Just Maybe, Maybe - Pattern in Ruby

The original code I posted was flawed.  The gist below is modified and I think would work but it's not as clear as I'd like it to be.  I'll follow up on this soon.

I recently released a tiny Gem called JustMaybe.  This caused me to do more reading on the subject.  I ran across the If you gaze into nil nil gazes also into you post.  In the JustMaybe readme I sort of mention this but I wanted to express it more clearly here.

I don't really think using Maybe mixed in with the logic of your application is a good idea.  As an example I think the problem presented in that article could be solved much more clearly with this bit of code.  Here's a solution that will never return nil.



My suggestion is that if you find yourself wrestling with possible nil return values .  See if you can turn the operation on it's head.  Don't ask your object about it's associations.  Ask the association about your object.  I often find this will clean up the logic.

I did however just release JustMaybe so obviously I think it's useful in some way?  For me that's almost strictly in presentation.  When you are presenting information you often have to reach deep down into the hierarchy of objects to extract the necessary bits.  Interestingly, at presentation, is when I think you should be deciding what to display when information is missing.  To me that creates a perfect situation for using the Maybe pattern.

Extending Date Parsing In Ruby 1.9.3

I had previously made posts about parsing user supplied dates in Ruby

• http://blog.michaelgreenly.com/2012/01/more-parsing-user-supplied-dates-and.html
• http://blog.michaelgreenly.com/2012/01/parsing-user-supplied-dates-and-times.html

Apparently in Ruby 1.9.3 the Date library was re-implemented in C for performance and of course that broke my extension.  I updated it to support Ruby 1.9.3 below




This may not work on Ruby versions prior to 1.9.3 I didn't try it.

I also found the american_date gem which apparently existed before I had originally written my extension and is almost certainly a better choice if you don't need any custom formats.

Generic "config/database.yml" for Postgres

I'm really lazy and always use the exact same "config/database.yml" file with my Rails projects....

More Parsing User Supplied Dates and Times in Rails

Rails does a pretty good job of making work with time zones feel transparent.  If you do everything it's way, including use ActionView::Helpers::DateHelper.  If you use a Javascript control that returns a text input field or if you allow the users to enter values directly or need to process uploaded content things get a bit more complicated.

If the string returned by the text input can be parsed by Date._parse your in luck.  Everything will still work as expected with no extra effort.  If on the other hand you want to do something as simple as display U.S. style dates, in the format "MM/DD/YYYY", you will have to manually parse the input.  Ruby's Date._parse understands 'YYYY/MM/DD' and 'DD/MM/YYYY' but not 'MM/DD/YYYY'.  In the case where you're processing the input from a Javascript control the most direct solution is to do something like this...



The example above does return a time value in the current timezone and will work as expected.  One big problem with strptime is that it isn't very flexible, the input string must match the provided format string exactly right down the the separator characters and spaces.  So this approach can work if your processing the output of a control but tends to fail if you need to process human input.

Another problem I have with the approach above is that it quickly becomes incredibly redundant if you're working with a large number of date/time fields in many different models.  To Dry that up a bit I decided to move my specialized processing into a modified version of Date._parse.  This means I don't have to litter my controllers with date time parsing code and I have a single entry point to test that all desired formats are supported.



Basically all this is doing is taking the input string and re-arranging the values into an order the original method will understand.  The biggest disadvantage to this problem is that it doesn't support per-locale input formats.  It guess I'll cross that bridge if I ever get to it.

Another Quick Rails/Postgres Tip

Assuming your local timezone is not 'UTC' you most likely will want to configure Postgres to use it anyway on your development machine so that your development environment behaves like your production environment.  To do this just find the 'timezone=' line in postgresql.conf and set it to 'UTC'.

Parsing User Supplied Dates and Times In Ruby

I'm in the middle of correcting some date/time handling in a Rails app. One of things that was wrong was the handling of user input date/time fields.  Obviously when users provide date/time values they're assuming we understand the information was provided from the perspective of their timezone.  Especially when we don't explicitly require them to provide timezone information.

The problem with this is that Ruby's DateTime.parse method always returns UTC based times.  You can see this clearly in the example below:
You can see the problem on the 6th line in the example above.  Even though the user input "2011/01/01T00:00" when that DateTime instance is converted to localtime the current timezone offset was subtracted.  In this case resulting in a time that's 6 hours earlier than desired.

The solution is straight forward of course. You just have to add the inverse of the users timezone offset to the value returned from DateTime.parse as demonstrated below.


On line 4 you can see it's correct even when converted from UTC to local time. 

As an alternative you could include the timezone information before parsing it but that's not always easy because you don't necessarily now the format of the user's input.


Of course all of this is generally pretty obvious but I shared just in case it can help some one less familiar with the issues.

Time Zones in Rails using Postgres

It appears that by default Rails creates it's timestamps in Postgres using timestamps without time zone information.  This means that if you're working directly with the database, using psql for example, and want to see local time representations you'll have to do something like this.

Specifically notice that I'm using the TIMEZONE function to re-cast the column to a timestamp with timezone.  Since the columns implied timezone is 'UTC' that's what's used.  Now that we have a timestamp with timezone we can convert that to local time using the "AT TIME ZONE" syntax.

More on Twitter Bootstrap in Rails

Yesterday I wrote about error handling in Rails while using Twitter Bootstrap.  One of the interesting things I was able to do with the custom form builder was move the active record attribute errors to the input fields they're associated with.  The example below is a bit simple but it should get the point across....

If you look at the template that created the form below you can see that that it doesn't include any special functionality.  All the work happens in the form builder.





If you look at the "text_field" method in the code below you can see that it just calls "input_wrapper" to add some html around the "text_field".  Since "input_wrapper" already had to test each active record attribute for errors so that it could add "error" to the classes of those elements it was pretty straight forward to tack an inline-help span on the end.





The code above is not elegant, it's just an experiment.  I certainly don't recommend using it as is but I wanted to share just in case some one is thinking about similar problems.


In general though I am fairly happy with the separation of concerns so far.  I think that by treating the multi-element fields bootstrap expects in it's forms as primitives I will be able keep my templates focused on the content and my presenters on the logic of the view.

Formatting Rails Errors for Twitter Bootstrap

I'm using Twitter's Bootstrap toolkit for the layout in a proof of concept application I'm working on.  So far my general impression is that I really like Bootstrap but there are a few obstacles to making it play nice with Rails.

One of those obstacles is that Rails formats fields with errors quite a bit differently than Bootstrap's CSS expects.  Below I describe a possible solution.  I'm not 100% sure this is the best approach but it seems to work well and I'm currently working on a proof of concept app that will be thrown out in several weeks so there's little risk if I'm wrong.

Rails default formatting looks like this...


Bootstrap's CSS expects this....


The solution I've arrived at is to first remove the "field_with_errors" div.  This is done by setting the field_error_proc to simply return the html passed in without wrapping it.


The next step was to create a FormBuilder subclass with the desired functionality.  The example below only demonstrates the 'text_field' method.  In practice I'll have to create a separate method for each Bootstrap form element I want to use in my application.


Then to use the custom form builder I'd do somthing like this....


I realize this wasn't exactly a detailed explanation of a solution but hopefully I managed to get the gist of the idea across.

Writing the Reader's Stories for the First BDD Blog Release

I continue with my plan to write user stories for my first release. In this post I'll cover features from the readers perspective.

1. Must be able to Create, Edit and View posts.
2. Must support per post permanent URLs

Here's the first draft of the readers browsing feature. We'll see where it evolves to in the future.

Feature: browsing
As a reader
I want to browse the site
So that I can read it's content


#
# scenarios that involve the reader navigating to the
# home page as the number of available posts vary
#
Scenario: 0 of 0
    Given that 0 posts exist
    When I go to the home page
    Then I should be on the error page
    And I should see "page not found"

Scenario: 1 of 1
    Given that 1 post exists
    When I go to the home page
    Then I should be on post page 1
    And I should not see a link to the next post
    And I should not see a link to the previous post

Scenario: 2 of 2
    Given that 2 posts exist
    When I go to the home page
    Then I should be on post page 2
    And I should see a link to the previous post
    And I should not see a link to the next post


#
# Scenarios that involve the reader navigating to a specific
# post; the first, middle and last.  These primarily demonstrate
# the behavior of the navigation links to previous and next
# post pages
#
Scenario: 1 of 2
    Given that 2 posts exist
    When I go to post page 1
    Then I should not see a link to the previous post
    And I should see a lnk to the next post

Scenario: 2 of 3
    Given that 3 posts exist
    When I go to post page 2
    Then I should see a link to the previous post
    And I should see a link to the next post


#
# navigation to specific posts is done through post
# page permalinks.  The permalinks are derived from
# the title of the post
#
Scenario: navigate to an non-existant post
    Given that no posts have been created
    When I go to the path /posts/get-rich-overnight
    Then I should be on the error page
    And I should see "page not found"
    And I should see a link to the home page

Scenario: navigate to an existing post
    Given that post 1 exists
    And that post 1 has the title "Get Rich Overnight"
    And that post 1 has the body lorem ispum
    When I go to /posts/get-rich-overnight
    Then I should see "Get Rich Overnight"
    And I should see lorem ipsum

Writing User Stories for the First BDDBlog Release

In a previous post I shared the basics of my release plan, crude as it was. Next I'm going to create user stories for the entire first release once that's done I'll plan and execute regular weekly iterations to complete all the release stories. The one thing I'm going to do a bit different is that I'm going to rough out my user stories directly to cucumber features then I will maintain a list of uncompleted features that can be prioritized as I burn it down.

The first task at hand is to turn my first release's outline into Cucumber features. As I do this I'm not going to get overly concerned about the wording of my features. I expect that I'll be massaging them as I implement them.

With all that out of the way It's time to burn down the items on my release list and create some features. The first two items in the list is the most basic;

1. Must be able to Create, Edit and View posts.
2. Must support per post permanent URLs

Next I'll look at this from each type of users role; Reader, Author and Admin but I'm going to do that in seperate posts.

Some Progress On My BDD Blog Project

I finally got around to doing some more work on my BDD Blog project over the weekend. Up to this point I had mostly been just experimenting (here, here and here), but this time I decided I was going to try and write and implement a real scenario.

My goal though is not so much to implement code. Instead I'm still trying to learn how to write features and specs. I want to get to the point where I'm comfortable enough with Cucumber, Webrat and RSpec that I'm confident I can express all the features for a release without wasting my time.

So lets start with the feature I'm going to try and implement. The feature it self is very straight forward. It walks through the process used by an author when creating a new post.

Scenario: save a draft post
  Given I am logged in as John Doe
  And I am on the post creation page
  And I have filled in the title with "Get Rich Overnight"
  And I have filled in the body with "lorem ipsum"
  When I click the "Create" button
  Then I should be on the post index page
  And I should see "Post successfully created"
  And I should see "1 to 1 of 1 posts"
  And there should only be 1 post listed
  And the first post's title should be "Get Rich Overnight"
  And the first post's author should be "John Doe"
  And the first post's status should be "draft"

I've already played around with AuthLogic and demonstrated it's use but there's a minor change here. I'm no longer testing authentication so I needed to reduce the login process to a single given statement.

Given I am logged in as John Doe

The code I used to do that is pretty straight foward.

Given /^I am logged in as (.+)$/ do |login|
  Given "I am the registered user #{login}"
  visit path_to("the login page")
  fill_in('Login', :with => login)
  fill_in('Password', :with => "password")
  click_button("Login")
end

The only thing even remotely interesting here is that I'm calling one of my previously defined 'Given' statements from within this one.

The next few statements are very straight forward and were implemented entirely in cucumber with the existing scaffold for the Post resource.

And I am on the post creation page
And I have filled in the title with "Get Rich Overnight"
And I have filled in the body with "lorem ipsum"
When I click the "Create" button
Then I should be on the post index page
And I should see "Post successfully created"

The next line in this scenario:

And I should see "1 to 1 of 1 posts"

I now realize didn't belong in this scenario (I'll refactor that out later) but since I learned something while implementing it I'm going to share.

Specifically, what I learned was this statement has no real purpose. I wanted to make sure that only the new post was present but all this really does it check to see if the "1 to 1 of 1 posts" text is present. It doesn't provide any real validation of the number of posts on the page. In the future I'll move this out to a new feature that's specifically written to test the pagination summary.

This line was not originally in my feature:

And there should only be 1 post listed

I added it when I realized my mistake with the previous line. This statement directly tests the number of posts on the page by counting the table rows in the post index table. It's implementation is very straight forward:

Then /^there should only be (\d+) posts? listed$/ do |count|
  response.should have_tag("tr" , :count =>(count.to_i + 1))
end

The next three lines:

And the first post's title should be "Get Rich Overnight"
And the first post's author should be "John Doe"
And the first post's status should be "draft"

These lines were all very similar. They are simply looking to make sure that the contents of specific columns in the post index table have the correct information. This statement is implemented with the following code:

Then /^the (.+) post's (.+) should be "([^\"]*)"$/ do |index, column, text|
  response.should have_selector("table>tr:nth-child(#{position_to(index) + 1})") do |tr|
    tr.should have_selector("td[class='#{column}']") do |td|
      td.should contain(text)
    end
  end
end

There's not much special there but it was my first attempt at really trying to use the CSS selector notation. Maybe after I'm more comfortable with it I'll devote a post to just that but for now I'll just explain this method.

In a nutshell it simply looks to make sure that nth row in the post index table has a td tag with a class matching the column name and that it's content matches the text passed in.

You may also notice I have a poorly named method called position_to in this method that simply turns the english words; first, second, third, etc... into the appropriate fixnum. Hopefully down the road I can find a library that does this but if not maybe I'll expand my approach into a full fledged gem.

This pretty much covered what I did with Cucumber and Webrat to fully implement this scenario but I did dip down into some view, controller and model specs to get the job done. I'll save that for another post.

Can't Use Ruby 1.9.1 Just Yet

I had wanted to switch to Ruby 1.9 for my current project. I'm using a fairly minimal number of dependencies and didn't expect any problems. Unfortunately for what ever reason RSpec fails silently under Ruby1.9.1 but works properly on Ruby1.8.7 (at least on my machine). Since I don't have the time to track it down right now it looks like I'm reverting back to Ruby1.8.7 for the time being.

Hopefully I'll have time to dig into it later in the week.

Testing Rails Authlogic with Cucumber

In early rails projects I rolled my own authentication system, later I used restful-authentication. This time I'm going to try out AuthLogic which seems to be fairly popular, cleanly packaged and actively developed. I may walk through an AuthLogic installation in a future post but today I just want to make sure it can play nicely with cucumber.

First I wrote a fairly basic login scenario. You'll notice I stuck with the declarative style in writing this and I followed the advice provided in "The RSpec Book". I use direct model access to create the registered user in the 'Given' but then use simulated browser access to do the login and verify the actions success.

Scenario: successful login
Given I am the registered user John Doe
And I am on the login page
When I login with valid credentials
Then I should be on the account page
And I should see "Login successful!"

This left me with three steps to implement:

Given /^I am the registered user (.+)$/ do |login|
  params = {
    "login"=> login,
    "password"=>"password",
    "password_confirmation"=>"password"
  }
  @user = User.create(params)
end

When /^I login with valid credentials$/ do
  fill_in('Login', :with => @user.login)
  fill_in('Password', :with => "password")
  click_button("Login")
end

Then /^I should be on ([^\"]*)$/ do |page_name|
response.request.path.should == path_to(page_name)
end

All fairly straight forward. I don't expect any problems using Cucumber with Authlogic.

Baby Step #2 with Rails, RSpec and Cucumber

In my previous post I took my first stab at Cucumber; first writing a failing Scenario, then getting it to pass, then writing a failing Spec and getting it to pass. The scenario in the previous post was overly simple, in fact I was able to complete it without writing any code at all. My goal this time is to continue my education with a slightly more realistic scenario.

One way to attack an application is in the same order it's used. In the case of a blog the author will need to add content before the site has any value, so that seems like a reasonable starting point.

# RAILS_ROOT/features/author_publish.feature
Feature: author publish
As an author
I want to publish a post
So that it can be read

Scenario: publish a post
  Given I am on the post creation page
  When I add a new post
  Then I should see a page for the new post

The narrative for this feature was easy but deciding on how to approach the first scenario wasn't. In the end after browsing around the web and reading tons of material it was another post from Ben Maybe called "Imperative vs Declarative Scenarios in User Stories" that seemed to help me decide.

With that feature in place running cucumber:

./script/cucumber features/author_publish.feature

Produces the following:

Feature: author publish
As an author
I want to publish a post
So that it can be read

Scenario: publish a post                    # features/author_publish.feature:6
  Given I am on the post creation page      # features/step_definitions/webrat_steps.rb:6
    Can't find mapping from "the post creation page" to a path. (RuntimeError)
    /home/mgreenly/Projects/blog/features/support/paths.rb:11:in `/^I am on (.+)$/'
    features/author_publish.feature:7:in `Given I am on the post creation page'
  When I add a new post                     # features/author_publish.feature:8
  Then I should see a page for the new post # features/author_publish.feature:9

1 scenario
1 failed step
2 undefined steps

You can implement step definitions for missing steps with these snippets:

When /^I add a new post$/ do
pending
end

Then /^I should see a page for the new post$/ do
pending
end

The problem here is cucumber doesn't know what to map the phrase 'the post creation page' to. This can be fixed by adding the following code to the navigation helper.

# RAILS_ROOT/features/support/paths.rb
def path_to(page_name)
  case page_name

  when /the homepage/
    root_path

  when /the post creation page/
    post_path("new")

  else
    raise "Can't find mapping from \"#{page_name}\" to a path."
  end
end

Now we run cucumber again:

./script/cucumber features/author_publish.feature

And it produces the following:

Feature: author publish
As an author
I want to publish a post
So that it can be read

Scenario: publish a post                    # features/author_publish.feature:6
  Given I am on the post creation page      # features/step_definitions/webrat_steps.rb:6
    undefined method `post_path' for # (NoMethodError)
    /home/mgreenly/Projects/blog/features/support/paths.rb:9:in `/^I am on (.+)$/'
    features/author_publish.feature:7:in `Given I am on the post creation page'
  When I add a new post                     # features/author_publish.feature:8
  Then I should see a page for the new post # features/author_publish.feature:9

1 scenario
1 failed step
2 undefined steps

You can implement step definitions for missing steps with these snippets:

When /^I add a new post$/ do
pending
end

Then /^I should see a page for the new post$/ do
pending
end

Now it's complaining that it can't find a mapping for the Post resource. That makes sense since in my previous round I only created a controller. Instead of laboring through the process to create the Post model why not cheat and use the built in generator.

./script/generate rspec_scaffold -f Post title:string body:text

Obvious, going this route means I've broken my previously completed feature, so out of curiosity lets check it out.

Running cucumber on the original feature:

./script/cucumber features/reader_browse.feature

Produces the following:

Feature: browsing
As a reader
I want to browse the site
So that I can read it's content

 Scenario: visit the homepage       # features/reader_browse.feature:6
   When I go to the homepage        # features/step_definitions/webrat_steps.rb:10
   Then I should see "Hello world!" # features/step_definitions/webrat_steps.rb:93
     expected the following element's content to include "Hello world!":
     Posts: index
     Listing posts
     Title
  Body
New post
      (Spec::Expectations::ExpectationNotMetError)
     features/reader_browse.feature:8:in `Then I should see "Hello world!"'

1 scenario
1 failed step
1 passed step

Broken as expected. The problem is that the index page no longer has the "Hello world!" text on it. No problem we'll fix that later, for now lets keep going on the new feature.

When I run cucumber now:

./script/cucumber features/author_publish.feature

I get:

Feature: author publish
As an author
I want to publish a post
So that it can be read

Scenario: publish a post                    # features/author_publish.feature:6
  Given I am on the post creation page      # features/step_definitions/webrat_steps.rb:6
  When I add a new post                     # features/author_publish.feature:8
  Then I should see a page for the new post # features/author_publish.feature:9

1 scenario
2 undefined steps
1 passed step

You can implement step definitions for missing steps with these snippets:

When /^I add a new post$/ do
pending
end

Then /^I should see a page for the new post$/ do
pending
end

So now the it's obviously finding the 'the post creation page' but the remaining When/Then are still pending.

So now I add the following code:

# RAILS_ROOT/features/step_definitions/publish_steps
When /^I add a new post$/ do
  fill_in "Title" , :with => "Some Title"
  click_button "Create"
end

Then I run cucumber:

./script/cucumber features/author_publish.feature

And get the following:

Feature: author publish
As an author
I want to publish a post
So that it can be read

 Scenario: publish a post                    # features/author_publish.feature:6
   Given I am on the post creation page      # features/step_definitions/webrat_steps.rb:6
   When I add a new post                     # features/step_definitions/posts_steps.rb:1
   Then I should see a page for the new post # features/author_publish.feature:9

1 scenario
1 undefined step
2 passed steps

You can implement step definitions for missing steps with these snippets:

Then /^I should see a page for the new post$/ do
 pending
end

Now I just need to satisfy the 'Then'

# RAILS_ROOT/features/step_definitions/publish_steps
Then /^I should see a page for the new post$/ do
  response.should contain("Post was successfully created")
end

Then I run cucumber and get the following:

Feature: author publish
As an author
I want to publish a post
So that it can be read

Scenario: publish a post                    # features/author_publish.feature:6
  Given I am on the post creation page      # features/step_definitions/webrat_steps.rb:6
  When I add a new post                     # features/step_definitions/publish_steps.rb:1
  Then I should see a page for the new post # features/step_definitions/publish_steps.rb:6

1 scenario
3 passed steps

Everything works, except the feature I broke, but I'll pick up there in the next post.

Baby Step #1 with Rails, RSpec and Cucumber

There are lots of resources on the web to help get started with RSpec and Cucumber but there's still a learning curve that I think only experience will overcome. So I decided just to dive in. I don't claim what I've done here is right, or even a good idea, but it's where I started.

My initial goal is just to walk through the 'outside in' red/green/refactor process of writing a feature then examples without spending to much time worrying if I selected the right story or if it's implemented it correctly.

This initial cycle is mostly about learning what the tools can do for me.

The simplest story that I can imagine goes like this:

as a reader I want to browse the site so that I can read it's content.

So with that story in hand It's time to write my first feature

 #  ./features/reader_browses.feature
Feature: browsing
 As a reader
 I want to browse the site
 So that I can view it's content

Scenario: visit the homepage
 When I go to the homepage
 Then I should see "Hello world!"

Next I run Cucumber, 'rake features', and get the following....

 Feature: browsing
 As a reader
 I want to browse the site
 So that I can view it's content

Scenario: visit the homepage       # features/reader_browses.feature:6
 When I go to the homepage        # features/step_definitions/webrat_steps.rb:10
   undefined method `root_path' for # (NoMethodError)
   /home/mgreenly/Projects/blog/features/support/paths.rb:6:in `/^I go to (.+)$/'
   features/reader_browses.feature:7:in `When I go to the homepage'
 Then I should see "Hello world!" # features/step_definitions/webrat_steps.rb:93

1 scenario
1 failed step
1 skipped step
rake aborted!

So the first error I get is because we don't have a default route, so lets fix it by adding the following line in "./config/routes.rb"

map.root :controller => 'posts'

Then I run Cucumber again, "rake features", and get the following....

 Feature: browsing
As a reader
I want to browse the site
So that I can read it's content

Scenario: visit the homepage       # features/reader_browse.feature:6
 When I go to the homepage        # features/step_definitions/webrat_steps.rb:10
   uninitialized constant PostsController (NameError)
   (eval):2:in `/^I go to (.+)$/'
   features/reader_browse.feature:7:in `When I go to the homepage'
 Then I should see "Hello world!" # features/step_definitions/webrat_steps.rb:93

1 scenario
1 failed step
1 skipped step
rake aborted!

So now it's telling me that the default route points to a PostsController and that controller doesn't exist. So lets create both the controller and it's rspec framework with the built in generator.

./script/generate rspec_controller posts

Now I'll run the spec, "rake spec", to see what I get

 ..

Finished in 0.045209 seconds

2 examples, 0 failures

It passes of course because the controller doesn't do anything and the spec doesn't make any requirements on it.

So lets see what Cucumber says now, "rake features"

 Feature: browsing
 As a reader
 I want to browse the site
 So that I can view it's content

Scenario: visit the homepage       # features/reader_browses.feature:6
 When I go to the homepage        # features/step_definitions/webrat_steps.rb:10
   No action responded to index. Actions:  (ActionController::UnknownAction)
   /usr/local/stow/ruby-1.8.7-p72/lib/ruby/1.8/benchmark.rb:308:in `realtime'
   (eval):2:in `/^I go to (.+)$/'
   features/reader_browses.feature:7:in `When I go to the homepage'
 Then I should see "Hello world!" # features/step_definitions/webrat_steps.rb:93

1 scenario
1 failed step
1 skipped step
rake aborted!

There's still an error but it's changed. Now it's complaining because the there's no 'index' action on the Posts controller. So lets drop back to the rspec example for the Posts controller and create a requirement for the index action. Remember first I want it to fail, then pass.

# in ./spec/controllers/posts_controller_spec.rb
describe "GET 'index'" do
it "should be successful" do
get 'index'
  response.should be_success
end
end

So now when I run "rake spec" I get....

.F.

1)
ActionController::UnknownAction in 'PostsController GET 'index' should be successful'
No action responded to index. Actions:
./spec/controllers/posts_controller_spec.rb:12:

Finished in 0.052123 seconds

3 examples, 1 failure
rake aborted!

Now that the spec is failing we can try to satisfy it. To do that I just create an empty file at "app/views/posts/index.erb" and then rerun "rake spec"

 ...

Finished in 0.049891 seconds

3 examples, 0 failures

So now that the spec passes I run Cucumber again, "rake features", and get ....

 Feature: browsing
 As a reader
 I want to browse the site
 So that I can view it's content

Scenario: visit the homepage       # features/reader_browses.feature:6
 When I go to the homepage        # features/step_definitions/webrat_steps.rb:10
 Then I should see "Hello world!" # features/step_definitions/webrat_steps.rb:93
   Could not parse document (RuntimeError)
   features/reader_browses.feature:8:in `Then I should see "Hello world!"'

1 scenario
1 failed step
1 passed step
rake aborted!

It's still failing but now it's upset because it can't find the "Hello world!" text on the index page. This is easy to fix, I just need to add the text to "app/views/posts/index.erb". Now I run "rake features".....

Feature: browsing
 As a reader
 I want to browse the site
 So that I can view it's content

Scenario: visit the homepage       # features/reader_browses.feature:6
 When I go to the homepage        # features/step_definitions/webrat_steps.rb:10
 Then I should see "Hello world!" # features/step_definitions/webrat_steps.rb:93

1 scenario
2 passed steps

It passes. Just to make sure I didn't screw anything up I run "rake spec" on more time.

...
Finished in 0.049713 seconds
3 examples, 0 failures

That makes one successfull pass through an outside-in red/green/refactor cycle using Cucumber and Rspec with Rails.

Start a BDD Rails Project

I was starting a new rails project today that I plan on developing with Cucumber and RSpec. I quickly realized that even though there are tons of tutorials out there none seemed to really describe the process of initializing a new project. So here it is in a nutshell.

$ rails blog
$ cd blog
$ ./script/generate rspec
$ ./script/generate cucumber
$ rake db:migrate

At this point you should be able to do a 'rake', 'rake spec' or 'rake features' command and not get any errors.

Rails Development On Ubuntu

There's no shortage of posts on the web describing how to setup Ubuntu for Rails development but I'm going to add one more anyway. I choose to write this because none of the posts I ran across were geared towards a current Rails, RSpec, Cucumber, Git setup.

I'm using Ubuntu 9.04 (JauntyJackalope) during this installation.

First install Ruby and it's associated packages.

sudo apt-get install ruby-full

Next install the native SQlite3 libraries and command line tools.

sudo apt-get install sqlite3 libsqlite3-dev

Next install some native xml processing libraries used by webrat.

sudo apt-get install libxml2-dev libxslt-dev

Next install Git

sudo apt-get install git-core

Next install the tools necessary to build native rubygem extensions.

sudo apt-get install ruby-dev build-essential

Next install RubyGems from source because the Ubuntu packages are usually not current. I also prefer to install RubyGems into my $HOME directory because mixing source packages in with distribution managed packages is generally a bad idea.

wget -c http://rubyforge.org/frs/download.php/45905/rubygems-1.3.1.tgz
tar xvzf rubygems-1.3.1
cd rubygems-1.3.1
ruby setup.rb --prefix=$HOME/.rubygems --no-format-executable

Once that's done you need to set some environment variables so that; Ruby can find the gems library, RubyGems can find the gem repository and executable gems will be in the $PATH. Add the following to the file $HOME/.bashrc

export RUBYLIB=$HOME/.rubygems/lib
export GEM_HOME=$HOME/.rubygems/gems
export PATH=$HOME/.rubygems/bin:$GEM_HOME/bin:$PATH

Then make sure the RubyGems library is up-to-date and install all the necessary gems.

gem update --system
gem install rake rails rspec-rails cucumber webrat sqlite3-ruby

If you want to double check your installation you can try the following...

ruby --version
gem --version
rails --version
spec --version
cucumber --version

Before Filtering HTTP Method Types

Previously I wrote the post 'Method Not Allowed' which is about how I decided to handle unsupported HTTP request method types in my actions. Today I realized I should DRY up my code and put this logic into a before filter.

Bzr to Git

I think it's safe to say that the Rails community has decided git is the distributed revision control tool of choice. Unfortunately for me I've been using bzr for the last year or so. Lucky for me it's easy migrate from bzr to git using svn as an intermediary.

svnadmin create --fs-type fsfs /path/to/svn/project
cd /path/to/bzr/project
bzr svn-push /path/to/svn/project/trunk
git-svn clone  -T /path/to/svn/project/trunk /path/to/git/project