As discussed earlier, controllers are where you write methods that interact with user input. Typically, controller methods respond to GET
and POST
requests 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/catalyst/MyApp/script/../lib/MyApp/Controller"
exists "/home/catalyst/MyApp/script/../t"
created "/home/catalyst/MyApp/script/../lib/MyApp/Controller/Books.pm"
created "/home/catalyst/MyApp/script/../t/controller_Books.t"
Then edit lib/MyApp/Controller/Books.pm (as discussed in Chapter 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::Book')->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. This is the "Catalyst Context object", and it is automatically passed as the second argument to all Catalyst action methods. It is used to pass information between components and provide access to Catalyst and plugin functionality.
Catalyst Controller 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 there can be an optional space between the colon and the attribute name; 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 method 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 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. You will also have to manually include $c
if you want access to the context object in the method vs. having Catalyst automatically include $c
in the argument list for you if it's a full-fledged action.)
There are five types of "special" built-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
(because of the leading slash). You can use :Args()
to specify how many arguments an action should accept. See "Action-types" in Catalyst::Manual::Intro for more information and 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 be automatically dispatched when servicing a single user request. See Catalyst::Manual::Tutorial::04_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
).
As mentioned in Chapter 2 of the tutorial, views are where you render output, typically for display in the user's web browser (but can generate other types of output such as PDF or JSON). 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 https://masonbook.houseabsolute.com/book/) and HTML::Template.
When using TT for the Catalyst view, the main helper script is Catalyst::Helper::View::TT. You may also come across references to Catalyst::Helper::View::TTSite, but its use is now deprecated.
For our book application, enter the following command to enable the TT
style of view rendering:
$ script/myapp_create.pl view HTML TT
exists "/home/catalyst/MyApp/script/../lib/MyApp/View"
exists "/home/catalyst/MyApp/script/../t"
created "/home/catalyst/MyApp/script/../lib/MyApp/View/HTML.pm"
created "/home/catalyst/MyApp/script/../t/view_HTML.t"
This creates a view called HTML
(the first argument) in a file called HTML.pm
that uses Catalyst::View::TT (the second argument) as the "rendering engine".
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/HTML.pm and you should see something similar to the following:
__PACKAGE__->config(
TEMPLATE_EXTENSION => '.tt',
render_die => 1,
);
And update it to match:
__PACKAGE__->config(
# Change default TT extension
TEMPLATE_EXTENSION => '.tt2',
render_die => 1,
);
This changes the default extension for Template Toolkit from '.tt' to '.tt2'.
You can also configure components in your application class. For example, Edit lib/MyApp.pm and you should see the default configuration above the call to _PACKAGE__->setup
(your defaults could be different depending on the version of Catalyst you are using):
__PACKAGE__->config(
name => 'MyApp',
# Disable deprecated behavior needed by old applications
disable_component_resolution_regex_fallback => 1,
);
Change this to match the following (insert a new __PACKAGE__->config
below the existing statement):
__PACKAGE__->config(
name => 'MyApp',
# Disable deprecated behavior needed by old applications
disable_component_resolution_regex_fallback => 1,
);
__PACKAGE__->config(
# Configure the view
'View::HTML' => {
#Set the location for TT files
INCLUDE_PATH => [
__PACKAGE__->path_to( 'root', 'src' ),
],
},
);
This changes the base directory for your template files from root
to root/src.
Please stick with the settings above for the duration of the tutorial, but feel free to use whatever options you desire in your applications (as with most things in Perl, there's more than one way to do it...).
Note: We will use root/src as the base directory for our template files, with 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).
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. -%]
[%- # Provide a title -%]
[% META title = 'Book List' -%]
[% # Note That the '-' at the beginning or end of TT code -%]
[% # "chomps" the whitespace/newline at that end of the -%]
[% # output (use View Source in browser to see the effect) -%]
[% # Some basic HTML with a loop to display books -%]
<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>
<td></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 (and essentially does nothing at the moment). Meanwhile, the FOREACH
loop iterates through each book
model object and prints the title
and rating
fields.
The [%
and %]
tags are used to delimit Template Toolkit 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 Template::Manual::Variables for details and examples). In addition to the usual Template::Toolkit module Pod documentation, you can access the TT manual at 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.) In Chapter 4 of the tutorial we will explore some extremely helpful and powerful features of DBIx::Class that allow you to pull code out of your views and controllers and place it where it rightfully belongs in a model class.
To test your work so far, first start the development server:
$ script/myapp_server.pl -r
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 chapter of the tutorial.
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 (https://www.sqlite.org), a popular database that is lightweight and easy to use. Be sure to get at least version 3. Open myapp01.sql in your editor and enter:
--
-- Create a very simple database to hold book and author information
--
PRAGMA foreign_keys = ON;
CREATE TABLE book (
id INTEGER PRIMARY KEY,
title TEXT ,
rating INTEGER
);
-- 'book_author' is a many-to-many join table between books & authors
CREATE TABLE book_author (
book_id INTEGER REFERENCES book(id) ON DELETE CASCADE ON UPDATE CASCADE,
author_id INTEGER REFERENCES author(id) ON DELETE CASCADE ON UPDATE CASCADE,
PRIMARY KEY (book_id, author_id)
);
CREATE TABLE author (
id INTEGER PRIMARY KEY,
first_name TEXT,
last_name TEXT
);
---
--- Load some sample data
---
INSERT INTO book VALUES (1, 'CCSP SNRS Exam Certification Guide', 5);
INSERT INTO book VALUES (2, 'TCP/IP Illustrated, Volume 1', 5);
INSERT INTO book VALUES (3, 'Internetworking with TCP/IP Vol.1', 4);
INSERT INTO book VALUES (4, 'Perl Cookbook', 5);
INSERT INTO book VALUES (5, 'Designing with Web Standards', 5);
INSERT INTO author VALUES (1, 'Greg', 'Bastien');
INSERT INTO author VALUES (2, 'Sara', 'Nasseh');
INSERT INTO author VALUES (3, 'Christian', 'Degu');
INSERT INTO author VALUES (4, 'Richard', 'Stevens');
INSERT INTO author VALUES (5, 'Douglas', 'Comer');
INSERT INTO author VALUES (6, 'Tom', 'Christiansen');
INSERT INTO author VALUES (7, 'Nathan', 'Torkington');
INSERT INTO author VALUES (8, 'Jeffrey', 'Zeldman');
INSERT INTO book_author VALUES (1, 1);
INSERT INTO book_author VALUES (1, 2);
INSERT INTO book_author VALUES (1, 3);
INSERT INTO book_author VALUES (2, 4);
INSERT INTO book_author VALUES (3, 5);
INSERT INTO book_author VALUES (4, 6);
INSERT INTO book_author VALUES (4, 7);
INSERT INTO book_author 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 3.7.3
Enter ".help" for instructions
Enter SQL statements terminated with a ";"
sqlite> select * from book;
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 book"
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.
Please note that here we have chosen to use 'singular' table names. This is because the default inflection code for older versions of DBIx::Class::Schema::Loader does NOT handle plurals. There has been much philosophical discussion on whether table names should be plural or singular. There is no one correct answer, as long as one makes a choice and remains consistent with it. If you prefer plural table names (e.g. you think that they are easier to read) then see the documentation in "naming" in DBIx::Class::Schema::Loader::Base (version 0.05 or greater).
For using other databases, such as PostgreSQL or MySQL, see Appendix 2.
Catalyst can be used with virtually any form of datastore available via Perl. For example, Catalyst::Model::DBI can be used to access databases through the traditional Perl DBI interface or you can use a model to access files of any type on the filesystem. However, most Catalyst applications use some form of object-relational mapping (ORM) technology to create objects associated with tables in a relational database, and Matt Trout's DBIx::Class (abbreviated as "DBIC") is the usual choice (this tutorial will use DBIx::Class).
Although DBIx::Class has included support for a create=dynamic
mode to automatically read the database structure every time the application starts, its use is no longer recommended. While it can make for "flashy" demos, the use of the create=static
mode we use below can be implemented just as quickly and provides many advantages (such as the ability to add your own methods to the overall DBIC framework, a technique that we see in Chapter 4).
Note: If you are not following along in the Tutorial Virtual Machine, please be sure that you have version 1.27 or higher of DBD::SQLite and version 0.39 or higher of Catalyst::Model::DBIC::Schema. (The Tutorial VM already has versions that are known to work.) You can get your currently installed version numbers with the following commands.
$ perl -MCatalyst::Model::DBIC::Schema\ 999
$ perl -MDBD::SQLite\ 999
Before you continue, make sure your myapp.db database file is in the application's topmost directory. Now use the model helper with the create=static
option to read the database with DBIx::Class::Schema::Loader and automatically build the required files for us:
$ script/myapp_create.pl model DB DBIC::Schema MyApp::Schema \
create=static dbi:SQLite:myapp.db \
on_connect_do="PRAGMA foreign_keys = ON"
exists "/home/catalyst/MyApp/script/../lib/MyApp/Model"
exists "/home/catalyst/MyApp/script/../t"
Dumping manual schema for MyApp::Schema to directory /home/catalyst/MyApp/script/../lib ...
Schema dump completed.
created "/home/catalyst/MyApp/script/../lib/MyApp/Model/DB.pm"
created "/home/catalyst/MyApp/script/../t/model_DB.t"
Please note the '\' above. Depending on your environment, you might be able to cut and paste the text as shown or need to remove the '\' character to that the command is all on a single line.
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 the lib/MyApp/Model directory.
DBIC::Schema
is the type of the model to create. This equates to Catalyst::Model::DBIC::Schema, the standard way to use a DBIC-based model inside of Catalyst.
MyApp::Schema
is the name of the DBIC schema file written to lib/MyApp/Schema.pm.
create=static
causes DBIx::Class::Schema::Loader to load the schema as it runs and then write that information out into lib/MyApp/Schema.pm and files under the lib/MyApp/Schema directory.
dbi:SQLite:myapp.db
is the standard DBI connect string for use with SQLite.
And finally, the on_connect_do
string requests that DBIx::Class::Schema::Loader create foreign key relationships for us (this is not needed for databases such as PostgreSQL and MySQL, but is required for SQLite). If you take a look at lib/MyApp/Model/DB.pm, you will see that the SQLite pragma is propagated to the Model, so that SQLite's recent (and optional) foreign key enforcement is enabled at the start of every database connection.
If you look in the lib/MyApp/Schema.pm file, you will find that it only contains a call to the load_namespaces
method. You will also find that lib/MyApp contains a Schema
subdirectory, which then has a subdirectory called "Result". This "Result" subdirectory then has files named according to each of the tables in our simple database (Author.pm, BookAuthor.pm, and Book.pm). These three files are called "Result Classes" (or "ResultSource Classes") in DBIx::Class 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::04_BasicCRUD).
The idea with the Result Source files created under lib/MyApp/Schema/Result 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_namespaces
in Schema.pm
will load each of the "Result Class" files from the lib/MyApp/Schema/Result subdirectory. The final outcome 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).
Additionally, the lib/MyApp/Schema.pm model can easily be loaded outside of Catalyst, for example, in command-line utilities and/or cron jobs. lib/MyApp/Model/DB.pm provides a very thin "bridge" between Catalyst and this external database model. Once you see how we can add some powerful features to our DBIC model in Chapter 4, the elegance of this approach will start to become more obvious.
NOTE: Older versions of Catalyst::Model::DBIC::Schema use the deprecated DBIx::Class load_classes
technique instead of the newer load_namespaces
. For new applications, please try to use load_namespaces
since it more easily supports a very useful DBIC technique called "ResultSet Classes." If you need to convert an existing application from "load_classes" to "load_namespaces," you can use this process to automate the migration, but first make sure you have version 0.39
of Catalyst::Model::DBIC::Schema and DBIx::Class::Schema::Loader version 0.05000
or later.
$ # Re-run the helper to upgrade for you
$ script/myapp_create.pl model DB DBIC::Schema MyApp::Schema \
create=static naming=current use_namespaces=1 \
dbi:SQLite:myapp.db \
on_connect_do="PRAGMA foreign_keys = ON"
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 [$c->model('DB::Book')->all]
and delete the next 2 lines):
=head2 list
Fetch all book objects and pass to books/list.tt2 in stash to be displayed
=cut
sub list :Local {
# Retrieve the usual Perl OO '$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::Book')->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::Book')
un-commented above written as $c->model('DB')->resultset('Book')
. 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 is 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::Book')->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.
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. Press Ctrl-C
to break out of the development server and enter:
$ export DBIC_TRACE=1
$ script/myapp_server.pl -r
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 a 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 -r
[debug] Debug messages enabled
[debug] Statistics enabled
[debug] Loaded plugins:
.----------------------------------------------------------------------------.
| Catalyst::Plugin::ConfigLoader 0.30 |
| Catalyst::Plugin::StackTrace 0.11 |
'----------------------------------------------------------------------------'
[debug] Loaded dispatcher "Catalyst::Dispatcher"
[debug] Loaded engine "Catalyst::Engine"
[debug] Found home "/home/catalyst/MyApp"
[debug] Loaded Config "/home/catalyst/MyApp/myapp.conf"
[debug] Loaded components:
.-----------------------------------------------------------------+----------.
| Class | Type |
+-----------------------------------------------------------------+----------+
| MyApp::Controller::Books | instance |
| MyApp::Controller::Root | instance |
| MyApp::Model::DB | instance |
| MyApp::Model::DB::Author | class |
| MyApp::Model::DB::Book | class |
| MyApp::Model::DB::BookAuthor | class |
| MyApp::View::HTML | 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 5.80020
HTTP::Server::PSGI: Accepting connections at http://0:3000
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::Author
, MyApp::Model::DB::BookAuthor
, and MyApp::Model::DB::Book
).
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 DBIx::Class used the following SQL to retrieve the data:
SELECT me.id, me.title, me.rating FROM book 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.
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.
In order to create a wrapper, you must first edit your TT view and tell it where to find your wrapper file.
Edit your TT view in lib/MyApp/View/HTML.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',
);
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 chapter 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 (https://www.jquery.com) or Dojo (https://dojotoolkit.org/).
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.
Hit "Reload" in your web browser and you should now see a formatted version of our basic book list. (Again, the development server should have automatically restarted when you made changes to lib/MyApp/View/HTML.pm. If you are not using the "-r" option, you will need to hit Ctrl-C
and manually restart it. Also note that the development server does NOT need to restart for changes to the TT and static files we created and edited in the root
directory -- those updates are handled on a per-request basis.)
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.
If you take a look at the Schema files automatically generated by DBIx::Class::Schema::Loader, you will see that it has already defined has_many
and belongs_to
relationships on each side of our foreign keys. For example, take a look at lib/MyApp/Schema/Result/Book.pm and notice the following code:
=head1 RELATIONS
=head2 book_authors
Type: has_many
Related object: L<MyApp::Schema::Result::BookAuthor>
=cut
__PACKAGE__->has_many(
"book_authors",
"MyApp::Schema::Result::BookAuthor",
{ "foreign.book_id" => "self.id" },
{ cascade_copy => 0, cascade_delete => 0 },
);
Each Book
"has_many" book_authors
, where BookAuthor
is the many-to-many table that allows each Book to have multiple Authors, and each Author to have multiple books. The arguments to has_many
are:
book_authors
- The name for this relationship. DBIC will create an accessor on the Books
DBIC Row object with this name.
MyApp::Schema::Result::BookAuthor
- The name of the DBIC model class referenced by this has_many
relationship.
foreign.book_id
- book_id
is the name of the foreign key column in the foreign table that points back to this table.
self.id
- id
is the name of the column in this table that is referenced by the foreign key.
See "has_many" in DBIx::Class::Relationship for additional information. Note that you might see a "hand coded" version of the has_many
relationship above expressed as:
__PACKAGE__->has_many(
"book_authors",
"MyApp::Schema::Result::BookAuthor",
"book_id",
);
Where the third argument is simply the name of the column in the foreign table. However, the hashref syntax used by DBIx::Class::Schema::Loader is more flexible (for example, it can handle "multi-column foreign keys").
Note: If you are using older versions of SQLite and related DBIC tools, you will need to manually define your has_many
and belongs_to
relationships. We recommend upgrading to the versions specified above. :-)
Have a look at lib/MyApp/Schema/Result/BookAuthor.pm and notice that there is a belongs_to
relationship defined that acts as the "mirror image" to the has_many
relationship we just looked at above:
=head1 RELATIONS
=head2 book
Type: belongs_to
Related object: L<MyApp::Schema::Result::Book>
=cut
__PACKAGE__->belongs_to(
"book",
"MyApp::Schema::Result::Book",
{ id => "book_id" },
{ join_type => "LEFT", on_delete => "CASCADE", on_update => "CASCADE" },
);
The arguments are similar, but see "belongs_to" in DBIx::Class::Relationship for the details.
Although recent versions of SQLite and DBIx::Class::Schema::Loader automatically handle the has_many
and belongs_to
relationships, many_to_many
relationship bridges (not technically a relationship) currently need to be manually inserted. To add a many_to_many
relationship bridge, first edit lib/MyApp/Schema/Result/Book.pm and add the following text below the # You can replace this text...
comment:
# many_to_many():
# args:
# 1) Name of relationship bridge, 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.
The many_to_many
relationship bridge is optional, but it makes it easier to map a book to its collection of authors. Without it, we would have to "walk" through the book_author
table as in $book->book_author->first->author->last_name
(we will see examples on how to use DBIx::Class objects in your code soon, but note that because $book->book_author
can return multiple authors, we have to use first
to display a single author). many_to_many
allows us to use the shorter $book->author->first->last_name
. Note that you cannot define a many_to_many
relationship bridge without also having the has_many
relationship in place.
Then edit lib/MyApp/Schema/Result/Author.pm and add the reverse many_to_many
relationship bridge for Author
as follows (again, be careful to put in above the 1;
but below the # DO NOT MODIFY THIS OR ANYTHING ABOVE!
comment):
# many_to_many():
# args:
# 1) Name of relationship bridge, 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_authors', 'book');
Run the Catalyst development 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 trace option just in case):
$ DBIC_TRACE=1 script/myapp_server.pl -r
Make sure that the application loads correctly and that you see the three dynamically created model classes (one for each of the Result Classes we created).
Then hit the URL http://localhost:3000/books/list with your browser and be sure that the book list still displays correctly.
Note: You will not see the authors yet because the view isn't taking advantage of these relationships. Read on to the next section where we update the template to do that.
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 and replace the "empty" table cell "<td></td>
" with the following:
...
<td>
[% # NOTE: See Chapter 4 for a better way to do this! -%]
[% # 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 doesn't return -%]
[% # a value, so nothing will be printed here. But, if you have something -%]
[% # in TT that does return a value and you don't want it printed, you -%]
[% # 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>
...
IMPORTANT NOTE: Again, you should keep as much "logic code" as possible out of your views. This kind of logic belongs in your model (the same goes for controllers -- keep them as "thin" as possible and push all of the "complicated code" out to your model objects). Avoid code like you see in the previous example -- we are only using it here to show some extra features in TT until we get to the more advanced model features we will see in Chapter 4 (see "EXPLORING THE POWER OF DBIC" in Catalyst::Manual::Tutorial::04_BasicCRUD).
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 DBIx::Class):
SELECT me.id, me.title, me.rating FROM book me:
SELECT author.id, author.first_name, author.last_name FROM book_author me
JOIN author author ON author.id = me.author_id WHERE ( me.book_id = ? ): '1'
SELECT author.id, author.first_name, author.last_name FROM book_author me
JOIN author author ON author.id = me.author_id WHERE ( me.book_id = ? ): '2'
SELECT author.id, author.first_name, author.last_name FROM book_author me
JOIN author author ON author.id = me.author_id WHERE ( me.book_id = ? ): '3'
SELECT author.id, author.first_name, author.last_name FROM book_author me
JOIN author author ON author.id = me.author_id WHERE ( me.book_id = ? ): '4'
SELECT author.id, author.first_name, author.last_name FROM book_author me
JOIN author 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 < and > 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 be found in the documentation for Template::Filters. (While we are on the topic of security and escaping of dangerous values, one of the advantages of using tools like DBIC for database access or HTML::FormFu for form management [see Chapter 9] is that they automatically handle most escaping for you and therefore dramatically increase the security of your app.)
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 script/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 Ctrl+C
out of the development server and then 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. You can even pipe this HTML text output to a text-based browser using a command like:
$ script/myapp_test.pl "/books/list" | lynx -stdin
And you should see a fully rendered text-based view of your page. (If you are following along in Debian 6, type sudo aptitude -y install lynx
to install lynx.) If you do start lynx, you can use the "Q" key to quit.
NOTE: The rest of this chapter of the tutorial is optional. You can skip to Chapter 4, Basic CRUD, if you wish.
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 perform 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 extend RenderView
in sub end
.
One of the nice features of RenderView
is that it automatically allows you to add dump_info=1
to the end of any URL for your application and it will force the display of the "exception dump" screen to the client browser. You can try this out by pointing your browser to this URL:
http://localhost:3000/books/list?dump_info=1
You should get a page with the following message at the top:
Caught exception in MyApp::Controller::Root->end "Forced debug -
Scrubbed output at /usr/share/perl5/Catalyst/Action/RenderView.pm line 46."
Along with a summary of your application's state at the end of the processing for that request. The "Stash" section should show a summarized version of the DBIC book model objects. If desired, you can adjust the summarization logic (called "scrubbing" logic) -- see Catalyst::Action::RenderView for details.
Note that you shouldn't need to worry about "normal clients" using this technique to "reverse engineer" your application -- RenderView
only supports the dump_info=1
feature when your application is running in -Debug
mode (something you won't do once you have your application deployed in production).
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::Book')->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 access the http://localhost:3000/books/list URL as before.
NOTE: 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 Chapter 2 and Chapter 9 of the Tutorial).
IMPORTANT: Make sure that you do not skip the following section before continuing to the next chapter 4 Basic CRUD.
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/HTML.pm.
Check the http://localhost:3000/books/list URL in your browser. It should look the same manner as with earlier sections.
You can jump to the next chapter of the tutorial here: Basic CRUD
Kennedy Clark, hkclark@gmail.com
Feel free to contact the author for any errors or suggestions, but the best way to report issues is via the CPAN RT Bug system at https://rt.cpan.org/Public/Dist/Display.html?Name=Catalyst-Manual.
Copyright 2006-2011, Kennedy Clark, under the Creative Commons Attribution Share-Alike License Version 3.0 (https://creativecommons.org/licenses/by-sa/3.0/us/).