Python implementation of Testerra API.
It is basically a wrapper for Selenium WebDriver and WebElement which brings some more comfort features. This is not a test framework, but it implements assertion features anyway.
The basic concept is, to identify WebElements on every action or property accessor to prevent StaleElementExceptions.
Quick start
import inject import paf.config from paf.locator import By from paf.page import FinderPage, PageFactory # Configure dependency injection inject.configure(paf.config.inject) # Instantiate the page factory page_factory = inject.instance(PageFactory) # Create a simple finder page page = page_factory.create_page(FinderPage) # Visit URL page.open("https://google.com") # Find element element = page.find("#q") # Perform actions element.type("Search") # Perform assertions element.expect.text.be("Search")
Prerequisites
- You need at least a local WebDriver installed.
brew|choco|apt install chromedriver
Feature list
- UiElements
- Locators
- Page objects
- Components
- Managing WebDrivers
- Execution controlling
- Inject listeners
Missing features (tdb)
- Rect assertions
- Drag & Drop over frames
Environment variables
PAF_BROWSER_SETTING=chrome:90: Sets the requested browser name and it's version.PAF_WINDOW_SIZE=1920x1080: Sets the browsers initial window size.PAF_WINDOW_POSITION=0x0: Place the initial window to a different location/screen.PAF_WINDOW_MAXIMIZE=False: Maximize the initial window.PAF_SCREENSHOTS_DIR=screenshots: Sets the screenshots' directory.PAF_SEQUENCE_WAIT_AFTER_FAIL=0.3: Wait in seconds whenever a sequence action fails.PAF_SEQUENCE_RETRY_COUNT=3: Retry count for every sequence action.PAF_DEMO_MODE=0: Enables the demo mode by highlighting actions and assertions.PAF_SELENIUM_SERVER_URL=http://127.0.0.1:4444: Uses Selenium server if set.PAF_DRIVER_PATH: Path to Webdriver binaryPAF_BINARY_PATH: Path to User agent's binary
Examples
Some real world examples.
- test_google.py: is a regular Google search, implemented with Page Objects and Components.
- test_todo_mvc.py: are re-implemented test cases from the Robot Framework TodoMVC example. It's IMHO developer friendly, better readable and less code.
- test_asyncio.py: Parallel execution via
asyncio. - test_monkeytype.py: Typing test on https://monkeytype.com
Comparison
Comparison of the syntax with other frameworks.
Pylenium
py.get("a[href='/about']").should().have_text("About") find("a[href='/about']").expect.text.be("About")
SeleniumBase
self.assert_text_not_visible("Thanks for your purchase.", "#app .success") find("#app .success").expect.text.not_be("Thanks for your purchase.")
Selene
browser.all('#rso>div').should(have.size_greater_than(5)) \ .first.should(have.text('Selenium automates browsers')) div = find("#rso>div") div.expect.count.greater_than(5).be(True) div.first.expect.text.be("Selenium automates browsers")
References: https://www.nextgenerationautomation.com/post/python-test-automation-frameworks
Developer area
Testing
Run the tests on your local machine
PAF_TEST_HEADLESS=1 PAF_TEST_LOCAL_SELENIUM=0 pytest --cov=paf -n=4 testBuild test base container (for use in GitHub Actions)
podman build -f ubuntu-base.Dockerfile --arch=amd64 -t paf-test-base:latest source build.env echo $DOCKER_CONTAINER_REGISTRY_TOKEN | podman login -u mreiche --password-stdin ghcr.io podman push paf-test-base:latest docker://ghcr.io/mreiche/paf-test-base:latest
Build test runner container (for testing)
Run tests within container
podman run -e PAF_TEST_HEADLESS=1 -e PAF_TEST_CONTAINER=1 -e PAF_TEST_LOCAL_SELENIUM=0 paf-test-base:latest pytest --cov=paf -n=4 testRun local selenium server
wget https://github.com/SeleniumHQ/selenium/releases/download/selenium-4.9.0/selenium-server-4.9.0.jar -O selenium-server.jar java -jar selenium-server.jar standalone --host 127.0.0.1
Debug XPaths in Browser's Developer Console
xpath = "//dt[.//text()='Title:']/following-sibling::dd[1]" snapshot = document.evaluate(xpath, document.body, null, XPathResult.ORDERED_NODE_SNAPSHOT_TYPE) snapshot.snapshotItem(0).textContent