Question

I am trying to use both Capybara and Cucumber in my Rails application. So, that’s what i did:

• Installed gems needed and added them to the Gemfile
• Ran the rails generate capybara:install --cucumber
• Created a feature and its steps definitions for cucumber

Here’s my feature (yeah, i am creating blog application):

Feature: Browse posts
    So that I can browse through the posts
    As a visitor
    I want to see all and/or particular posts

    Scenario: Viewing all the posts
        Given Posts exist
        When I navigate to the website home
        Then I should see all the posts

    Scenario: Viewing particular post
        Given Post #1 exists
        When I navigate to /posts/view/1
        Then I should see the post with id=1

And here’s its step definitions:

Given /^Posts exist$/ do
    assert (not Post.all.empty?), "No posts found at all"
end

Given /^Post #(\d+) exists$) do |id|
    assert Post.find_by_id(id).valid?, "Post ##{ id } was not found at all"
end

When /^I navigate to (.+)$/ do |url|
    if url =~ /^the website home$/ then
        visit '/'
    else
        visit url
    end
end

Then /^I should see all(.+)posts$/ do |delimiter|
    posts = Post.all

    posts.each do |post|
        page.should have_content post.title
    end
end

Then /^I should see the post with id([^\d]{1,})(\d+)$/ do |delimiter, id|
    post = Post.find_by_id id

    page.should have_content post.title
end

And when running rake cucumber i get this message:

Then I should see all the posts     # features/step_definitions/browsing_posts.rb:13
  undefined method `should' for #<Capybara::Session> (NoMethodError)
  ./features/step_definitions/browsing_posts.rb:17:in `block (2 levels) in <top (required)>'
  ./features/step_definitions/browsing_posts.rb:16:in `each'
  ./features/step_definitions/browsing_posts.rb:16:in `/^I should see all(.+)posts$/'
  features/browsing_posts.feature:9:in `Then I should see all the posts'

What am I doing wrong? And yeah, are there any mistakes in my feature?

Answer 1

Replace page.should have_content post.title with assert page.has_content?(post.title). If that works, apply that similarly to the other have_content statements

Edit: Those statements involving should are based on rspec expectations, and should be used only if you are already using rspec or some other testing framework that responds_to should. Coming from the test::unit angle, this is probably just another kind of assert.