PyFuncT stands for Python Functional Testing and it is a small framework that aims to help writing functional automated tests with python for testing web applications.

Principles:

• Organization: Provides a clean workflow to store element selectors, pages and configurations, helping you to keep your tests organized and reusable.

• Flexibility: PyFunct tests are fully built with python, giving you total flexibility. With test frameworks that provide natural-language programming, sometimes it can get really tricky to perform simple actions (like opening two browsers at the same time)

• Freedom: Most of the framework is customizable and optional, it means you can choose not to use some parts of it or make a few tweaks, if necessary.

PyFunct includes:

• Pages (which holds element selectors and URLs)
• Actions
• Config (global configuration easily manageable)
• Splinter driver compatibility, which includes selenium, phantomJS, zopetest and more

This should get you started with the basic functionality. There are more examples in the examples folder.

To install pyfunct, all you need to do is run: pip install pyfunct

Or

easy_install pyfunct, if you must.

Here is a code snippet with two basic tests that do the same thing: A wikipedia search. The first one uses pages and the second one uses pages and actions. Both concepts will be explained ahead.

from pyfunct import FunctTestCase

class MyTestCase(FunctTestCase):

    def test_searching_a_wiki(self):
        # This goes to the page and loads its elements.
        self.browser.open_page('wikipedia index')

        # Fill the search input with "Functional testing"
        self.browser.type('search input', 'Functional testing')

        # Submit the search by clicking the button
        self.browser.click_and_wait('search button')

        page_title = self.browser.page_title
        expected_title = 'Functional testing - Wikipedia, the free encyclopedia'
        self.assertIn(expected_title, page_title)

    def test_searching_a_wiki_using_actions(self):
        self.actions.perform_search(self.browser, 'Functional testing')

        expected_title = 'Functional testing - Wikipedia, the free encyclopedia'

        self.actions.assert_title_contains(self.browser, expected_title)

In the above code, we wrote a testcase that inherits from FunctTestCase. All PyFunct tests must inherit from it. FunctTestCase is just a testcase that inherits from unittest.TestCase (python's native unittest testcase) with a few shortcuts. It means you can use its methods (such as assertIn).

Before a test suite runs, a browser instance is created. That same browser instance will be reused by all test cases from that class, although it's the typical and default behavior, you can still disable it by setting the flag reuse_browser to False in your TestCase. The browser instance will be closed once all tests from the class finish running. You can also create as many browsers as you want by calling self.create_browser().

In Step 1 we've made references to wikipedia index, search input and search button. These are aliases that were defined in a Page class. To create it, you should do the following:

from pyfunct import Page

class IndexPage(Page):

    page_name = 'wikipedia index'

    def get_url(self):
        return '/'

    @property
    def elements_selectors(self):
        return (
            ('search input', "//input[@name='search']", 'xpath'),
            ('search button', "#searchButton", 'css')
        )

All classes that inherit from Page and provide a page_name will be accessible by the browser.

In the second test (test_searching_a_wiki_using_actions), we've used two actions: perform_search and assert_title_contains. And with that, we've made the same thing as the first test, but in a simpler and more reusable way. To write these actions and have them accessible by actions, from a FunctTestCase, you need to use the @action decorator, as follows:

from pyfunct import action

@action
def perform_search(browser, query):
    # This goes to the page and loads its elements.
    browser.open_page('wikipedia index')

    # Fill the search input with "Functional testing"
    browser.type('search input', 'Functional testing')

    # Submit the search by clicking the button
    browser.click_and_wait('search button')

@action
def assert_title_contains(browser, expected_title):
    page_title = browser.page_title
    assert expected_title in page_title, "The expected title was not found in the page title"

Until now, we did not define either the browser driver or the base url we should use. Pyfunct comes with a simple class-based configuration, which sets the global configuration attributes of your choice. Check it out:

from pyfunct import BaseConfig

class WikipediaConfig(BaseConfig):
    base_url = 'http://en.wikipedia.org'

That's it, we've just set the global config to have wikipedia as the base_url, since we are testing the wikipedia page. There's no need to change the default_driver_name, since it is using splinter by default. Unless you would like to use another browser. In that case, please take a look at the documentation.

Currently, PyFunct does not provide a test runner and you can run it as you wish. A good choice for it is nose.