PullMonkey Blog

11 Feb

Blog plugin tutorial for Ruby on Rails


Update: Bloggity does not require the Engines plugin to run if you are using Rails 2.3 or above (where the Engines plugin is baked in). -- Noted below by Bill.

Update: Added the plugin to github - simple_blog. It is not production ready or really all that usable quite yet.

Ok, so this a rant and I am sorry for that - but as simple as it is, I have been looking for a blog plugin lately. The problem with the plugins I find is that I don't want to have to deal with the engines plugin or have the controllers, models, views, etc ... all extracted into my applications code. I want it all external (hence a plugin) but let it be minimally configurable.

So in my recent search for a blog plugin for rails, I came across two that look very useful, but each with their flaws:
1) bloget - Everything is extracted to my code space. Why? Yes, I realize that it is most likely because I will want to override things, but get out of my space and keep to yourself! 🙂
Provide me a way to override things that I would need to (there really shouldn't be too many), after all it is ruby.
2) bloggity - Uses the engines plugin! I have nothing against the engines plugin (I think it is well written and documented) but for a freaking blog plugin?!? Why?

Is there a third option?
Glad you asked - yes, there is a third option - I hate to say it, but do it right! There's your third option.
Ok, but really, if there is a third option (a third plugin), I would love to hear about it.

Ok, so all that to lead up to a little plugin tutorial? Well, it got your attention didn't it?

Starting from scratch

Ok, I guess I will start from scratch. So let's get started.

Creating a plugin

1
2
3
4
5
6
7
8
9
10
11
12
13
14

pullmonkey$ ./script/generate plugin simple_blog
      create  vendor/plugins/simple_blog/lib
      create  vendor/plugins/simple_blog/tasks
      create  vendor/plugins/simple_blog/test
      create  vendor/plugins/simple_blog/README
      create  vendor/plugins/simple_blog/MIT-LICENSE
      create  vendor/plugins/simple_blog/Rakefile
      create  vendor/plugins/simple_blog/init.rb
      create  vendor/plugins/simple_blog/install.rb
      create  vendor/plugins/simple_blog/uninstall.rb
      create  vendor/plugins/simple_blog/lib/simple_blog.rb
      create  vendor/plugins/simple_blog/tasks/simple_blog_tasks.rake
      create  vendor/plugins/simple_blog/test/simple_blog_test.rb
      create  vendor/plugins/simple_blog/test/test_helper.rb

Create the app directories for your plugin

1
2
3
4
5
6
7
8
9
10

pullmonkey$ cd vendor/plugins/simple_blog/  # pretty important
pullmonkey$ ls
init.rb  install.rb  lib  MIT-LICENSE  Rakefile  README  tasks  test  uninstall.rb
pullmonkey$ mkdir app
pullmonkey$ mkdir -p app/models
pullmonkey$ mkdir -p app/controllers
pullmonkey$ mkdir -p app/views
pullmonkey$ mkdir -p app/helpers
pullmonkey$ ls app/
controllers  helpers  models  views

Well that was easy, so let's move on.

Models, Views, Controllers and Helpers - Living as one in my plugin

Models

Ok, so we have a clear path for where our models, controllers, views, and helpers should live, right?
For simplicity, let's just have a post and comment model - you have all seen this a billion times.

Models: vendor/plugins/simple_blog/app/models/post.rb

1
2
3

class Post < ActiveRecord::Base
  has_many :comments
end

Models: vendor/plugins/simple_blog/app/models/comment.rb

1
2
3

class Comment < ActiveRecord::Base
  belongs_to :post
end

And there you have it.
So what do you do to tell your rails application about your models?
Simple - inside vendor/plugins/simple_blog/init.rb - add these lines

1
2
3

model_path = File.join(directory, 'app', 'models')
$LOAD_PATH << model_path
ActiveSupport::Dependencies.load_paths << model_path

Ok, so let's test it out.

Step 1 - we will need some default migrations for the model to use.
Post migration:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

pullmonkey$ ./script/generate migration post
# This is what mine looks like
class Post < ActiveRecord::Migration
  def self.up
    create_table :posts do |t|
      t.string :subject
      t.text   :body
      t.timestamps
    end
  end

  def self.down
    drop_table :posts
  end
end

And then the comment migration:

1
2
3
4
5
6
7
8
9
10
11
12
13
14

pullmonkey$ ./script/generate migration comment
# This is what mine looks like
  def self.up
    create_table :comments do |t|
      t.string :username
      t.text   :body
      t.references :post
      t.timestamps
    end
  end

  def self.down
    drop_table :comments
  end

Run the migrations:



pullmonkey$ rake db:migrate

That was all just setup - now for the actual testing:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

pullmonkey$ ./script/console
Loading development environment (Rails 2.2.2)
>> Comment.new
=> #<Comment id: nil, username: nil, body: nil, post_id: nil, created_at: nil, updated_at: nil>
>> Post.new
=> #<Post id: nil, subject: nil, body: nil, created_at: nil, updated_at: nil>
>> p = Post.create(:subject => "Test 1", :body => "My Body")
=> #<Post id: 1, subject: "Test 1", body: "My Body", created_at: "2009-02-11 19:09:25", updated_at: "2009-02-11 19:09:25">
>> p.body
=> "My Body"
>> p.subject
=> "Test 1"
>> p.new_record?
=> false
>> p.comments
=> []
>> c = Comment.create(:username => 'pullmonkey', :body => "this is simple")
=> #<Comment id: 1, username: "pullmonkey", body: "this is simple", post_id: nil, created_at: "2009-02-11 19:10:01", updated_at: "2009-02-11 19:10:01">
>> p.comments << c
=> [#<Comment id: 1, username: "pullmonkey", body: "this is simple", post_id: 1, created_at: "2009-02-11 19:10:01", updated_at: "2009-02-11 19:10:06">]
>> p.comments
=> [#<Comment id: 1, username: "pullmonkey", body: "this is simple", post_id: 1, created_at: "2009-02-11 19:10:01", updated_at: "2009-02-11 19:10:06">]
>> Post.first.comments
=> [#<Comment id: 1, username: "pullmonkey", body: "this is simple", post_id: 1, created_at: "2009-02-11 19:10:01", updated_at: "2009-02-11 19:10:06">]

That's probably good enough. We have a working model and relationships. The best part is that all the code is still in the plugin.
What does my code space contain?

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

pullmonkey$ ls -l app/**
app/controllers:
total 4
-rw-rw-r--  1 pullmonkey pullmonkey 720 Feb 11 11:00 application.rb

app/helpers:
total 4
-rw-rw-r--  1 pullmonkey pullmonkey 115 Feb 11 11:00 application_helper.rb

app/models:
total 0

app/views:
total 4
drwxrwxr-x  2 pullmonkey pullmonkey 4096 Feb 11 11:00 layouts

Just the defaults - neat 🙂

Controllers

In much the same way as models, we can easily use controllers from our plugin. No extracting, no engines plugin.

Controllers: vendor/plugins/simple_blog/app/controllers/posts_controller.rb

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

class PostsController < ApplicationController
  def index
    @posts = Post.all
  end

  def show
    @post = Post.find(params[:id])
  end

  def new
    @post = Post.new
  end

  def create
    if @post = Post.create(params[:post])
      flash[:notice] = "Post Created"
      redirect_to :action => 'index'
    else
      flash[:error] = "Post Not Created"
      render :action => 'new'
    end
  end
  #.... more code
end

Controllers: vendor/plugins/simple_blog/app/controllers/comments_controller.rb

1
2
3
4
5
6

class CommentsController < ApplicationController
  def index
    @comments = Comment.find_all_by_post_id(params[:post_id])
  end
  #.... more code
end

Now, to register the controllers, add the following to vendor/plugins/simple_blog/init.rb:

1
2
3
4

controller_path = File.join(directory, 'app', 'controllers')
$LOAD_PATH << controller_path
ActiveSupport::Dependencies.load_paths << controller_path
config.controller_paths << controller_path

Ok, before we can really test this we will need to do the views, so keep going.

Views

Create your view directories:

1
2

pullmonkey$ mkdir -p app/views/posts
pullmonkey$ mkdir -p app/views/comments

Create your views:
For this example, I will just create one, then we will test it.

Views: vendor/plugins/simple_blog/app/views/posts/index.html.erb

1
2
3
4
5
6
7
8
9
10
11

<h1>Posts</h1>
<% @posts.each do |post| -%>
  <h2><%= h post.subject %></h2>
  <%= post.body %>
  <h3>Comments</h3>
  <% post.comments.each do |comment| -%>
    <b>by <%= comment.username %></b><br/>
    <%= comment.body %><br/>
    <br/>
  <% end -%>
<% end -%>

Append your view paths:

If you don't do this next step, you will very likely see an error message like this:
Missing template posts/index.erb in view path /home/pullmonkey/rails_projects/simple_blog/app/views:

So let's add it.
There are at least two ways to do this. 1) Added to your controllers individually or 2) Add to application controller globally.
I prefer the less obtrusive, so let's go with number 1.
For this test, we will just work with the posts controller, so open it up again and add this line:



self.append_view_path(File.join(File.dirname(__FILE__), '..', 'views'))

So your file should look like this now:

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

class PostsController < ApplicationController
  self.append_view_path(File.join(File.dirname(__FILE__), '..', 'views'))

  def index
    @posts = Post.all
  end

  def show
    @post = Post.find(params[:id])
  end

  def new
    @post = Post.new
  end

  def create
    if @post = Post.create(params[:post])
      flash[:notice] = "Post Created"
      redirect_to :action => 'index'
    else
      flash[:error] = "Post Not Created"
      render :action => 'new'
    end
  end
end

Time to test

Start your web server - ./script/server
Browse to http://localhost:3000/posts
You should see the post we created up above via Post.create(...) and its associated comment that we also created above.

Note:Feel free to overwrite any of the views. This can be done simply for the posts index view by creating the same file under RAILS_ROOT/app/views/posts/index.html.erb and doing what you'd like.

That's it for part 1

Ok, so that's part 1. The goal was to keep everything external and I think we succeeded (aside from migrations).
No offense to those that use engines or extract files into one's application's space, we all have our ways - the above is what I prefer.

Part 2 will consist mainly of filling this out a bit more and further discussion on adding helpers, routes and migrations to your plugin without interfering in the application's code space.

As always, have fun and good luck!


Filed under: development, Home, rails, tutorials Tags: , , , ,


11 Responses to “Blog plugin tutorial for Ruby on Rails”

  1. By Bryan on Feb 11, 2009 | Reply

    Cool stuff Charlie. Plugins are still new to me. 🙂

    In your init.rb file, where does ‘directory’ come from in model_path = File.join(directory, ‘app’, ‘models’)? Is that a helper method Rails provides? I assume ‘config’ is a helper method also?

    Don’t you need to migrate your database with the migrations you created before you do the testing you describe?

    —
    Bryan

  2. By charlie on Feb 11, 2009 | Reply

    @Bryan – directory is a variable for the current directory of the file you are running the File.join method from. It is a ruby method from File.

    Yes, must run migrations – thanks for pointing that out – updated the article 🙂

  3. By Ravi on Feb 11, 2009 | Reply

    Great Post. Waiting on Part 2…

  4. By Jamie on Feb 11, 2009 | Reply

    You’re a life-saver. I couldn’t for the life of me figure out how to add load paths from within a gem I am writing, it’s just this bit:

    <filter:code attributes=lang="ruby">
    model_path = File.join(directory, ‘app’, ‘models’)
    $LOAD_PATH << model_path
    ActiveSupport::Dependencies.load_paths << model_path
    < /filter:code>

  5. By Bill on Aug 30, 2009 | Reply

    I think it’s great for people to have options on how to get their blog rolling — it definitely shouldn’t be a hard thing for people to do.

    That said, the record should indicate that Bloggity does not require the Engines plugin to run if you are using Rails 2.3 or above (where the Engines plugin is baked in). Please update accordingly.

    Thanks,
    Bill

  6. By charlie on Aug 30, 2009 | Reply

    Thanks Bill, updated.

  7. By Bill Bartmann on Sep 2, 2009 | Reply

    Cool site, love the info.

  8. By SonyaSunny on Sep 4, 2009 | Reply

    Ugh, I liked! So clear and positively.

  9. By Forex Forum on Sep 15, 2009 | Reply

    Great blog, reading it through RSS feed as well

  10. By Subin Varghese on Oct 16, 2009 | Reply

    Hi,

    Great tutorial. First one which actually worked for me. Was Googling the whole web for a decent tutorial to learn to build a rails engine. Went through few other tutorials for this but got struck or got some unknown error when following them. This is the first one which went off quite well.

    Thanks a lot.

    Could you provide the second part of the tutorial. Really need them. Please… 🙂

  11. By Aguswgs on Apr 21, 2011 | Reply

    Great article, I was thinking the same, migrate all my clients to Ruby and Rails and use this experience to really learn the language, but I was not pretty sure to do it, now, I am agree with you, I has been in computers for more than 20 years, I love delphi, php, and javascripts, I can do what I want with this languages… thank your your article…

Sorry, comments for this entry are closed at this time.