Selenium Ruby tutorial

1. 🤖 How it works

Applitools SDKs work with existing test frameworks to take screenshots of pages, elements, regions or iframes and upload them along with DOM snapshots to our Eyes server. Our AI then compares them with previous test executions' screenshots (aka Baselines) and reports if there is a bug or not. It's that simple!

Applitools AI with RCA picture

1.1 Baseline vs. Checkpoint images

When you first run the test, our A.I. server stores those first set of screenshots as Baseline images. When you run the same test again (and everytime there after), the A.I. server compares the new set of screenshots, aka Checkpoint images, with the corresponding Baseline images and highlights differences in a pink color.

Baseline vs Checkpoint
The picture above is showing the Side-by-Side view of the baseline and checkpoint images

1.2 Marking the test as "Pass" or "Fail"

When the AI compares the baseline and the checkpoint image, if it finds a legitimate difference, it will mark the test as Unresolved. This is because the AI doesn't know if the difference is because of a new feature or a real bug and will wait for you to manually mark it as a Pass/Fail for the 1st time.

If you mark the Unresolved checkpoint image as "Failed", it'll only mark the current test result as Failed.

Mark the checkpoint as a fail
The picture above is showing how to mark the checkpoint image as Failed

Note:

To automatically mark the checkpoint as a "Fail" in the future test runs, you need to do the following:

  1. Annotate at least one of differences as a "bug region"
  2. Select the "Fail tests" checkbox in the popup window
  3. Press "Thumbs Up" (not "Thumbs Down") button in the checkpoint image's toolbar (Note: this is counter-intuitive. But what happens is that, we now create a new Baseline along with this bug and "Failed" metadata. So if the same image with the exact bug appears, it'll fail again)
  4. Press "Save" in the main toolbar

If you mark the Unresolved checkpoint image as a "Pass", then it means that the difference is due to a new feature so we set the new checkpoint image as the new baseline and mark the current test as Pass. Going forward we'll compare any future tests with this new baseline.

Mark the checkpoint as a Pass
The picture above is showing how to mark the checkpoint image as Passed

Note:

  • Applitools AI has been trained with 100s of millions of images. It doesn't do pixel-to-pixel comparison as this can lead to a lot of false positives. It instead simulates human eyes that ignore differences that humans can't detect and highlight differences that humans can detect.

  • ACCURACY: Our A.I.'s current accuracy rate is 99.9999%! Which means for most applications the odds that you'll see false-positives are 1 in a million!

A powerful test results dashboard

We provide a state-of-the-art dashboard that makes it very easy for you to analyze differences, report bugs and much more. For more information on the Applitools dashboard check out these articles.

Seeing test result summary
The picture above is showing the summary view
2. 🖼 Analyzing differences

The following Gifs show various tools Applitools provides to easily analyze various differences

Highlight differences between the baseline and checkpoint

Highlight diffs

Zoom into differences

Zoom into diffs

Toggle between baseline and checkpoint

Toggle differences between baseline and checkpoint

Show both the baseline and checkpoint side-by-side

Show both baseline and checkpoint side-by-side
3. 🐞 Reporting bugs (straight into Jira or Github)

You can select a section of the image and directly file a bug in Jira or Github. No need to manually take screenshots, write steps and explain things! To read more about bug regions check out this article.

Reporting bugs to Jira or Github
4. ✅ Prerequisites
  1. Create a free Applitools account and get the Applitools API KEY API-key

  2. Ensure Ruby is installed on your machine

  3. Install ChromeDriver on your machine and make sure it's is in the PATH.

    Here are some resources from the internet that'll help you.

Note:

Yoy may store the Applitools API Key in the environment variable. While this works for running the tests in the command line, if you are reading it from Eclipse or IntelliJ, then you need to launch those editors from the Terminal to read them. Otherwise, these editors won't load the environment variables!

  • Mac: export APPLITOOLS_API_KEY='YOUR_API_KEY'

  • Windows: set APPLITOOLS_API_KEY='YOUR_API_KEY'

5. 🚀 Run Tests

There are two main ways in which you can run tests.

The Applitools Ultrafast Grid (option 1) Test across all browsers and viewports 30-70 times faster than traditional test clouds. Using this approach, your tests will execute on a local browser once with the DOM, CSS, and additional artifacts captured by the SDK and sent to the Ultrafast Grid for parallel rendering and Visual AI comparison. This is the preferred execution method, however please note that it is a premium feature.

The second approach is to Run Local (option 2). This is the classic test execution in which your tests execute locally and each call to eyes via the SDK will send a screenshot to Eyes for Visual AI comparison.

While running tests using the classic approach is a great start - we recommend running tests on the Ultrafast Grid, the fastest way to ensure quality and consistency across all browsers, viewports, and devices.

Option 1 - Run With The Ultrafast Grid

1. Introduction To the Ultrafast Grid

In real world testing, you'll need to test your app on multiple browsers, viewports and mobile devices. So you'll need to run and re-run ALL the tests in different browsers and multiple viewports to ensure things are fine - which could take hours and hours to complete. You may also encounter browsers not opening properly, or hanging and what not.

What if you could run in just one browser, just once, and still do cross-browser tests across various browsers, viewports and multiple mobile emulators? That's where Ultrafast Grid comes in.

TL;DR

When you execute this tutorial, it'll run the test locally in Chrome browser but will show results in the following 5 combinations:

  1. Chrome browser (800 X 600 viewport)
  2. Chrome browser (700 X 500 viewport)
  3. Firefox browser (1200 X 800 viewport)
  4. Firefox browser (1600 X 1200 viewport)
  5. An iPhone4 emulator

... all in ~30 seconds!⏰🚀

2. What Is The Ultrafast Grid?

The Ultrafast Grid provides an elegant and an ultra fast way to render pages in different browsers, viewports and mobile simulators and then take screenshots and then send those screenshots to Applitools AI server. With Ultrafast Grid you just need to run your test in a single browser but perform end-to-end and cross-browser tests in various browsers such as Chrome,Firefox, IE, and also numerous combinations of viewports and in mobile simulators -- all in under a minute. Without Ultrafast Grid, you'll be wasting a lot of time and money running those tests in multiple browsers which could take hours and are error-prone.

And since Ultrafast Grid is integrated with our Eyes AI, you'll actually find a lot more functional and visual bugs than you'd normally find.

Further, you can use our Root-Cause-Analysis (RCA) feature to figure out exactly which DOM element or CSS caused difference! So you can go backwards, from the screenshot to the code!

Best part: All you need to do is to simply make eyes.check calls in your existing Selenium tests after adding our SDK and instantiating Ultrafast Grid. And that's all there is to it!

How it works

Once you add and instantiate the Applitools SDK, simply add eyes.check calls wherever you want to check if the state of the page looks OK. Then run the Selenium test as usual. And while running the tests, anytime it encounters eyes.check (or similar methods), Applitools SDK uploads the DOM resource of that page to our Ultrafast Grid behind the scenes and in batches. The Ultrafast Grid then renders every page in different browsers and mobile emulators (in parallel) and takes screenshots as per the configuration you set. Finally, it sends those screenshots to our AI Server.

When you first run the test, our AI server simply stores those 1st set of screenshots as Baseline images. When you run the same test again, the AI server will compare the new set of screenshots, aka Checkpoint images, with the corresponding Baseline images and highlights bugs.

Ultrafast Grid architecture picture

3. Getting Started With The Ultrafast Grid

Pre-requisites

  1. Ruby is installed on your machine.
  2. Chrome Webdriver is on your machine and is in the PATH. Here are some resources from the internet that'll help you.

Steps to run this example

Start with the ready-to-run Gitnub repo code

  1. Git clone this repo
    • git clone https://github.com/applitools/tutorial-selenium-ruby-ultrafastgrid.git
  2. Open the folder tutorial-selenium-ruby-ultrafastgrid
  3. Run command gem install bundler && bundle install. This will install necessary ruby gems and dependencies
  4. Set the APPLITOOLS_API_KEY. Login to Applitools > Click on the Person icon > My API Key
  5. Run the test: bundle exec ruby simple_test_script.rb# tutorial-selenium-ruby-visualgrid

Existing project

Adding Applitools to an existing project

gem install eyes_selenium
1

Source code

require 'eyes_selenium'

# Create a new chrome web driver
web_driver = Selenium::WebDriver.for :chrome

# Create a runner with concurrency of 1
visual_grid_runner = Applitools::Selenium::VisualGridRunner.new(1)

# Create Eyes object with the runner, meaning it'll be a Visual Grid eyes.
eyes = Applitools::Selenium::Eyes.new(runner: visual_grid_runner)

# Initialize eyes Configuration
eyes.configure do |conf|
  #  You can get your api key from the Applitools dashboard
  conf.api_key = 'APPLITOOLS_API_KEY'
  # create a new batch info instance and set it to the configuration
  conf.batch = Applitools::BatchInfo.new("Ultrafast Batch")
  conf.app_name = 'Demo App'
  conf.test_name = 'Ultrafast grid demo'
  conf.viewport_size = Applitools::RectangleSize.new(800, 600)
  # Add browsers with different viewports
  conf.add_browser(800, 600, BrowserTypes::CHROME)
      .add_browser(700, 500, BrowserTypes::FIREFOX)
      .add_browser(800,600, BrowserTypes::SAFARI)
      .add_browser(1600,1200, BrowserTypes::IE_11)
      .add_browser(1024,768, BrowserTypes::EDGE_CHROMIUM)
  #  Add mobile emulation devices in Portrait mode
  conf.add_device_emulation(Devices::IPhoneX, Orientations::PORTRAIT)
      .add_device_emulation(Devices::Pixel2, Orientations::PORTRAIT)
end

 # ⭐️ Note to see visual bugs, run the test using the above URL for the 1st run.
 # but then change the above URL to https://demo.applitools.com/index_v2.html
 # (for the 2nd run)
begin

  # Call Open on eyes to initialize a test session
  driver = eyes.open(driver: web_driver)

  # Navigate to the url we want to test
  driver.get('https://demo.applitools.com/index.html')

  # check the login page with fluent api, see more info here
  # https://applitools.com/docs/topics/sdk/the-eyes-sdk-check-fluent-api.html
  eyes.check('Login page', Applitools::Selenium::Target.window.fully)

  # Click the 'Log In' button
  driver.find_element(:id, 'log-in').click

  # Check the app page
  eyes.check('App Page', Applitools::Selenium::Target.window.fully)

  # Call Close on eyes to let the server know it should display the results
  eyes.close_async
rescue => e
  puts e.message
  # If the test was aborted before eyes.close / eyes.close_async was called, ends the test as aborted.
  eyes.abort_async
ensure
  # Close the browser
  driver.quit

  # we pass false to this method to suppress the exception that is thrown if we
  # find visual differences
  results = visual_grid_runner.get_all_test_results
  puts results
end
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
Option 2 - Run Locally

Run the demo app

  1. Get the code
    • Option 1: git clone https://github.com/applitools/tutorial-selenium-ruby-basic.git
    • Option 2: Download it as a Zip file
  2. Open the folder tutorial-selenium-ruby-basic
  3. Run command gem install bundler && bundle install. This will install necessary ruby gems and dependencies
  4. Set the APPLITOOLS_API_KEY. Login to Applitools > Click on the Person icon > My API Key
  5. Run the test: bundle exec ruby simple_test_script.tb

Add Applitools to an existing project

Install the SDK

gem install eyes_selenium
1

Sample test

require 'eyes_selenium'

runner = Applitools::ClassicRunner.new
eyes = Applitools::Selenium::Eyes.new(runner: runner)
web_driver = Selenium::WebDriver.for :chrome

eyes.batch = Applitools::BatchInfo.new("Demo Batch")

eyes.configure do |conf|
  conf.app_name = 'Demo App'
  conf.test_name = 'Smoke Test'
  conf.viewport_size = Applitools::RectangleSize.new(800, 600)
end

begin
  # Call Open on eyes to initialize a test session
  driver = eyes.open(driver: web_driver)

  # Navigate to the url we want to test
  driver.get('https://demo.applitools.com')

  # Note to see visual bugs, run the test using the above URL for the 1st run.
  # but then change the above URL to https://demo.applitools.com/index_v2.html (for the 2nd run)

  # check the login page
  eyes.check_window('Login window')

  # Click the 'Log In' button
  driver.find_element(:id, 'log-in').click

  # Check the app page
  eyes.check_window('App window')
  eyes.close_async
rescue => e
  puts e.message
  eyes.abort_if_not_closed
ensure
  # Close the browser
  driver.quit
  # Get and print all test results
  puts runner.get_all_test_results
end

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
6. ⚙️ Troubleshooting Common Issues
  1. Forgetting to set your API key (or getting 401 exception).

  2. Applitools is not opening the app (appears to crash).

    • In order to make sure the screenshots are consistent across different tests runs. Applitools uses Viewport as an input. A viewport is simply the size of the screenshot you are taking (width and height). Applitools tries to resize the browser to that size before taking the screenshot. If you are running the app on a small laptop and/or have set the Viewport much larger than the laptop’s screen, then you’ll see this error.
    • Solution: Reduce the height or the width. For more see here.
  3. Not properly loading the API key from the environment variable into your IDE (like Eclipse).

    • After setting the APPLITOOLS_API_KEY in the environment variable to hold your Applitools API key, open the IDE from the command line terminal (and not from the IDE directly). On Mac, it’d look like this: Open a Terminal and then type: open ~/Applications/Eclipse.app. This will load Eclipse with all the environment variables.
  4. Trying to run the test but nothing happens

    • Check if you the APPLITOOLS_API_KEY is null
  5. Incompatibility between Chrome browser and Chrome driver.

  6. Debug logs

    • See this article to enable debug logs to help file support ticket

Resources




Terms & Conditions Privacy Policy GDPR© 2020 Applitools. All rights reserved.

Last Updated: 5/22/2020, 8:21:31 PM