Blogger 2
In this project you’ll create a simple blog system and learn the basics of Ruby on Rails including:
- Models, Views, and Controllers (MVC)
- Data Structures & Relationships
- Routing
- Migrations
- Views with forms, partials, and helpers
- RESTful design
- Adding gems for extra features
This tutorial is open source. If you notice errors, typos, or have questions/suggestions, please submit them to the project on GitHub.
I0: Up and Running
Part of the reason Ruby on Rails became popular quickly is that it takes a lot of the hard work off your hands, and that’s especially true in starting up a project. Rails practices the idea of "sensible defaults" and will, with one command, create a working application ready for your customization.
Setting the Stage
First we need to make sure everything is set up and installed. See the Environment Setup page for instructions on setting up and verifying your Ruby and Rails environment.
This tutorial targets Rails 4.0.0, and may need slight adaptations for other versions. Let us know if you run into something strange!
From the command line, switch to the folder that will store your projects. For instance, I use /Users/jcasimir/projects/
. Within that folder, run the following command:
Terminal
$
|
|
Use cd blogger
to change into the directory, then open it in your text editor. If you’re using Sublime Text you can do that with subl .
.
Project Tour
The generator has created a Rails application for you. Let’s figure out what’s in there. Looking at the project root, we have these folders:
app
- This is where 98% of your effort will go. It contains subfolders which will hold most of the code you write including Models, Controllers, Views, Helpers, JavaScript, etc.bin
- This is where your app’s executables are stored:bundle
,rails
,rake
, andspring
.config
- Control the environment settings for your application. It also includes theinitializers
subfolder which holds items to be run on startup.db
- Will eventually have amigrations
subfolder where your migrations, used to structure the database, will be stored. When using SQLite3, as is the Rails default, the database file will also be stored in this folder.lib
- This folder is to store code you control that is reusable outside the project.log
- Log files, one for each environment (development, test, production)public
- Static files can be stored and accessed from here, but all the interesting things (JavaScript, Images, CSS) have been moved up toapp
since Rails 3.1test
- If your project is using the defaultTest::Unit
testing library, the tests will live heretmp
- Temporary cached filesvendor
- Infrequently used, this folder is to store code you do not control. With Bundler and Rubygems, we generally don’t need anything in here during development.
Configuring the Database
Look in the config
directory and open the file database.yml
. This file controls how Rails’ database connection system will access your database. You can configure many different databases, including SQLite3, MySQL, PostgreSQL, SQL Server, and Oracle.
If you were connecting to an existing database you would enter the database configuration parameters here. Since we’re using SQLite3 and starting from scratch, we can leave the defaults to create a new database, which will happen automatically. The database will be stored in db/development.sqlite3
Starting the Server
Let’s start up the server. From your project directory:
Terminal
$ |
|
You’re ready to go!
Viewing the App
Open any web browser and enter the address http://0.0.0.0:3000
. You can also use http://localhost:3000
or http://127.0.0.1:3000
– they are all "loopback" addresses that point to your machine.
You’ll see the Rails’ "Welcome Aboard" page. Click the "About your application’s environment" link and you should see the versions of various gems. As long as there’s no big ugly error message, you’re good to go.
Getting an Error?
If you see an error here, it’s most likely related to the database. You are probably running Windows and don’t have either the SQLite3 application installed or the gem isn’t installed properly. Go back to Environment Setup and use the Rails Installer package. Make sure you check the box during setup to configure the environment variables. Restart your machine after the installation and give it another try.
Creating the Article Model
Our blog will be centered around "articles," so we’ll need a table in the database to store all the articles and a model to allow our Rails app to work with that data. We’ll use one of Rails’ generators to create the required files. Switch to your terminal and enter the following:
Terminal
$
|
|
Note that we use bin/rails
here but we used rails
previously. The rails
command is used for generating new projects, and the bin/rails
command is used for controlling Rails.
We’re running the generate
script, telling it to create a model
, and naming that model Article
. From that information, Rails creates the following files:
db/migrate/(some_time_stamp)_create_articles.rb
: A database migration to create thearticles
tableapp/models/article.rb
: The file that will hold the model codetest/models/article_test.rb
: A file to hold unit tests forArticle
test/fixtures/articles.yml
: A fixtures file to assist with unit testing
With those files in place we can start developing!
Working with the Database
Rails uses migration files to perform modifications to the database. Almost any modification you can make to a DB can be done through a migration. The killer feature about Rails migrations is that they’re generally database agnostic. When developing applications developers might use SQLite3 as we are in this tutorial, but in production we’ll use PostgreSQL. Many others choose MySQL. It doesn’t matter – the same migrations will work on all of them! This is an example of how Rails takes some of the painful work off your hands. You write your migrations once, then run them against almost any database.
Migration?
What is a migration? Let’s open db/migrate/(some_time_stamp)_create_articles.rb
and take a look. First you’ll notice that the filename begins with a mish-mash of numbers which is a timestamp of when the migration was created. Migrations need to be ordered, so the timestamp serves to keep them in chronological order. Inside the file, you’ll see just the method change
.
Migrations used to have two methods, up
and down
. The up
was used to make your change, and the down
was there as a safety valve to undo the change. But this usually meant a bunch of extra typing, so with Rails 3.1 those two were replaced with change
.
We write change
migrations just like we used to write up
, but Rails will figure out the undo operations for us automatically.
Modifying change
Inside the change
method you’ll see the generator has placed a call to the create_table
method, passed the symbol :articles
as a parameter, and created a block with the variable t
referencing the table that’s created.
We call methods on t
to create columns in the articles
table. What kind of fields does our Article need to have? Since migrations make it easy to add or change columns later, we don’t need to think of everything right now, we just need a few to get us rolling. Let’s start with:
title
(a string)body
(a "text")
That’s it! You might be wondering, what is a "text" type? This is an example of relying on the Rails database adapters to make the right call. For some DBs, large text fields are stored as varchar
, while others like Postgres use a text
type. The database adapter will figure out the best choice for us depending on the configured database – we don’t have to worry about it.
Add those into your change
like this:
1 2 3 4 5 6 7 8 |
|
Timestamps
What is that t.timestamps
doing there? It will create two columns inside our table named created_at
and updated_at
. Rails will manage these columns for us. When an article is created its created_at
and updated_at
are automatically set. Each time we make a change to the article, the updated_at
will automatically be updated.
Running the Migration
Save that migration file, switch over to your terminal, and run this command:
Terminal
$
|
|
This command starts the rake
program which is a ruby utility for running maintenance-like functions on your application (working with the DB, executing unit tests, deploying to a server, etc).
We tell rake
to db:migrate
which means "look in your set of functions for the database (db
) and run the migrate
function." The migrate
action finds all migrations in the db/migrate/
folder, looks at a special table in the DB to determine which migrations have and have not been run yet, then runs any migration that hasn’t been run.
In this case we had just one migration to run and it should print some output like this to your terminal:
Terminal
$ |
|
It tells you that it is running the migration named CreateArticles
. And the "migrated" line means that it completed without errors. When the migrations are run, data is added to the database to keep track of which migrations have already been run. Try running rake db:migrate
again now, and see what happens.
We’ve now created the articles
table in the database and can start working on our Article
model.
Working with a Model in the Console
Another awesome feature of working with Rails is the console
. The console
is a command-line interface to your application. It allows you to access and work with just about any part of your application directly instead of going through the web interface. This will accelerate your development process. Once an app is in production the console makes it very easy to do bulk modifications, searches, and other data operations. So let’s open the console now by going to your terminal and entering this:
Terminal
$
|
|
You’ll then just get back a prompt of >>
. You’re now inside an irb
interpreter with full access to your application. Let’s try some experiments. Enter each of these commands one at a time and observe the results:
IRB
2.1.1 :001> 2.1.1 :002> 2.1.1 :003> |
|
The first line was just to demonstrate that you can run normal Ruby, just like irb
, within your console
. The second line referenced the Article
model and called the all
class method which returns an array of all articles in the database – so far an empty array. The third line created a new article object. You can see that this new object had attributes id
, title
, body
, created_at
, and updated_at
.
Looking at the Model
All the code for the Article
model is in the file app/models/article.rb
, so let’s open that now.
Not very impressive, right? There are no attributes defined inside the model, so how does Rails know that an Article should have a title
, a body
, etc? The answer is a technique called reflection. Rails queries the database, looks at the articles table, and assumes that whatever columns that table has should be the attributes for the model.
You’ll recognize most of them from your migration file, but what about id
? Every table you create with a migration will automatically have an id
column which serves as the table’s primary key. When you want to find a specific article, you’ll look it up in the articles table by its unique ID number. Rails and the database work together to make sure that these IDs are unique, usually using a special column type in the DB called "serial".
In your console, try entering Article.all
again. Do you see the blank article that we created with the Article.new
command? No? The console doesn’t change values in the database until we explicitly call the .save
method on an object. Let’s create a sample article and you’ll see how it works. Enter each of the following lines one at a time:
IRB
2.1.1 :001> 2.1.1 :002> 2.1.1 :003> 2.1.1 :004> 2.1.1 :005> |
|
Now you’ll see that the Article.all
command gave you back an array holding the one article we created and saved. Go ahead and create 3 more sample articles.
Setting up the Router
We’ve created a few articles through the console, but we really don’t have a web application until we have a web interface. Let’s get that started. We said that Rails uses an "MVC" architecture and we’ve worked with the Model, now we need a Controller and View.
When a Rails server gets a request from a web browser it first goes to the router. The router decides what the request is trying to do, what resources it is trying to interact with. The router dissects a request based on the address it is requesting and other HTTP parameters (like the request type of GET or PUT). Let’s open the router’s configuration file, config/routes.rb
.
Inside this file you’ll see a LOT of comments that show you different options for routing requests. Let’s remove everything except the first line (Blogger::Application.routes.draw do
) and the final end
. Then, in between those two lines, add resources :articles
so your file looks like this:
1 2 3 |
|
This line tells Rails that we have a resource named articles
and the router should expect requests to follow the RESTful model of web interaction (REpresentational State Transfer). The details don’t matter right now, but when you make a request like http://localhost:3000/articles/
, the router will understand you’re looking for a listing of the articles, and http://localhost:3000/articles/new
means you’re trying to create a new article.
Looking at the Routing Table
Dealing with routes is commonly very challenging for new Rails programmers. There’s a great tool that can make it easier on you. To get a list of the routes in your application, go to a command prompt and run rake routes
. You’ll get a listing like this:
Terminal
$ |
|
Experiment with commenting out the resources :articles
in routes.rb
and running the command again. Un-comment the line after you see the results.
These are the seven core actions of Rails’ REST implementation. To understand the table, let’s look at the first row as an example:
Terminal
|
The left most column says articles
. This is the prefix of the path. The router will provide two methods to us using that name, articles_path
and articles_url
. The _path
version uses a relative path while the _url
version uses the full URL with protocol, server, and path. The _path
version is always preferred.
The second column, here GET
, is the HTTP verb for the route. Web browsers typically submit requests with the verbs GET
or POST
. In this column, you’ll see other HTTP verbs including PUT
and DELETE
which browsers don’t actually use. We’ll talk more about those later.
The third column is similar to a regular expression which is matched against the requested URL. Elements in parentheses are optional. Markers starting with a :
will be made available to the controller with that name. In our example line, /articles(.:format)
will match the URLs /articles/
, /articles.json
, /articles
and other similar forms.
The fourth column is where the route maps to in the applications. Our example has articles#index
, so requests will be sent to the index
method of the ArticlesController
class.
Now that the router knows how to handle requests about articles, it needs a place to actually send those requests, the Controller.
Creating the Articles Controller
We’re going to use another Rails generator but your terminal has the console currently running. Let’s open one more terminal or command prompt and move to your project directory which we’ll use for command-line scripts. In that new terminal, enter this command:
Terminal
$
|
|
The output shows that the generator created several files/folders for you:
app/controllers/articles_controller.rb
: The controller file itselfapp/views/articles
: The directory to contain the controller’s view templatestest/controllers/articles_controller_test.rb
: The controller’s unit tests fileapp/helpers/articles_helper.rb
: A helper file to assist with the views (discussed later)test/helpers/articles_helper_test.rb
: The helper’s unit test fileapp/assets/javascripts/articles.js.coffee
: A CoffeeScript file for this controllerapp/assets/stylesheets/articles.css.scss
: An SCSS stylesheet for this controller
Let’s open up the controller file, app/controllers/articles_controller.rb
. You’ll see that this is basically a blank class, beginning with the class
keyword and ending with the end
keyword. Any code we add to the controller must go between these two lines.
Defining the Index Action
The first feature we want to add is an "index" page. This is what the app will send back when a user requests http://localhost:3000/articles/
– following the RESTful conventions, this should be a list of the articles. So when the router sees this request come in, it tries to call the index
action inside articles_controller
.
Let’s first try it out by entering http://localhost:3000/articles/
into your web browser. You should get an error message that looks like this:
1 2 3 |
|
The router tried to call the index
action, but the articles controller doesn’t have a method with that name. It then lists available actions, but there aren’t any. This is because our controller is still blank. Let’s add the following method inside the controller:
1 2 3 |
|
Instance Variables
What is that "at" sign doing on the front of @articles
? That marks this variable as an "instance level variable". We want the list of articles to be accessible from both the controller and the view that we’re about to create. In order for it to be visible in both places it has to be an instance variable. If we had just named it articles
, that local variable would only be available within the index
method of the controller.
A normal Ruby instance variable is available to all methods within an instance.
In Rails’ controllers, there’s a hack which allows instance variables to be automatically transferred from the controller to the object which renders the view template. So any data we want available in the view template should be promoted to an instance variable by adding a @
to the beginning.
There are ways to accomplish the same goals without instance variables, but they’re not widely used. Check out the Decent Exposure gem to learn more.
Creating the Template
Now refresh your browser. The error message changed, but you’ve still got an error, right?
1 2 3 |
|
The error message is pretty helpful here. It tells us that the app is looking for a (view) template in app/views/articles/
but it can’t find one named index.erb
. Rails has assumed that our index
action in the controller should have a corresponding index.erb
view template in the views folder. We didn’t have to put any code in the controller to tell it what view we wanted, Rails just figures it out.
In your editor, find the folder app/views/articles
and, in that folder, create a file named index.html.erb
.
Naming Templates
Why did we choose index.html.erb
instead of the index.erb
that the error message said it was looking for? Putting the HTML in the name makes it clear that this view is for generating HTML. In later versions of our blog we might create an RSS feed which would just mean creating an XML view template like index.xml.erb
. Rails is smart enough to pick the right one based on the browser’s request, so when we just ask for http://localhost:3000/articles/
it will find the index.html.erb
and render that file.
Index Template Content
Now you’re looking at a blank file. Enter in this view template code which is a mix of HTML and what are called ERB tags:
1 2 3 4 5 6 7 8 9 |
|
ERB is a templating language that allows us to mix Ruby into our HTML. There are only a few things to know about ERB:
- An ERB clause starts with
<%
or<%=
and ends with%>
- If the clause started with
<%
, the result of the ruby code will be hidden - If the clause started with
<%=
, the result of the ruby code will be output in place of the clause
Save the file and refresh your web browser. You should see a listing of the articles you created in the console. We’ve got the start of a web application!
Adding Navigation to the Index
Right now our article list is very plain, let’s add some links.
Looking at the Routing Table
Remember when we looked at the Routing Table using bin/rake routes
from the command line? Look at the left-most column and you’ll see the route names. These are useful when creating links.
When we create a link, we’ll typically use a "route helper" to specify where the link should point. We want our link to display the single article which happens in the show
action. Looking at the table, the name for that route is article
and it requires a parameter id
in the URL. The route helper we’ll use looks like this:
1
|
|
For example, article_path(1)
would generate the string "/articles/1"
. Give the method a different parameter and you’ll change the ID on the end.
Completing the Article Links
Back in app/views/articles/index.html.erb
, find where we have this line:
1
|
|
Instead, let’s use a link_to
helper:
1
|
|
The first part of this helper after the link_to
, in this case article.title
, is the text you want the link to say. The next part is our route helper.
When the template is rendered, it will output HTML like this:
1
|
|
New Article Link
At the very bottom of the template, let’s add a link to the "Create a New Article" page.
We’ll use the link_to
helper, we want it to display the text "Create a New Article"
, and where should it point? Look in the routing table for the new
action, that’s where the form to create a new article will live. You’ll find the name new_article
, so the helper is new_article_path
. Assemble those three parts and write the link in your template.
But wait, there’s one more thing. Our stylesheet for this project is going to look for a certain class on the link to make it look fancy. To add HTML attributes to a link, we include them in a Ruby hash style on the end like this:
1
|
|
Or, if you wanted to also have a CSS ID attribute:
1 2 |
|
Use that technique to add the CSS class new_article
to your "Create a New Article" link.
1 2 3 4 5 6 7 8 9 10 11 |
|
Review the Results
Refresh your browser and each sample article title should be a link. If you click the link, you’ll get an error as we haven’t implemented the show
method yet. Similarly, the new article link will lead you to a dead end. Let’s tackle the show
next.
Creating the SHOW Action
Click the title link for one of your sample articles and you’ll get the "Unknown Action" error we saw before. Remember how we moved forward?
An "action" is just a method of the controller. Here we’re talking about the ArticlesController
, so our next step is to open app/controllers/articles_controller.rb
and add a show
method:
1 2 3 |
|
Refresh the browser and you’ll get the "Template is Missing" error. Let’s pause here before creating the view template.
A Bit on Parameters
Look at the URL: http://localhost:3000/articles/1
. When we added the link_to
in the index and pointed it to the article_path
for this article
, the router created this URL. Following the RESTful convention, this URL goes to a SHOW method which would display the Article with ID number 1
. Your URL might have a different number depending on which article title you clicked in the index.
So what do we want to do when the user clicks an article title? Find the article, then display a page with its title and body. We’ll use the number on the end of the URL to find the article in the database.
Within the controller, we have access to a method named params
which returns us a hash of the request parameters. Often we’ll refer to it as "the params
hash", but technically it’s "the params
method which returns a hash".
Within that hash we can find the :id
from the URL by accessing the key params[:id]
. Use this inside the show
method of ArticlesController
along with the class method find
on the Article
class:
1
|
|
Back to the Template
Refresh your browser and we still have the "Template is Missing" error. Create the file app/views/articles/show.html.erb
and add this code:
1 2 3 |
|
Refresh your browser and your article should show up along with a link back to the index. We can now navigate from the index to a show page and back.
Styling
This is not a CSS project, so to make it a bit more fun we’ve prepared a CSS file you can drop in. It should match up with all the example HTML in the tutorial.
Download the file from http://tutorials.jumpstartlab.com/assets/blogger/screen.css and place it in your app/assets/stylesheets/
folder. It will be automatically picked up by your project.
Saving Your Work On GitHub
Now that we have completed our first feature, it’s a great time to start thinking about how to save our project.
If you have not already installed git, please follow the instructions on installation here.
Git tracks changes in code throughout time, and is a great tool once you have started working collaboratively. First you need to create a GitHub account.
Next, create a repository for the project and on the command line do;
Terminal
$ $ $ $ $ |
|
Congratulations! You have pushed the code to your GitHub repository. At any time in the future you can backtrack to this commit and refer to your project in this state. We’ll cover this in further detail later on.
I1: Form-based Workflow
We’ve created sample articles from the console, but that isn’t a viable long-term solution. The users of our app will expect to add content through a web interface. In this iteration we’ll create an HTML form to submit the article, then all the backend processing to get it into the database.
Creating the NEW Action and View
Previously, we set up the resources :articles
route in routes.rb
, and that told Rails that we were going to follow the RESTful conventions for this model named Article. Following this convention, the URL for creating a new article would be http://localhost:3000/articles/new
. From the articles index, click your "Create a New Article" link and it should go to this path.
Then you’ll see an "Unknown Action" error. The router went looking for an action named new
inside the ArticlesController
and didn’t find it.
First let’s create that action. Open app/controllers/articles_controller.rb
and add this method, making sure it’s inside the ArticlesController
class, but outside the existing index
and show
methods:
1 2 3 |
|
Starting the Template
With that defined, refresh your browser and you should get the "Template is Missing" error.
Create a new file app/views/articles/new.html.erb
with these contents:
1
|
|
Refresh your browser and you should just see the heading "Create a New Article".
Why the name new.html.erb
? The first piece, new
, matches the name of the controller method. The second, html
, specifies the output format sent to the client. The third, erb
, specifies the language the template is written in. Under different circumstances, we might use new.json.erb
to output JSON or new.html.haml
to use the HAML templating language.
Writing a Form
It’s not very impressive so far – we need to add a form to the new.html.erb
so the user can enter in the article title and body. Because we’re following the RESTful conventions, Rails can take care of many of the details. Inside that erb
file, enter this code below your header:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 |
|
What is all that? Let’s look at it piece by piece:
form_for
is a Rails helper method which takes one parameter, in this case@article
and a block with the form fields. The first line basically says "Create a form for the object named@article
, refer to the form by the namef
and add the following elements to the form…"- The
f.label
helper creates an HTML label for a field. This is good usability practice and will have some other benefits for us later - The
f.text_field
helper creates a single-line text box namedtitle
- The
f.text_area
helper creates a multi-line text box namedbody
- The
f.submit
helper creates a button labeled "Create"
Does it Work?
Refresh your browser and you’ll see this:
1 2 3 |
|
Huh? We didn’t call a method model_name
?
We didn’t explicitly, but the model_name
method is called by form_for
. What’s happening here is that we’re passing @article
to form_for
. Since we haven’t created an @article
in this action, the variable just holds nil
. The form_for
method calls model_name
on nil
, generating the error above.
Setting up for Reflection
Rails uses some of the reflection techniques that we talked about earlier in order to set up the form. Remember in the console when we called Article.new
to see what fields an Article
has? Rails wants to do the same thing, but we need to create the blank object for it. Go into your articles_controller.rb
, and inside the new
method, add this line:
1
|
|
Then refresh your browser and your form should come up. Enter in a title, some body text, and click CREATE.
The create
Action
Your old friend pops up again…
1 2 |
|
We accessed the new
action to load the form, but Rails’ interpretation of REST uses a second action named create
to process the data from that form. Inside your articles_controller.rb
add this method (again, inside the ArticlesContoller
class, but outside the other methods):
1 2 3 |
|
Refresh the page and you’ll get the "Template is Missing" error.
We Don’t Always Need Templates
When you click the "Create" button, what would you expect to happen? Most web applications would process the data submitted then show you the object. In this case, display the article.
We already have an action and template for displaying an article, the show
, so there’s no sense in creating another template to do the same thing.
Processing the Data
Before we can send the client to the show
, let’s process the data. The data from the form will be accesible through the params
method.
To check out the structure and content of params
, I like to use this trick:
1 2 3 |
|
The fail
method will halt the request allowing you to examine the request
parameters.
Refresh/resubmit the page in your browser.
Understanding Form Parameters
The page will say "RuntimeError".
Below the error information is the request information. We are interested in the parameters (I’ve inserted line breaks for readability):
1 2 3 |
|
What are all those? We see the {
and }
on the outside, representing a Hash
. Within the hash we see keys:
utf8
: This meaningless checkmark is a hack to force Internet Explorer to submit the form using UTF-8. Read more on StackOverflowauthenticity_token
: Rails has some built-in security mechanisms to resist "cross-site request forgery". Basically, this value proves that the client fetched the form from your site before submitting the data.article
: Points to a nested hash with the data from the form itselftitle
: The title from the formbody
: The body from the form
commit
: This key holds the text of the button they clicked. From the server side, clicking a "Save" or "Cancel" button look exactly the same except for this parameter.action
: Which controller action is being activated for this requestcontroller
: Which controller class is being activated for this request
Pulling Out Form Data
Now that we’ve seen the structure, we can access the form data to mimic the way
we created sample objects in the console. In the create
action, remove the
fail
instruction and, instead, try this:
1 2 3 4 5 |
|
If you refresh the page in your browser you’ll still get the template error. Add one more line to the action, the redirect:
1
|
|
Refresh the page and you should go to the show for your new article. (NOTE: You’ve now created the same sample article twice)
More Body
The show
page has the title, but where’s the body? Add a line to the create
action to pull out the :body
key from the params
hash and store it into @article
.
Then try it again in your browser. Both the title
and body
should show up properly.
Fragile Controllers
Controllers are middlemen in the MVC framework. They should know as little as necessary about the other components to get the job done. This controller action knows too much about our model.
To clean it up, let me first show you a second way to create an instance of Article
. You can call new
and pass it a hash of attributes, like this:
1 2 3 4 5 6 7 |
|
Try that in your app, if you like, and it’ll work just fine.
But look at what we’re doing. params
gives us back a hash, params[:article]
gives us back the nested hash, and params[:article][:title]
gives us the string from the form. We’re hopping into params[:article]
to pull its data out and stick it right back into a hash with the same keys/structure.
There’s no point in that! Instead, just pass the whole hash:
1 2 3 4 5 |
|
Test and you’ll find that it… blows up! What gives?
For security reasons, it’s not a good idea to blindly save parameters sent into us via the params hash. Luckily, Rails gives us a feature to deal with this situation: Strong Parameters.
It works like this: You use two new methods, require
and permit
.
They help you declare which attributes you’d like to accept. Most of
the time, they’re used in a helper method. Add the below code to app/helpers/articles_helper.rb
.
1 2 3 |
|
Now on your articles_controller.rb add: ‘include ArticlesHelper’ directly below your class name.
You then use this method instead of the params
hash directly:
1
|
|
Go ahead and add this helper method to your code, and change the arguments to new
. It should look like this, in your articles_controller.rb file, when you’re done:
1 2 3 4 5 6 7 8 9 10 11 |
|
Now in your articles_helper.rb file it should look like this:
1 2 3 4 5 6 7 |
|
We can then re-use this method any other time we want to make an
Article
.
Deleting Articles
We can create articles and we can display them, but when we eventually deliver this to less perfect people than us, they’re going to make mistakes. There’s no way to remove an article, let’s add that next.
We could put delete links on the index page, but instead let’s add them to the show.html.erb
template. Let’s figure out how to create the link.
We’ll start with the link_to
helper, and we want it to say the word "delete" on the link. So that’d be:
1
|
|
But what should some_path
be? Look at the routes table with rake routes
. The destroy
action will be the last row, but it has no name in the left column. In this table the names "trickle down," so look up two lines and you’ll see the name article
.
The helper method for the destroy-triggering route is article_path
. It needs to know which article to delete since there’s an :id
in the path, so our link will look like this:
1
|
|
Go to your browser, load the show page, click the link, and observe what happens.
REST is about Path and Verb
Why isn’t the article being deleted? If you look at the server window, this is the response to our link clicking:
Terminal
|
Compare that to what we see in the routes table:
Terminal
|
|
The path "articles/3"
matches the route pattern articles/:id
, but look at the verb. The server is seeing a GET
request, but the route needs a DELETE
verb. How do we make our link trigger a DELETE
?
You can’t, exactly. While most browsers support all four verbs, GET
, PUT
, POST
, and DELETE
, HTML links are always GET
, and HTML forms only support GET
and POST
. So what are we to do?
Rails’ solution to this problem is to fake a DELETE
verb. In your view template, you can add another attribute to the link like this:
1
|
|
Through some JavaScript tricks, Rails can now pretend that clicking this link triggers a DELETE
. Try it in your browser.
The destroy
Action
Now that the router is recognizing our click as a delete, we need the action. The HTTP verb is DELETE
, but the Rails method is destroy
, which is a bit confusing.
Let’s define the destroy
method in our ArticlesController
so it:
- Uses
params[:id]
to find the article in the database - Calls
.destroy
on that object - Redirects to the articles index page
Do that now on your own and test it.
Confirming Deletion
There’s one more parameter you might want to add to your link_to
call in your show.html.erb
:
1
|
|
This will pop up a JavaScript dialog when the link is clicked. The Cancel button will stop the request, while the OK button will submit it for deletion.
Creating an Edit Action & View
Sometimes we don’t want to destroy an entire object, we just want to make some changes. We need an edit workflow.
In the same way that we used new
to display the form and create
to process that form’s data, we’ll use edit
to display the edit form and update
to save the changes.
Adding the Edit Link
Again in show.html.erb
, let’s add this:
1
|
|
Trigger the edit_article
route and pass in the @article
object. Try it!
Implementing the edit
Action
The router is expecting to find an action in ArticlesController
named edit
, so let’s add this:
1 2 3 |
|
All the edit
action does is find the object and display the form. Refresh and you’ll see the template missing error.
An Edit Form
Create a file app/views/articles/edit.html.erb
but hold on before you type anything. Below is what the edit form would look like:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 |
|
In the Ruby community there is a mantra of "Don’t Repeat Yourself" – but that’s exactly what I’ve done here. This view is basically the same as the new.html.erb
– the only change is the H1. We can abstract this form into a single file called a partial, then reference this partial from both new.html.erb
and edit.html.erb
.
Creating a Form Partial
Partials are a way of packaging reusable view template code. We’ll pull the common parts out from the form into the partial, then render that partial from both the new template and the edit template.
Create a file app/views/articles/_form.html.erb
and, yes, it has to have the underscore at the beginning of the filename. Partials always start with an underscore.
Open your app/views/articles/new.html.erb
and CUT all the text from and including the form_for
line all the way to its end
. The only thing left will be your H1 line.
Add the following code to that view:
1
|
|
Now go back to the _form.html.erb
and paste the code from your clipboard.
Writing the Edit Template
Then look at your edit.html.erb
file. You already have an H1 header, so add the line which renders the partial.
Testing the Partial
Go back to your articles list and try creating a new article – it should work just fine. Try editing an article and you should see the form with the existing article’s data – it works OK until you click "Update Article."
Implementing Update
The router is looking for an action named update
. Just like the new
action sends its form data to the create
action, the edit
action sends its form data to the update
action. In fact, within our articles_controller.rb
, the update
method will look very similar to create
:
1 2 3 4 5 6 |
|
The only new bit here is the update
method. It’s very similar to Article.new
where you can pass in the hash of form data. It changes the values in the object to match the values submitted with the form. One difference from new
is that update
automatically saves the changes.
We use the same article_params
method as before so that we only
update the attributes we’re allowed to.
Now try editing and saving some of your articles.
Adding a Flash
Our operations are working, but it would be nice if we gave the user some kind of status message about what took place. When we create an article the message might say "Article ‘the-article-title’ was created", or "Article ‘the-article-title’ was removed" for the remove action. We can accomplish this with the flash
.
The controller provides you with accessors to interact with the flash
. Calling flash.notice
will fetch a value, and flash.notice = "Your Message"
will store the string in the flash
.
Flash for Update
Let’s look first at the update
method we just worked on. It currently looks like this:
1 2 3 4 5 6 |
|
We can add a flash message by inserting one line:
1 2 3 4 5 6 7 8 |
|
Testing the Flash
Try editing and saving an article through your browser. Does anything show up?
We need to add the flash to our view templates. The update
method redirects to the show
, so we could just add the display to our show template.
However, we will use the flash in many actions of the application. Most of the time, it’s preferred to add it to our layout.
Flash in the Layout
If you look in app/views/layouts/application.html.erb
you’ll find what is called the "application layout". A layout is used to wrap multiple view templates in your application. You can create layouts specific to each controller, but most often we’ll just use one layout that wraps every view template in the application.
Looking at the default layout, you’ll see this:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 |
|
The yield
is where the view template content will be injected. Just above that yield, let’s display the flash by adding this:
1
|
|
This outputs the value stored in the flash
object in the attribute :notice
.
More Flash Testing
With the layout modified, try changing your article, clicking save, and you should see the flash message appear at the top of the show
page.
Adding More Messages
Typical controllers will set flash messages in the update
, create
, and destroy
actions. Insert messages into the latter two actions now.
Test out each action/flash, then you’re done with I1.
An Aside on the Site Root
It’s annoying me that we keep going to http://localhost:3000/
and seeing the Rails starter page. Let’s make the root show our articles index page.
Open config/routes.rb
and right above the other routes (in this example, right above resources :articles
) add in this one:
1
|
|
Now visit http://localhost:3000
and you should see your article list.
Another Save to GitHub.
The form-based workflow is complete, and it is common to commit and push changes after each feature. Go ahead and add/commit/push it up to GitHub:
Terminal
$ $ $ |
|
If you are not happy with the code changes you have implemented in this iteration, you don’t have to throw the whole project away and restart it. You can use GitHub’s reset –hard functionality to roll back to your first commit, and retry this iteration from there. To do so, in your terminal, type in:
Terminal
$ $ |
|
I2: Adding Comments
Most blogs allow the reader to interact with the content by posting comments. Let’s add some simple comment functionality.
Designing the Comment Model
First, we need to brainstorm what a comment is…what kinds of data does it have…
- It’s attached to an article
- It has an author name
- It has a body
With that understanding, let’s create a Comment
model. Switch over to your terminal and enter this line:
Terminal
$
|
|
We’ve already gone through what files this generator creates, we’ll be most interested in the migration file and the comment.rb
.
Setting up the Migration
Open the migration file that the generator created,
db/migrate/some-timestamp_create_comments.rb
. Let’s see the fields that were
added:
1 2 3 |
|
Once that’s complete, go to your terminal and run the migration:
Terminal
$
|
|
Relationships
The power of SQL databases is the ability to express relationships between
elements of data. We can join together the information about an order with the
information about a customer. Or in our case here, join together an article in
the articles
table with its comments in the comments
table. We do this by
using foreign keys.
Foreign keys are a way of marking one-to-one and one-to-many relationships. An article might have zero, five, or one hundred comments. But a comment only belongs to one article. These objects have a one-to-many relationship – one article connects to many comments.
Part of the big deal with Rails is that it makes working with these
relationships very easy. When we created the migration for comments we started
with an references
field named article
. The Rails convention for a
one-to-many relationship:
- the objects on the "many" end should have a foreign key referencing the "one" object.
- that foreign key should be titled with the name of the "one" object, then an underscore, then "id".
In this case one article has many comments, so each comment has a field named article_id
.
Following this convention will get us a lot of functionality "for free." Open your app/models/comment.rb
and check it out:
1 2 3 |
|
A comment relates to a single article, it "belongs to" an article. We then want to declare the other side of the relationship inside app/models/article.rb
like this:
1 2 3 |
|
Now an article "has many" comments, and a comment "belongs to" an article. We have explained to Rails that these objects have a one-to-many relationship.
Testing in the Console
Let’s use the console to test how this relationship works in code. If you don’t have a console open, go to your terminal and enter rails console
from your project directory. If you have a console open already, enter the command reload!
to refresh any code changes.
Run the following commands one at a time and observe the output:
IRB
2.1.1 :001> 2.1.1 :002> 2.1.1 :003> 2.1.1 :004> 2.1.1 :005> |
|
When you called the comments
method on object a
, it gave you back a blank array because that article doesn’t have any comments. When you executed Comment.new
it gave you back a blank Comment object with those fields we defined in the migration.
But, if you look closely, when you did a.comments.new
the comment object you got back wasn’t quite blank – it has the article_id
field already filled in with the ID number of article a
. Additionally, the following (last) call to a.comments
shows that the new comment object has already been added to the in-memory collection for the a
article object.
Try creating a few comments for that article like this:
IRB
2.1.1 :001> 2.1.1 :002> 2.1.1 :003> 2.1.1 :004> 2.1.1 :005> |
|
For the first comment, c
, I used a series of commands like we’ve done before. For the second comment, d
, I used the create
method. new
doesn’t send the data to the database until you call save
. With create
you build and save to the database all in one step.
Now that you’ve created a few comments, try executing a.comments
again. Did your comments all show up? When I did it, only one comment came back. The console tries to minimize the number of times it talks to the database, so sometimes if you ask it to do something it’s already done, it’ll get the information from the cache instead of really asking the database – giving you the same answer it gave the first time. That can be annoying. To force it to clear the cache and lookup the accurate information, try this:
IRB
2.1.1 :001> 2.1.1 :002> |
|
You’ll see that the article has associated comments. Now we need to integrate them into the article display.
Displaying Comments for an Article
We want to display any comments underneath their parent article. Open app/views/articles/show.html.erb
and add the following lines right before the link to the articles list:
1 2 |
|
This renders a partial named "comment"
and that we want to do it once for each element in the collection @article.comments
. We saw in the console that when we call the .comments
method on an article we’ll get back an array of its associated comment objects. This render line will pass each element of that array one at a time into the partial named "comment"
. Now we need to create the file app/views/articles/_comment.html.erb
and add this code:
1 2 3 4 |
|
Display one of your articles where you created the comments, and they should all show up.
Web-Based Comment Creation
Good start, but our users can’t get into the console to create their comments. We’ll need to create a web interface.
Building a Comment Form Partial
The lazy option would be to add a "New Comment" link to the article show
page. A user would read the article, click the link, go to the new comment form, enter their comment, click save, and return to the article.
But, in reality, we expect to enter the comment directly on the article page. Let’s look at how to embed the new comment form onto the article show
.
Just above the "Back to Articles List" in the articles show.html.erb
:
1
|
|
This is expecting a file app/views/comments/_form.html.erb
, so create the app/views/comments/
directory with the _form.html.erb
file, and add this starter content:
1 2 |
|
Look at an article in your browser to make sure that partial is showing up. Then we can start figuring out the details of the form.
In the ArticlesController
First look in your articles_controller.rb
for the new
method.
Remember how we created a blank Article
object so Rails could figure out which fields an article has? We need to do the same thing before we create a form for the Comment
.
But when we view the article and display the comment form we’re not running the article’s new
method, we’re running the show
method. So we’ll need to create a blank Comment
object inside that show
method like this:
1 2 |
|
Due to the Rails’ mass-assignment protection, the article_id
attribute
of the new Comment
object needs to be manually assigned with the id
of the Article
. Why do you think we use Comment.new
instead of
@article.comments.new
?
Improving the Comment Form
Now we can create a form inside our comments/_form.html.erb
partial like this:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
|
Trying the Comment Form
Save and refresh your web browser and you’ll get an error like this:
1 2 3 |
|
The form_for
helper is trying to build the form so that it submits to article_comments_path
. That’s a helper which we expect to be created by the router, but we haven’t told the router anything about Comments
yet. Open config/routes.rb
and update your article to specify comments as a sub-resource.
1 2 3 |
|
Then refresh your browser and your form should show up. Try filling out the comments form and click SUBMIT – you’ll get an error about uninitialized constant CommentsController
.
Did you figure out why we aren’t using @article.comments.new
? If you want, edit the show
action and replace @comment = Comment.new
with @comment = @article.comments.new
. Refresh the browser. What do you see?
For me, there is an extra empty comment at the end of the list of comments. That is due to the fact that @article.comments.new
has added the new Comment
to the in-memory collection for the Article
. Don’t forget to change this back.
Creating a Comments Controller
Just like we needed an articles_controller.rb
to manipulate our Article
objects, we’ll need a comments_controller.rb
.
Switch over to your terminal to generate it:
Terminal
$
|
|
Writing CommentsController.create
The comment form is attempting to create a new Comment
object which triggers the create
action. How do we write a create
?
You can cheat by looking at the create
method in your articles_controller.rb
. For your comments_controller.rb
, the instructions should be the same just replace Article
with Comment
.
There is one tricky bit, though! We need to assign the article id to our comment like this:
1 2 3 4 5 6 7 8 9 10 11 12 |
|
After Creation
As a user, imagine you write a witty comment, click save, then what would you expect? Probably to see the article page, maybe automatically scrolling down to your comment.
At the end of our create
action in CommentsController
, how do we handle the redirect? Instead of showing them the single comment, let’s go back to the article page:
1
|
|
Recall that article_path
needs to know which article we want to see. We might not have an @article
object in this controller action, but we can find the Article
associated with this Comment
by calling @comment.article
.
Test out your form to create another comment now – and it should work!
Cleaning Up
We’ve got some decent comment functionality, but there are a few things we should add and tweak.
Comments Count
Let’s make it so where the view template has the "Comments" header it displays how many comments there are, like "Comments (3)". Open up your article’s show.html.erb
and change the comments header so it looks like this:
1
|
|
Form Labels
The comments form looks a little silly with "Author Name". It should probably say "Your Name", right? To change the text that the label helper prints out, you pass in the desired text as a second parameter, like this:
1
|
|
Change your comments/_form.html.erb
so it has labels "Your Name" and "Your Comment".
Add a Timestamp to the Comment Display
We should add something about when the comment was posted. Rails has a really neat helper named distance_of_time_in_words
which takes two dates and creates a text description of their difference like "32 minutes later", "3 months later", and so on.
You can use it in your _comment.html.erb
partial like this:
1
|
|
With that, you’re done with I2!
Time to Save to GitHub Again!
Now that the comments feature has been added push it up to GitHub:
Terminal
$ $ $ |
|
I3: Tagging
In this iteration we’ll add the ability to tag articles for organization and navigation.
First we need to think about what a tag is and how it’ll relate to the Article model. If you’re not familiar with tags, they’re commonly used in blogs to assign the article to one or more categories.
For instance, if I write an article about a feature in Ruby on Rails, I might want it tagged with all of these categories: "ruby", "rails" and "programming". That way if one of my readers is looking for more articles about one of those topics they can click on the tag and see a list of my articles with that tag.
Understanding the Relationship
What is a tag? We need to figure that out before we can create the model. First, a tag must have a relationship to an article so they can be connected. A single tag, like "ruby" for instance, should be able to relate to many articles. On the other side of the relationship, the article might have multiple tags (like "ruby", "rails", and "programming" as above) - so it’s also a many relationship. Articles and tags have a many-to-many relationship.
Many-to-many relationships are tricky because we’re using an SQL database. If an Article "has many" tags, then we would put the foreign key article_id
inside the tags
table - so then a Tag would "belong to" an Article. But a tag can connect to many articles, not just one. We can’t model this relationship with just the articles
and tags
tables.
When we start thinking about the database modeling, there are a few ways to achieve this setup. One way is to create a "join table" that just tracks which tags are connected to which articles. Traditionally this table would be named articles_tags
and Rails would express the relationships by saying that the Article model has_and_belongs_to_many
Tags, while the Tag model has_and_belongs_to_many
Articles.
Most of the time this isn’t the best way to really model the relationship. The connection between the two models usually has value of its own, so we should promote it to a real model. For our purposes, we’ll introduce a model named "Tagging" which is the connection between Articles and Tags. The relationships will setup like this:
- An Article
has_many
Taggings - A Tag
has_many
Taggings - A Tagging
belongs_to
an Article andbelongs_to
a Tag
Making Models
With those relationships in mind, let’s design the new models:
- Tag
name
: A string
- Tagging
tag_id
: Integer holding the foreign key of the referenced Tagarticle_id
: Integer holding the foreign key of the referenced Article
Note that there are no changes necessary to Article because the foreign key is stored in the Tagging model. So now lets generate these models in your terminal:
Terminal
$ $ $ |
|
Expressing Relationships
Now that our model files are generated we need to tell Rails about the relationships between them. For each of the files below, add these lines:
In app/models/article.rb
:
1
|
|
In app/models/tag.rb
:
1
|
|
After Rails had been around for awhile, developers were finding this kind of relationship very common. In practical usage, if I had an object named article
and I wanted to find its Tags, I’d have to run code like this:
1
|
|
That’s a pain for something that we need commonly.
An article has a list of tags through the relationship of taggings. In Rails we can express this "has many" relationship through an existing "has many" relationship. We will update our article model and tag model to express that relationship.
In app/models/article.rb
:
1 2 |
|
In app/models/tag.rb
:
1 2 |
|
Now if we have an object like article
we can just ask for article.tags
or, conversely, if we have an object named tag
we can ask for tag.articles
.
To see this in action, start the bin/rails console
and try the following:
IRB
2.1.1 :001> 2.1.1 :002> 2.1.1 :003> 2.1.1 :004> |
|
An Interface for Tagging Articles
The first interface we’re interested in is within the article itself. When I write an article, I want to have a text box where I can enter a list of zero or more tags separated by commas. When I save the article, my app should associate my article with the tags with those names, creating them if necessary.
Add the following to our existing form in app/views/articles/_form.html.erb
:
1 2 3 4 |
|
With that added, try to create an new article in your browser and your should see this error:
1 2 3 |
|
An Article doesn’t have an attribute or method named tag_list
. We made it up in order for the form to display related tags, but we need to add a method to the article.rb
file like this:
1 2 3 |
|
Back in your console, find that article again, and take a look at the results of tag_list
:
IRB
2.1.1 :001> 2.1.1 :002> 2.1.1 :003> |
|
That is not quite right. What happened?
Our array of tags is an array of Tag instances. When we joined the array Ruby called the default #to_s
method on every one of these Tag instances. The default #to_s
method for an object produces some really ugly output.
We could fix the tag_list
method by:
- Converting all our tag objects to an array of tag names
- Joining the array of tag names together
1 2 3 4 5 |
|
Another alternative is to define a new Tag#to_s
method which overrides the default:
1 2 3 4 5 6 7 8 9 |
|
Now, when we try to join our tags
, it’ll delegate properly to our name
attribute. This is because #join
calls #to_s
on every element of the
array.
Your form should now show up and there’s a text box at the bottom named "Tag list". Enter content for another sample article and in the tag list enter ‘ruby, technology’. Click save. It…. worked?
But it didn’t. Click ‘edit’ again, and you’ll see that we’re back to the #<Tag...
business, like before. What gives?
1 2 3 4 5 |
|
Unpermitted parameters? Oh yeah! Strong Parameters has done its job, saving us from parameters we don’t want. But in this case, we do want that parameter. Open up your app/helpers/articles_helper.rb
and fix the article_params
method:
1 2 3 |
|
If you go back and put "ruby, technology" as tags, and click save, you’ll get this new error:
1 2 |
|
What is this all about? Let’s start by looking at the form data that was posted when we clicked SAVE. This data is in the terminal where you are running the rails server. Look for the line that starts "Processing ArticlesController#create", here’s what mine looks like:
1 2 |
|
The field that’s interesting there is the "tag_list"=>"technology, ruby"
. Those are the tags as I typed them into the form. The error came up in the create
method, so let’s peek at app/controllers/articles_controller.rb
in the create
method. See the first line that calls Article.new(article_params)
? This is the line that’s causing the error as you could see in the middle of the stack trace.
Since the create
method passes all the parameters from the form into the Article.new
method, the tags are sent in as the string "technology, ruby"
. The new
method will try to set the new Article’s tag_list
equal to "technology, ruby"
but that method doesn’t exist because there is no attribute named tag_list
.
There are several ways to solve this problem, but the simplest is to pretend like we have an attribute named tag_list
.
We can define the tag_list=
method inside article.rb
like this: (do not delete your original tag_list method)
1 2 3 |
|
Just leave it blank for now and try to resubmit your sample article with tags. It goes through!
Not So Fast
Did it really work? It’s hard to tell. Let’s jump into the console and have a look.
IRB
2.1.1 :001> 2.1.1 :002> |
|
I bet the console reported that a
had []
tags – an empty list. (It also probably said something about an ActiveRecord::Associations::CollectionProxy
😉 ) So we didn’t generate an error, but we didn’t create any tags either.
We need to return to the Article#tag_list=
method in article.rb
and do some more work.
The Article#tag_list=
method accepts a parameter, a string like "tag1, tag2, tag3" and we need to associate the article with tags that have those names. The pseudo-code would look like this:
- Split the tags_string into an array of strings with leading and trailing whitespace removed (so
"tag1, tag2, tag3"
would become["tag1","tag2","tag3"]
- For each of those strings…
- Ensure each one of these strings are unique
- Look for a Tag object with that name. If there isn’t one, create it.
- Add the tag object to a list of tags for the article
- Set the article’s tags to the list of tags that we have found and/or created.
The first step is something that Ruby does very easily using the String#split
method. Go into your console and try "tag1, tag2, tag3".split. By default it split on the space character, but that’s not what we want. You can force split to work on any character by passing it in as a parameter, like this: "tag1, tag2, tag3".split(",")
.
Look closely at the output and you’ll see that the second element is " tag2"
instead of "tag2"
– it has a leading space. We don’t want our tag system to end up with different tags because of some extra (non-meaningful) spaces, so we need to get rid of that. The String#strip
method removes leading or trailing whitespace – try it with " my sample ".strip
. You’ll see that the space in the center is preserved.
So first we split the string, and then trim each and every element and collect those updated items:
IRB
2.1.1 :001>
|
|
The String#split(",")
will create the array with elements that have the extra spaces as before, then the Array#collect
will take each element of that array and send it into the following block where the string is named s
and the String#strip
and String#downcase
methods are called on it. The downcase
method is to make sure that "ruby" and "Ruby" don’t end up as different tags. This line should give you back ["programming", "ruby", "rails"]
.
Lastly, we want to make sure that each and every tag in the list is unique. Array#uniq
allows us to remove duplicate items from an array.
IRB
2.1.1 :001>
|
|
Now, back inside our tag_list=
method, let’s add this line:
1
|
|
So looking at our pseudo-code, the next step is to go through each of those tag_names
and find or create a tag with that name. Rails has a built in method to do just that, like this:
1
|
|
And finally we need to collect up these new or found new tags and then assign them to our article.
1 2 3 4 5 |
|
Testing in the Console
Go back to your console and try these commands:
IRB
2.1.1 :001> 2.1.1 :002> |
|
You should get back a list of the two tags. If you’d like to check the other side of the Article-Tagging-Tag relationship, try this:
IRB
2.1.1 :001> 2.1.1 :002> |
|
And you’ll see that this Tag is associated with just one Article.
Adding Tags to our Display
According to our work in the console, articles can now have tags, but we haven’t done anything to display them in the article pages.
Let’s start with app/views/articles/show.html.erb
. Right below the line that displays the article.title
, add these lines:
1 2 3 4 5 6 |
|
Refresh your view and…BOOM:
1 2 3 |
|
The link_to
helper is trying to use tag_path
from the router, but the router doesn’t know anything about our Tag object. We created a model, but we never created a controller or route. There’s nothing to link to – so let’s generate that controller from your terminal:
Terminal
$
|
|
Then we need to add tags as a resource to our config/routes.rb
, it should look like this:
1 2 3 4 5 6 7 8 9 |
|
Refresh your article page and you should see tags, with links, associated with this article.
Listing Articles by Tag
The links for our tags are showing up, but if you click on them you’ll see our old friend "No action responded to show." error.
Open app/controllers/tags_controller.rb
and define a show action:
1 2 3 |
|
Then create the show template app/views/tags/show.html.erb
:
1 2 3 4 5 6 7 |
|
Refresh your view and you should see a list of articles with that tag. Keep in mind that there might be some abnormalities from articles we tagged before doing our fixes to the tag_list=
method. For any article with issues, try going to its edit
screen, saving it, and things should be fixed up. If you wanted to clear out all taggings you could do Tagging.destroy_all
from your console.
Listing All Tags
We’ve built the show
action, but the reader should also be able to browse the tags available at http://localhost:3000/tags
. I think you can do this on your own. Create an index
action in your tags_controller.rb
and an index.html.erb
in the corresponding views folder. Look at your articles_controller.rb
and Article index.html.erb
if you need some clues.
Now that we can see all of our tags, we also want the capability to delete them.
I think you can do this one on your own too. Create a destroy
action in your
tags_controller.rb
and edit the index.html.erb
file you just created. Look
at your articles_controller.rb
and Article show.html.erb
if you need some
clues.
With that, a long Iteration 3 is complete!
Saving to GitHub.
Woah! The tagging feature is now complete. Good on you. Your going to want to push this to the repo.
Terminal
$ $ $ |
|
I4: A Few Gems
In this iteration we’ll learn how to take advantage of the many plugins and libraries available to quickly add features to your application. First we’ll work with paperclip
, a library that manages file attachments and uploading.
Using the Gemfile to Set up a RubyGem
In the past Rails plugins were distributed in zip or tar files that got stored into your application’s file structure. One advantage of this method is that the plugin could be easily checked into your source control system along with everything you wrote in the app. The disadvantage is that it made upgrading to newer versions of the plugin, and dealing with the versions at all, complicated.
These days, all Rails plugins are now ‘gems.’ RubyGems is a package management system for Ruby, similar to how Linux distributions use Apt or RPM. There are central servers that host libraries, and we can install those libraries on our machine with a single command. RubyGems takes care of any dependencies, allows us to pick any options if necessary, and installs the library.
Let’s see it in action. Go to your terminal where you have the rails server running, and type Ctrl-C
. If you have a console session open, type exit
to exit. Then open up Gemfile
and look for the lines like this:
1 2 3 4 5 6 7 8 |
|
These lines are commented out because they start with the #
character. By specifying a RubyGem with the gem
command, we’ll tell the Rails application "Make sure this gem is loaded when you start up. If it isn’t available, freak out!" Here’s how we’ll require the paperclip gem, add this near those commented lines:
1
|
|
Terminal
$
|
|
When you’re writing a production application, you might specify additional parameters that require a specific version or a custom source for the library. With that config line declared, go back to your terminal and run rails server
to start the application again. You should get an error like this:
Terminal
$ |
|
The last line is key – since our config file is specifying which gems it needs, the bundle
command can help us install those gems. Go to your terminal and:
Terminal
$
|
|
It should then install the paperclip RubyGem with a version like 3.5.2. In some projects I work on, the config file specifies upwards of 18 gems. With that one bundle
command the app will check that all required gems are installed with the right version, and if not, install them.
Note: You may need to reload your rails server for the paperclip methods to work.
Now we can start using the library in our application!
Setting up the Database for Paperclip
We want to add images to our articles. To keep it simple, we’ll say that a single article could have zero or one images. In later versions of the app maybe we’d add the ability to upload multiple images and appear at different places in the article, but for now the one will show us how to work with paperclip.
First we need to add some fields to the Article model that will hold the information about the uploaded image. Any time we want to make a change to the database we’ll need a migration. Go to your terminal and execute this:
Terminal
$
|
|
That will create a file in your db/migrate/
folder that ends in _add_paperclip_fields_to_article.rb
. Open that file now.
Remember that the code inside the change
method is to migrate the database forward, and Rails should automatically figure out how to undo those changes. We’ll use the add_column
and remove_column
methods to setup the fields paperclip is expecting:
1 2 3 4 5 6 7 8 |
|
Then go to your terminal and run rake db:migrate
. The rake command should show you that the migration ran and added columns to the database.
Adding to the Model
The gem is loaded, the database is ready, but we need to tell our Rails application about the image attachment we want to add. Open app/models/article.rb
and just below the existing has_many
lines, add these lines:
1 2 |
|
This has_attached_file
method is part of the paperclip library. With that declaration, paperclip will understand that this model should accept a file attachment and that there are fields to store information about that file which start with image_
in this model’s database table.
As of version 4.0, all attachments are required to include a content_type validation, a file_name validation, or to explicitly state that they’re not going to have either. Paperclip raises MissingRequiredValidatorError error if you do not do this. So, we add the validates_attachment_content_type line so that our model will validate that it is receiving a proper filetype.
We also have to deal with mass assignment! Modify your app/helpers/articles_helper.rb
and update the article_params
method to permit an :image
as:
1 2 3 |
|
Modifying the Form Template
First we’ll add the ability to upload the file when editing the article, then we’ll add the image display to the article show template. Open your app/views/articles/_form.html.erb
view template. We need to make two changes…
In the very first line, we need to specify that this form needs to accept "multipart" data. This is an instruction to the browser about how to submit the form. Change your top line so it looks like this:
1
|
|
Then further down the form, right before the paragraph with the save button, let’s add a label and field for the file uploading:
1 2 3 4 |
|
Trying it Out
If your server isn’t running, start it up (rails server
in your terminal). Then go to http://localhost:3000/articles/
and click EDIT for your first article. The file field should show up towards the bottom. Click the Choose a File
and select a small image file (a suitable sample image can be found at http://hungryacademy.com/images/beast.png). Click SAVE and you’ll return to the article index. Click the title of the article you just modified. What do you see? Did the image attach to the article?
When I first did this, I wasn’t sure it worked. Here’s how I checked:
- Open a console session (
rails console
from terminal) - Find the ID number of the article by looking at the URL. In my case, the url was
http://localhost:3000/articles/1
so the ID number is just1
- In console, enter
a = Article.find(1)
- Right away I see that the article has data in the
image_file_name
and other fields, so I think it worked. - Enter
a.image
to see even more data about the file
Ok, it’s in there, but we need it to actually show up in the article. Open the app/views/articles/show.html.erb
view template. Before the line that displays the body, let’s add this line:
1
|
|
Then refresh the article in your browser. Tada!
Improving the Form
When first working with the edit form I wasn’t sure the upload was working because I expected the file_field
to display the name of the file that I had already uploaded. Go back to the edit screen in your browser for the article you’ve been working with. See how it just says "Choose File, no file selected" – nothing tells the user that a file already exists for this article. Let’s add that information in now.
So open that app/views/articles/_form.html.erb
and look at the paragraph where we added the image upload field. We’ll add in some new logic that works like this:
- If the article has an image filename *Display the image
- Then display the
file_field
button with the label "Attach a New Image"
So, turning that into code…
1 2 3 4 5 6 7 |
|
Test how that looks both for articles that already have an image and ones that don’t.
When you "show" an article that doesn’t have an image attached it’ll have an ugly broken link. Go into your app/views/articles/show.html.erb
and add a condition like we did in the form so the image is only displayed if it actually exists.
Now our articles can have an image and all the hard work was handled by paperclip!
Further Notes about Paperclip
Yes, a model (in our case an article) could have many attachments instead of just one. To accomplish this you’d create a new model, let’s call it "Attachment", where each instance of the model can have one file using the same fields we put into Article above as well as an article_id
field. The Attachment would then belong_to
an article, and an article would have_many
attachments.
Paperclip supports automatic image resizing and it’s easy. In your model, you’d add an option like this:
1
|
|
This would automatically create a "medium" size where the largest dimension is 300 pixels and a "thumb" size where the largest dimension is 100 pixels. Then in your view, to display a specific version, you just pass in an extra parameter like this:
1
|
|
If it’s so easy, why don’t we do it right now? The catch is that paperclip doesn’t do the image manipulation itself, it relies on a package called imagemagick. Image processing libraries like this are notoriously difficult to install. If you’re on Linux, it might be as simple as sudo apt-get install imagemagick
. On OS X, if you have Homebrew installed, it’d be brew install imagemagick
. On windows you need to download and copy some EXEs and DLLs. It can be a hassle, which is why we won’t do it during this class.
If you do manage to get imagemagick installed, be advised that the custom sizes will only take affect on those images uploaded after the imagemagick installation. In otherwords, when the image is uploaded - Paperclip will use Imagemagick to create the customized sizes specified on the has_attached_file
line. This also means that if you change your sizes as a later time, any images that had been previously uploaded won’t have versions at those new sizes.
A Few Sass Examples
All the details about Sass can be found here: http://sass-lang.com/
We’re not focusing on CSS development, so here are a few styles that you can copy & paste and modify to your heart’s content.
Place the following styles in a new file and save it as styles.css.scss in app/assets/stylesheets/
.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 |
|
If you refresh the page, it should look slightly different! But we didn’t add a reference to this stylesheet in our HTML; how did Rails know how to use it? The answer lies in Rails’ default layout.
Working with Layouts
We’ve created about a dozen view templates between our different models. Imagine that Rails didn’t just figure it out. How would we add this new stylesheet to all of our pages? We could go into each of those templates and add a line like this at the top:
1
|
|
Which would find the Sass file we just wrote. That’s a lame job, imagine if we had 100 view templates. What if we want to change the name of the stylesheet later? Ugh.
Rails and Ruby both emphasize the idea of "D.R.Y." – Don’t Repeat Yourself. In the area of view templates, we can achieve this by creating a layout. A layout is a special view template that wraps other views. Rails has given us one already: app/views/layouts/application.html.erb
.
Check out your app/views/layouts/application.html.erb
:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
|
Whatever code is in the individual view template gets inserted into the layout where you see the yield
. Using layouts makes it easy to add site-wide elements like navigation, sidebars, and so forth.
See the stylesheet_link_tag
line? It mentions ‘application.’ That means it should load up app/assets/stylesheets/application.css
… Check out what’s in that file:
1 2 3 4 5 6 7 8 9 10 11 12 13 |
|
There’s that huge comment there that explains it: the require_tree .
line automatically loads all of the stylesheets in the current directory, and includes them in application.css
. Fun! This feature is called the asset pipeline
, and it’s pretty new to Rails. It’s quite powerful.
Now that you’ve tried out a plugin library (Paperclip), Iteration 4 is complete!
Saving to GitHub.
Terminal
$ $ $ |
|
I5: Authentication
Authentication is an important part of almost any web application and there are several approaches to take. Thankfully some of these have been put together in plugins so we don’t have to reinvent the wheel.
There are two popular gems for authentication: One is named AuthLogic and I wrote up an iteration using it for the Merchant tutorial, but I think it is a little complicated for a Rails novice. You have to create several different models, controllers, and views manually. The documentation is kind of confusing, and I don’t think my tutorial is that much better. The second is called Devise, and while it’s the gold standard for Rails 3 applications, it is also really complicated.
Sorcery is a lightweight and straightforward authentication service gem. It strikes a good balance of functionality and complexity.
Installing Sorcery
Sorcery is just a gem like any other useful package of Ruby code, so to use it in our Blogger application we’ll need to add the following line to our Gemfile:
1
gem 'sorcery'
When specifying and installing a new gem you will need to restart your Rails Server
Then at your terminal, instruct Bundler to install any newly-required gems:
Terminal
$
|
|
Once you’ve installed the gem via Bundler, you can test that it’s available with this command at your terminal:
Terminal
$
|
|
If you receive a LoadError like `cannot load such file – bcrypt`, add this to your Gemfile: `gem ‘bcrypt-ruby’`, and then run `bundle` again.
Somewhere in the middle of the output you should see the following:
Terminal
$ |
|
If it’s there, you’re ready to go!
Running the Generator
This plugin makes it easy to get up and running by providing a generator that creates a model representing our user and the required data migrations to support authentication. Although Sorcery provides options to support nice features like session-based "remember me", automatic password-reset through email, and authentication against external services such as Twitter, we’ll just run the default generator to allow simple login with an email and password.
One small bit of customization we will do is to rename the default model created by Sorcery from "User" to "Author", which gives us a more domain-relevant name to work with. Run this from your terminal:
Terminal
$
|
|
Take a look at the output and you’ll see roughly the following:
Terminal
|
Let’s look at the SorceryCore migration that the generator created before we migrate the database. If you wanted your User models to have any additional information (like "department_name" or "favorite_color") you could add columns for that, or you could create an additional migration at this point to add those fields.
For this tutorial, you will need to add the username column to the Author model. To do that, open the migration file *_sorcery_core.rb
file under db/migrate
and add make sure your file looks like this:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 |
|
So go to your terminal and enter:
Terminal
$
|
|
Let’s see what Sorcery created inside of the file app/models/author.rb
:
1 2 3 |
|
We can see it added a declaration of some kind indicating our Author class authenticates via the sorcery gem. We’ll come back to this later.
Creating a First Account
First, stop then restart your server to make sure it’s picked up the newly generated code.
Though we could certainly drop into the Rails console to create our first user, it will be better to create and test our form-based workflow by creating a user through it.
We don’t have any create, read, update, and destroy (CRUD) support for our Author model. We could define them again manually as we did with Article. Instead we are going to rely on the Rails code controller scaffold generator.
Terminal
$ |
|
Rails has two scaffold generators: scaffold and scaffold_controller. The scaffold generator generates the model, controller and views. The scaffold_controller will generate the controller and views. We are generating a scaffold_controller instead of scaffold because Sorcery has already defined for us an Author model.
As usual, the command will have printed all generated files.
The generator did a good job generating most of our fields correctly, however,
it did not know that we want our password field and password confirmation field
to use a password text entry. So we need to update the authors/_form.html.erb
:
1 2 3 4 5 6 7 8 |
|
When we created the controller and the views we provided a password
field and
a password_confirmation
field. When an author is creating their account we
want to ensure that they do not make a mistake when entering their password, so
we are requiring that they repeat their password. If the two do not match, we
know our record should be invalid, otherwise the user could have mistakenly set
their password to something other than what they expected.
To provide this validation when an author submits the form we need to define this relationship within the model.
1 2 3 4 |
|
The password
and password_confirmation
fields are sometimes referred to as
"virtual attributes" because they are not actually being stored in the
database. Instead, Sorcery uses the given password along with the automatically
generated salt
value to create and store the crypted_password
value.
Visiting http://localhost:3000/authors at this
moment we will find a routing error. The generator did not add a resource for
our Authors. We need to update our routes.rb
file:
1 2 3 4 |
|
With this in place, we can now go to http://localhost:3000/authors/new and we should see the new user form should popup. Let’s enter in "admin@example.com" for email, and "password" for the password and password_confirmation fields, then click "Create Author". We should be taken to the show page for our new Author user.
Now it’s displaying the password and password_confirmation text here, lets delete that! Edit your app/views/authors/show.html.erb
page to remove those from the display.
If you click Back, you’ll see that the app/views/authors/index.html.erb
page also shows the hash and salt. Edit the file to remove these as well.
We can see that we’ve created a user record in the system, but we can’t really tell if we’re logged in. Sorcery provides a couple of methods for our views that can help us out: current_user
and logged_in?
. The current_user
method will return the currently logged-in user if one exists and false
otherwise, and logged_in?
returns true
if a user is logged in and false
if not.
Let’s open app/views/layouts/application.html.erb
and add a little footer so the whole <body>
chunk looks like this:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 |
|
The go to http://localhost:3000/articles/
and you should see "Logged out" on the bottom of the page.
Logging In
How do we log in to our Blogger app? We can’t yet! We need to build the actual endpoints for logging in and out, which means we need controller actions for them. We’ll create an AuthorSessions controller and add in the necessary actions: new, create, and destroy.
First, let’s generate the AuthorSessions controller:
Terminal
$
|
|
Now we’ll add new
, create
, and destroy
methods to app/controllers/author_sessions_controller.rb
:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 |
|
As is common for Rails apps, the new
action is responsible for rendering the related form, the create
action accepts the submission of that form, and the destroy
action removes a record of the appropriate type. In this case, our records are the Author objects that represent a logged-in user.
Let’s create the template for the new
action that contains the login form, in app/views/author_sessions/new.html.erb
: (you may have to make the directory)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 |
|
The create
action handles the logic for logging in, based on the parameters passed from the rendered form: email and password. If the login is successful, the user is redirected to the articles index, or if the user had been trying to access a restricted page, back to that page. If the login fails, we’ll re-render the login form. The destroy
action calls the logout
method provided by Sorcery and then redirects.
Next we need some routes so we can access those actions from our browser. Open up config/routes.rb
and make sure it includes the following:
1 2 3 4 |
|
Terminal
$ |
|
Our Author Sessions are similar to other resources in our system. However, we only want to open a smaller set of actions. An author is able to be presented with a login page (:new), login (:create), and logout (:destroy). It does not make sense for it to provide an index, or edit and update session data.
The last two entries create aliases to our author sessions actions.
Externally we want our authors to visit pages that make the most sense to them:
- http://localhost:3000/login
- http://localhost:3000/logout
Internally we also want to use path and url helpers that make the most sense:
- login_path, login_url
- logout_path, logout_url
Now we can go back to our footer in app/views/layouts/application.html.erb
and update it to include some links:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 |
|
Now we should be able to log in and log out, and see our status reflected in the footer. Let’s try this a couple of times to confirm we’ve made it to this point successfully. (You may need to restart the rails server to successfully log in.)
Securing New Users
It looks like we can create a new user and log in as that user, but I still want to make some more changes. We’re just going to use one layer of security for the app – a user who is logged in has access to all the commands and pages, while a user who isn’t logged in can only post comments and try to login. But that scheme will breakdown if just anyone can go to this URL and create an account, right?
Let’s add in a protection scheme like this to the new users form:
- If there are zero users in the system, let anyone access the form
- If there are more than zero users registered, only users already logged in can access this form
That way when the app is first setup we can create an account, then new users can only be created by a logged in user.
We can create a before_filter
which will run before the new
and create
actions of our authors_controller.rb
. Open that controller and put all this code in:
1 2 3 4 5 6 7 8 |
|
The first line declares that we want to run a before filter named zero_authors_or_authenticated
when either the new
or create
methods are accessed. Then we define that filter, checking if there are either zero registered users OR if there is a user already logged in. If neither of those is true, we redirect to the root path (our articles list) and return false. If either one of them is true this filter won’t do anything, allowing the requested user registration form to be rendered.
With that in place, try accessing authors/new
when you’re logged in and when you’re logged out. If you want to test that it works when no users exist, try this at your console:
IRB
2.1.1 :001>
|
|
Then try to reach the registration form and it should work! Create yourself an account if you’ve destroyed it.
Securing the Rest of the Application
The first thing we need to do is sprinkle before_filters
on most of our controllers:
- In
authors_controller
, add a before filter to protect the actions besidesnew
andcreate
like this:before_filter :require_login, except: [:new, :create]
- In
author_sessions_controller
all the methods need to be accessible to allow login and logout - In
tags_controller
, we need to prevent unauthenticated users from deleting the tags, so we protect justdestroy
. Since this is only a single action we can use:only
like this:before_filter :require_login, only: [:destroy]
- In
comments_controller
, we never implementedindex
anddestroy
, but just in case we do let’s allow unauthenticated users to only accesscreate
:before_filter :require_login, except: [:create]
- In
articles_controller
authentication should be required fornew
,create
,edit
,update
anddestroy
. Figure out how to write the before filter using either:only
or:except
Now our app is pretty secure, but we should hide all those edit, destroy, and new article links from unauthenticated users.
Open app/views/articles/show.html.erb
and find the section where we output the "Actions". Wrap that whole section in an if
clause like this:
1 2 3 |
|
Look at the article listing in your browser when you’re logged out and make sure those links disappear. Then use the same technique to hide the "Create a New Article" link. Similarly, hide the ‘delete’ link for the tags index.
Your basic authentication is done, and Iteration 5 is complete!
Extra Credit
We now have the concept of authenticated users, represented by our Author
class, in our blogging application, and it’s authors who are allowed to create and edit articles. What could be done to make the ownership of articles more explicit and secure, and how could we restrict articles to being edited only by their original owner?
Saving to GitHub.
Terminal
$ $ $ |
|
That is the last commit for this project! If you would like to review your previous commits, you can do so with the git log command in terminal:
Terminal
$ |
|
I6: Extras
Here are some ideas for extension exercises:
- Add a site-wide sidebar that holds navigation links
- Create date-based navigation links. For instance, there would be a list of links with the names of the months and when you click on the month it shows you all the articles published in that month.
- Track the number of times an article has been viewed. Add a
view_count
column to the article, then in theshow
method ofarticles_controller.rb
just increment that counter. Or, better yet, add a method in thearticle.rb
model that increments the counter and call that method from the controller. - Once you are tracking views, create a list of the three "most popular" articles
- Create a simple RSS feed for articles using the
respond_to
method and XML view templates