The Rails Cookbook
A Ruby on Rails Crash Course
The Rails Cookbook: A Ruby on Rails Crash Course
Brian Hogan Published 2008
Dedication
This book is dedicated to the Web Development students at the University of Wisconsin-Eau Claire, past, present, and future.
Table of Contents
1. Introduction ... 1
1. Rails ... 1
2. Basic Concepts ... 1
2.1. MVC Design Pattern ... 1
2.2. Agile Programming Techniques ... 2
2.3. Other features and components of Rails ... 2
3. Installing Ruby on Rails ... 3
2. Your First Rails Project - A Cookbook ... 4
1. Configuring your Database ... 4
1.1. Working with SQLite3 ... 5
2. Creating the Recipes interface using Scaffolding ... 5
2.1. Migrations ... 7
2.2. The migration file ... 7
2.3. Creating the table from the migration ... 7
3. Test it out ... 8
4. How did we get all that? ... 8
3. Validating User Input ... 9
1. Validations ... 9
2. Unit Tests ... 10
2.1. How Tests Work ... 11
2.2. Fixtures ... 11
2.3. Running the Test ... 12
3. Providing Feedback to Users ... 12
4. Cleaning Up the Scaffolding ... 14
1. Cleaning up the List page ... 14
1.1. Helpers ... 15
2. Cleaning up the Show page ... 15
3. Using Partials to share common code ... 16
4. Where’s the rest of my HTML? ... 17
5. Adding Categories ... 18
1. Create a category model and table ... 18
2. Adding some default records with Rake ... 18
3. Modifying the Recipes table ... 19
4. Creating an Association Between a Recipe and a Category ... 19
4.1. Lazy vs. Eager Loading ... 20
5. Adding categories to the controllers and views ... 20
5.1. The New and Edit forms ... 20
5.2. The Show view ... 21
5.3. The List view ... 22
6. Test it out ... 22
6. Other Rails Tidbits ... 23
1. Writing Documentation with RDoc ... 23
2. Annotating Models ... 24
3. Debugging and Exploring with Console ... 24
4. Logging ... 25
5. Writing your own SQL statements ... 25
7. Development and Deployment ... 27
1. Exploring Development Tools ... 27
2. Deploying Rails Applications ... 27
8. Where To Go Next? ... 28
The Rails Cookbook
2. Online Resources ... 28
9. Homework and Exploration ... 29
10. Support This Tutorial ... 30
List of Figures
3.1. User Feedback as provided by Rails validations ... 13
4.1. The modified Index page ... 15
4.2. The Show Page ... 16
6.1. RDoc output in HTML ... 23
6.2. Annotations added to a model ... 24
Chapter 1. Introduction
Get ready to forget everything you know about web developent, because Ruby on Rails will change your world. Web development with Ruby on Rails is one of the fastest ways to build quality applications in a fraction of the time, and all it takes is a little patience and a little time. If you've ever wanted to get started with this technology, this guide is for you.
This simple yet detailed tutorial will guide you through the steps to build a working cookbook application using the Ruby on Rails framework. This guide is meant to expose you to the Rails framework and is not meant to explain all of the concepts in depth. When you’re finished with this tutorial, you will have a better understanding of the basic concepts behind Rails and you should have a basic understanding of how basic Rails applications are structured.
1. Rails
Ruby on Rails is the hot new framework for building database-driven web applications. The Rails frame-work provides developers with the need to create web applications using Agile programming methodolo-gies such as rapid prototyping, test-driven development, and easy refactoring.
Ruby is a dynamically-typed fully object-oriented scripting language used primarily in Japan until it drew the attention of a developer from a small company called 37Signals. The developer, David Heinemeier Hansson, was working on a new project called Basecamp. David felt very limited by the languages he was using to develop the project and so he started working with Ruby, taking advantage of all of the built-in features of the language.
In July 2004, David released the Rails framework which he extracted from his work on Basecamp. Several versions later, Rails has burst onto the scene and attracted the attention of many development shops and industry leaders including Amazon, Oracle, Boeing, Thoughtworks, and many others.
2. Basic Concepts
Let's take a bit and look at the basic concepts of the Rails framework.
2.1. MVC Design Pattern
The MVC or Model-View-Controller pattern explicitly splits up your code and separates business logic from presentation and flow control logic. It also provides mechanisms for easy reuse of code. Traditional web applications written in ASP, PHP, or ColdFusion tend to have scripts intermingled with business logic. Developers can avoid this but without a design pattern such as MVC, the process tends to be trial and error.
2.1.1. The components of MVC
Models contain all business rules and data interaction. All database-related CRUD (Create, Read, Update,
and Delete) code belongs in the model as well. If this pattern is followed correctly, you’ll never write a select statement anywhere outside of the model. Instead, you will access your data by calling a method on the model.
Views are what your end users will see. They are the web pages in a web application, or the user screens
in a desktop application. They should contain very little presentation logic and should be optimized for reuse. Views should never interact directly with models.
Controllers are the glue of the application. Controllers receive user requests, retrieve data from the models,
Introduction
2.1.2. Rails-specific MVC components
The Rails framework divides the View layer into three separate pieces.
Layouts contain your overall template and can be mapped to a controller or a specific view. Instead of
placing and repeating your entire layout HTML in each view page, you centralize it in the layout. There’s no need to split a layout into various pieces like a header and footer either. This makes layouts easy to design.
Helpers provide a mechanism to store presentation-related code. Helpers are very similar to a
“code-behind” page. Rails comes with hundreds of pre-written helpers built in, but Rails makes it easy to define your own. This way you can avoid lots of messy logic code in your view pages.
Partials are pieces of views that need to be used over and over. The web form that lets a user create a new
entry might contain the same fields as the form that the user will use to update or edit that entry. Partials are used to centralize that code for easy reuse.
2.2. Agile Programming Techniques
2.2.1. Test-driven development
Test-driven development (TDD) is a programming methodology where you write tests to prove that your code actually works. In an ideal world, you will write your tests first and then write enough application code to make your tests pass.
For example, a developer will create a unit test to create a new user. The test will fail until the developer actually writes the code that creates the user. The developer writes the code and continues to run the tests until they pass.. If they don’t pass, the developer knows that his or her functional code is wrong. If you write your tests first, the tests are always guaranteed to be current. New tests are added when new features are added, and tests are changed to reflect new requirements.
2.2.2. Refactoring
According to Martin Fowler, a senior consultant at Thoughtworks, the basic idea is that you make small changes to your code to improve your design, making it easier to understand and to modify. Refactoring enables you to evolve your code slowly over time, to take an iterative and incremental approach to pro-gramming. Martin's refactoring site, www.refactoring.com, is a good online resource.
Rails makes refactoring easy. Because of the strict adherence to the MVC pattern, it’s trivial for expe-rienced developers to take an entire section of an application and rewrite it without breaking the entire application.
2.3. Other features and components of Rails
Generate scripts help you create Rails models, views, and controllers.Destroy scripts help you remove Rails models, views, and controllers.
Mongrel is a Ruby-based web server which is used in development and deployment so you can test your
application without having to jump through deployment hoops.
Runner is a script that allows you to execute your application code from outside of the Rails application.
This is useful if you want to use a task scheduling program to periodically invoke some application code.
Introduction
Functional tests contain code you use to test your controllers and views
Migrations allow you to define a database-agnostic schema incrementally, with the ability to roll back your
changes. You can use Migrations to easily develop on SQLite, stage to MySQL, and deploy on Oracle.
Plugins allow developers to add new features to the framework at almost any level. Plugins can be used
to introduce new features, override existing ones, modify the Ruby core classes, or even share common models and controllers across multiple applications. Plugins are easy to write and seamless to use. They allow the Rails core team to deny many feature requests, keeping the Rails framework small and agile
Rake is a Ruby program that runs user-defined tasks. Rails includes many tasks that help you manage
your application.
Finally, Rails provides the ability to “freeze” your application to the current version of Rails that you used for development. A simple Rake task bundles Rails with your application, ensuring that your application remains safe when the Rails framework is updated.
3. Installing Ruby on Rails
Warning
These instructions are for Windows users.
Download the One Click Ruby Installer1 for Windows and run the installer, accepting all of the defaults. Next, open a command prompt and type
gem update --system gem install rails gem install mongrel
This will get your environment set up using the current stable release of Rails. The Gem package manage-ment tool helps you install Ruby libraries that you want to use system-wide. You have to have administra-tive privileges though, so if you don’t, you’ll need to have an administrator run those commands for you.
Working with Older Versions of Rails
This document is written with Rails 2.0.2 in mind. It is possible that the version of Rails that's currently provided by Rubygems is newer. This may cause some problems as things change rapidly with Rails and things are deprecated or moved into plugins. However, it's easy to have multiple versions of the Rails framework installed on your machine, thanks to Rubygems.
The command gem list rails will quickly tell you which versions of Rails you have installed. To install a specific version of Rails, you simply issue this command:
gem install rails -v=2.0.2
To install Rails 1.2.3 in order to follow along with some older books and tutorials, install it with
gem install rails -v=1.2.3
Chapter 2. Your First Rails Project - A
Cookbook
We're going to build a simple cookbook that will let us keep track of our favorite recipes. Each recipe will have ingredients, instructions, and a title. Eventually you'll be able to associate a category to the recipe. Create a new Rails project by opening a command window and typing
rails cookbook
Generating a project using a specific Rails version
If you have multiple versions of Rails installed via RubyGems, you can choose which version of Rails you wish to use for your application. For example, if you need to use Rails 2.0.2 to follow along with this book, you can install Rails 2.0.2 with
gem install rails -v=2.0.2
and then create the cookbook project with
rails _2.0.2_ cookbook
The _2.0.2_ parameter tells RubyGems which version to use. To use Rails 1.2.3 to follow along with the famous Depot application in Agile Web Development with Rails - Second Edition, you simply create the project with
rails _1.2.3_ depot
That's all there is to it.
This command will create the folder cookbook and place a whole bunch of framework files in that folder. As you move through this tutorial, you’ll notice that Rails has many generators that build files for you. Generators are great for file creation and laying down code that’s generic.
1. Configuring your Database
Open the file config/database.yml and review the contents of the file. It should look something like this: 1 development: 2 adapter: sqlite3 3 database: db/cookbook_dev.db 4 5 test: 6 adapter: sqlite3 7 database: db/cookbook_test.db
This is a YAML file. (rhymes with camel) It’s a structured configuration format that maps directly to nested hashes in Ruby and is very common for configurations in Ruby on Rails. Tabs must not be used in YAML files. Instead, two spaces are used for each indentation.
Your First Rails Project - A Cookbook
• Adapter is the database adapter that we want to use. Examples are mysql, sql_server, oracle, postresql, sqlite3, and sqlite. We’re using sqlite3 because it’s easy for beginners, requires no setup, and is the default database for a new Rails project.
• Database is the name of the database. In this case, it’s the path to the database file. Other complex adapters would have you specify the database name or Oracle TNSNames entry here, and then you would have host, username, and password fields as well.
1.1. Working with SQLite3
SQLite 3 is the default database in Rails 2.0. With SQLite3, the database does not need to exist before we start the project; it will be created for you when you run your first migration (but don’t worry about that just yet!) Other databases like MySQL or Microsoft SQL Server require that the database (or schema) exist and that the appropriate privileges are applied. Since we’re trying to get you excited about Rails, we want to keep the momentum going. Using SQLite3 as a database makes it really simple to create a working rapid prototype. You can then move to a different database later, because you’ll define your database tables using pure Ruby code instead of database-specific SQL DDL statements.
1.1.1. Installing SQLite3 on Windows
1. Download the files sqlite3.exe and sqlite3.dll and place them in your c:\ruby\bin folder so that they are available on your path.
Get sqlite3.exe from http://www.sqlite.org/sqlite-3_5_7.zip
Get sqlite3.dll from http://www.sqlite.org/sqlitedll-3_5_7.zip
Extract these files to c:\ruby\bin
Note
If you're on Mac OSX, you already have SQLite3 installed! 2. Open a command prompt and type
gem install sqlite3-ruby
When it finishes installing, you’re good to go!
2. Creating the Recipes interface using
Scaf-folding
Since we’re writing a cookbook, the logical place to start would be with recipes.
Rails uses generators to help you do some common tasks. We’ll use a generator to create a scaffold.
Scaffolding is the much talked about but poorly understood feature of Rails. It’s meant to be a starting
point… to give you a quick interface where you can do some simple testing and get some of the mundane repetitive code written for you. However, scaffolding can only take you so far and is not meant for use in production, hence the name “scaffolding”. There are some steps you’ll need to take to clean up the scaffolding.
Your First Rails Project - A Cookbook
Let’s create a simple interface that will allow us to manage recipes in the system. The scaffold generator creates a model, controller, a set of views, and a migration, or a table definition. At the command prompt, move into your cookbook project folder
cd cookbook
The generate scaffold command takes several parameters. The first parameter is the model name. Model names are singular. The generator will use this model name to create a controller and a definition for a database table. Both of these, by convention, will be pluralized. The second parameter is a string that defines your database table structure. Each field can be specified along with its data type. The scaffold generator uses this information to build the web forms your users will see. They won’t be pretty but they will work.
Type (all on one line)
ruby script/generate scaffold recipe title:string ingredients:text instructions:text
The generator runs, creating the following output:
exists app/models/ exists app/controllers/ exists app/helpers/ create app/views/recipes exists app/views/layouts/ exists test/functional/ exists test/unit/ create app/views/recipes/index.html.erb create app/views/recipes/show.html.erb create app/views/recipes/new.html.erb create app/views/recipes/edit.html.erb create app/views/layouts/recipes.html.erb create public/stylesheets/scaffold.css dependency model exists app/models/ exists test/unit/ exists test/fixtures/ create app/models/recipe.rb create test/unit/recipe_test.rb create test/fixtures/recipes.yml create db/migrate create db/migrate/001_create_recipes.rb create app/controllers/recipes_controller.rb create test/functional/recipes_controller_test.rb create app/helpers/recipes_helper.rb
route map.resources :recipes
The generator created a recipe model, and it also created a controller called
recipes_controller.rb. The controller will contain all of the logic that handles user requests and interacts with the models. In fact, if we look in there, it’s already written it for us! It’s got code to handle creating, editing, listing, viewing, and deleting of recipes. Because these are all common tasks, the gener-ators can do a pretty solid job of handling this for us.
One major problem with scaffolding is that it’s not dynamic. Now that we’ve generated these, we can’t rely on scaffolding any more. Any manual changes we make would be destroyed if we attempted to run the scaffold generator again. That means that if we change the table, we’ll need to modify the views. That’s okay though because we already have a good starting point.
The model we just created requires a database table called “recipes”. Normally, you’d go and create that database table using some sort of SQL statement or visual tool. In Rails, we use migrations.
Your First Rails Project - A Cookbook
2.1. Migrations
Migrations are used to modify your database. You use them to execute DDL statements against your database system. One of the best things about them is that they allow you to define your database as it changes; you can roll your changes back if they don’t work without worrying about goofing up your database.
They’re also an invaluable tool when moving to production. Migrations are supported by all of the Rails database adapters. This means that you can change database systems and apply the migration to the new database which will create your structures for you. This eliminates the need to know the various dialects of data definition languages that may change across database systems. Developers can test with SQLite3, develop with MySQL, and deploy to Oracle.
2.2. The migration file
Open the file db\migrate\001_create_recipes.rb. It’s contents should resemble this:
1 class CreateRecipes < ActiveRecord::Migration 2 def self.up
3 create_table :recipes do |t| 4 t.string :title
5 t.text :ingredients, :instructions 6 t.timestamps
7 end
8 end
9
10 def self.down
11 drop_table :recipes 12 end
13 end
Rails uses the information in this file to create a ‘recipes’ table in the database. Note that the above defi-nition does not include a primary key field. Unless you specify otherwise, Rails will create an ‘id’ column automatically, and will mark it as a primary key.
Why is the table name “recipes” and not “recipe”? Remember that by default, Rails likes table names to be the plural form of your model. It’s pretty smart too because it can do things like person => people and category => categoies. This isn’t mandatory but if we follow these conventions, we can save a few lines of code and skip a few steps. Rails will automatically look for the recipes table when we access the Recipe model in our code.
2.3. Creating the table from the migration
At this point, the table doesn’t actually exist in our database—we just have the “blueprint” for it in Ruby code. To execute the migration, we’ll run the command
rake db:migrate
from our command line. Run that command and you’ll see feedback stating that our recipes table was created.
(in C:/rails/workspace/cookbook)
== 1 CreateRecipes: migrating ================================================ -- create_table(:recipes)
Your First Rails Project - A Cookbook
== 1 CreateRecipes: migrated (0.0310s) =======================================
3. Test it out
Would you believe that’s all you have to do to get a simple application written with Rails? Start the internal server and test out your application. At the command prompt, enter
ruby script/server
and wait a few seconds until you see that the server has in fact started.
Navigate to http://localhost:3000/recipes/ and you should be able to enter a few recipes into the system. Once you’ve entered a few recipes, continue with the tutorial. The application works, but it’s a long way from good.
4. How did we get all that?
When you generated the Recipe model, it created a new class that extends a class called ActiveRecord::Base. This parent class contains all of the functionality needed to create, read, update, and delete records from a database. This parent class makes all kinds of dynamic assumptions about your database connection. As we discovered before, it uses the class name (recipe) and pluralizes it to figure out what database table to use, and It uses your database.yml to find out what database server to use. The first time you access a model in a Rails application, the application connects to the associated database server and queries the database for the information about the table. It uses this information to dynamically build methods for data storage and retrieval.
The Scaffold generator you ran uses that technique to build an interface that will let you create, read, update, delete, and list the rows of the recipes table. It uses the information your model has obtained and then builds HTML forms for data entry, using the data types specified by your database table’s configuration. So instead of having to write code to connect to a database and then build data entry forms, you can use the scaffolding feature of Rails as a starting point. This is just the beginning though. There's a lot more to Rails than just scaffolding an application from a single table.. In fact, most professional Rails developers don’t use scaffolding at all.
Chapter 3. Validating User Input
You may notice that if you entered recipes into the system without filling in any fields, the data was still placed into the database. You will definitely want to require that people enter certain fields into the database.
We’re going to do this with some simple validations and a simple unit test to ensure that our validations work.
In Rails, validation is done in the model. This is actually good practice because the models could be used outside of the Web. The MVC pattern standard is to place all business logic and business rules in the models and have only presentation logic in the views and controllers.
A word about models
Our model class has very little code. In fact, it really has no code at all. As you learned earlier, its parent class does all of the work.
Normally, Ruby classes use methods for accessing instance variables in a class. ActiveRecord, the ORM library used by Rails, attempts to build methods dynamically based on the metadata in your database. The class is mapped to a table and upon first use, ActiveRecord creates an instance of the class and builds the accessor methods (and many many other methods). In production mod, this reflection is only done once until the application is restarted by a server admin.
Each field in our table becomes an accessor (a Ruby term meaning that it can be read and written to). For example, we said a recipe has a title. Our Recipe class will have a method called title
which can be used as a getter and a setter.
1. Validations
Validations are special methods provided by the Validations feature in Active Record. There are quite a few built-in methods we can use to validate data. For simplicity, we’ll use only validates_presence_of on our Recipe model.
Open app/models/recipe.rb and change the contents to
1 class Recipe < ActiveRecord::Base
2 validates_presence_of :title, :ingredients, :instructions 3 end
4
Notice that we use symbols as the parameters to the validates_presence_of method. Symbols are a special type of string used in Ruby to denote a value. One of the interesting things about Ruby is that methods can be defined so they take any number of parameters. In this case, the validates_presence_of
method is taking in an array of symbols which represent the database fields that need to be validated. This simple line of code is all we need to make sure that users enter something for the title, ingredients, and instructions for a recipe. It’s not foolproof, but it’s a good start.
If you want to see what other types of validations are out there, take a look at this page in the Rails API: http://api.rubyonrails.com/classes/ActiveRecord/Valida-tions/ClassMethods.html
Validating User Input
2. Unit Tests
We need to put on the brakes here and write a quick unit test to ensure that our validation works. Unit tests are built right into the Rails framework and are designed to test the functionality of your individual models. Unit tests become vitally important to your development process because they allow you to test your business logic and prove that things are working at the model level. This way you don’t have to keep using a browser to track down bugs that aren’t related to the display.
Rails automatically generated a unit test skeleton for recipe when we generated the recipe model. Open the file test/unit/recipe_test.rb and change the contents to the following:
1 require File.dirname(__FILE__) + '/../test_helper'
2
3 class RecipeTest < ActiveSupport::TestCase 4
5 def test_should_create_valid_record 6 recipe = Recipe.new
7 recipe.title = "Ice water"
8 recipe.ingredients = ["one glass","water","ice"].join("<br>")
9 recipe.instructions = "Combine all ingredients into the glass and let sit for two minutes. Serve immediately."
10 assert_kind_of Recipe, recipe 11 assert recipe.save
12 end
13
14 def test_should_not_save_unless_title_exists 15 recipe = Recipe.new
16 assert !recipe.save # save should fail because there are errors.
17 assert_equal "can't be blank", recipe.errors.on(:title) 18 end
19
20 def test_should_not_save_unless_ingredients_exists 21 recipe = Recipe.new
22 assert !recipe.save # save should fail because there are errors.
23 assert_equal "can't be blank", recipe.errors.on(:ingredients) 24 end
25
26 def test_should_not_save_unless_instructions_exists 27 recipe = Recipe.new
28 assert !recipe.save # save should fail because there are errors.
29 assert_equal "can't be blank", recipe.errors.on(:instructions) 30 end
31 32 end
33
That might look complicated, but it’s really not. We have four methods there…
test_should_create_valid_record,
test_should_not_save_unless_title_exists,
test_should_not_save_unless_ingredients_exists, and
test_should_not_save_unless_instructions_exists.
test_should_create_valid_record simply creates a new instance of Recipe, sets the values for the recipe, and then saves the record. Save should return true, so we use assert to evaluate the value of save. If it evaluates to true, this test passes. If not, it fails. This one needs to pass all the time, as it’s the baseline test. If this test starts failing, that’s an indication that the program’s logic has changed.
Validating User Input
The other three tests simply attempt to save the record without setting one of the required fields. We expect these to fail because of our validations—in this test we haven’t actually provided any of the required fields, but we are testing for only one error at a time to avoid making our tests too complicated.. We’re also asserting the inverse of true for the save. (assert that Recipe.save is not true. Then we assert that the error messages are set for each field. Each validation has its own message format. In this case, the validates_presence_of validation stores “can’t be blank” in the errors collection, under a key for each invalid attribute. If the title isn’t blank, you’ll find the error message for that in the errors collection, under the :title key. How Tests Work
2.1. How Tests Work
Tests actually work by using the test database you defined in database.yml earlier. A test file starts by dumping everything in your database and then updating it so it’s consistent with your development database. Never use the same database for production, development, and tetsting!!!!
Each test file is independent of the others. You can feel free to delete as many records as you want in a test and they will be recreated when you start the test again.
Each method that starts with ‘test_’ will be run as part of the test suite.
2.2. Fixtures
Tests can get data from fixtures. Fixtures are loaded into each test by the fixtures method. You’ll need a fixture for each table in your database, not each model in your system.
Modify the fixture for the recipes table by editing /test/fixtures/recipes.yml and add a few recipes.
Note
Be careful not to use tabs and also be sure to leave a space after each colon! YAML is a tricky little format.
1 ice_water: 2 id: 1
3 title: Ice Cream
4 ingredients: 3 scoops vanilla ice cream<br/chocolate syrup
5 instructions: Scoop ice cream into the bowl and pour chocolate syrup on top. 6 toast:
7 id: 2
8 title: Toast
9 ingredients: bread, butter, jelly
10 instructions: Place bread in the toaster for 1 minute. Remove from toaster and apply butter to each piece.
When the test is run, this data gets loaded into the recipes table and is available within the test. The tests you currently have in your test don’t need these fixtures, but you may write future tests that depend on having data in the test database. For example, you may want to write a test to make sure that there can only be one recipe called “Toast”. That test might look something like this:
1 def test_should_only_have_one_recipe_called_toast 2 @recipe = Recipe.new(:title =>"Toast")
Validating User Input
3 @recipe.valid?
4 Assert @recipe.errors.on(:title).include?("must be unique") 5 end
2.3. Running the Test
To run the test, we need to first prepare the test database. We do that by running the rake task
rake db:test:clone
This task takes the structures from our development database and creates a new test database that we can use over and over again.
Our test is actually a standalone Ruby application. We can run the test directly using Ruby.
ruby test\unit\recipe_test.rb
Everything should work well. You should get no errors or failures.
Note
You can run all of the unit tests by running rake test:units
3. Providing Feedback to Users
Now that we know our validation works, we should see what it looks like when users attempt to leave fields blank.
Active Record’s Validations places error messages in an array which can be accessed by a helper method in a view. The helper method error_messages_for takes in an instance of an Active Record model and attempts to display the error messages in a nice friendly manner. Built in helper methods for text boxes, select boxes, and text areas also are validation-aware. They will become styled automatically, providing additional visual cues. The styles are applied using CSS, so you can modify the way they look. Take a look at Figure 3.1, “User Feedback as provided by Rails validations” to see the results.
Validating User Input
Figure 3.1. User Feedback as provided by Rails validations
This task alone could take a web developer a few hours to get right. We’ve created a working solution with a unit test in only a few minutes. These validations work for new entries and existing entries. There are numerous plugins available for Rails to change how validation works at the web page level. Plugins are available at >http://www.agilewebdevelopment.com/plugins
Chapter 4. Cleaning Up the Scaffolding
As stated before, the scaffolding needs some work before it can be used in production. As you become more familiar with the way Rails works, you may find yourself relying less and less on scaffolding to create your pages.
1. Cleaning up the List page
The list page is pretty good for starters, but one thing that’s kinda rough about it is that it requires Javascript to be enabled in order to delete records! Yuck!.
It might be nice to show just the recipe name and when it was last updated instead of the ingredients and instructions. Remember, the list page is built using whatever fields you specified in your scaffold command.
Replace the contents of the page views/recipes/index.erb.html with the following code:
1 <h1>Listing recipes</h1> 2 3 <table> 4 <tr> 5 <th>Title</th> 6 <th>Last Updated</th> 7 <th colspan="2"> </th> 8 </tr> 9
10 <% for recipe in @recipes %> 11 <tr>
12 <td><%= link_to h(recipe.title), recipe %></td> 13 <td><%= time_ago_in_words recipe.updated_at.to_time %> 14 <td><%= link_to 'Edit', edit_recipe_path(recipe) %></td>
15 <td><%= button_to 'Destroy', recipe, :confirm => 'Are you sure?', :method => :delete %></td> 16 </tr> 17 <% end %> 18 </table> 19 20 <br /> 21
22 <%= link_to 'New recipe', new_recipe_path %>
Refresh the index page in your browser. Your new page should resemble something like the one in Fig-ure 4.1, “The modified Index page”.
Cleaning Up the Scaffolding
Figure 4.1. The modified Index page
1.1. Helpers
Rails provides many helper methods to make displaying pages easier. This example shows a few helpers: • link_to : Used to create a hyperlink. We use this because Rails can manage links for us so we don’t need to worry about relative or absolute paths. Rails has a routing system that makes linking things together a snap.
• h : This simple helper escapes HTML and Javascript. It helps to sanitize output, in case someone put malicious JavaScript into one of your input fields.
• time_ago_in_words : This method takes a time and tells you in words how long ago it was. This is how we get those neat “posted five minutes ago” messages on blogs.
The Rails documentation 1 has more information on helpers.
Why Button_to instead of Link_to
Notice that the delete feature is a button now instead of links? For safety reasons, we want all links to destructive actions like deletes to be called via post methods. The button_to tag does this for us. We can then style it using CSS so it looks nicer later on.
2. Cleaning up the Show page
Take a look at app/controller/recipes_controller.rb and find the show section
1 def show
2 @recipe = Recipe.find(params[:id]) 3 end
This code retrieves the recipe from the database. The id sent via the url is sent to the database. This actually generates the sql statement “select * from recipes where id = 1”. The find method is a class method of Recipe and returns an instance of a recipe object. Models in Active Record handle the retrieval and representation of the data.
Cleaning Up the Scaffolding
Here we see that it stores the resulting Recipe instance into an instance variable (The @ means instance variable in Ruby.) The instance variable is passed on to the show.html.erb file. Knowing that, we can easily display the information about our recipe.
Open app/views/recipe/show.html.erb and replace the contents with
1 <h2><%=h @recipe.title %> [<%= link_to 'Edit', edit_recipe_path(@recipe) %> ]</h2> 2 3 <h3>Ingredients:</h3> 4 <p><%=h @recipe.ingredients %></p> 5 6 <h3>Instructions</h3> 7 <p><%=h @recipe.instructions %></p> 8
9 <%= link_to 'Back', recipes_path %>
Your page should look something like Figure 4.2, “The Show Page”
Figure 4.2. The Show Page
3. Using Partials to share common code
The New and Edit forms are virtually identical except for their headings. The actual form could be shared across both files using a partial which is similar to an include file in PHP.. This way, if you add a field to the form, you only need to add it to one file instead of two. In Rails 1.x, this was handled by the scaffold generator, but in Rails 2.0, this is no longer the case. Let’s add it back.
Create a new file in app/views/recipes called _form.html.erb. The filename should begin with an underscore because that’s how Rails distinguishes partials from regular views. Open up the
new.html.erb file and find this code:
1 <p> 2 <b>Title</b><br /> 3 <%= f.text_field :title %> 4 </p> 5 6 <p> 7 <b>Ingredients</b><br /> 8 <%= f.text_area :ingredients %> 9 </p> 10 11 <p> 12 <b>Instructions</b><br /> 13 <%= f.text_area :instructions %> 14 </p>
Cleaning Up the Scaffolding
Copy that code to your clipboard and paste it into _form.html.erb. Once it's pasted, remove the code from the new.html.erb file.. Open edit.html.erb and locate the same code. Remove it from the file.
Now add this line to edit.html.erb, in place of the code you just removed:
1 <%= render :partial=>"form", :locals=>{:f => f} %>
The :locals => {:f => f} option allows you to pass variables into the partial. Since the variable
f for the form is a local variable, it needs to be passed to the partial so the partial can “see” it and access it. The edit.html.erb file should now look like this:
1 <h1>Editing recipe</h1> 2 3 <%= error_messages_for :recipe %> 4 5 <% form_for(@recipe) do |f| %> 6
7 <%= render :partial=>"form", :locals=>{:f => f} %> 8 9 <p> 10 <%= f.submit "Update" %> 11 </p> 12 <% end %> 13
14 <%= link_to 'Show', @recipe %> | 15 <%= link_to 'Back', recipes_path %>
Add the same line of code to new.html.erb. in place of the code you previously removed. You’re now sharing code between two views using a partial.
4. Where’s the rest of my HTML?
You’ve probably noticed that our view pages have not included any required HTML elements, nor have they mentioned anything about a style sheet. Yet we can see styles being rendered and we see a lot of HTML markup if we view the source. Where’s this coming from?
Rails has a wonderfully complex yet simple template mechanism. Look in the app/vies/layout folder and you’ll see a file called recipes.html.erb. This file was created during your scaffold operation and is automatically linked to the recipes controller.
This file wraps any view files you render from within the recipes controller. If you want one single layout for your entire application, you can rename this file to application.html.erb and every controller will use it.
This file can use any variables set in the controller, including variables created within the view files them-selves, because it is read last before being sent to the browser. This means you can set the page title in the individual controller actions or even in the .erb files themselves.
Chapter 5. Adding Categories
A recipe should be able to be categorized. For this example, we’ll just say that a recipe can only belong to one category, and a category can have many recipes. We’ll say it that way because that’s how Active Record allows us to express relationships.
1. Create a category model and table
We don’t need to create a full interface to manage categories at this time, so create the new model by dropping to a command prompt and typing
ruby script/generate model Category
Open the created migration file at db/migrate/002_create_categories.rb and change it to
1 class CreateCategories < ActiveRecord::Migration 2 def self.up
3 create_table :categories do |t| 4 t.string :name 5 t.timestamps 6 end 7 8 end 9
10 def self.down
11 drop_table :categories 12 end
13 end
Run the migration by executing the command rake db:migrate from the command line. It will create the new table.
2. Adding some default records with Rake
Sometimes it’s nice to have your database pre-populated with records. You saw how fixtures can do that with test data, but that’s not always a good choice. Migrations could be used to insert data into your database but that can be volatile as well. The best approach is to use rake, which is the same tool you’ve been using to run your migrations.
Rake is an automation language. To use it, you simply create a file with some tasks and then execute it via the rake command.
Rails projects look for Rake tasks in files with the .rake extension in the project’s lib/tasks folder. Create a new file in that folder called import.rake. Place this code in the file:
1 namespace :db do
2
3 desc "Puts default categories in the database"
4 task :import_categories => :environment do
5
6 Category.create :name =>"Beverages"
7 Category.create :name =>"Deserts"
8 Category.create :name =>"Appetizers"
9 Category.create :name =>"Entrees"
Adding Categories
11 Category.create :name =>"Sandwiches"
12
13 end
14 end
A namespace is just a container for code like in any other language. When you issued the command rake db:migrate you called the migrate task within the db namespace. We’ll follow that same convention here. To import the records into your database ,issue the command
rake db:import_categories
3. Modifying the Recipes table
Since we want to have a relationship between categories and recipes, we have to place a foreign key in the recipes table so it can be associated with a recipe. We’ll do this with a migration too.
Create a new migration. From the command line, execute the command
ruby script/generate migration recipe_category
This will create a new migration file db/migrate/003_RecipeCategory.rb. Open the file and replace it with the following code:
1 class RecipeCategory < ActiveRecord::Migration 2 def self.up
3 add_column :recipes, :category_id, :integer 4 end
5
6 def self.down
7 remove_column :recipes, :category_id 8 end
9 end
Run the migration to alter the database. (rake db:migrate).
At this point you will need to stop and restart your web server. This is only necessary with SQLite3, and only because you changed a table's structure.. Other databases can be modified without restarting the web server. Press CTRL+BREAK to stop Mongrel and then restart it by executing the command
ruby script/server
4. Creating an Association Between a Recipe
and a Category
Associations allow objects to interact. Associations are methods that map the primary keys of one table to the foreign keys of another; the relational mapping part of “object-relational mapping”.
Open app/models/recipe.rb and modify its contents with the following code:
1 class Recipe < ActiveRecord::Base 2 belongs_to :category
3 validates_presence_of :title, :ingredients, :instructions 4 end
Adding Categories
5
The belongs_to method takes in a symbol name of a class with which we wish to associate. Rails needs no further information because it will assume that :category references a class called Category, that the table name will be categories, and that this table (recipes) has a foreign key column called category_id
that will reference the id column in the categories table. Of course, we can override these assumptions, but it’s easier just to follow convention.
This association adds some new methods to an instance of Recipe. We can now access the name of the recipe directly.
recipe = Recipe.find(1) # gets recipe with id of 1 recipe.category.name # gets the associated category name
4.1. Lazy vs. Eager Loading
The above code will do the retrieval using lazy loading, meaning that two SQL statements will be called. This could be bad if we were retrieving all recipes and displaying the category for each one. If you have 200 recipes, the above code would generate 201 SQL statements!
Thankfully there is a solution for situations like this… eager loading. Rails will generate proper left joins for us if we specify the objects to include.
recipe = Recipe.find(1)
becomes
recipe = Recipe.find(1, :include => [:category])
and now only one statement is sent to the database.
5. Adding categories to the controllers and
views
In order to add the category selection to the forms and views, we need to do some work in the controller.
5.1. The New and Edit forms
Open app/controller/recipes_controller.rb and locate the new method. Modify it so it retrieves the categories into an instance variable called @categories. Remember that a controller’s instance variables are then accessible in the view pages.
1 def new
2 @recipe = Recipe.new
3 @categories = Category.find :all 4 end
Now find the edit method and modify it so it also retrieves the categories into @categories.
Adding Categories
2 @recipe = Recipe.find(params[:id]) 3 @categories = Category.find :all 4 end
Open app/views/recipes/_form.html.erb and add the following block at the end of the file:
1 <p>
2 <b>Category</b><br />
3 <%= f.select :category_id, @categories.collect{|c| [c.name, c.id] }, :include_blank => true %>
4 </p>
This code adds a select box which will contain all of the categories so that your users can place a recipe into the category chosen by the dropdown. This is an example of the select helper.
The .collect method iterates through all the categories and returns an array. In this case we’re returning an array of arrays which the select helper can use to build the form. This is an example of a block and you’ll see lots of these in Rails applications.
We’re also including a blank option so that a user doesn’t have a category already selected when they view the page.
5.2. The Show view
When we display a recipe, we want to now show the category for the recipe. We can do that easily thanks to the way the belongs_to association works. When we associated a category to a recipe using that association, it added a method to our Recipe model called category, which returns an instance of the associated category record.
Locate the show action in recipes_controller and add eager loading for the category using the
:include option for find.
1 def show
2 @recipe = Recipe.find(params[:id], :include=>[:category]) 3
4 respond_to do |format| 5 format.html # show.html.erb
6 format.xml { render :xml => @recipe } 7 end
8 end
Open app/views/recipes/show.html.erb and add
<p>Category: <%= h(@recipe.category.name) rescue “No category found” %></p>
somewhere on the page. When you refresh, you'll see the category displayed.
Tip
The rescue statement catches a possible exception that could be thrown if the recipe does not yet have an assigned category. Your recipes don't all have the category assigned yet, and without this rescue statement, this page would fail to render.
Adding Categories
5.3. The List view
Find the index action in recipes_controller.rb and add the :include option to the find to eager-load the category information just like the previous example.
1 def index
2 @recipes = Recipe.find(:all, :include=>[:category]) 3
4 respond_to do |format|
5 format.html # index.html.erb
6 format.xml { render :xml => @recipes } 7 end
8
9 end
Open app/views/recipes/index.html.erb and modify it so you can see the category name in the table. You’ll need to add a column heading as well as the data cell itself. Remember to use the association to retrieve the category name just like you did on the show page!
6. Test it out
With the associations in place and the views fixed up, go ahead and play with your application. Notice how you can now add categories to your recipes, and when you edit an existing recipe, its associated category automatically shows up in the dropdown list.
Chapter 6. Other Rails Tidbits
There's a lot more to the Rails framework than what we covered here. Let's explore a few other features.
1. Writing Documentation with RDoc
Documenting code is one of the most useful things a developer can do. Unfortunately it’s often done poorly if it’s even done at all.
Ruby on Rails aims to change how developers write documentation by making use of RDoc. RDoc is a program that can parse Ruby files for comments and convert these comments to HTML pages or other formats. It generates very clean and nice-looking documentation and is so easy to use that developers quickly come to actually enjoying documentation.
Any comments located directly above a class or method declaration will be interpreted by the RDOC parser to be the comments for that given block of code.
Here’s an example of some commented code.
1 #=Recipes
2 # Recipes are added, removed, maintained, and viewed using
3 # the actions in this controller.
4 #==Authentication
5 # There is no authentication on this controller
6 class RecipesController < ApplicationController 7
8 # This action handles the default document (Index) and
9 # simply redirects users to the list action.
10 def index 11 list
12 render :action => 'list'
13 end
14 15 end
When we run the command rake appdoc, our HTML documentation will be created for us. See Figure 6.1, “RDoc output in HTML”
Other Rails Tidbits
2. Annotating Models
One of the things that can get confusing with Rails is the fact that the models don’t have much code. Because the methods are generated for you, you can’t look at a model and tell what database fields you refer to unless you were to comment them yourself.
Thankfully we can handle this simply by installing a plugin called annotate_models written by the creator of RDoc, Dave Thomas. Install the plugin by executing this command from the command line
ruby script/plugin install http://repo.pragprog.com/svn/Public/plug-ins/annotate_models/
Note
You’ll need to have the Subversion client tools installed on your machine for this to work and the svn command needs to be on your path. Visit http://subversion.tigris.org/ servlets/ProjectDocumentList?folderID=91 to get the latest version if you don't have it..
After installation, be sure to close any open command prompt windows. Then re-open your command prompt window and navigate back to the root of your project.)
Once installed, run the command rake annotate_models
Your models will now be commented with the schema definition. This will then be placed in your docu-mentation the next time you generate your docs. You can see the output in Figure 6.2, “Annotations added to a model”
Figure 6.2. Annotations added to a model
3. Debugging and Exploring with Console
One of the ways you can debug a Rails application is with unit tests. However, often times you might not know what to write. That’s where console comes in. Console lets you load the entire Rails application into the Interactive Ruby environment so you can execute commands and interact with your application. From the root of your project, execute the command
Other Rails Tidbits
ruby script/console
to enter the console. Once the console is loaded, you can start experimenting with your objects, as shown in Figure 6.3, “Using the Rails Console to work with objects”
Figure 6.3. Using the Rails Console to work with objects
In the above example, I use the console to create a new recipe and save it to the database. I then retrieve the id for the recipe. Then I use the find method on Recipe to locate the recipe again. Then I see if it has a category. Of course, it doesn’t so I fetch a category from the database and assign it to my instance. The association is not saved until I execute the save method of my instance.
This is just the beginning, but it shows how you can use the console to learn more about how the methods on the classes work without having to write any view pages or controller code.
4. Logging
Rails applications automatically log requests and responses to the various logs. One log you should really keep an eye on is your development.log file. It contains a lot of useful information such as the parameters sent on each request as well as the SQL statements created by Rails and sent to the database.
This is the place you’ll want to look to tune your application. Not only can you see if you’re executing too many SQL statements for the job at hand, but you can also see how long it took Rails to serve the request to your client.
5. Writing your own SQL statements
At first glance, Rails may seem limited. We’ve gone through this entire project without writing any SQL. A lot of the time we won’t have to worry about it. However, it is still very possible for us to get into the code and do what we need to do.
For example, one of the methods in Active Record is called find_by_sql which allows us to look records up using our own custom SQL statement.
@results = Recipe.find_by_sql "select r.title, c.name from recipes r
Other Rails Tidbits
on r.category_id = c.id"
=> [#<Recipe:0x3750478 @attributes={"name"=>"Beverages", "title"=>"Test"}>]
You have to understand Ruby to understand what this example returns, so I’ll help out. The square brackets ([]) surrounding the result means that you're dealing with an Array. The #<Recipe piece means it’s a Recipe object. So when you use the find_by_sql method, you receive an array of objects which you can then iterate over.
@results.each do |recipe|
puts recipe.name # print to STDOUT
puts recipe.title # print to STDOUT
end
Note that a new method title has been created in the instance of the class. Active Record inspected the column names that came back from the database and dynamically created accessor methods for us to use.
Warning
Never use puts in your Rails application directly. It can cause problems that you may not find later on. It’s only to be used in tests and in the console to help you debug. If you’re wondering, it’s equivalent to system.out.println or echo.
There are many other features in Rails that make it extremely flexible. Don’t get fooled by the hype about ORM and scaffolding. Rails is much more than that!
Chapter 7. Development and
Deployment
1. Exploring Development Tools
There are several different tools you can use to easily build Rails applications.
• If you’re on a Mac, you have to check out TextMate. While not an IDE, it is built with Rails development in mind and is just a dream to work with when you see it. It’s the featured editor in all of the Rails screencasts. See http://macromates.com/ for more information.
• Those interested in a full-blown IDE for Rails development will love NetBeans Ruby IDE from Sun (yup, the Java guys). Snag your copy at http://download.netbeans.org/netbeans/6.0/final/ and be sure to grab the one for Ruby. You will need the Java SDK as well to run this, but it’s worth it.
• Windows users might like to give E-Texteditor a try. It’s a clone of Textmate, and while not nearly as feature-rich yet, it’s one of the best editors available on Windows so far.
• Advanced users will definitely want to look at Cream with the rails.vim plugin Cream: http://cream.sourceforge.net/
Rails.vim plugin: http://www.vim.org/scripts/script.php?script_id=1567
2. Deploying Rails Applications
Deploying Rails applications is no trivial task. Rails applications require more setup work than PHP ap-plications and there are lots of things that can go wrong when you’re first starting out. Do not attempt deployment until you are very comfortable with the Rails framework.
Applications can be deployed using FastCGI and Lighttpd or Apache, or by using the Mongrel web server alongside a load-balancing mechanism like Apache 2.2, Pound, Pen, or Nginx.
There are many web hosting companies that support Rails Shared hosts such as Dreamhost, Site5, Textdrive, RailsPlayground, and Bluehost keep the cost low by sharing space and memory with other users. This is a great low-cost way to launch your application.
If you need high availability, you can look at RailsMachine and EngineYard, two solutions that promise to host large Rails sites with ease. They are significantly more expensive than shared hosting plans. En-gineYard is incredible, but you can’t afford it unless you are really making money. It’s worth every penny though.
If you just want to set things up yourself on dedicated virtual servers, you could look at LayeredTech, Linode, and Rimuhosting. Rimuhosting provides good support for Rails on their virtual servers, as does Linode.
More information on deployment can be found in the book Deploying Rails Applications from the Prag-matic Programmers.
Chapter 8. Where To Go Next?
Hopefully this exercise gave you enough of an overview of the Ruby on Rails framework to see the po-tential impact it has on rapid application development. From here, you should be able to extend this as-signment by doing the homework or explore further by coming up with your own application, using this as a model.
1. Books
• Learn to Program (Chris Pine) http://www.pragprog.com/titles/fr_ltp
• Agile Web Development with Rails (Dave Thomas)
http://www.pragmaticprogrammer.com/titles/rails/index.html • Ruby for Rails (David A. Black)
http://www.manning.com/black/ • Programming Ruby (Dave Thomas)
http://www.pragprog.com/titles/ruby3 or older version online for free at http://www.rubycentral.com/ book/
• Deploying Rails Applications (Ezra Zygmuntowicz, Bruce Tate, Clinton Begin, Geoffrey Grosenbach, and Brian Hogan)
http://www.pragprog.com/titles/fr_deploy
• Rails for Java Developers (Stuart Halloway and Justin Gehtland) http://www.pragprog.com/titles/fr_r4j
• Rails for PHP Developers (by Derek DeVries and Mike Naberezny) http://www.pragprog.com/titles/ndphpr
• The Rails Way (Obie Fernandez)
http://www.amazon.com/Rails-Way-Addison-Wesley-Professional-Ruby/dp/0321445619
2. Online Resources
• Ruby on Rails Discussion group (http://groups.google.com/group/rubyonrails-talk) • #rubyonrails IRC channel (http://wiki.rubyonrails.org/rails/pages/IRC)
• Peepcode (http://peepcode.com/) • Railscasts (http://railscasts.com/)
Chapter 9. Homework and Exploration
If you want to take the exercise further, you can try some of the ideas below. The answers are not provided but the solutions for each problem can be found by looking at similar exercises in the preceding tutorial. 1. Document some other methods in your controllers and models and then regenerate the docs. Here are
some simple formatting symbols: • # is the comment
• = is a large header • == is a level 2 header • --- is a horizontal rule • * is a bullet
2. Create a has_many association from Category to :recipes
• Write a unit test that tests this relationship. (You’ll need to make fixtures for categories and recipes and you’ll need to load both of these in the test file.)
• See http://api.rubyonrails.org/classes/ActiveRecord/Associations/ClassMethods.html for details on has_many
3. Create a controller and views to manage the categories. Use the recipe controller and views as an ex-ample. If you choose to scaffold this, you may need to undo some work you’ve done so be careful. It’s easier to write it by hand.
• When you display a category, display the recipes associated with that category, making use of the has_many association you created.
• On the Show page for a recipe, make a link to the category
<%=link_to @recpie.category, show_category(@recipe.category) %>
4. Extra Credit: When a category is deleted, set all associated recipes to a nil category using a before_destroy ActiveRecord callback.
Chapter 10. Support This Tutorial
This Rails crash-course has been supported by many wonderful donors just like you. If you found this useful, please send a small donation via Paypal by visiting http://www.napcs.com/re-sources/rails/cookbook. The latest version of this book is always available there as well. Thank you for your support.
Index
, 15A
accessor, 9 Array, 26 Associations, 19 belongs_to, 20C
Console, 24 Controllers, 1D
database.yml, 4 DatabasesAdding default data, 18 Eager Loading, 20 Documentation
Annotating Models with schema info, 24 generating documentation with RDOC, 23
E
Exception Handling rescue, 21F
Fixtures, 11G
Generators, 5 scaffold, 6H
Helper methods link_to, 15 Helpers, 2L
Layouts, 2, 17 default layout, 17M
Migrations, 7 running, 7 Model-View-Controller pattern, 1 Models, 1 database reflection, 8 dynamic method creation, 9P
Partials, 2, 16
passing variables into, 17
R
Rake
automating tasks with, 18 Rake tasks
Cloning the test database from the development database, 12 RDoc, 23 example, 23 Generating documentation, 23 Refactoring, 2 Relationships, 19 see Associations, 19 Rubygems Gems, 3 multiple versions, 3
S
Scaffolding, 5 issues with, 6 script/server, 8 SQLstatements in the Rails logs, 25 writing your own, 25
SQLite3 automatic creation, 5 installation of, 5 Symbols, 9
