The Sauce Labs Cookbook

Sauce Headless

Front End Performance Testing

Insights

External Resources

More Info


Page tree
Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 6 Next »

With Sauce Labs you can run automated tests of your web apps against a wide variety of device/operating system/browser combinations.  This tutorial will show you how to get started with testing with Python on Sauce. 

Prerequisites

Code Example

This Python example script executes a simple test to open a browser, navigate to the Sauce Labs demo web app, and then close the browser. It also incorporate a number of best practices, such as using environment variables for authentication credentials. You can clone this script directly from the Sauce Labs Python training repo on GitHub and modify it as you want to test your authentication credentials, set desired capabilities, and run tests against your own web app. 


 Click here to view the example script

import pytest
import os
from selenium import webdriver
from _pytest.runner import runtestprotocol

@pytest.fixture
def driver(request):
    sauce_username = os.environ["SAUCE_USERNAME"]
    sauce_access_key = os.environ["SAUCE_ACCESS_KEY"]
    remote_url = "https://ondemand.saucelabs.com:443/wd/hub"

    sauceOptions = {
        'screenResolution': '1280x768',
        'seleniumVersion': '3.141.59',
        # best practices involve setting a build number for version control
        'build': 'Onboarding Sample App - Python + Pytest',
        'name': '4-best-practices',
        'username': sauce_username,
        'accessKey': sauce_access_key,
        # tags to filter test reporting.
        'tags': ['instant-sauce', 'pytest', 'module4'],
        # setting sauce-runner specific parameters such as timeouts helps
        # manage test execution speed.
        'maxDuration': 1800,
        'commandTimeout': 300,
        'idleTimeout': 1000,
    }

    chromeOpts = {
        'platformName': 'Windows 10',
        'browserName': 'chrome',
        'browserVersion': 'latest',
        'goog:chromeOptions': {'w3c': True},
        'sauce:options': sauceOptions
    }

    browser = webdriver.Remote(command_executor=remote_url, desired_capabilities=chromeOpts)
    yield browser
    browser.quit()


def test_should_open_chrome(driver):
    driver.get("https://www.saucedemo.com")
    actual_title = driver.title
    expected_title = "Swag Labs"
    assert expected_title == actual_title

Running the Example Code

  1. Copy the example code and save it into a file called first_test.py.
    Make sure your username and access key are included in the URL passed through to the command_executor.
  2. Open terminal and navigate to the directory where the file is located.
  3. Execute the test:

    python first_test.py
python first_test.py

Check your dashboard and you will see that your test has just run on Sauce!

Note: The implicitly wait method tells the browser to wait a set amount of time (in seconds) for elements to appear on the page before giving up. Without it, slow loading DOMs could cause our tests to fail when they might otherwise pass.

Let's look at the test a little closer so you can write one like it, or set your existing tests to run on Sauce.

 Running Tests on Sauce

If you wanted to run Selenium locally, you might initiate a driver for the browser that you want to test on like so:

driver = webdriver.Firefox()

However, to run on Sauce we use the general webdriver.Remote() instead of the specific webdriver.Firefox(), and then pass it two paramaters: command_executor and desired_capabilties.

(Note: webdriver.Remote() is a standard Selenium interface, so you can do anything that you could do with a local Selenium test. The only code specific to Sauce Labs was the URL that makes the test run using a browser on the Sauce Labs servers.)

# this is how you set up a test to run on Sauce Labs
desired_cap = {
    'platform': "Mac OS X 10.9",
    'browserName': "chrome",
    'version': "31",
}
driver = webdriver.Remote(
    command_executor='http://philg:45753ef0-4aac-44a3-82e7-b3ac81af8bca@ondemand.saucelabs.com:80/wd/hub',
    desired_capabilities=desired_cap)

 Command Executor

The command_executor code tells our script to use browsers in the Sauce Labs cloud instead of a local browser. Here we simply pass in a URL that contains your Sauce username and access key, which can be found on your dashboard.

 Testing on Different Platforms

Because we are not specifying webdriver.Firefox() as we were before, we must use desired capabilities to specify what browser/OS combination(s) to spin up and execute against.

If you want to run against different platforms, simply update your desired capabiltlies to something new, like so:

desired_cap = {'browserName': "chrome", 'platform': "Windows 8.1", 'version': "42.0"}

The desired capabilities are a set of keys and values that will be sent to the Selenium server running in the Sauce Labs cloud. These keys and values tell the Selenium server the specifications of the automated test that you will be running. Using our Platforms Configurator you can easily determine the correct desired capabilities for your test.

 Running Tests Against Local Applications

If your test application is not publicly available, you will need to use Sauce Connect so that Sauce can reach it. 

Sauce Connect is a tunneling app that allows you to execute tests securely when testing behind firewalls or on localhost. For more detailed information, please visit see the Sauce Connect docs

 Reporting to the Sauce Labs Dashboard

 Recording Pass/Failure Results

"Wait," you might be asking, "My test says 'Complete' but what happens if it fails?"

Unfortunately, Sauce has no way to determine whether your test passed or failed automatically, since it is determined entirely by your business logic. We can, however, tell Sauce about the results of our tests automatically using the Sauce python client:

pip install sauceclient

Then add this to your test:

# this authenticates you 
from sauceclient import SauceClient
sauce_client = SauceClient("philg", "45753ef0-4aac-44a3-82e7-b3ac81af8bca")

# this belongs in your test logic
sauce_client.jobs.update_job(driver.session_id, passed=True)

 Setting a Build Number

Now, you may want to associate this with a build id in your Continuous Integration pipeline. To do this, just include a build number in your desired capabilities:

desired_cap = {
    'platform': "Mac OS X 10.9",
    'browserName': "chrome",
    'version': "31",
    'build': "build-1234",
}

This is important for organizing tests in our new dashboard. Please see our docs for more information here

 Tagging Your Tests

You can also create custom tags for your tests that you can use to search on the archives page:

desired_cap = {
    'platform': "Mac OS X 10.9",
    'browserName': "chrome",
    'version': "31",
    'build': "build-1234",
    'tags': [ "tag1", "tag2", "tag3" ]
}

 Running Tests in Parallel

Now that you're running tests on Sauce, you may wonder how you can run your tests more quickly. Running tests in parallel is the answer! To do this we will need to use a third party test runner. The most popular third party test runners for Python are py.test and nose.

 Parallel Testing with py.test

First install py.test:

pip install selenium sauceclient pytest pytest-xdist

Now, let's use the platform configurator to add two platforms that we want to test on. Replace "desired_cap" in our example script above with an array of configs like so:

browsers = [{
    "platform": "Windows 10",
    "browserName": "internet explorer",
    "version": "11"
}, {
    "platform": "OS X 10.11",
    "browserName": "safari",
    "version": "8.1"
}]

Just add and remove browsers here as you see fit.

Then create a decorator and add your tests like in this example below. You can actually just add your test cases to this script:

import os
import unittest
import sys
import new
from selenium import webdriver
from sauceclient import SauceClient

browsers = [{
    "platform": "Windows 10",
    "browserName": "internet explorer",
    "version": "11"
}, {
    "platform": "OS X 10.11",
    "browserName": "safari",
    "version": "8.1"
}]

username = 'philg'
access_key = '45753ef0-4aac-44a3-82e7-b3ac81af8bca'

# This decorator is required to iterate over browsers
def on_platforms(platforms):
    def decorator(base_class):
        module = sys.modules[base_class.__module__].__dict__
        for i, platform in enumerate(platforms):
            d = dict(base_class.__dict__)
            d['desired_capabilities'] = platform
            name = "%s_%s" % (base_class.__name__, i + 1)
            module[name] = new.classobj(name, (base_class,), d)
    return decorator

@on_platforms(browsers)
class FirstSampleTest(unittest.TestCase):

    # setUp runs before each test case
    def setUp(self):
        self.desired_capabilities['name'] = self.id()
        self.driver = webdriver.Remote(
           command_executor="http://%s:%s@ondemand.saucelabs.com:80/wd/hub" % (username, access_key),
           desired_capabilities=self.desired_capabilities)

    # verify google title
    def test_google(self):
        self.driver.get("http://www.google.com")
        assert ("Google" in self.driver.title), "Unable to load google page"

    # type 'Sauce Labs' into google search box and submit
    def test_google_search(self):
        self.driver.get("http://www.google.com")
        elem = self.driver.find_element_by_name("q")
        elem.send_keys("Sauce Labs")
        elem.submit()

    # tearDown runs after each test case
    def tearDown(self):
        self.driver.quit()
        sauce_client = SauceClient(username, access_key)
        status = (sys.exc_info() == (None, None, None))
        sauce_client.jobs.update_job(self.driver.session_id, passed=status)

Since we are running on two platforms, use -n2 like so to run your tests in parallel:

py.test -n10 first_test.py

Visit your dashboard and you should see four tests running at once! Use the py.test docs to learn more about how to properly use py.test for running tests in parallel.

 Parallel Testing with nose

First, install nose:

pip install selenium sauceclient nose==1.1.0

Now, let's use the platform configurator to add two platforms that we want to test on. Replace "desired_cap" in our example script above with an array of configs like so:

browsers = [{
    "platform": "Windows 10",
    "browserName": "internet explorer",
    "version": "11"
}, {
    "platform": "OS X 10.11",
    "browserName": "safari",
    "version": "8.1"
}]

Putting it all together, here is a working script that will run tests in parallel on Sauce with the browsers we have assigned above:

import os
import sys
import inspect
from nose.tools import with_setup
from selenium import webdriver
from sauceclient import SauceClient

browsers = [{
    "platform": "Windows 10",
    "browserName": "internet explorer",
    "version": "11"
}, {
    "platform": "OS X 10.11",
    "browserName": "safari",
    "version": "8.1"
}]

username = 'philg'
access_key = '45753ef0-4aac-44a3-82e7-b3ac81af8bca'

def launchBrowser(caps):
    caps['name'] = inspect.stack()[1][3]
    return webdriver.Remote(
            command_executor = "http://%s:%s@ondemand.saucelabs.com:80/wd/hub" % (username, access_key),
            desired_capabilities = caps);

def teardown_func():
    global driver
    driver.quit()
    sauce_client = SauceClient(username, access_key)
    status = sys.exc_info() == (None, None, None)
    sauce_client.jobs.update_job(driver.session_id, passed=status)

# Will generate a test for each browser and os configuration
def test_generator_verify_google():
    for browser in browsers:
        yield verify_google, browser

@with_setup(None, teardown_func)
def verify_google(browser):
    global driver
    driver = launchBrowser(browser)
    driver.get("http://www.google.com")
    assert ("Google" in driver.title), "Unable to load google page"
    elem = driver.find_element_by_name("q")
    elem.send_keys("Sauce Labs")
    elem.submit()

Now run with the following command: 

nosetests --processes=8 --process-timeout=120

Note: This script references the sauceclient and selenium so make sure you have those installed.

Visit your dashboard and you should see four tests running at once! Use the nose docs to learn more about how to properly use nose for running tests in parallel.

  • No labels