Extract Capybara setup to a helper file

Any Capybara-related test setup has been moved to `capybara_helper.rb`.
Now we just must require this file in order to run tests with Capybara.

I also moved the `#console_log` helper to this file to showcase how
other global helper methods could be defined in this file.

test/integration/test_js_errors.rb

@@ -1,9 +1,4 @@
-require File.expand_path(File.join(File.dirname(__FILE__), '..', 'helper'))
-require 'selenium-webdriver'
-require 'capybara/dsl'
-
-Capybara.default_driver = ::CAPYBARA_DRIVER
-Capybara.server = :webrick
+require_relative '../capybara_helper'
 
 def console_log(page, level = :severe)
   page.driver.browser.logs.get(:browser).select{|log| log.level == level.to_s.upcase }
@@ -21,7 +16,7 @@ end
 
 context 'Frontend with mathjax' do
   include Capybara::DSL
-  
+
   setup do
     @path = cloned_testpath("examples/lotr.git")
     @wiki = Gollum::Wiki.new(@path)
@@ -29,21 +24,21 @@ context 'Frontend with mathjax' do
     Precious::App.set(:wiki_options, {mathjax: true})
     Capybara.app = Precious::App
   end
-  
+
   test 'no unexpected errors on /' do
     visit '/'
     log = console_log(page)
     assert_only_expected_errors(log)
   end
-  
+
   test 'no unexpected errors on /create/' do
     visit '/create/Foobar'
     log = console_log(page)
     assert_only_expected_errors(log)
   end
-  
+
   teardown do
     Capybara.reset_sessions!
     Capybara.use_default_driver
   end
-end
+end