pytest_mozwebqa

pytest_mozwebqa is a plugin for py.test that provides additional features needed for Mozilla's WebQA projects.

Requires:

• py.test
• selenium
• requests

Continuous Integration

Installation

$ python setup.py install

Running tests with pytest_mozwebqa

Usage: py.test [options] [file_or_dir] [file_or_dir] [...]

For full usage details run py.test --help.

Configuration

You can also create a mozwebqa.cfg file that will be used to set defaults. This is so that projects can keep this alongside the tests to simplify running them via the command line. The options are currently limited to those that could be project specific.

[DEFAULT]
baseurl: 'http://www.example.com'

Examples

Run tests against a local webdriver using Firefox:

$ py.test --baseurl=http://example.com --driver=firefox --firefoxpath=/Applications/Firefox.app/Contents/MacOS/firefox-bin

Run tests against a local webdriver using Google Chrome:

$ py.test --baseurl=http://example.com --driver=chrome --chromepath=/Applications/chromedriver

Run tests against a remote webdriver server either directly or via grid:

$ py.test --baseurl=http://example.com --browsername=firefox --browserver=5 --platform=mac

Writing tests for pytest_mozwebqa

You will need to include the mozwebqa in the method signature for your tests, and pass it when constructing page objects.

Example

def test_new_user_can_register(self, mozwebqa):
    home_pg = home_page.HomePage(mozwebqa)
    home_pg.go_to_home_page()
    home_pg.login_region.click_sign_up()

    registration_pg = registration_page.RegistrationPage(mozwebqa)
    registration_pg.register_new_user()
    Assert.equal(registration_pg.page_title, "Sign Up Complete!")

Destructive tests

In order to prevent accidentally running destructive tests, only tests marked as nondestructive will run by default. If you want to mark a test as nondestructive then add the appropriate marker as shown below:

Example (mark test as nondestructive)

import pytest
@pytest.mark.nondestructive
def test_safely(self, mozwebqa):
    ...

If you want to run destructive tests then you can specify the --destructive command line option.

Sensitive environments

If running against a sensitive (production) environment any destructive tests will be skipped with an appropriate error message. You can specify a regular expression that matches your sensitive environments using the --sensitiveurl command line option.

Setting capabilities

It's possible to specify additional capabilities on the command line:

Example (accept SSL certificates)

--capability=acceptSslCerts:true

Setting Firefox preferences

If you're using Firefox it's possible to set custom preferences:

Example (disable addon compatibility checking)

--firefoxpref=extensions.checkCompatibility.nightly:false

Specifying a Firefox profile

If you're using Firefox it's possible to specify an existing Firefox profile to use when starting Firefox.

Example (use the profile located at /path/to/profile_directory)

--profilepath='/path/to/profile_directory'

Installing Firefox extensions

If you're using Firefox it's possible to install extensions when starting the browser.

Example (install the extensions located at /path/to/ext1/ext1.xpi and /path/to/ext2/ext2.xpi)

--extension='/path/to/ext1/ext1.xpi' --extension='/path/to/ext2/ext2.xpi'

Setting Google Chrome options

If you're using Google Chrome then you can set various options on the command line using a JSON string.

Valid keys are:

• arguments: a list of command-line arguments to use when starting Google Chrome.
• binary_location: path to the Google Chrome executable to use.

For more details on Google Chrome options see: http://code.google.com/p/chromedriver/wiki/CapabilitiesAndSwitches

Example (set initial homepage)

--chromeopts='{"arguments":["homepage=http://www.example.com"]}'

Installing Google Chrome extensions

If you're using Google Chrome it's possible to install extensions when starting the browser.

Example (install the extensions located at /path/to/ext1/ext1.crx and /path/to/ext2/ext2.crx)

--extension='/path/to/ext1/ext1.crx' --extension='/path/to/ext2/ext2.crx'

HTML report

If the pytest-html plugin is installed then the HTML reports will include additional information such as the failing URL, screenshot, and page source. For Sauce Labs, a link to the job and inline video will also be included. Check the pytest-html documentation for how to install the plugin and generate HTML reports.

Appium

To use Appium instead of Selenium, simply specify the appropriate capabilities. For example, to run web application tests against Safari on an iPhone simulator:

py.test --browsername=Safari \
--capability=platformName:iOS \
--capability=platformVersion:8.2 \
--capability="deviceName:iPhone Simulator" \
--capability=appiumVersion:1.3.7

Sauce Labs integration

To run your automated tests using Sauce Labs, you must provide a valid username and API key. This can be done either by creating a setup.cfg file with a [saucelabs] section or by setting the SAUCELABS_USERNAME and SAUCELABS_API_KEY environment variables.

Configuration

Below is an example setup.cfg showing the configuration options:

[saucelabs]
username = username
api-key = secret
tags = tag1, tag2
privacy = public

The tags entry is an optional comma separated list of tags that can be used to filter the jobs in the Sauce Labs dashboard.

The privacy entry is used to determine who you share your Sauce Labs jobs with. Check the documentation for the accepted values. If not set, this defaults to public restricted.

Running tests

To run your automated tests, simply specify SauceLabs as your driver:

py.test --driver=SauceLabs --browsername=Firefox --platform="Windows 8"

See the supported platforms to help you with your configuration. Additional capabilities can be set using the --capability comand line arguments. See the test configuration documentation for full details of what can be configured.

Privacy

You can specify the job sharing level for individual tests by setting a mark on the test method. This takes priority over the privacy entry in the configuration file:

import pytest
@pytest.mark.privacy('public')
def test_public(self, mozwebqa):
    home_pg = home_page.HomePage(mozwebqa)

You can also explicitly mark the test as private:

import pytest
@pytest.mark.privacy('private')
def test_private(self, mozwebqa):
    home_pg = home_page.HomePage(mozwebqa)

For the full list of accepted values, check the Sauce Labs documentation.

BrowserStack integration

To run your automated tests using BrowserStack, you must provide a valid username and access key. This can be done either by creating a setup.cfg file with a [browserstack] section or by setting the BROWSERSTACK_USERNAME and BROWSERSTACK_ACCESS_KEY environment variables.

Configuration

Below is an example setup.cfg showing the configuration options:

[browserstack]
username = username
access-key = secret

Running tests

To run your automated tests, simply specify BrowserStack as your driver:

py.test --driver=BrowserStack --browsername=firefox --platform=WIN8

See the capabilities documentation for additional configuration that can be set using --capability command line arguments.

Using a proxy server

If you want the browser launched to use a proxy (currently only supported by Firefox and Google Chrome) you must specify the --proxyhost and --proxyport command line arguments.

Example (proxy is running on localhost port 8080)

--proxyhost=localhost --proxyport=8080