I'm a firm believer in the importance of top-down and behavior-driven development. I often start writing an integration test as the first step to implementing a story. When I started doing Rails development, the expressiveness of Ruby encouraged me to start building a DSL to easily express the way I most-often wrote integration tests. In the pre-RSpec days, this was just a subclass of ActionController::IntegrationTest that encapsulated the session management code to simplify authoring tests from the perspective of a single user. As the behavior-driven development idea started taking hold, I adapted the DSL to more-closely match those concepts, and finally integrated it with RSpec. The result of this effort was Spectie (rhymes with necktie).

The primary goal of Spectie is to provide a simple, straight-forward way for developers to write BDD-style integration tests for their projects in a way that is most natural to them, using existing practices and idioms of the Ruby language.

Here is a simple example of the Spectie syntax in a Rails integration test:

Feature "Compelling Feature" do
  Scenario "As a user, I would like to use a compelling feature" do
    Given :i_have_an_account, :email => "ryan@kinderman.net"
    And   :i_have_logged_in

    When  :i_access_a_compelling_feature

    Then  :i_am_presented_with_stunning_results
  end

  def i_have_an_account(options)
    @user = create_user(options[:email])
  end

  def i_have_logged_in
    log_in_as @user
  end

  def i_access_a_compelling_feature
    get compelling_feature_path
    response.should be_success
  end 

  def i_am_presented_with_stunning_results
    response.should have_text("Simply stunning!")
  end
end

Install

Spectie is available on GitHub, Gemcutter, and RubyForge. The following should get it installed quickly for most people:

% sudo gem install spectie

For more information on using Spectie, visit http://github.com/ryankinderman/spectie.

Why not Cucumber or Coulda?

At the time that this is being written, Cucumber is the new hotness in BDD integration testing. My reasons for sticking with Spectie instead of switching to Cucumber like the rest of the world are as follows:

• Using regular expressions in place of normal Ruby method names seems like a potential maintenance nightmare, above and beyond the usual potential.
• The layer of indirection that is created in order to write tests in plain text doesn't seem worth the cost of maintenance in most cases.
• Separating a feature from its "step definitions" seems mostly unnecessary. I like keeping my scenarios and steps in one file until the feature becomes sufficiently big that it warrants extra organizational consideration.

These reasons are more-or-less the same as those given by Evan Light, who recently published Coulda, which is his solution for avoiding the cuke. What sets Spectie apart from Coulda is its reliance on and integration with RSpec. The Spectie 'Feature' statement has the same behavior as an RSpec 'describe' statement, and the 'Scenario' statement is the same as the RSpec 'example' and 'it' statements. By building on RSpec, Spectie can take advantage of the contextual nesting provided by RSpec, and rely on RSpec to provide the BDD-style syntax within what I've been calling a scenario statement (the words after the Given/When/Thens). Coulda is built directly on Test::Unit. I'm a firm believer in code reuse, and RSpec is the de facto standard for writing BDD-style tests. Spectie, then, is a feature-driven skin on top of RSpec for writing BDD-style integration tests. To me, it only makes sense to do things that way; as RSpec evolves, so will Spectie.

The Short

I've written a plugin for Ruby on Rails that allows you to test SSL-dependent application behavior that is driven by the ssl_requirement plugin without the need to install and configure a web server with SSL.

Learn more

The Long

A while back, I wanted the Selenium tests for a Ruby on Rails app I was working on to cover the SSL requirements and allowances of certain controller actions in the system, as defined using functionality provided by the ssl_requirement plugin. I also wanted this SSL-dependent behavior to occur when I was running the application on my local development machines. I had two options:

1. Get a web server configured with SSL running on my development machines, as well as on the build server.

2. Patch the logic used by the system to determine if a request is under SSL or not, as well as the logic for constructing a URL under SSL, so that the system can essentially mimic an SSL request without a server configured for SSL.

Since I had multiple Selenium builds on the build server, setting up an SSL server involved adding a host name to the loopback for each build, so that Apache could switch between virtual hosts for the different server ports. I also occasionally ran web servers on my development machines on ports other than the default 3000, as did everyone else on the team, so that we'd all have to go through the setup process for multiple servers on those machines as well. We would need to do all of this work in order to test application logic that, strictly speaking, didn't even require the use of an actual SSL server. Given that the only thing that I was interested in testing was that the requests to certain actions either redirected or didn't, depending on their SSL requirements, all I really needed was to make the application mimic an SSL request.

To mimic an SSL request in conjunction with using the ssl_requirement plugin without an SSL server consisted of patching four things:

1. ActionController::UrlRewriter#rewrite_url - Provides logic for constructing a URL from options and route parameters

   If provided, the :protocol option normally serves as the part before the :// in the constructed URL.

   The method was patched so that the constructed URL always starts with "http://". If :protocol is equal to "https", this causes an "ssl" key to be added to the query string of the constructed URL, with a value of "1".

2. ActionController::AbstractRequest#protocol - Provides the protocol used for the request.

   The normal value is one of "http" or "https", depending on whether the request was made under SSL or not.

   The method was patched so that it always returns "http".

3. ActionController::AbstractRequest#ssl? - Indicates whether or not the request was made under SSL.

   The normal value is determined by checking if request header HTTPS is equal to "on" or HTTP_X_FORWARDED_PROTO is equal to "https".

   The method was patched so that it checks for a query parameter of "ssl" equal to "1".

4. SslRequirement#ensure_proper_protocol - Used as the before_filter on a controller that includes the ssl_requirement plugin module, which causes the redirection to an SSL or non-SSL URL to occur, depending on the requirements defined by the controller.

   This method was patched so that, instead of replacing the protocol used on the URL with "http" or "https", it either adds or removes the "ssl" query parameter.

For more information, installation instructions, and so on, please refer to the plugin directly at:

http://github.com/ryankinderman/mimic_ssl

If you use ActiveRecord observers in your application and are concerned about the isolation of your model unit tests, you probably want some way to disable/enable observers. Unfortunately, Rails doesn't provide an easy way to do this. So, here's some code I threw together a while ago to do just that.

module ObserverTestHelperMethods
  def observer_instances
    ActiveRecord::Base.observers.collect do |observer|
      observer_klass = \
        if observer.respond_to?(:to_sym)
          observer.to_s.camelize.constantize
        elsif observer.respond_to?(:instance)
          observer
        end
      observer_klass.instance
    end
  end

  def observed_classes(observer=nil)
    observed = Set.new
    (observer.nil? ? observer_instances : [observer]).each do |observer|
      observed += (observer.send(:observed_classes) + observer.send(:observed_subclasses))
    end
    observed
  end

  def observed_classes_and_their_observers
    observers_by_observed_class = {}
    observer_instances.each do |observer|
      observed_classes(observer).each do |observed_class|
        observers_by_observed_class[observed_class] ||= Set.new
        observers_by_observed_class[observed_class] << observer
      end
    end
    observers_by_observed_class
  end

  def disable_observers(options={})
    except = options[:except]
    observed_classes_and_their_observers.each do |observed_class, observers|
      observers.each do |observer|
        unless observer.class == except
          observed_class.delete_observer(observer)
        end
      end
    end
  end

  def enable_observers(options={})
    except = options[:except]
    observer_instances.each do |observer|
      unless observer.class == except
        observed_classes(observer).each do |observed_class|
          observer.send :add_observer!, observed_class
        end
      end
    end
  end
end

Include this in a Test::Unit::TestCase or 'include' in your RSpec configuration, whatever rocks your boat. Here's a stupid example:

class SomethingCoolTest < Test::Unit::TestCase
  include ObserverTestHelperMethods

  def setup
    disable_observers
  end

  def teardown
    enable_observers
  end

  def test_without_observers
    # ...
  end

end

When you go to test the behavior of the observer itself, simply disable/enable like the following to disable/enable all observers except the one you're testing:

class DispassionateObserverTest < Test::Unit::TestCase
  include ObserverTestHelperMethods

  def setup
    disable_observers :except => DispassionateObserver
  end

  def teardown
    enable_observers :except => DispassionateObserver
  end

  def test_without_observers_except_dispassionate_observer
    # ...
  end

end

In Rails, hooking up an ActiveRecord model to use a value object to aggregate over a set of database fields is a piece of cake. With the accessor methods that are created for a composed_of association, you can now deal exclusively with the composed_of field on your model, instead of directly manipulating or querying the individual database fields that it aggregates. Or can you? As long as all you're doing with the aggregate field is getting and setting its value, your aggregated database fields remain encapsulated. However, if you want to retrieve instances of your model from the database through a call to a finder method, you must do so on the individual database fields.

Consider the following ActiveRecord model definition:

class Customer < ActiveRecord::Base
  composed_of :balance, :class_name => "Money", :mapping => %w(balance_amount amount)
end

Given such a model, we can do something like this with no problem:

customer = Customer.new
customer.balance = Money.new(512.08)
customer.balance                      # returns #<Money:abc @amount=512.08>
customer.save!

However, now that we've saved the record, we might want to get that record back from the database at some point with code that looks something like:

customer = Customer.find(:all, :conditions => { :balance => Money.new(512.08) })

or like:

customer = Customer.find_all_by_balance( Money.new(512) )

This would provide full encapsulation of the aggregated database fields for the purposes of both record creation and retrieval. The problem is, at the time of my posting this article, it doesn't work. Instead, you have to do this:

customer = Customer.find(:all, :conditions => { :balance_amount => 512.08 })

To deal with this problem, I've submitted a ticket, which is currently scheduled to be available in Rails 2.1.

If you need this functionality, but your project is using a pre-2.1 release of Rails, I've also created a plugin version of the changes I submitted in the aforementioned ticket. To install:

script/plugin install http://svn.ryankinderman.net/find_conditions_with_aggregation

Addendum: The patch has been committed to changeset 8671. Yay!

I recently talked to a number of Rails developers about their general approach to testing some new functionality they're about to code. I asked these developers if they found it to be more useful to start testing from the bottom-up or top-down. I suggested to them that, since Rails uses the MVC pattern, it's easy to think of the view, or user interface, as the "top", and the model as the "bottom". Surprisingly, nearly every developer that I asked this question of answered that they prefer to start from the bottom, or model, and test upwards. Nearly every one! I expected that I'd get a much more mixed response than I have. In fact, I think that the correct place to start testing is precisely at the highest level possible, to reduce the risk of building software based on incorrect assumptions of how best to solve a user requirement.

Bottom-up Testing

Bottom-up testing implies bottom-up design in TDD. In bottom-up design, a developer would probably consider the high-level objectives and break them up into manageable components that interact with each other to provide the desired functionality. The developer thinks about how each component will be used by its client components, and tests accordingly.

The problem with the bottom-up approach is that it's difficult to really know how a component needs to be used by its clients until the clients are implemented. To consider how the clients will be implemented, the developer must also think about how those clients will be used by their clients. This thought process continues until we reach the summit of our mighty design! Hopefully, when the developer is done pondering, they can write a suite of tests for a component which directly solves the needs of its client components. In my experience, however, this is rarely the case. What really happens is that the lower-level components tend either to do too much, too little, or the right amount in a way that is awkward or complicated to make use of.

The advantage of bottom-up testing is that, since we're starting with the most basic, fundamental components, we guarantee that we'll have some working software fairly quickly. However, since the software being written may not be closely associated with the high-level user requirements, it may not produce results that are necessarily valuable to the user. A simple client could quickly be written which demonstrates how the components work to the user, but that's besides the point unless the application being developed is a simple application. In such a case, the bottom-level of components are probably close enough to the top-level ones that there is little risk involved in choosing either the bottom-up or top-down approach.

Unless you're writing a small application, the code is probably going to have to support unforeseen use cases. When this comes as a result of ungrounded assumptions about the software that's already been written, this can mean a lot of rework. I can tell you from experience, once you realize that your lower-level components don't fit the bill for the higher levels in the system, it can be quite a chore to go back and fix, remove, or replace all of that unnecessary or incorrect code.

Top-down Testing

Top-down testing implies top-down design in TDD. Following the top-down approach, the developer will pick the highest level of the system to be tested; that is to say, the part of the system that has the closest correlation to the user requirements. This approach is sometimes referred to as Behavior Driven Development. Whatever it's called, the point is that you test the most critical parts of the application first.

Since software is often written for human users, the most critical parts usually involve the front-end as it relates to the value being provided by the system being developed. When testing from the top-down, the effort is the inverse of bottom-up testing: Instead of spending a lot of time thinking about how the components to be developed will be used by other components to be developed, the focus is on how the user needs to interact with the system. Testing involves proving that the system supports the required usability. For an application with a graphical front-end, this might involve testing for a minimal version of that front-end.

The disadvantage of top-down testing is that you can end up with a lot of stubbed or mocked code that you then have to go back and implement. This means it might take longer before you have software that actually does something besides pass tests. However, there are ways that you can minimize this sort of recursive development problem.

One way to minimize the time between starting development of a feature and demonstrating functionality that is valuable to the user is to focus on a thin slice of the overall architectural pie of the application. For example, there may be a number of views that need to be implemented before the system provides some major piece of functionality. However, the developer can focus on one view at a time, or one part of the view. That way, the number of components that need to be implemented before the system does something useful is small; ideally, one component in each architectural layer that I need build out, and often times only a part of the overall functionality of each component.

Another way to minimize the amount of time before the system does something useful is to code a small bit of functionality without worrying about breaking the problem up into classes until you have some tested, working code to analyze. You can then use established methods for refactoring to bring the code to an acceptable level of quality.

The advantage of top-down testing is that you write functionality that solves the most critical functionality first. This generally means starting development at a high level. When the system eventually does something besides pass tests, what it does will provide value to its users. Additionally, because development starts at a high level, the code that is written is based on the current understanding of the problem, and not on assumptions. This guarantees that the tests and code that are written are not superfluous.

Conclusion

The challenge with top-down testing is that you must be highly disciplined to ensure that the code you write is being refactored and is properly evolving into a cohesive domain model for the application. This is compared with bottom-up testing, where you start with the domain model and build your system around it. Either way, you're going to be refactoring code. The difference is in where the time in refactoring is spent. In my experience, when doing bottom-up testing, more time is spent correcting incorrect assumptions about how the domain model will be used than on actually improving code that already works to solve the user requirements. In order to avoid making assumptions about the code being written, it must be written at the level that is closest to providing actual value to the end-user. In so doing, the developer focuses on continuous refinement of code that already provides value, as opposed to speculative design and development.

For those of you using the composite_primary_keys gem as of version 0.9.0, you may encounter an issue if you try to do something like:

class Reading < ActiveRecord::Base
  belongs_to :reader, :class_name => "User"
end

When a User is loaded up from the database via the reader association, the CPK modification to ActiveRecord::Reflection::AssociationReflection#primary_key_name incorrectly returns "user_id" as the primary key name. If you encounter this issue, I've submitted a patch against revision 124 that can be obtained here.

Hopefully this will get fixed in the next release. More hopefully, I won't need to care by then.

Sometimes, it's such a hassle to get up and turn on a light when I'm in the middle of coding. Other times, I don't have control over the lighting, like in a dim pub or cafe. In dim lighting, it's a strain on my eyes to be looking at black text on a white background. I've tried dimming the brightness, but that decreases the contrast, which is also a strain on my eyes. Then, I discovered Control+Option+Command+8, which is a keyboard shortcut to the "Reverse black and white" Universal Access feature. Contrary to it's name, it does more than reverse black and white, it reverses all of the colors on the screen. This, however, is acceptable and requires only a slight mental adjustment for me. Now I use a nice black-on-white color scheme in TextMate for coding, and when I'm in dim lighting, I simply reverse the colors and it's white-on-black, still full-contrast. I'm happy, except for one thing.

I often have a number of Terminal windows open when I'm working, and I open and close them frequently. Although I'm happy with a black-on-white color scheme for TextMate, I can't tolerate anything but white-on-black for my normal Terminal windows. So, what happens when I throw the screen colors in reverse is that all of my lovely white-on-black Terminal windows go to black-on-white. Unacceptable! Thus began my first foray into AppleScript.

After referencing here, here, and here, along with the Terminal "dictionary", I came up with this:

tell application "System Events"
  tell application processes
    key code 28 using {command down, option down, control down}
  end tell
end tell

tell application "Terminal"
  if default settings is equal to settings set "White on Black" then
    set settingsSet to "Black on White"
  else if default settings is equal to settings set "Black on White" then
    set settingsSet to "White on Black"
  end if
  set default settings to settings set settingsSet
  repeat with w in every window
    set current settings of w to settings set settingsSet
  end repeat
end tell

This script first invokes the Control+Option+Command+8 keyboard shortcut to reverse the screen colors. It then toggles the default "settings set" for Terminal to one of two pre-defined sets I've created: "White on Black" and "Black on White". Hopefully the names are self-explanatory. Then, the script iterates over all currently-open Terminal windows and changes their current "settings set" to whatever the new default is.

Pretty simple. I named the script "Reverse Screen Colors" and dropped it in /Users/[user account]/Library/Scripts. I invoke it via the keyboard with QuickSilver.

Today, I was trying to get Selenium to click an anchor tag that was created with the "link" plugin in a TinyMCE editor. I was able to verify that the link was present with something as simple as verifyElementPresent('link=Link Text'). However, when I tried calling clickAndWait('link=Link Text'), it gave a "Window does not exist" error. A quick Google search yielded the answer: a bug in Selenium Core.

When the TinyMCE "link" plugin creates a link that doesn't open in a new window, it sets the "target" attribute on the anchor tag to "_self". Selenium Core versions prior to 0.8.4 (which hasn't been released yet) don't respond to links with "target" set to "_self".

If you're doing Rails development and using the selenium_on_rails plugin, it uses an old version of Selenium Core (0.7.something) as of this posting. To fix the anchor tag problem, I replaced the contents of the selenium-core directory under vendor/plugins/selenium_on_rails with that of the core directory of the Selenium Core 0.8.3 release, then applied the patch described in the bug spec linked to above. This seems to have fixed the problem.

Hopefully this saves you all some time and muddling.

As I mentioned previously, the MetaKoans Ruby Quiz (#67) is a great way to learn meta-programming. However, it had some shortcomings. I've used MetaKoans as a training tool, and something I hear a lot is that it's unclear why certain things make a koan, or set of koans, pass. One reason for this confusion is that, often times while puzzling through the solution, a student will do something that causes multiple koans to pass at once. Due to the way that the koans are structured, I wasn't able to find a way to make a single koan pass at a time.

I've addressed this shortcoming by restructuring the koans so that the problem can be solved incrementally, one koan at a time. While restructuring the koans, I wrote a solution to each in turn, and saved that solution to its own knowledge file. Each file is a small refactoring from the one before it, ultimately building up to the final solution.

For the purposes of future training sessions, I've also started adding documentation to each refactoring, explaining how it changed from the previous one, and why it changed the way it did. The documentation isn't complete yet, but it's a start.

The restructured MetaKoans, along with the individual refactorings of my solution, can be found at http://svn.ryankinderman.net/metakoans_training. Feel free to check it out.

A little explanation

You'll notice that the knowledge files in my solution follow the pattern: knowledge_for_koan_XX_Y.rb. The XX number is the koan that the knowledge is a solution for. The Y number is the ordered refactoring index, with 1 being the first, most straight-forward solution, and subsequent indices being refinements of the original.

The reason for this structuring is that, often times, the straight-forward, brute-force solution to a koan isn't always the optimal solution. So, I'd make refactorings to show how the code could, IMHO, be improved.

Finally

Thanks go to ara.t.howard for coming up with the original MetaKoans quiz. It's been an extremely informative tool for myself and many others.

For those Typo users out there, this might be helpful.

As usually happens once every 3 or 4 blog posts, a few days ago I discovered a bug in Typo that cripples my blog and causes me to have to bring it down for a while. I'm not sure exactly which revision it was; one from version 4.0.2 I believe. Finding bugs so frequently probably has something to do with the fact that my idea of "upgrading" is doing an svn up from trunk and running db:migrate RAILS_ENV=production. Anyways, each time this happens, I end up spending at least 2 hours getting into "Typo debug mode", once getting passed "oh crap my blog is screwed mode", and fixing the problem. I don't mind debugging Typo. Having to do it now and then wouldn't be a problem if the types of problems I found were edge cases; unfortunately, they're usually not. These are things that, most definitely, should be covered by unit tests somewhere and caught when they get broken. This time around, I found three issues of this type.

The first issue was as simple as forgetting to update the values in the state column of the contents table in or before database migration 058. These values apparently were valid at some point with values such as ContentState::Ham, ContentState::Spam, ContentState::Published, and so on. However, they are now supposed to be simply ham, spam, and published, respectively. Not updating this data causes migration 058 to fail with Content complaining about an invalid state value. I had to go into mysql and update the values manually.

The second problem is also a simple oversight. After running migration 058, all comments are now stored in a table called feedback. Fine. However, whereas before the comments were displayed in the blog, after running the migration, you'll notice that they're not. This is because there is a published field in the feedback table that indicates whether a comment should be shown or not. After running migration 058, the value of this field for all comments on the blog is 0 (zero), which causes a comment to not be shown. Simply set this value to 1, and the comments are back again.

The third issue, which I discovered by accident from having a tail -f log/production.log running whilst fixing the first bug. I noticed that, every once in a while, a request would come in for Parameters: {"action"=>"show", "id"=>"some_category", "controller"=>"categories"}, which would result in an ActiveRecord::RecordNotFound error to be raised from Category.find_by_permalink. I noticed that the value of params[:id] was actually the name of a tag, not a category. So, I traced the error to Tag.permalink_url, which looked like this:

File: typo/app/models/tag.rb

def permalink_url(anchor=nil, only_path=true)
  blog = Blog.find(1) # remove me...

  blog.url_for(
    :controller => 'categories',
    :action => 'show',
    :id => permalink
  )
end

Woops! I guess we don't want to construct a URL for CategoriesController as the permalink for a Tag. I just changed :controller => 'categories' to :controller => 'tags' and the problem was fixed.

Addendum: Updated this entry to include a description of and resolution to the issue with the feedback.published database field.