1. 🤖 How it works
Applitools SDKs work with existing test frameworks and takes screenshots of the page, an element, a region or an iframe and uploads 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!
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.
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.
To automatically mark the checkpoint as a "Fail" in the future test runs, you need to do the following:
- Annotate at least one of differences as a "bug region"
- Select the "Fail tests" checkbox in the popup window
- 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)
- 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.
Applitools AI has been trained with 100s of millions of images. It doesn't do pixel-to-pixel comparison because this can lead to a lot of false positives. It instead simulates human eyes that ignore differences 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.
2. 🖼 Analyzing differences
The following Gifs show various tools Applitools provides to easily analyze various differences
Highlight differences between the baseline and checkpoint
Zoom into differences
Toggle between baseline and checkpoint
Show both the 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.
4. ✅ Prerequisites
Create a free Applitools account and get your Applitools API KEY
- You may skip this step if you want to hard code the API KEY inside the tutorial project's code.
- It's better to store APPLITOOLS_API_KEY in the system variables (in Windows) or in the
~/.bash_profile(in Mac) so that it is accessible from all Terminal shells
Install git from https://git-scm.com
gitis optional. You need this mainly to clone the demo project from the Github repository. Instead of installing
git, you can simply download the Zip file from the repo. Further, If you are using Mac OSX, you already have
Ruby is installed on your machine
Chrome Webdriver is on your machine and in the PATH
- Download ChromeDriver from here
Ensure you download the ChromeDriver version that matches your Chrome browser version
5.1 🚀 - Run the existing demo app
Get the code:
- Option 1:
git clone https://github.com/applitools/tutorial-watir-ruby-basic
- Option 2: Download it as a Zip file and unzip it.
- Option 1:
Open the folder
gem install bundler && bundle installthis will install necessary ruby gems and dependencies.
Run the test:
bundle exec ruby simple_test_script.rb
5.2 🤓 - Add Applitools to an existing project
Install the SDK
gem install eyes_selenium
Example Watir Test
require 'eyes_selenium' require 'watir' runner = Applitools::ClassicRunner.new eyes = Applitools::Selenium::Eyes.new(runner: runner) browser = Watir::Browser.new(:chrome) Applitools::EyesLogger.log_handler = Logger.new(STDOUT).tap do |l| l.level = Logger::INFO end sconf = Applitools::Selenium::Configuration.new.tap do |conf| conf.api_key = ENV['APPLITOOLS_API_KEY'] conf.app_name = 'Demo App' conf.test_name = 'Ruby Watir demo' conf.viewport_size = Applitools::RectangleSize.new(800, 600) end eyes.config = sconf begin # Call Open on eyes to initialize a test session driver = eyes.open(driver: browser.driver) # Navigate to the url we want to test driver.get('https://demo.applitools.com/index.html') # 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('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) eyes.close(false) rescue => e puts e.message eyes.abort ensure # Close the browser driver.quit # Get and print all test results puts runner.get_all_test_results end