So I've been struggling with this for quite some time now and I can't seem to figure our what's going wrong, and couldn't find much on what could possibly cause this issue.
I'm relatively new to Ruby and Rails, as well as test/behavior driven development and am trying to write some acceptance (browser) tests using PhantomJS through Poltergeist, using Rspec and Capybara. I believe some people also call this integration tests (they may be from some perspective), but that's a whole other discussion.
I have a really simple feature that I can't get to do what I want:
require 'feature_helper'
feature 'Logging in', :js => true do
scenario 'with incorrect credentials' do
visit '/login'
puts page.html
save_and_open_page
page.driver.render('_screenshot.png', :full => true)
page.html.should have_selector("title", :text => "hi")
end
end
So. Simple, right. It should just go to /login
and throw the HTML content at me, as well I want to see the page using save_and_open_page
, and I want it to take a screenshot. I added a simple should have_selector
in order to have the test fail in an attempt to get more feedback.
The relative contents of my feature_helper.rb
:
require 'spec_helper'
require 'capybara/rspec'
require 'capybara/rails'
require 'capybara/poltergeist'
include Capybara::DSL
Capybara.register_driver :poltergeist do |app|
Capybara::Poltergeist::Driver.new(app, {
:debug => true,
:inspector => true
})
end
Capybara.default_driver = :poltergeist
Capybara.javascript_driver = :poltergeist
FakeWeb.allow_net_connect = %r[^https?://(127.0.0.1|localhost)] # allow phantomjs/poltergeist requests
DatabaseCleaner.strategy = :truncation
RSpec.configure do |config|
config.before :each do
# Set the hostname to something with test
@host = "test.iome:3003"
host! @host
Capybara.default_host = Capybara.app_host = "http://#{@host}/"
Capybara.server_port = 3003
Capybara.reset_sessions!
# Start the database cleaner
config.use_transactional_fixtures = false
DatabaseCleaner.start
end
config.after :each do
DatabaseCleaner.clean
end
end
And also my spec_helper.rb
:
# This file is copied to spec/ when you run 'rails generate rspec:install'
ENV["RAILS_ENV"] ||= 'test'
require File.expand_path("../../config/environment", __FILE__)
require 'rspec/rails'
require 'pry'
require 'fakeweb'
FakeWeb.allow_net_connect = false
It's all fairly simple.
Now, in my console I see the following:
{"name"=>"visit", "args"=>["http://test.iome:3003//login"]}
{"response"=>{"status"=>"fail"}}
{"name"=>"body", "args"=>[]}
{"response"=>"<html><head></head><body></body></html>"}
<html><head></head><body></body></html>
{"name"=>"body", "args"=>[]}
{"response"=>"<html><head></head><body></body></html>"}
{"name"=>"render", "args"=>["_screenshot.png", true]}
{"response"=>true}
{"name"=>"body", "args"=>[]}
{"response"=>"<html><head></head><body></body></html>"}
Also, the screenshot is just a white and empty page. When I tail my log/test.log
file, I don't see that a request is being performed. I've tried changing the method visit
to get
, and that'll make the request, but won't change any of the results.
I've completely run out of ideas of what this could be and it's rather frustrating :(
Final information then about versions:
Unfortunately we're still at ruby 1.8.7, but are working on bumping that version up. Still, I think this shouldn't influence the tests.
Any help would be greatly appreciated!
So eventually I got help from a colleague, and we managed to fix it. We used the lvh.me
domain for this, as any request to that domain will resolve in localhost, allowing you to use subdomains without a problem. You could probably also use hostname.127.0.0.1.xip.io
for this.
Our spec_helper.rb
now looks like this:
# Use capybara in combination with poltergeist for integration tests
require 'capybara/rails'
require 'capybara/rspec'
require 'capybara/poltergeist'
require 'rack_session_access/capybara'
Capybara.default_driver = :poltergeist
Capybara.always_include_port = true
Capybara.app_host = 'http://application-test.lvh.me' # Any lvh.me domain resolves to localhost
Capybara.default_wait_time = 8 # How long capybara should look for html elements
require 'vcr'
VCR.configure do |config|
config.cassette_library_dir = 'spec/vcr_cassettes'
config.hook_into :fakeweb
config.ignore_localhost = true
config.configure_rspec_metadata!
config.ignore_hosts 'codeclimate.com'
end
require 'fakeweb'
FakeWeb.allow_net_connect = false
Because we hooked in VCR to record any requests going out during the first run of the integration tests, all your integration tests, or features, should contain this code:
before(:all) do
FakeWeb.allow_net_connect = true
end
after(:all) do
FakeWeb.allow_net_connect = false
end
If you want to change the subdomain during your specs, you can use the following:
before(:each) do
@original_host = Capybara.app_host
Capybara.app_host = 'http://does-not-exist.lvh.me'
visit '/login'
end
after(:each) do
Capybara.app_host = @original_host
end
Making screenshots can now be done using page.save_screenshot
during specs. Hope this helps.
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With