SeleniumRecord

Selenium Record is a DSL for easily writing acceptance tests. It is a wrapper over Selenium ruby bindings to let you easily apply the well known page object pattern.

Within your web app's UI there are areas that your tests interact with. A Selenium Record object simply models these as objects within the test code. This reduces the amount of duplicated code and means that if the UI changes, the fix need only be applied in one place.

selenium_record:install

Generates scaffolding for selenium objects

bundle exec rake selenium_record:install --test_framework=rspec

Options:

• 'test_framework': [String] Test framework to use. Possible values: 'rspec', 'cucumber', 'test_unit'.
• 'object_module': [String] Base module for selenium objects. Default: 'SeleniumObject'
• 'navigation_components': [Array] The names of the navigation components expected. Default: ['pages', 'tab']

Everytime you create a Selenium object it is fired the lookup process. This mean that based on selected strategy is defined the value for root_el (an instance of Selenium::WebDriver::Element). What really matters about this is that all find_element or find_elements operations will be done always in the scope of the root_el.

The lookup process defines two hooks:

• before_load_dom: This hook allows you to do some actions before the lookup process. For example, SeleniumRecord::NavigationItem takes this hook to click on a menu page and start lookup only once the new page it is loaded.

• after_load_dom: This hook allows you to do some actions after lookup process. For example, SeleniumRecord::Base takes this hook to inject components defined.

# selenium_record/component_autoload.rb

# Inject components after loading dom, providing reader methods for
# accessing them
# @param names [Array<Symbol>] component names. Valid formats Regex:
# /(?<component_name>.*)_(?<component_type>view|tab|pill|modal|panel)$/
def component_reader(*names)
  names.each { |name| create_component_reader name }
  define_method :after_load_dom do
    load_components names
  end
end

REMEMBER: All selenium objects should define their lookup strategy through the class method lookup_strategy.

In the next example it is search in first place element relative to element for locator returned by title_locator instance method. Once found this element it is searched the root_el using a relative xpath.

# spec/support/selenium_objects/panels/image_panel.rb
module SeleniumObjects
  module Panels
    # Selenium Page Object to interact with areas sections for a page
    class ImagePanel < Base::ApplicationView
      lookup_strategy :relative_title, locator: { xpath: '../../..' }
      
      # ...

This strategy searchs exactly for the element which the locator passed as parameter.

# spec/support/selenium_objects/panels/image_panel.rb
module SeleniumObjects
  module Panels
    # Selenium Page Object to interact with areas sections for a page
    class ImagePanel < Base::ApplicationView
      lookup_strategy :matching, locator: { xpath: ".//div[@class='todo-list']" }
      # ...

This strategy sets as root_el the element associated with html body tag

module SeleniumRecord
  # Base model to be extended by all Selenium page objects
  class NavigationItem < Base
    lookup_strategy :root
    # ...

When you call component_reader class method you pass a list of components that will be injected into the current component once it is instantiated. They will be available through the related accessor methods (Example: filter_panel).

# spec/support/selenium_objects/panels/search/filter_panel
module SeleniumObjects
  module Panels
    class FilterPanel < Base::ApplicationPanel
      # ...
      def search
        # Do search stuff
      end

# spec/support/selenium_objects/views/search_view
module SeleniumObjects
  module Views
    class SearchView < Base::ApplicationView
      navigate_to :labels
      component_reader :filter_panel, :results_panel
# ...

In addition all method calls to instance methods of components will be proxied through the container component if this component doesn't define the proper method. In the previous example:

search_view # => SearchView instance
search_view.search  # => equal to search_view.filter_panel.search

When you call action_component_reader class method you can pass a list of components that will perform an action once they are instantiated. Because of that, when you create an instance of ConceptView it will be generated only the accessors methods to create the related instance of DetailsTab (equal to other tabs).

As DetailsTab is a component of ConceptView, the attribute parent_el of DetailsTab will match the root_el of ConceptView.

# spec/support/selenium_objects/tabs/details_tab
module SeleniumObjects
  module Tabs
    class DetailsTab < Base::ApplicationView
      navigate_to :details

# spec/support/selenium_objects/views/concept_view.rb
module SeleniumObjects
  module Views
    class ConceptView < Base::ApplicationView
      lookup_strategy :root
      action_component_reader :main_tab, :details_tab, :audit_tab

If we focus on DetailsTab:

    class DetailsTab < Base::ApplicationView
      navigate_to :details

The method navigate_to will perform a click on link with localized string "details".

Of course you can customize to fit your needs this approach using a code like this one. In the next example we change from expected trans key to trans "txt.views.layouts.sections.#{key}"

# spec/support/selenium_objects/base/application_tab.rb
module SeleniumObjects
  module Base
    # Base class for all selenium objects representing an application tab
    class ApplicationTab < ::SeleniumRecord::NavigationItem

      def self.navigate_to(key)
        define_method :before_navigate do
          @title = trans "txt.views.layouts.sections.#{key}"
        end
      end

      # ...

Scopes allow you to execute code inside a block in the scope of the Selenium::WebDriver::Element associated to the scope definition. To define a scoped block you use the syntax: scope :my_symbol { #stuff }. The code will be executed in the context of the element with locator defined through method my_symbol_locator.

# spec/support/selenium_objects/views/concept_view.rb
module SeleniumObjects
  module Views
    class ConceptView < Base::ApplicationView
      lookup_strategy :root

      def create_new_version
        scope :fieldset_relations do
          # Methods from Actions module are executed relative to the element
          # specified by fieldset_relations_locator
          # See `selenium_record/actions` for more details
        end
      end

      def fieldset_relations_locator
        { xpath: ".//div[@class='relations']"
      end
      
      # ...

Everytime you call methods like find or find_elements, it will be returned the Selenium::WebDriver::Element instances with methods from modules:

• SeleniumRecord::Axiable

WARNING: Currently this behaviour it is only implemented for find

As a good practice, all stuff related to a explicit javascript library should be package inside plugins folder. As an example:

# spec/support/selenium_objects/plugins/jquery_autocomplete.rb
module SeleniumObjects
  module Plugins
    class JQueryAutocomplete < ::SeleniumRecord::Base
      attr_accessor :input_el
      # @param browser [Selenium::WebDriver::Driver]
      # @param locator [Hash] The locator of the input text autocomplete element
      def initialize(browser, locator)
        @root_el = browser
        @input_el = find(locator)
      end

      def perform(text)
        @text = text
        search_text
        select_from_pulldown_menu
        wait_for_tag_created
      end

      # ...

After running rake selenium_record::install, you should include a module in Rspec with:

module SeleniumRecordHelpers
  def browser
    @browser ||= page.driver.browser
  end

  def create_page(page_sym)
    klass = page_sym.to_s.camelize
    "SeleniumObjects::Pages::#{klass}Page".constantize.new(browser).load_dom!
  end
end

This gem is still under development! As this gem was born while I was trying to right acceptance tests in a more maintainable and productive way, it is not already tested!! Check the roadmap for upcoming updates. All pending tasks should be completed for '1.0.0' version. Keep up to date!

SeleniumRecord is on its way towards 1.0.0. Please refer to issues for details.

Add this line to your application's Gemfile:

gem 'seleniumrecord'

And then execute:

$ bundle

Or install it yourself as:

$ gem install seleniumrecord

TODO: Write usage instructions here

1. Fork it ( https://github.com/dsaenztagarro/selenium-record/fork )
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin my-new-feature)
5. Create a new Pull Request

• Selenium Wiki Page
• "Selenium 2 Testing Tools" by David Burns

Thanks to Hola Internet for let me right this kind of tools