Our Rumble App: What Does this Error Mean?


Oct 20, 2008

Francois, Daniel, and I (and Mat, in spirit) spent the weekend rumbling. It was a great time coding with these two superstars, but you don't care about that.

Our app is called what does this error mean?. We all see error messages, and until now, the best way to look for solutions to those error messages was doing a google search. The problem with google searches, though, is that the results are ordered by the quality of the site, not by the quality of the solution. What does this error mean solves that problem, and a few more.

But, really, reading sucks. So, watch our screencast to learn all about wdtem.

Rails & Merb Integration

As part of our rumble project, we built plugins for both rails and merb that override their default development mode error messages. With our plugin installed, you'll see our logo below the error message. Simply click on the logo to automatically jump to a what does this error mean search!

For rails:

$ script/plugin install git://github.com/giraffesoft/what_does_this_error_mean-rails.git

For merb:

sudo gem install what_does_this_error_mean-merb

Please Vote for Us!

If you think our app is cool, please consider voting for us once rails rumble voting starts! We don't know the url yet, but have been told to point people to our team page.


Watch François Beausoleil Train for Rails Rumble


Oct 15, 2008

My Rails Rumble team is all set to win, because we trained the hardest. Check out the video François made of some of his hard work preparing for the competition.


RailsRumble 2008 Training from François Beausoleil on Vimeo.


Blank: A Starter App for r_c and shoulda Users


Oct 10, 2008

Bort was released recently. Peter Cooper speculated that "...it could well catch on as the de facto bare bones / generic Rails application". But, what about us non-RSpec users? There are dozens of us, I tell you. Dozens!

We build a lot of apps at GiraffeSoft — we love to experiment with whatever ideas excite us on any given day. We're all sick of editing restful_auth code, and moving tests over to Shoulda and the one assertion per test pattern. Bort doesn't suit our needs. So, blank was born.

Right now, it's pretty simple. It has authentication, and forgot password. That's about it. But, it's no biggie. Since blank creates your new app as a git repo that shares history with blank's repo, you can pull in changes we make at any time. So, when we finally get around to implementing openid support, you'll get it for free, if you start with blank.

Vendored

All of our standard tools (and rails) are vendored:

  • active_presenter
  • andand
  • attribute_fu
  • hoptoad
  • mocha
  • rake
  • restful_authentication
  • ruby-openid
  • will_paginate

Installation

Installing blank is as easy as running a rake task. Except that blank uses thor instead, because it’s the new hotness, and it supports remote tasks.

Just install thor:

$ sudo gem install thor

…then install blank’s thor tasks:

$ thor install http://github.com/giraffesoft/blank/tree/master%2Fthorfiles%2Fblank.thor?raw=true

…then you’re ready to create a new app with blank:

$ thor blank:new_app the_name_of_my_app the@git.repo.it.will.live.in

That’s it! The thorfile will display a couple of rake notes where you should replace blank app with your app's name. Also, you'll want to fill in your hoptoad API key in config/initializers/hoptoad.rb.

If we improve the thor file, all you have to do is run:

$ thor update blank

before creating your next app, and you’ll get the changes automagically.

Development

All development will be done at the github repo. Fork away :)

Credits

Blank was created by me, with contributions from Daniel Haran.


Introducing ActivePresenter: The presenter library you already know.


Jul 27, 2008

Presenters were a hot topic in the rails community last year. A lot of prominent bloggers wrote about using them, and the implementations they had come up with. Oddly, though, when I needed one a couple of weeks ago, I was unable to find a suitable implementation. Lots of articles — no code.

Let's answer the question on everybody's mind before we move on. Feel free to skip ahead if you already know the answer.

WTF is a presenter?!

In its simplest form, a presenter is an object that wraps up several other objects to display, and manipulate them on the front end. For example, if you have a form that needs to manipulate several models, you'd probably want to wrap them in a presenter.

Indeed, attribute_fu solves this problem for some cases. However, when you're dealing with unrelated models, or, really, any situation other than a parent saving its children, you're probably better off using a presenter.

Presenters take the multi model saving code out of your controller, and put it in to a nice object. Because presenter logic is encapsulated, it's reusable, and easy to test.

Want one?

ActivePresenter

Daniel and I wrote most of this on the train ride over to RubyFringe. It is an ultra-simple presenter base class that is designed to wrap ActiveRecord models (and, potentially, others that act like them).

ActivePresenter works just like an ActiveRecord model, except that it works with multiple models at the same time. Let me show you.

Imagine we've got a signup form that needs to create a new User, and a new Account. We'd create a presenter that looks like this.

class SignupPresenter < ActivePresenter::Base
  presents :user, :account
end

Then, we'd write a new action like this one:

def new
  @signup_presenter = SignupPresenter.new
end

And a form:

<%= error_messages_for :signup_presenter %>

<%- form_for @signup_presenter, :url => signup_url do |f| -%>
  <%= f.label :account_subdomain, "Subdomain" %>
  <%= f.text_field :account_subdomain %>
  <%= f.label :user_login, "Login" %>
  <%= f.text_field :user_login %>
<%- end -%>

A create action:

def create
  @signup_presenter = SignupPresenter.new(params[:signup_presenter])
  
  if @signup_presenter.save
    redirect_to dashboard_url
  else
    render :action => "new"
  end
end

Lastly, an update action:

def update
  @signup_presenter = SignupPresenter.new(:user => current_user, :account => current_account)
  
  if @signup_presenter.update_attributes(params[:signup_presenter])
    redirect_to dashboard_url
  else
    render :action => "edit"
  end
end

Seem familiar?

If you're using r_c, most of this comes for free with:

class SignupsController < ResourceController::Base
  model_name :signup_presenter
end

For more on complex forms in rails, and the presenter pattern, see my upcoming PeepCode screencast!

Organization

I have been sticking my presenters in app/presenters. If you want to do the same, you'll need to add a line like this to your environment.rb:

config.load_paths += %W( #{RAILS_ROOT}/app/presenters )

Get It!

As a gem:

$ sudo gem install active_presenter

As a rails gem dependency:

config.gem 'active_presenter'
Or get the source from github:
$ git clone git://github.com/giraffesoft/active_presenter.git

(or fork it at http://github.com/giraffesoft/active_presenter)

Also, check out the RDoc.


Business Logic Bleeding in to Views and Controllers


Jul 16, 2008

I've been doing a fair bit of training recently — both while bringing Mat up to speed on the latest and greatest best practices, and going and speaking to clients' teams. A few concepts keep coming up, so I'm going to try to blog about them as they do. Here's the first one.

I see this all over client projects, and admittedly, some of my older ones:

<%- if @post.creator == current_user -%>
  <%= link_to "Edit", edit_post_path(@post) %>
<%- end -%>

Seems innocuous — the post can only be be edited by its creator. But, that is a business logic rule, so it belongs in your model (and your tests).

context "Posts" do
  setup do
    @creator      = create_user
    @post         = @creator.posts.create!(hash_for_post)
    @another_user = create_user
  end

  should "be editable by their creator" do
    assert @post.can_be_edited_by?(@user)
  end
  
  should "not be editable by another random user" do
    assert !@post.can_be_edited_by?(@another_user)
  end
end
class Post < ActiveRecord::Base
  def can_be_edited_by?(user)
    user == creator
  end
end

Why? A few reasons. First, because your model classes should represent a complete picture of your data's structure, and business logic rules. Second, a rule like that deserves testing, even if it's as simple as the one in this example. Finally, because later on, you might want admins to be able to edit posts, too.

context "Posts" do
  setup do
    @creator      = create_user
    @post         = @creator.posts.create!(hash_for_post)
    @another_user = create_user
    @admin        = create_user :admin => true
  end

  should "be editable by their creator" do
    assert @post.can_be_edited_by?(@user)
  end
  
  should "be editable by an admin" do
    assert @post.can_be_edited_by?(@admin)
  end
  
  should "not be editable by another random user" do
    assert !@post.can_be_edited_by?(@another_user)
  end
end
class Post < ActiveRecord::Base
  def can_be_edited_by?(user)
    user == creator || user.admin?
  end
end

If we're using the can_be_edited_by? method all over our controllers and views, a change to the access policy shouldn't entail anything other than editing a couple of lines of code in our models and unit tests.

Subtler

Here's another one I see all the time (especially in my code). This one tends to be harder to spot.

class MessagesController
  before_filter :login_required
  
  protected
    def project
      @project ||= current_user.projects.find(params[:project_id])
    end
  
    def messages
      project.messages.find(:all)
    end
end

The idea here is that we're using the association as an implicit access control mechanism. If the user row is associated with the project row, the user has access to that project. I know that there are several r_c users who have used this pattern, since it's so easy to implement with r_c. I've even recommended it. Ouch.

The problem with this pattern is lack of encapsulation. There are a good handful of situation in which you might want the access rule to change such that you'd have to go and change every call to current_user.projects. Which is ugly.

What if you decided to make ProjectMembership a join model, for example? Rather than actually deleting membership rows, you decide you'd like, for record keeping purposes, to have a revoked_at field which denotes a former membership made invalid. You might think — no problem, I can just change the conditions of the association.

class User
  has_many :projects, :through => :project_memberships, :conditions => "project_memberships.revoked_at IS NULL"
end

Aside from ambiguous naming, this remains an incomplete solution. At some point, you're going to decide that admins have access to all projects. You could add that condition to your SQL, too, but that approach is problematic for the same reason we're talking about, here: the definition of an admin may change. So, we need a different strategy.

class Project
  has_many :project_memberships
  has_many :users, :through => :project_memberships
  
  named_scope :with_active_membership_for, lambda { |user| { :include => :project_memberships, :conditions => ['project_memberships.user_id = ? AND project_memberships.revoked_at IS NULL', user.id] } }
  
  def self.for(user)
    user.admin? ? self : self.with_active_membership_for(user)
  end
end

This isn't a perfect solution, since the definition of a revoked membership is living in the SQL; ideally that would be defined in ProjectMembership. But, with a join model, I'm not sure of any other way. So, at least in this instance, we can use the Project.for method in our controllers, and views, and not have to worry about a change in the project access rules causing a need for major refactoring later on.

class MessagesController
  before_filter :login_required
  
  protected
    def project
      @project ||= Project.for(current_user).find(params[:project_id])
    end
  
    def messages
      project.messages.find(:all)
    end
end

But, Not Always

One could look at these examples and decide to engage in reductio ad absurdum. Why not encapsulate the messages collection, they'd ask?

The rule I tend to stick to is that if I know that there's any kind of access control policy being enforced on an object or collection, it gets encapsulated, and tested in the model.

In the fictional examples above, the assumption was that provided access to the project, a user would be allowed to access all of messages within. In that case, I wouldn't have encapsulated, because there would have been no rule to encapsulate.

As soon as there is a policy, though, it belongs in tested methods in your models.

The controller's responsibility in all of this is to control access to a resource in the sense that it actually performs the check, and redirects the user to a login screen, or shows an access denied message if they are not entitled to perform said action on said resource. The controller is not, however, responsible for deciding who it is that gets access; that's the model's job.

Like a bouncer at the door of a nightclub, your controller shouldn't make the rules, it should only enforce them. Luckily for you, cute 17 year old girls won't have the same effect on your controller that they might on a bouncer.


r_c, meet RSpec


Jun 20, 2008

Jonathan Barket just sent me a patch that gets resource_controller's scaffold_resource generator to spit out RSpec specs! I quickly merged it in to master, and wanted to announce it to the world, since it's something I've been hoping to get added to the plugin for a while.

r_c's scaffold_resource generator now supports vanilla test/unit, Shoulda, and RSpec for tests, and erb, and haml for templating. It will sense which of those plugins or gems you have installed in your app, and generate accordingly — it's absolutely 0 configuration.

So, if you're an RSpec user, grab the latest master from github, and get generating! :)


Introducing has_browser: Parameterized browse interfaces for your AR models.


May 19, 2008

Pretty soon after starting to use has_finder, I needed to write a browse interface. That is, given a set of parameters, return all the models that match. With has_finder, it is possible to compartmentalize those parameters in to small chunks, each with names that make sense in your model's domain. You can then chain them together to get their combined scope. The only problem iswas calling them when you aren't aware, in advance, of which finders need to be called.

Having had this problem a few times, and being sick of solving it over and over again, I decided to extract a general solution from a project I'm working on.

has_browser

It's a simple plugin, with a simple syntax. Using the canonical blog example, let's imagine we want to create a browse interface to posts. We'd want the user to be able to browse by category, author, or tags, but not to be able to access any of the other finders on the Post model for obvious security reasons. To set up has_browser, we'd do something like this:

has_browser :category, :tags, :author

Then, assuming the has_finders are already written, the posts can be browsed as follows:

Post.browse(:category => 'Testing', :tags => 'activerecord', :author => 'james')

In that example, each of the finders requires an argument; has_browser also supports finders that don't. As long as the argumentless finder is present in the browse hash, it will be called:

has_browser :category, :tags, :author, :without_args => [:order_by_date, :order_by_number_of_comments]

Post.browse(:category => 'Testing', :tags => 'activerecord', :author => 'james', :order_by_number_of_comments => 'true')

Browse can also be called from association_proxies. For a multi-blog platform, we could easily restrict browsing of posts to the current blog:

@blog.posts.browse(:category => 'Testing', :tags => 'activerecord', :author => 'james', :order_by_number_of_comments => 'true')

Since has_browser returns the same proxy as has_finder, it is possible to further restrict the results of a browse by chaining finders after the browse call. With our blog, for example, we'd probably want to restrict browsing to published posts.

@blog.posts.browse(:category => 'Testing', :tags => 'activerecord', :author => 'james', :order_by_number_of_comments => 'true').published

Note: It is not possible to chain finders before the browse call.

Finally, like has_finder, has_browser is compatible with will_paginate out of the box.

Get It

Releases of has_browser will be available as a gem, which you can freeze in to your rails app (init.rb included):

$ sudo gem install has_browser

Development, of course, is at github.


Experimental resource_controller Feature: Custom Action Discovery


May 08, 2008

A few weeks ago, Nate Wiger emailed me to ask whether I was interested in a patch for r_c. Evidently, it is possible to determine which controller actions need to be created by examining the routes that are pointing to it. Nate had also noticed that most controller actions follow a pattern:

  1. Load or build an object or a collection (i.e. Post.find(1) or Post.build(params[:post])).
  2. Optionally do something with the object(s) (i.e. @post.save).
  3. Optionally set the flash.
  4. Render a response.

Put those pieces together, and it's possible to create a controller abstraction that is aware of what's routed to it, and sets up a set of sensible defaults, even for custom actions. As you might have guessed, Nate and I have done exactly that.

For the 7 default RESTful actions, things haven't changed much. The only real difference is that you can now change step 2 (refer to list above), by passing a block to action.action.

create.action do
  @post.my_custom_save_method
end

The action block's return value determines whether the success or failure callbacks (i.e. success.flash vs failure.wants) are triggered.

For custom actions, you'll now be able to use the same DSL you're used to for customizing default actions. For example, if you had a custom action called update_collection, you might put something like this in your routes:

map.connect '/posts', :controller => 'posts', :action => 'update_collection', :conditions => {:method => :put}

Based on that route, r_c will automatically be aware of your action, and create a basic skeleton for it, following the pattern of the list above. For a collection update method, you might want to customize your action like this:

update_collection do
  before { @posts.each { |p| p.update_attributes params[:posts] } }
  update_collection.action { @posts.save }
end

That's all you'd have to do. r_c would automatically load the objects (including any parent, respecting polymorphism), and render using its internal helpers.

Note: It is also possible to modify loading by setting a block for the build accessor, which corresponds to step 1.

Highly Experimental

This is highly experimental software. I'm not sure how much I love it, since it's a bit on the magical side for my tastes. However, it is kinda neat to have your controller be aware of which actions to create, based on what is routed to it. I have been having a bit of fun playing around with this.

So, please give it a shot, and let us know how you like it on the mailing list.

Get it from github (checkout the automatic_route_discovery branch). If you aren't using git, you can get a tarball here.


Rails Training w/ James Golick & Other Rails Ninjas


Apr 29, 2008

This past December, a friend of mine, Peter, wanted to improve his rails skills. I had been asked by a few people about teaching some rails training sessions, but wanted to give it a beta test first. When I was in Toronto over the holidays, Peter and I gave it a trial run, which went great. Since then, I've been dying to offer ruby / rails training to a wider audience.

How It'll Work

The sessions will be based on the format that I tried out in my beta session with Peter. We'll start by covering some advanced rails fundamentals, with time for questions, and plenty of time to go off on long tangents about whatever you might want to learn (like, how to write a plugin, or how some of the rails internals work). The remainder of the session will be spent coding.

During the coding portion of the sessions, the instructor(s) will pair program with each of the participants in a rotation. Think of it as an opportunity to work extremely closely with a rails expert. We'll be able to apply what we've learned in the earlier portion of the session. We'll work on your real problems; that way, we'll be teaching a custom course that's custom tailored to you, and you'll walk away from the experience with some code, to boot!

Participants will be asked to bring two projects: one in progress (or finished), and one idea.

  • The project in progress serves as a tool for analysis of where you're at, and what kind of coding practices you have in general. We'll work out some possible refactorings for that project, and talk about how you'd make use of advanced rails techniques to improve its readability, maintainability, etc.
  • Next, we'll work on your project idea. The rest of the session will be spent architecting, and actually coding this project. We'll help you pick the right plugin set, model your data, and anything else.

Logistics

  • I'd like to offer the first session some time in early July, most likely over a weekend.
  • The sessions will be held in Montreal, unless there's a compelling reason to offer them somewhere else, like, a group of participants who all live in the same city.
  • Price is still TBD; I've got some potential sponsors (which would offset the price for participants), but nothing confirmed yet. There would certainly be a discount for groups or companies.

Finally, I intend to keep the instructor to participant ratio extremely low to support plenty of one-on-one time. So, if there is sufficient interest, I'll get some of my fellow Montreal rails ninjas in as additional instructors. I've got some awesome people in mind. Who knows — you might even get scouted by one of the local rails shops!

If you're interested, send me an email (top right) for more info.


Introducing Action Messager: Dead simple IM notifications for your app!


Apr 07, 2008

Sometimes email is the wrong choice for webapp notifications. Our inboxes are becoming increasingly cluttered, and especially for those of us who carry PDAs, it can be a pain to scroll through twelve facebook notifications just to get to the mail that we actually need to read. What's more, email just isn't that great of a tool for receiving short messages, since you have to open them each individually. At the very least, it's nice to have another choice.

Particularly with shorter notifications, instant messaging can be a great alternative to email. The messages don't clutter up your inbox, or need to be opened individually. Most people already use it on a daily basis. The only problem is implementation.

ActionMessager

ActionMessager is a simple framework for creating IM-based notifiers. It is structured like ActionMailer, so it's got virtually zero learning curve for most rails developers. What's more, because the syntax is the same, it's pretty easy to create a delegate class that acts like your mailer, but sends IMs as well. That means drop-in replacement!

All you have to do start sending IM notifications to your users is subclass ActionMessager::Base. Then, create a method that sets an array of recipients, and returns the message you'd like to send:

class JabberNotifier < ActionMessager::Base
  def friendship_request(friendship_request)
    @recipients = friendship_request.receiver.jabber_contact
    
    "You have received a friendship request from #{friendship_request.sender.name}! Click here to accept or decline: #{friendship_request.url}"
  end
end

Then, wherever you'd like to send the notification:

JabberNotifier.deliver_friendship_request(friendship_request)

That's all there is to it.

Caveats

Currently, only jabber is supported. It is possible to access other IM services over jabber, but I'm not 100% clear how it works, and I don't yet have need for it, so that may or may not come later.

The bigger caveat, though, is speed. Sending a notification takes anywhere from 1-2s (whether you send 1 or 5). It may be possible to improve performance by using a different jabber library, and I'll probably investigate that in the near future. Even with a performance boost, though, you should still take your notifications out of the request cycle, by using something like workling to process them asynchronously.

Get It!

Get action_messager in gem form:

$ sudo gem install action_messager

Or get the source, from github:

$ git clone git://github.com/giraffesoft/action_messager.git

attribute_fu and jQuery shake hands


Apr 03, 2008

This blog has been quiet of late, because I'm working on a couple of exciting, but still top-secret projects. Anyway, I recently moved one of those projects over to jQuery, because of its speed, syntax, and general awesomeness. Tonight, when I went to create a multi-model form with attribute_fu, I was stopped dead in my tracks by its heavy dependency on prototype. A few minutes of hacking later, a_f and jQuery are playing rather nicely together.

Get it from the jquery branch of a_f's git repo. If you aren't using git yet, this might be just the excuse you need to check it out! Or, download a tarball from github. After you install the plugin, you'll have to copy its one javascript dependency (jQuery templates) from the javascripts directory over in to your public/javascripts and require it in your layout.


Plugins I've Known and Loved #3: Ultrasphinx


Mar 03, 2008

If you've ever implemented searching in your rails app, you probably noticed that it's a major hassle (unless, of course, you already know about Ultrasphinx). Ferret is probably the most commonly used indexing solution, but, it isn't anywhere near production ready. acts_as_solr, another option, is so full of show stopper bugs that it really isn't even worth bothering with. The worst part about both of those solutions is that they usually work great in your development environment. Try to put them in to production, though, and you're in for big problems.

Enter Ultrasphinx, by Evan Weaver. Sphinx is a super high performance search daemon, designed to suck information out of a mysql or pgsql database (though, it isn't limited to that). Normally, one would configure Sphinx with SQL queries that it uses to fetch the data. Apparently, the configuration file can be a major hassle to work with — I wouldn't know. Ultrasphinx provides you with a declarative API that it ultimately uses to generate the Sphinx configuration file for you, making the process quick and painless. A few lines of code, a couple of rake commands, and you're searching.

The Sphinx approach has several advantages over ferret, and acts_as_solr. First of all, Sphinx is extremely stable. They're nearing a 1.0 release, and its stability certainly merits that. You don't even really need to worry about monitoring the search daemon, because it just isn't going to crash. Also, if you're running any rails apps in production, you know that long running ActiveRecord callbacks can lead to your app performing very poorly. The ferret and solr solutions both rely on active record callbacks to inform the indexer of new data. Moreover, if the search daemon goes down for any period of time, or, say, the index becomes corrupted (ferret, I'm looking at you), your entire app is going to be down for the count. You set the Sphinx indexer to run every half hour on a cron job, and since it works with the db directly, its performance and stability characteristics have absolutely zero impact on your app.

Trying it Out

So, let's implement a search for our blog. First, we'll want to edit the paths to the index and logs in the default config file. Note that Ultrasphinx uses two types of config files: one that the programmer edits, and one that it generates; both are necessary on all machines that are accessing the index (seriously, not having the generated config on a slave machine caused me some trouble). Then, we'll want to declare our post model as indexed (note that you must declare the fields as strings, not symbols):

class Post < ActiveRecord::Base
  is_indexed :fields => ['title', 'body']
end

Then, we'll need to ask Ultrasphinx to generate a configuration file. You've got to re-run this rake task any time you make changes to the definition of your models' is_indexed declaration. Make sure to .gitignore (svn:ignore for people still stuck in svn land) the generated file in development. Evan recommends checking the production file in to version control, but I have it set to generate automatically on deploy with a cap recipe. That way, if I make changes to the indexing, I won't forget to re-generate the config file. All you have to do is run:

$ rake ultrasphinx:configure

Then, since it's our first time, we've got to run the indexer:

$ rake ultrasphinx:index

Finally, we'll need to start the search daemon:

$ rake ultrasphinx:daemon:start

Now, we can start searching.

@search = Ultrasphinx::Search.new(:query => params[:query])
@search.run
@search.results

So, it's really easy to build a basic search engine using Ultrasphinx. There are some gotchas, though.

Gotchas & Notes

  • More complex indexing with Ultrasphinx can be slightly more verbose, and SQL-focused than it would be with a solution that relies on AR callbacks to do its indexing.
  • Transforming data with Ruby is impossible; the data you index must be in your database (unless you use a stored procedure, which can actually be written in ruby if you're using pgsql, but I digress).
  • Ultrasphinx preloads your indexed models when it is initialized. So, if your models depend on any monkey patches that live in your app's lib directory, they must be loaded before the Ultrasphinx (in my experience, this has meant pluginizing my monkey patches). Because of the way that exceptions are caught in the preloading routine, you won't see the actual error that is stopping your model from loading. Instead, you'll just get a constant loading error, or a name error, or something. If you see something like that, and your models load fine without Ultrasphinx installed, look for dependency issues.
  • Ultrasphinx attempts to preload your indexed models using a regex that doesn't respect commenting (at the time of writing). If you're struggling with issues mentioned in the last gotcha, you'll probably try commenting out the is_indexed call to see whether that's what's causing the problem. That won't work. You can either delete the is_indexed call entirely when debugging, or pull from my git repo, where I've modified the regex to respect #-style commenting (but not the =begin/=end style).
  • If you see DivideByZeroErrors in production, it's probably because you're missing the generated configuration file on one or more of your app server machines.

Check it Out

You'll need to grab Sphinx. Then, the plugin...

Get it from svn:

$ svn export svn://rubyforge.org/var/svn/fauna/ultrasphinx/trunk vendor/plugins/ultrasphinx

Or pull from my git repo (for the change described in the gotchas section):

git clone git://github.com/giraffesoft/ultrasphinx.git

To Sum Up

Ultrasphinx is by far the most effective rails searching solution I've come across. Unlike most of the other options, the search daemon is incredibly stable, and the index never seems to become corrupted (I'm running it in a relatively high load production environment with absolutely zero trouble so far). Also, since Ultrasphinx doesn't rely on AR callbacks for indexing, your application isn't quite as coupled to your search daemon; if it dies, search functionality will break, but the rest of your app will still function. It's not without problems, and complex indexing can be trickier, but Ultrasphinx's stability and performance make the choice a no-brainer.


Plugins I've Known and Loved #2: has_finder


Feb 25, 2008

Plugins I've Known and Loved started out as a would-be series of presentations at Montreal on Rails. I've been bad about getting to the second presentation, so I thought I'd continue PLIKaL here. No, I'm not going to talk about any of my plugins. Instead, over the next week or two, I'd like to tell you about some wonderful plugins that I only recently discovered (though, none of them are particularly new). Today — a plugin that belongs in everybody's toolbox: has_finder.

The basic idea of this school of plugins is creating reusable scopes for your models. For instance, if you had a blogging application with a post model, you might want to query for published posts. To that end, you might write something like this in your controller:

Post.find(:all, :conditions => {:published => true})

Hopefully, though, you're familiar with the skinny controller, fat model best practice, so you'd write it like this:

class Post < ActiveRecord::Base
  def self.published
    find(:all, :conditions => {:published => true})
  end
end

That's a lot better, but far from perfect. If you were to write several such methods, for example, you would not be able to combine them easily. So, looking for all published posts written by James would require writing a second method; not very DRY. Of course, ideally, you'd use an association proxy to accomplish that goal anyhow. That would work with our hand-written finder, but we'd lose all the benefits of the association_proxy like nested finds (...published.find(:all, :order => 'created_at DESC')), and we certainly wouldn't be able to chain two of them together (...published.order_by_recent). has_finder solves those problems, and more.

I Can Has Finder?

In order to reproduce the finder we wrote earlier, you'd write the following:

has_finder :published, :conditions => {:published => true}

Should you want to find all published blog posts by James, you could then make use of association proxy:

User.find_by_name('James').published.find(:all, :order => 'created_at DESC')

I have come across a good number of associations used for similar purposes, but defined on the associated model. For instance (omitting irrelevant details):

has_many :published_posts, :conditions => {:published => true}

Having discovered the wonders of has_finder, I now consider this to be an anti-pattern. The has_finder method is definitely DRYer. Far more importantly, however, the definition of what makes a post published (in our case, :published => true) belongs in the post model. The same rule applies to ordering. Following this best practice ensures there's only one point of change for refactoring, and that your models and their tests tell a more complete story about the data they represent.

Finally, has_finder supports parameterization of finders. You can wrap your conditions hash in a lambda that accepts an arbitrary number of arguments. Continuing with our blog post example, you might wish to query your posts table for all posts this week, this month, or this year, depending on the situation. To keep DRY, you could define your finder as follows:

has_finder :since, lambda { |date| {:conditions => ['created_at > ?', date]} }

Using it like this:

Post.since(1.week.ago)

To put it all together, let's query for all James's published posts in the last week:

User.find_by_name('James').posts.published.since(1.week.ago)

It reads just like a sentence!

Oh yeah, and all of this is compatible out of the box with will_paginate. I heard a rumor that it's going to be sucked down in to rails, too. Really, how could you not check it out (get it here)?


resource_controller 0.2: maintenance release - no more edge_compatible branch


Feb 15, 2008

This is mainly a maintenance release.

  • The helpers have been broken out in to four separate files internally, to help with managing the deep nesting branch.
  • There have also been a few little refactorings in preparation for some new features to come shortly.
  • The biggest thing to note for users is that there is no longer an edge_compatible branch. Since rails 1.2.6 generates the same style of named routes as 2.0.2 (edit_tag_photo_path instead of tag_edit_photo_path), there is no longer a need to continue two separate streams of development (yay)!.
  • The generator has been updated to spit out the right filenames for templates (rhtml/haml vs html.erb/html.haml), and old-style migrations (t.column instead of t.string) for any users still stuck on 1.2.6, so the transition shouldn't be a problem.

Get It!

You can get the new version by exporting from svn:

svn export http://svn.jamesgolick.com/resource_controller/tags/stable vendor/plugins/resource_controller

Or, if you're using piston, you may need to switch to the new url if you were previously on edge compatible (this is untested, so it may be slightly wrong).

  piston switch http://svn.jamesgolick.com/resource_controller/tags/stable
  piston update

As always, everybody is encouraged to come join the discussion on the increasingly lively mailing list.


A Response to Dreamhost


Jan 12, 2008

Dreamhost sure stirred things up with their article rant about rails deployment and performance issues. Their complaint is roughly that they are having to work too hard to support rails; it needs performance and deployment improvements. Dreamhost voiced all of these complaints on the company's blog.

As they put it...

What I do have personal knowledge of is how difficult it can be to get a Rails application up and running and to keep it running. DreamHost has over 10 years of experience running applications in most of the most popular web programming frameworks and Rails has and continues to be one of the most frustrating.
...the solution from the Rails community itself was quite honestly, stupid.
That suggestion shows either a complete lack of understanding of how web hosting works, or an utter disregard for the real world.
Ruby on Rails needs to be a helluva lot faster.
Ruby on Rails needs to more or less work in ANY environment.

Forgive me for chopping up their arguments. If you have the time, please read the article. These quotes don't quite do it justice; they serve merely to provide the tone of the piece.

Frankly, I think DHH responded well: rails core team is there to serve their own purposes. Ruby on Rails doesn't need to do anything, despite what a Dreamhost blogger might suggest. The people hacking on rails core don't target platforms like Dreamhost's. So, in fact, rails is not designed to run on shared hosting.

It's not just shared hosting, either. Dreamhost is trying to support rails on massively oversold servers. Configuring rails in a shared environment (particularly under apache) is difficult; configuring, and maintaining their servers to support rails probably costs Dreamhost a lot of money. Rails doesn't perform well in that environment, which probably costs them even more money in support calls from frustrated customers. So, their profit margins on rails service are likely narrowing.

Why is that the rails community's problem to solve? Dreamhost has a business challenge. Rails deployment isn't perfect — it has more than its fair share of problems; but, there are plenty of rails apps running just fine in production. This blog, a rails app, runs on a 20$ per month plan from slicehost. The identical app barely ran at Dreamhost. It's anecdotal evidence, sure, but the point is that rails deployment (at least on this scale) is far from impossible. Dreamhost just can't seem to squeeze it in to their oversold offerings. And, they want somebody else to fix the problem for them.

There are plenty of shared hosting offerings available that support rails nicely. Those companies don't seem to be having any trouble creating an environment that works. None of them have published angry articles pointing the finger at "the rails community". But, I digress.

If it were an individual complaining — a noobie who was having a tough time deploying his rails app on a server he could afford — then perhaps there would be a reason to look at this differently. But, Dreamhost is a corporation; they are simply "...looking to capitalize on a framework that's driving a lot of demand...", to borrow some words from DHH.

If this had been a PC manufacturer, complaining loudly (and rudely) that linux doesn't run on cheap enough hardware for them to sell PCs for 50$/each... If Linksys had posted a rant on their company blog, complaining that linux needed to be faster, because their routers weren't performing well enough... If IBM had posted a rant on their company blog, complaining that the linux community showed "...either a complete lack of understanding of how web hosting works, or an utter disregard for the real world."... Everybody would have said the same thing: so, fix it.

Plenty of businesses are capitalizing on open source technologies. But, part of the deal in the real world is that when something doesn't work right, you might have to fix it yourself; but you can, and that's part of what's so great about open source. Engine Yard just hired the entire rubinius team (and some people to hack merb, so I hear). Linux kernel developments are facilitated in large part by corporate sponsorship, bounties, etc. The system works, because companies can solve their own problems, while leveraging the work of countless others. That's how it works in the real world.


Noisy Backtraces Got You Down?


Dec 01, 2007

Bothered by all the noise in my tests' backtraces, I was thrilled when I first saw a thread on the shoulda mailing list with some discussion around making them a little bit easier on the eyes. Assuming that creating such a filter would be a long and tedious process of monkey-patching test/unit, I forgot about the idea, assuming the job better left for somebody with more time to spare than myself.

When Dan Croak revived the thread with some sample code, cooked up at a Boston.rb hackfest, it occurred to me that the job was far more manageable than I had originally conceived. I quickly fired Dan an email asking whether he'd be interested in a pluginization of their concept. With a resounding yes! from Dan, we set off to create quiet_stacktracebacktrace.

The rest of this post is cross-posted on GIANT ROBOTS

90% of this typical backtrace will not help you hunt and kill bad code:

1) Failure:
test: logged in on get to index should only show projects for the user's account. (ProjectsControllerTest)
[/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/assertions.rb:48:in `assert_block'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/assertions.rb:500:in `_wrap_assertion'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/assertions.rb:46:in `assert_block'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/assertions.rb:63:in `assert'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/assertions.rb:495:in `_wrap_assertion'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/assertions.rb:61:in `assert'
test/functional/projects_controller_test.rb:31:in `__bind_1196527660_342195'
/Users/james/Documents/railsApps/projects/vendor/plugins/shoulda/lib/shoulda/context.rb:98:in `call'
/Users/james/Documents/railsApps/projects/vendor/plugins/shoulda/lib/shoulda/context.rb:98:in `test: logged in on get to index should only show projects for the user's account. '
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/testcase.rb:78:in `__send__'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/testcase.rb:78:in `run'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/testsuite.rb:34:in `run'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/testsuite.rb:33:in `each'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/testsuite.rb:33:in `run'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/testsuite.rb:34:in `run'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/testsuite.rb:33:in `each'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/testsuite.rb:33:in `run'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/ui/testrunnermediator.rb:46:in `run_suite'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/ui/console/testrunner.rb:67:in `start_mediator'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/ui/console/testrunner.rb:41:in `start'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/ui/testrunnerutilities.rb:29:in `run'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/autorunner.rb:216:in `run'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit/autorunner.rb:12:in `run'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/test/unit.rb:278
test/functional/projects_controller_test.rb:36]:
one or more projects shown does not belong to the current user's account.
&lt;false&gt; is not true.

Noisy backtraces must be ruthlessly silenced like political dissidents in Stalinist Russia. This much is clear.

Quiet Backtrace

Install the gem:

  sudo gem install quietbacktrace

Require quietbacktrace:

## test_helper.rb
require 'quietbacktrace'

Run your Test::Unit tests:

1) Failure:
test: logged in on get to index should only show projects for the user's account. (ProjectsControllerTest)
[test/functional/projects_controller_test.rb:31
vendor/plugins/shoulda/lib/shoulda/context.rb:98
vendor/plugins/shoulda/lib/shoulda/context.rb:98
test/functional/projects_controller_test.rb:36]:
one or more projects shown does not belong to the current user's account.
&lt;false&gt; is not true.

Ooh la la! Now we’re cooking with gas. However, those shoulda-related lines are cluttering an otherwise perfect backtrace. Luckily, Quiet Backtrace is designed to be extended by calling two types of blocks that yield one line of the backtrace at a time.

Silencers and filters

  1. Silencers let you specify conditions that, if true, will remove the line from the backtrace.
  2. Filters let you use Ruby’s succulent set of String methods to modify a line by slicing and stripping and chomping away at anything you deem ugly and unnecessary. (such as the :in `__bind_1196527660_342195’ in the original example)

Say you want to remove Shoulda-related lines… you create a new silencer and add it the Array of backtrace_silencers:

class Test::Unit::TestCase
  self.new_backtrace_silencer :shoulda do |line| 
    line.include? 'vendor/plugins/shoulda'
  end
  
  self.backtrace_silencers << :shoulda
end

Re-run your tests and bask in the sweet sounds of silence:

1) Failure:
test: logged in on get to index should only show projects for the user's account. (ProjectsControllerTest)
[test/functional/projects_controller_test.rb:31
test/functional/projects_controller_test.rb:36]:
one or more projects shown does not belong to the current user's account.
&lt;false&gt; is not true.

Exquisitely sparse. Quiet Backtrace clears distractions from the “getting to green” TDD process like a Buddhist monk keeping his mind clear during meditation.

Getting noisy again

On occasion, you’ll want to see the noisy backtrace. Easy:

class Test::Unit::TestCase
  self.quiet_backtrace = false
end

You can set Test::Unit::TestCase.quiet_backtrace to true or false at any level in your Test::Unit code. Stick it in your test_helper.rb file or get noisy in an individual file or test. More flex than a rubber band.

Using Quiet Bactrace with Rails

After you have installed the gem on your local machine, I recommend using the excellent gemsonrails plugin to freeze it to your vendor/gems directory and automatically add it to your load path.

Install gemsonrails and add it your Rails app if you don’t already have it:

gem install gemsonrails
cd rails-app-folder
gemsonrails

Then freeze quietbacktrace:

rake gems:freeze GEM=quietbacktrace

Quiet Backtrace will now work with your tests, but because this gem is meant to work on any Ruby project with Test::Unit, it does turn any Rails-specific silencers or filters on by default. However, there is one of each, ready to be switched on, that remove the most dastardly lines.

Add these lines to your /test/test_helper.rb file to get perfectly clean Rails backtraces:

class Test::Unit::TestCase
  self.backtrace_silencers << :rails_vendor
  self.backtrace_filters   << :rails_root
end

Ongoing development

Bug reports and patches are welcome at RubyForge. Talk about it on the mailing list. It’s is a safe place. A place where we can feel free sharing our feelings. A nest in a tree of trust and understanding.


Introducing AttributeFu: Multi-Model Forms Made Easy


Dec 01, 2007

Ever needed to create a form that works with two models? If you have, you know that it's a major pain. It really seems like it should be easier, doesn't it? Powerful, SQL-free associations join our models; helpers build our forms in fewer keystrokes than it takes to call somebody a fanboy in a digg comment. But, as soon as you want to put the pieces together — as soon as you seek multi-model form bliss, the whole system falls apart.

Ryan Bates has popularized a few recipes for cooking up just such a dish (we all owe that man a debt of gratitude, don't we?). Great solutions (thanks Ryan!). There are only two flaws to speak of.

  • AJAX-friendliness: What is this - web1? Maybe we should all build our sites in frames, and pepper them with blinking text.
  • It's not a plugin: You mean I have to write code!? I thought rails was going to wash my dishes, walk my dog, and build my websites, while I sat outside on the porch smoking cigars, praying to DHH.

Being the compulsive pluginizer that I am, I simply couldn't resist this great opportunity.

attribute_fu

attribute_fu makes multi-model forms so easy you'll be telling all of your friends about it. Unless your friends are serious geeks, though, most of them will probably hang up on you. So, I recommend resisting that urge, unless you're looking for a good bedtime story.

To get started, enable attributes on your association (only has_many for now).

class Project < ActiveRecord::Base
  has_many :tasks, :attributes => true
end

Your model will then respond to task_attributes, which accepts a hash; the format of the which is a little bit different from what Ryan describes in his tutorials. Here's an example.

@project.task_attributes => { task_one.id => {:title => "Wash the dishes"},
                              task_two.id => {:title => "Go grocery shopping"},
                              :new => {
                                "0" => {:title => "Write a blog post about AttributeFu"}
                              }}

Follow the rules, though, and attribute_fu will shield you from the ugly details.

Form Helpers

The plugin comes with a set of conventions (IM IN UR PROJECT, NAMIN UR PARTIALZ). Follow them, and building associated forms is as easy as it should be. Take note of the partial's name, and the conspicuously absent fields_for call.

## _task.html.erb
<p class="task">
  <label for="task_title">Title:</label><br/>
  <%= f.text_field :title %>
</p>

Rendering one or more of the above forms is just one call away.

## _form.html.erb
<div id="tasks">
  <%= f.render_associated_form(@project.tasks) %>
</div>

You may want to add a few blank tasks to the bottom of your form — formerly requiring an extra line in your controller.

<%= f.render_associated_form(@project.tasks, :new => 3) %>

This being web2.0, removing elements from the form with Javascript (but your boss calls it AJAX) is a must.

## _task.html.erb
<p class="task">
  <label>Task</label><br/>
  <%= f.text_field :title %>
  <%= f.remove_link "remove" %>
</p>

Adding elements to the form via Javascript doesn't require any heavy lifting either.

## _form.html.erb
<%= f.add_associated_link('Add Task', @project.tasks.build) %>

It's that easy.

Last but not least (okay, maybe least), if you're one of those people who just has to be different - who absolutely refuses to follow the rules, you can still use attribute_fu (I guess...). See the RDoc for how.

Plugins, Plugins, Get Your Plugins

$ piston import http://svn.jamesgolick.com/attribute_fu/tags/stable vendor/plugins/attribute_fu

Check out the RDoc for more details.

Come join the discussion on the mailing list.


resource_controller 0.1.6: API Enhancements


Nov 26, 2007

It's been a while since the last r_c release; I have been busy. I'm still busy, so this isn't a huge release, but these changes have been in SVN for a while, so I thought I'd share them.

There was a cool thread over at RefactorMyCode.com where we discussed some potential API improvements for r_c (it's still open - come share your ideas!). Here's what made it in to this release...

respond_to

respond_to now accepts symbols, and has a few extra names. All of these examples have the same effect...

create.response :html do |format|
  format.js
end

create.respond_to :html, :js

create.responds_to do |format|
  format.html
  format.js
end

create.wants.html
create.wants.js

Callbacks Accept Symbols

I've always liked that the filters, and callbacks in rails accept the name of a method or a block. Now, all the callbacks in resource_controller accept them both, too!

create.before { @post.author = current_user }

create.before :set_user

private
  def set_user
    @post.author = current_user
  end

Configuration-Style Syntax for model_name, route_name, and object_name

Now, rather than overriding a method to provide alternate model, object or route names, you can set them through a configuration-style interface (though, you can still override the methods if you prefer).

model_name  :post_tag
route_name  :tag
object_name :tag

Note: resource_name is gone as of this release. If you were overriding that in your controllers, you'll have to use a combination of the other helpers to get the same effect.

LOTS More Helper Methods Exposed

If you've had a problem with r_c helper methods not being exposed in your views, it should be solved now. Just about every possible r_c method is now available in your views!

Get It!

1.2.3+ Compatible

svn export http://svn.jamesgolick.com/resource_controller/tags/stable vendor/plugins/resource_controller

Edge/Rails 2.0 Compatible

svn export http://svn.jamesgolick.com/resource_controller/tags/edge_compatible/stable vendor/plugins/resource_controller

If you're looking for older versions, try browsing http://svn.jamesgolick.com/resource_controller/tags.

For more info, see the resource_controller section of my blog.

Come join the discussion on the mailing list.


Looking for Rails Work?


Nov 09, 2007

I've posted this in a few places, and gotten some great responses. I don't know why I didn't think of posting it here earlier. (Of course, I give preference to anybody who reads my blog :)). Our little job blurb looks like this:

GiraffeSoft is looking to add an awesome rails developer to our team. If you live and breathe ruby & rails, are test(or spec)-obsessed (this one is absolutely 100% critical), want to work on a team that is actually agile, and get paid to work on our open-source projects (and maybe some of your own), we want to talk to you. We work from home, so, self-motivation, and management skills are critical. Where you are, and when you do the work, however, are not, so long as things get done on time. It will start out as a contract position, with the potential to become full-time down the line. If you're interested, send a short note, resume, and code samples to noticeme at giraffesoft dot ca.

Looking forward to hearing from you.


resource_controller: Redesign My API at refactormycode.com


Nov 05, 2007

Now that I've gotten a few maintenance releases out the door, I'm starting to feel like r_c's feature set is pretty solid; I think most people's needs should be covered (still a few things left to include, but we're getting there). The next step in improving resource_controller is evolving the API. What better place to hold the discussion than Marc-André's refactormycode.com?

First Refactoring designing

In the comments of the last release post, Luigi Montanez brought up an issue that's been bugging me since I first released the plugin. Currently, some customizations (before/after callbacks, response blocks, etc) are made with a "DSL"-style API. Some customizations (like alternate model names), however, are made by overriding helper methods. It's not that I don't like overriding methods, but I find the API a bit inconsistent in that way.

Currently, providing an alternate model name looks like this...

class PostsController < ResourceController::Base
  private
    def model_name
      'blog_post'
    end
end

It would be really neat, though, if it was possible to set the model_name the same way you set the flash...

class PostsController < ResourceController::Base
  model_name :blog_post
end

Please, come join the discussion at RmC, and put some of your ideas on the table for resource_controller's API.


resource_controller 0.1.5: Another Week, Another Release


Nov 01, 2007

Since last week, I've had some really great user feedback from the community. From that feedback, four new features made it in to this week's release. Here they are.

Support for Custom Route Names

Osh let me know that r_c didn't really support non-standard resource names. That is, it wasn't possible to have a controller whose name didn't match its model. Now, not only does r_c support non-standard resource names, but it supports every configuration of non-standard resources imaginable.

Osh's example was the simplest, and probably the most common. He has a resource with a name that doesnt't match the model. In that case, all you have to do is override model_name.

class TagsController < ResourceController::Base
  private
    def model_name
      'photo_tag'
    end
end

In that example, the variables, and form_builders available in your views would be named photo_tag(s). If you wanted to change that to match the name of your model, you'd override object_name.

class TagsController < ResourceController::Base
  private
    def model_name
      'photo_tag'
    end
    
    def object_name
      'photo_tag'
    end
end

Your controller might have some non-standard name, too. If it does, just override route_name.

## routes.rb
map.resources :tags, :controller => "photo_tags"

## photo_tags_controller.rb
class PhotoTagsController < ResourceController::Base
  private
    def route_name
      'tag'
    end
end

All of those new helpers default to the value of the resource_name helper, which is derived from the name of the controller.

New Urligence Syntax

In order to support non-standard routes in resource_controller, Urligence has gained some new syntax. smart_url infers the name of a resource from an object's class name, so a non-standard route name was impossible. Now, if you provide a 2-element array parameter to smart_url, the first element (a symbol) is used as the resource name, and the second element is the object that is passed to the url helper. It looks like this...

smart_url([:tag, @photo_tag])

All of the old syntax is still, of course, valid, and can be mixed and matched with the new syntax.

hash_for, path, and url

On the topic of Urligence, Hsiu-Fan raised the issue that smart_url would only return paths. Not really an accurate method name, and not convenient if you want a URL, or a hash. Hsiu-Fan was also nice enough to send in a patch, which I refactored a bit, and now we have four methods.

smart_url
smart_path
hash_for_smart_url
hash_for_smart_path

Important: The smart_url method now outputs URLs, not paths. If you are depending on paths coming out of smart_url anywhere in your code, or tests, this may be a code breaking change.

r_c has also gained helpers for all of these methods...

# object
[new_|edit_]object_url
[new_|edit_]object_path
hash_for_[new_|edit_]object_url
hash_for_[new_|edit_]object_path

# collection
collection_url
collection_path
hash_for_collection_url
hash_for_collection_path

New Instantiation Syntax

Another thing Osh brought up with me, during our back and forth, was his concern with r_c's inheritance syntax. Since there isn't any particular reason for that syntax other than the fact that I like it, I have added an alternative. Just say resource_controller, and you'll get the exact same effect as inheriting from ResourceController::Base.

class PhotosController < ApplicationController
  resource_controller
end

Just make sure that you call the resource_controller method before you use any other r_c features.

Get It!

1.2.3+ Compatible

svn export http://svn.jamesgolick.com/resource_controller/tags/stable vendor/plugins/resource_controller

Edge/Rails 2.0 Compatible

svn export http://svn.jamesgolick.com/resource_controller/tags/edge_compatible/stable vendor/plugins/resource_controller

If you're looking for older versions, try browsing http://svn.jamesgolick.com/resource_controller/tags.

For more info, see the resource_controller section of my blog.

Come join the discussion on the mailing list.


resource_controller update: Bug Fix, Generators, and Google Group!


Oct 25, 2007

A week after I released resource_controller, there are a few issues that needed to be resolved.

Specifying Actions

Hampton pointed out the fact that I hadn't mentioned any way of specifying which actions a controller should respond to. You might want to have a controller that only responds to show, and index, for example. It was possible (and documented - see the RDocs), but somewhat broken. This release fixes that bug.

You can either provide a list of actions, or use :all, :except => syntax.

class PostsController < ResourceController::Base
  actions :show, :index
  ## .. or ..
  actions :all, :except => :index
end

Generators

resource_controller comes with a scaffold_resource generator now. In the 1.2.3+ stream, it will override the built-in version. I figure - if you're using resource_controller, you probably won't have much need for the default one.

It's pretty clever, too. If you have Haml installed, it generates Haml. If you have Shoulda installed, it spits out functional tests that make use of Shoulda's wonderful should_be_restful macro (note: I'm not necessarily opposed to doing the same for RSpec, if there's demand - let me know). Otherwise, it spits out erb, and the same functional tests as the generator included with rails.

The syntax is exactly the same as the built-in generator.

script/generate scaffold_resource post title:string body:text

Get It

1.2.3+ Compatible

svn export http://svn.jamesgolick.com/resource_controller/tags/stable vendor/plugins/resource_controller

Edge/Rails 2.0 Compatible

svn export http://svn.jamesgolick.com/resource_controller/tags/edge_compatible/stable vendor/plugins/resource_controller

If you're looking for older versions, try browsing http://svn.jamesgolick.com/resource_controller/tags.

For more info, see the release announcement.

Community

I also wanted to announce that I've created a google group for resource_controller.

Come join it, and discuss what can be done to make r_c even better!


Introducing resource_controller: Focus on what makes your controller special.


Oct 19, 2007

If you're using rails, and you've kept up with what's happening in the community, and in the framework, you're probably writing RESTful apps. You may have noticed, like I did, that most of your controllers follow the same basic pattern - the one that the scaffold_resource (just scaffold in edge/2.0) generator spits out.

I wanted a great way to hide that pattern, and describe the unique features of my controllers through a more domain-specific API. Since I tend to write a lot of nested, and polymorphic resources, I also wanted to DRY up all of that code, and make generating urls a lot more intuitive. Finally, I wanted all of that code to be well-tested, so that I could count on hacking it up when I needed new features, without breaking things.

So, I created it, and here it is.

resource_controller

A basic, out of the generator resource just requires inheriting from ResourceController::Base.
class PostsController < ResourceController::Base  
end

API

Before/after callbacks, responses, and flash messages are managed through an extremely flexible API. I tried my best, here, to make syntax succinct when your needs are succinct, and allow API calls to be organized in an intuitive way. What I came up with was a scoping system, where calls to the various API methods can be chained, or scoped with a block.

Take the create action, for example. You might want to add a before block, to assign the author attribute of your post model to the current user.
class PostsController < ResourceController::Base
  create.before do
    @post.author << current_user
  end
end
As your application matures, you may want to add an RJS response to your create method, for some AJAXy goodness. Easy.
class PostsController < ResourceController::Base
  create.before do
    @post.author << current_user
    
    response do |wants|
      wants.html
      wants.js
    end
  end
end
Actually, there's even a shortcut for adding responses. You can omit the response block, if you just want to add responses to what's already provided by resource_controller (just HTML) or has been added elsewhere.
class PostsController < ResourceController::Base
  create do
    before do
      @post.author << current_user
    end
    
    wants.js
  end
end
Since I commonly use the same RJS template for several actions, I've found this short-form really handy, because I can use a one-liner to add the same response to several actions.
class PostsController < ResourceController::Base
  [create, update].each { |action| action.wants.js {render :template => "post.rjs"} }
end
Oh yeah, and you can change the flash too.
class PostsController < ResourceController::Base
  create.flash "Wow! RESTful controllers are so easy with resource_controller!"
end
For more API details, and examples, see the RDocs.

Helpers

Helpers are used internally to manage objects, generate urls, and manage parent resource associations.

If you want to customize certain controller behaviour, like member-object, and collection fetching, overriding helper methods is all it takes.

Note:For certain resource_controller functionality to work properly, user-defined helpers must make use of the other r_c helpers. The following examples do not follow that convention for clarity purposes - see the docs for more details.

If you wanted, for example, to use a permalink for your post, you'd need to alter the way that posts are fetched. Just override the object method.
class PostsController < ResourceController::Base
  private
    def object
      @object ||= Post.find_by_permalink(param[:id])
    end
end
You'll probably also want to add pagination to your index method. In the same way we altered member object fetching, we can change collection fetching behavior.
class PostsController < ResourceController::Base
  private
    def collection
      @collection ||= Post.find(:all, :page => {:size => 10, :current => params[:page]})
    end
end
Details and examples in the RDocs.

Namespaced Resources

...are handled automatically, and any namespaces are available, symbolized, in array form from the namespaces helper method.

Nested Resources

Again, handled automatically. This can be a real pain, and it's a lot easier with r_c. With an ActiveRecord-like syntax, just say belongs_to :model, and resource_controller takes care of the associations for you.
class CommentsController < ResourceController::Base
  belongs_to :post
end

Polymorphic Resources

This is a concept that can be found in a lot of my apps. Prior to resource_controller, it was a real pain on various levels. I solved some of my problems with urligence, and took things the rest of the way with resource_controller. It really does pretty much everything for you.

Just use the belongs_to syntax in your controller, and r_c infers whichever association (if any) is present when an action is called. The arguments passed to belongs_to are single possible parents; there is no support for deeply nested resources (although, I'm not necessarily opposed to adding it, if there is demandwe).
class CommentsController < ResourceController::Base
  belongs_to :post, :product
end
In the above example, the controller will automatically infer the presence of either a parent Post, or Product, and scope all the comments to that parent. The controller will also respond without a parent if you have that in your routes.

Thanks to urligence, generating urls in your polymorphic controller's views is really easy. The object_url, and collection_url helpers will maintain your parent resource's scope automatically.
# /posts/1/comments
object_url          # => /posts/1/comments/#{@comment.to_param}
object_url(comment) # => /posts/1/comments/#{comment.to_param}
edit_object_url     # => /posts/1/comments/#{@comment.to_param}/edit
collection_url      # => /posts/1/comments

# /products/1/comments
object_url          # => /products/1/comments/#{@comment.to_param}
object_url(comment) # => /products/1/comments/#{comment.to_param}
edit_object_url     # => /products/1/comments/#{@comment.to_param}/edit
collection_url      # => /products/1/comments

# /comments
object_url          # => /comments/#{@comment.to_param}
object_url(comment) # => /comments/#{comment.to_param}
edit_object_url     # => /comments/#{@comment.to_param}/edit
collection_url      # => /comments
More in the RDocs.

Getting it

resource_controller is available under the MIT License.

It is currently available in two streams:

1.2.3+ Compatible

Install it:

svn export http://svn.jamesgolick.com/resource_controller/tags/stable vendor/plugins/resource_controller

SVN (stable): http://svn.jamesgolick.com/resource_controller/tags/stable

SVN (ongoing): http://svn.jamesgolick.com/resource_controller/trunk

Note: If you want to run the tests, cd in to the test directory, and type rake test.

Edge/Rails 2.0 Compatible

Install it:

svn export http://svn.jamesgolick.com/resource_controller/tags/edge_compatible/stable vendor/plugins/resource_controller

SVN (stable): http://svn.jamesgolick.com/resource_controller/tags/edge_compatible/stable

SVN (ongoing): http://svn.jamesgolick.com/resource_controller/branches/edge_compatible

Note: If you want to run the tests in the edge-compatible version, cd in to the test directory, and type rake rails:freeze:edge. I didn't want people to have to download all of edge rails just to get the plugin.

Also, check out the rdoc

Feedback

Drop me a line and let me know what you think. I'd love to hear your suggestions for API or implementation improvements, or anything else!


Rails polymorphic url generation sucks. Here's something better.


Oct 11, 2007
If you've ever written a rails app with a lot of polymorphic resources, you've probably noticed that generating urls for your polymorphic routes is a real pain. In edge rails, some of the problems have been solved by the new polymorphic url methods. Unfortunately, those solutions are both inadequate, and "not recommended for use by end users" (see DHH's comment at the bottom). The problem is that if you want to use the same views and controller code without a lot of ugly conditionals to generate your urls, you're out of luck.

The Problem

## routes.rb

map.resources :photos
map.resources :users, :has_many => :photos
map.resources :tags, :has_many => :photos

## photos_controller.rb

class PhotosController
  ## in the create method
  wants.html { redirect_to url_for(url_helper_params) } # I know DHH doesn't want us to use this, but how else can I get this effect?

  ## in the destroy method
  wants.html { redirect_to photos_url } # unfortunately, this will drop any parent paths we may have
                                        # a user who called DELETE /users/james/photos/1 ends up at /photos instead of /users/james/photos)

  private
    # Note: I actually use a modification to make_resourceful to do this far more cleanly, but wanted to be as complete as possible in this example.
    def parent_object
      case
        when params[:user_id] then User.find_by_id(params[:article_id])
        when params[:tag_id] then Tag.find_by_id(params[:news_id])
      end    
    end
  
    def url_helper_params
      [parent_object, @photo].reject { |obj| obj.nil? } # we have to get rid of any the parent_object if it's nil, or else this will fail :(
    end
end

# index.rhtml

<%= link_to "New Photo", new_photo_path %> # same problem as our destroy redirect
We have routes like /users/james/photos, and /tags/montrealonrails/photos. We want to seamlessly maintain the top level of the path without resorting to conditionals or repetition. That is, if you browse to /users/james/photos, and click on the show link, you should go to /users/james/photos/1, not /photos/1. There should also be a polymorphic plural url helper. We have url_for for members (/photos/1), but for collections (/photos), we're stuck resorting to calling the url helpers themselves. So, to create a collection url, we'd have to check what the parent object was, and then call the appropriate helper based on that. And, all of this is further complicated when there are namespaces (/cms/products). Wouldn't it be great if there was a better way?

Stupid Simple Urls

Try saying that three times fast. Done? Good, because it's not called that - I was just messing with you.

URLigence

Intelligent url building.
## routes.rb

map.resources :photos
map.resources :users, :has_many => :photos
map.resources :tags, :has_many => :photos

map.namespace :admin do |admin|
  admin.resources :products, :has_many => :options
  admin.resources :options
end

##

assert_equal "/photos", smart_url(:photos)
assert_equal "/photos", smart_url(nil, nil, nil, :photos) # in case your parent objects aren't there
assert_equal "/users/james/photos", smart_url(User.find_by_name(:james), :photos)
assert_equal "/users/james/photos/1", smart_url(User.find_by_name(:james), Photo.find_by_id(1))
assert_equal "/photos/1", smart_url(nil, Photo.find_by_id(1)) # in the case that your parent object isn't there
assert_equal "/photos/1/edit", smart_url(:edit, Photo.find_by_id(1))

assert_equal "/admin/products", smart_url(:admin, :products)
assert_equal "/admin/products/1", smart_url(:admin, Product.find_by_id(1))
assert_equal "/admin/products/1/options", smart_url(:admin, Product.find_by_id(1), :options)
assert_equal "/admin/products/1/edit", smart_url(:edit, :admin, Product.find_by_id(1))
assert_equal "/admin/options", smart_url(:admin, nil, :options) # in the case that your parent object isn't there
It's pretty straightforward. Just call smart_url with any parameters that might be there, including ones that might be nil, and symbols at the beginning for namespaces (simple_url/cms/products), member actions (/products/1/edit), or both (/cms/products/1/edit), objects in the middle, and a symbol at the end for a resource's index method (/products). That's it.

Get It

Get version 0.1 at http://svn.jamesgolick.com/urligence/tags/stable. Oh, and don't take the name of the tag too seriously - I just wrote it tonight.

svn export http://svn.jamesgolick.com/urligence/tags/stable vendor/plugins/urligence

Let me know what you think of it.


Montreal on Rails #3


Oct 03, 2007
With our new digs, things were better than ever at last night's MoR. It's really great to see so many people sharing their knowledge, and just participating in the Montreal community. Aside from my presentation, Gary, and Francois both gave outstanding talks.

Javascripting for Rails Devs

Gary Haran talked to us about how to use prototype.js to ease some of the pains of writing javascript, and get yourself a beautiful ruby-like API. Since Gary taught me everything I know about Javascript while I was working at ZipLocal, I have been pushing him to share his knowledge with the masses, and he finally did with last night's presentation. Javascript (especially when combined with prototype) is awesome, and I hope this will be the first of many presentations from Gary on how to harness the awesome power in JS with less of the pain.

Piston: Vendor branch management made simple

François Beausoleil, the author of Piston, the excellent vendor branch management gem, was nice enough to come all the way to Montreal, and talk to us about his creation. In a nutshell, Piston aims to solve all of the problems that you get, when you depend on svn:externals to manage plugins in your rails app. Things like a new revision of a plugin breaking all your tests, and blocking deployment of a critical bug fix come to mind. I had heard a lot about Piston before his talk, but I am now fully sold, and will spend some time converting my current apps to use it this afternoon.

After all the talks were over, I had some great conversations with some of the awesome Montrealers who attend these great events. I talked Giraffes with Mat, iPhone(s) with Carl, number puzzles with Daniel, debated test macros with Marc-André, and discussed the latest and greatest in photography world with Alain Pilon. All in all, a great evening. Thanks to Fred and Ben for the space!

For more on last night's event, check out the photos, and my post about my presentation. As I said in my other post, video of last night's presentations will be available very soon.


Plugins I've Known and Loved #1


Oct 03, 2007
Last night at Montreal on Rails, I gave a talk about two of my favourite Rails plugins: make_resourceful & shoulda. For those of you who were unable to attend, the videos will be available very soon for your consumption.

make_resourceful

By encapsulating the standard RESTful controller pattern, make_resourceful allows you to focus on what's really important in your controllers, and saves you a ton of keystrokes along the way. The stable version is available here. Although, I prefer to live on the edge. Also, please join the discussion.

shoulda

Shoulda is the awesome set of test macros, and extensions to test/unit by the folks over at thoughtbot. A major time and keystroke saver for test-driven developers. (You are writing tests, right? No? Start here.) So, get your shoulda here, and be sure to check out the rDocs and tutorials.

Also, please check out the presentation slides, and the source code from my demo (note: see here for where I started, and HEAD of trunk is roughly where I ended). Also, see my review of MoR#3.


association_proxy.is_a?(Array), acts like an array, but doesn't save like an array


Sep 26, 2007
So, how many times have you had the urge to write:
@post.comments.reject!(&:spam?)
@post.save
Lots? Me too. So, I think it's worth clearing the air on our good friend association proxy. He really, really wants you to believe he's an array. He'll even tell you he is.
assert @post.comments.is_a?(Array)
But, don't believe him, because a lot of his array functionality is useless to us. Your favourite rubyisms (reject!, select!, etc) will modify the collection, but will not modify the associations. This test passes:
def test_association_proxy_really_looks_like_an_array_but_it_isnt
  assert Post.find(1).comments.is_a?(Array)
  
  assert Post.find(1).comments.select(&:spam?).length > 0
    
  post = Post.find(1)
  post.comments.reject!(&:spam?)
  post.save!
  
  assert_equal [], post.comments.select(&:spam?)
  assert Post.find(1).comments.select(&:spam?).length > 0
end
Instead, of course, you have to use the delete method.
post.comments.delete(post.comments.select(&:spam?))
Association proxy is tricky.


Promiscuity for Fun and Profit


Aug 30, 2007
Sometimes, we advanced rails users take our knowledge of the framework for granted, have unfair expectations for less advanced rails coders. The worst examples of such unreasonable demands can be found on a daily basis in the IRC channel (#rubyonrails on Freenode, for the uninitiated). And, while the IRC channel can certainly be harsh, even those of us who don't throw around the term "noob" like it's going out of style sometimes (often) complain more than we help. As one of my peers likes to remind me, I'm present in that category more often than I'd like to admit (if I didn't have to). As an aside: While Gary is often the one I'm ranting to, his code is never the subject of the rant. I'd really like to turn my negative energy in to something positive. This will be the first of many articles that seek to add to the pool of resources on rails best practices, shortcuts, tips, and tricks. I hope you like it.

Today's Topic: Polymorphic Associations

It was a check-in today that sparked the round of complaining that sparked the round of Gary telling me to shut up that sparked the desire to write this article. The check-in added a bunch of tables to a project, with a schema that looked (something) like this:
create_table :cities do |t|
  t.column :name, :string
  ...
end

create_table :city_aliases do |t|
  t.column :city_id, :integer
  t.column :alias, :string
end

create_table :provinces do |t|
  t.column :name, :string
  ...
end

create_table :province_aliases do |t|
  t.column :province_id, :integer
  t.column :alias, :string
end
Well, it's definitely not the worst schema ever, but it can be improved drastically with a polymorphic association. What the heck is that, you ask?

Promiscuous Models

Essentially, a polymorphic association is a model that can belong_to any other table. Polymorphic associations are the way that a lot of popular plugins work. Ever wonder how acts_as_taggable just magically adds tagging to any model, without any migration? Well, it's super easy. Just replace your duplicate tables with one of these babies:
create_table :name_aliases do |t|
  t.column :aliasable_type, :string
  t.column :aliasable_id, :integer
  t.column :name, :string
end
You'll need a model, too:
class NameAlias < ActiveRecord::Base
  belongs_to :aliasable, :polymorphic => true
end
Note: I chose the name NameAlias because alias is a reserved word in ruby, and it may cause you some problems if you try to name a model that.

Going back to our example, we can redefine our models as follows:
class City < ActiveRecord::Base
  has_many :name_aliases, :as => :aliasable
end

class Province < ActiveRecord::Base
  has_many :name_aliases, :as => :aliasable
end

Now What?

A polymorphic association works just like any normal association. That means you can use all of your favorite AssociationProxy methods, and everything.
City.find_by_name("montreal").name_aliases.create :name => "mtl"
...will work just as it always has.

How'd you do that?

When you tell rails that :polymorphic => true, in your belongs_to options, it knows that it needs to store the id and the type of the associated class. You give the association a generic name like aliasable or taggable, to let rails know what the fields are named in the database. For example, if you call your association aliasable, you need to have the fields aliasable_type and aliasable_id. Make sure aliasable_type is a long enough string column to store the name of your longest model in it. In production, both of these fields should probably be indexed. When you associate a class with your polymorphic model, you have to tell it to associate as your generic association name, instead of its own, to let rails know how to join the tables up properly: :as => :aliasable.

Name Aliases that Rock

Because this example is a particularly useful one, it's worth showing you a method that I write often. It will work with any association, but I often use it with aliases. It adds an extra finder, which allows you to find_by_name_alias.
class Person
  has_many :name_aliases, :as => :aliasable
    
  def self.find_by_name_alias(name)
      Person.find(:all, :include => :name_aliases, :conditions => ["name_aliases.name = ?", name])
  end
end

Feedback, please!

So, there you have it: polymorphic associations. This is where you come in. I have some questions:
  • How'd you like it?
  • What could be improved?
  • Should I use a different format (like a screencast)?
  • Are there any topics you'd like to see?
Thanks for reading.


3 ways to improve your bullshit methodology


Aug 10, 2007
Marc Cournoyer writes a great post, detailing 5 ways to know whether your methodology is working (or whether its bullshit), specifically re: TDD/XP.

I have been trying to improve my TDD practice for some time now. I am slowly getting better at writing tests first (and just writing tests, of course), but it does represent quite a significant shift in thinking. And, when you're used to writing the code first, as Marc says, that's where you're naturally going to go when the pressure is on. So, how do we stop this behavior? How do we get in to the test-first zone?

Here are 3 things that have started working for me:

1. When you're stuck on what to test, make a list of possible inputs and selected outputs


One of the biggest challenges for me has been overcoming my tendency towards doing something like exploratory testing of my own code, as I write it. This was the bad habit of not knowing what my code was going to do, before I wrote it. I'd spend some time fiddling around with a few lines that I thought might accomplish what I wanted, and looking at output, until it looked right (sound familiar?). With TDD, you have to start by thinking about what your code will output.
Take a second before you write any tests, and make a list of input parameters, and output expectations. Once you have this list, you'll see that it is much easier to know what you need to test for, and it will even help you write your code afterwards, too. This is an easy one, but it illustrates the point:
# PostsController#show
#
# Inputs:
#   params[:id]
# Outputs:
#   @post <-- contains the Post which corresponds to the params[:id] input parameter
#   OR
#   throws ActiveRecord::RecordNotFound if Post w/id == params[:id] does not exist

2. Make it an exercise and practice, practice, practice


Take 2 hours at home, in your spare time, and give yourself too much to do. Outline more features than you can realistically implement in that timeframe, and go for it. Racing the clock helps, because that's what you'll be facing on a real project. It sounds cheesy, but it has really worked well for me.

3. Use autotest

(the direct link is here, but it seems to be down right now)

The easier, and more comfortable testing is, the more likely you are to do it. Autotest watches all of your files, and when one changes, it runs the appropriate tests. All you have to do is save the relevant file, and look over at your terminal window to see the results. No more hitting refresh in your browser, or even running tests manually.
Marc also told me about CruiseControl.rb. I haven't had a chance to play with it yet, but it looks very cool. The idea is that if somebody checks something in to source control that breaks any tests, they are alerted immediately. Anything that makes testing easier is probably better.

What methodology-improvement tips do you have?


When to test


Aug 10, 2007

Always

A lot of people seem to struggle with knowing when to test their code. The simple answer is always, and everything.

Always test every line of code that you write.

Know that your code works (and interacts) as expected


First and foremost, writing tests provides you with a safety net. Making changes and refactoring code becomes a calculated maneuver, instead of a guessing game. Stop asking yourself: Did I break something? And know that if the tests didn't break, you're safe.

When you first write a block of code, manual testing might be sufficient to ensure that it works correctly, because its purpose is fresh in your mind. But, what about all the code it interacts with? Can you possibly remember test for all of the things you might have affected? Can the next user of your code remember to test for all of the things they might have affected? Of course not, and that's exactly why it's up to you to build those tests for them.

Living proof


Take a look at this snippet that got checked in to svn today:
belongs_to :user, :company
Had the author of this code written tests for his code, or even run the generated test suite for his model, he might have realized that he'd made a typo, and that the belongs_to method doesn't accept multiple models as its first argument, and that the correct code would have looked like this:
belongs_to :user
belongs_to :company
Test first, test always.