Back to Repositories

Validating URL State Management in Capybara Test Framework

This test suite examines URL handling and navigation behavior in Capybara, focusing on current URL, path, and host tracking. It verifies how Capybara manages application state across different server instances and navigation scenarios, ensuring proper URL management during user interactions.

Test Coverage Overview

The test suite provides comprehensive coverage of URL-related functionality in Capybara:

Direct page navigation and URL verification
Relative and absolute link handling
Form submission impact on URLs
Fragment handling in redirects
History API integration (pushState/replaceState)
iframe URL behavior

Implementation Analysis

The testing approach utilizes RSpec’s structured syntax with Capybara’s navigation methods. It implements multiple test servers to validate cross-origin behavior and employs helper methods for consistent URL verification. The suite demonstrates proper isolation of test scenarios and thorough state management.

Technical Details

Uses Capybara::Server for test environment setup
Implements custom helper methods for URL validation
Leverages RSpec expectations for assertions
Handles both HTTP and HTTPS protocols
Tests JavaScript-dependent features with specific requirements

Best Practices Demonstrated

The test suite exemplifies several testing best practices:

Proper test isolation with before hooks
Comprehensive edge case coverage
DRY principle through helper methods
Clear test descriptions and organization
Proper cleanup of global state

teamcapybara/capybara

lib/capybara/spec/session/current_url_spec.rb

            
# frozen_string_literal: true

require 'capybara/spec/test_app'

Capybara::SpecHelper.spec '#current_url, #current_path, #current_host' do
  before :all do # rubocop:disable RSpec/BeforeAfterAll
    @servers = Array.new(2) { Capybara::Server.new(TestApp.new).boot }
    # sanity check
    expect(@servers[0].port).not_to eq(@servers[1].port) # rubocop:disable RSpec/ExpectInHook
    expect(@servers.map(&:port)).not_to include 80 # rubocop:disable RSpec/ExpectInHook
  end

  def bases
    @servers.map { |s| "http://#{s.host}:#{s.port}" }
  end

  def should_be_on(server_index, path = '/host', scheme = 'http')
    # Check that we are on /host on the given server
    s = @servers[server_index]

    expect(@session).to have_current_path("#{scheme}://#{s.host}:#{s.port}#{path}", url: true)

    expect(@session.current_url.chomp('?')).to eq("#{scheme}://#{s.host}:#{s.port}#{path}")
    expect(@session.current_host).to eq("#{scheme}://#{s.host}") # no port
    expect(@session.current_path).to eq(path.split('#')[0])
    # Server should agree with us
    expect(@session).to have_content("Current host is #{scheme}://#{s.host}:#{s.port}") if path == '/host'
  end

  def visit_host_links
    @session.visit("#{bases[0]}/host_links?absolute_host=#{bases[1]}")
  end

  it 'is affected by visiting a page directly' do
    @session.visit("#{bases[0]}/host")
    should_be_on 0
  end

  it 'returns to the app host when visiting a relative url' do
    Capybara.app_host = bases[1]
    @session.visit("#{bases[0]}/host")
    should_be_on 0
    @session.visit('/host')
    should_be_on 1
    Capybara.app_host = nil
  end

  it 'is affected by setting Capybara.app_host' do
    Capybara.app_host = bases[0]
    @session.visit('/host')
    should_be_on 0
    Capybara.app_host = bases[1]
    @session.visit('/host')
    should_be_on 1
    Capybara.app_host = nil
  end

  it 'is unaffected by following a relative link' do
    visit_host_links
    @session.click_link('Relative Host')
    should_be_on 0
  end

  it 'is affected by following an absolute link' do
    visit_host_links
    @session.click_link('Absolute Host')
    should_be_on 1
  end

  it 'is unaffected by posting through a relative form' do
    visit_host_links
    @session.click_button('Relative Host')
    should_be_on 0
  end

  it 'is affected by posting through an absolute form' do
    visit_host_links
    @session.click_button('Absolute Host')
    should_be_on 1
  end

  it 'is affected by following a redirect' do
    @session.visit("#{bases[0]}/redirect")
    should_be_on 0, '/landed'
  end

  it 'maintains fragment' do
    @session.visit("#{bases[0]}/redirect#fragment")
    should_be_on 0, '/landed#fragment'
  end

  it 'redirects to a fragment' do
    @session.visit("#{bases[0]}/redirect_with_fragment")
    should_be_on 0, '/landed#with_fragment'
  end

  it 'is affected by pushState', requires: [:js] do
    @session.visit('/with_js')
    @session.execute_script("window.history.pushState({}, '', '/pushed')")
    expect(@session.current_path).to eq('/pushed')
  end

  it 'is affected by replaceState', requires: [:js] do
    @session.visit('/with_js')
    @session.execute_script("window.history.replaceState({}, '', '/replaced')")
    expect(@session.current_path).to eq('/replaced')
  end

  it "doesn't raise exception on a nil current_url", requires: [:driver] do
    allow(@session.driver).to receive(:current_url).and_return(nil)
    @session.visit('/')
    expect { @session.current_url }.not_to raise_exception
    expect { @session.current_path }.not_to raise_exception
  end

  context 'within iframe', requires: [:frames] do
    it 'should get the url of the top level browsing context' do
      @session.visit('/within_frames')
      expect(@session.current_url).to match(/within_frames\z/)
      @session.within_frame('frameOne') do
        expect(@session.current_url).to match(/within_frames\z/)
      end
    end
  end
end