I spent a little time working on some tests for a legacy web app that we plan to re-write in Ruby. Before the big re-write, I thought it would be wise to write some integration tests to call the app via http and verify responses. I wanted to use vanilla http and not have the overhead of launching a browser and was frustrated that Capybara didn’t have a driver to suit my needs.

I started down the path of writing tests from scratch using Net::HTTP, but figured that there had to be someone else who had run into this before. @kakutani told me about capybara-mechanize which seems to be exactly what I need.

Here’s my first really basic experiment using RSpec to test a well-known web site:

require 'capybara/rspec'
require 'capybara/mechanize'
Capybara.default_driver = :mechanize

describe 'web app' do
  include Capybara

  before do
    Capybara.app_host = "http://www.google.com"
  end

  it "has a copyright" do
    visit '/'
    page.should have_content('© 2011')
  end

end

One thought on “legacy testing with capybara/mechanize

  1. > I wanted to use vanilla http and not have the overhead of launching a browser and was frustrated that Capybara didn’t have a driver to suit my needs.

    Hopes this helps you….

    Akephalos[0]: “Akephalos is a full-stack headless browser for integration testing with Capybara. It is built on top of HtmlUnit, a GUI-less browser for the Java platform, but can be run on both JRuby and MRI with no need for JRuby to be installed on the system.”

    [0]: https://github.com/bernerdschaefer/akephalos

Leave a reply

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong> 

required