WTF - Web Test Framework

https://github.com/wiredrive/wtframework

Selenium does a good job at providing APIs for automating the browser. What the Web Test Framework (referred to as WTF for short) is to take that base, and build on it to provide a structured test framework for testing a Web Applications in a maintainable manner. The goal is to provide the project structures and tools to help QA/SDET professionals quickly setup and develop acceptance level web tests which are configurable, robust, and easy to use. The ultimate goal is to build a common framework on top of Selenium to provide test professionals the tools to build web tests as easily as how Rails and Django makes it easy to build web applications.

Installation

Requirements

• Python 2.7 - http://www.python.org/download/
• PyPi (pip) - http://www.pip-installer.org/en/latest/

Installation via PYPI

pip install wtframework

Setting up your project

Run the following command to initialize an empty project structure for a WTF test.

wtf_init.py YourProjectName --withexamples

Windows Note: .py files may not be executable, you may have to prefix these commands with the python command. python wtf_init.py YourProject --withexamples

This will create an the folders and packages of your project. You'll see something like:

/YourProjectName
	/assets - place non-code files used in your tests here.
	/configs - location of config files.
	/data - data files (like CSV files) goes here.
	/reference-screenshots - if enabled, reference screenshots are placed here.
	/reports - test result XML files will go here when you run tests.
	/screenshots - screenshots taken on test failures will go here.
	/tests - top level package for your test code.
		/flows - high level reuseable multipage flows.
		/models - data models go here. (like DataBase ORM code)
		/pages - Your page objects go here.
		/support - reuseable support utility functions go here.
		/testdata - custom code for working with test data.
		/tests - Your high level tests will go here.

Now the directory structure and your python path is setup to run tests in WTF framework.

Running your tests

Run your tests using nosetests.

./runtests.py [-c ConfigFile]

This will execute the unit tests in the tests/tests folder. The test results will be written to reports/, any screenshots taken during errors will be stored in the screenshots/ folder.

Configuring Eclipse/PyDev Environment

1. Download/Install Eclipse. http://www.eclipse.org/
2. Install the PyDev plugin. http://pydev.org/
3. Goto Eclipse -> Preferences (Or on windows, this Window -> Preferences )
4. Goto PyDev -> Interpretor Python then open the Libraries tab.
5. Add you python site-packages (where pip installs packages to) At this point your PyDev enviornment should be able to recognize your installed packages.
6. In Eclipse, goto "File" and create a new PyDev project.
7. Fill out the required fields and use your generated project structure as your Project folder. This should create the PyDev project files necessary to allow you to work on this project as a PyDev project.
8. Open the project settings, then under "PyDev - PYTHONPATH" settings, add your project base directory as a source folder. Then save.

At this point, you should be able to right click, select "Run As" and execute your test cases as PyUnit test case.

Installing the WTF PageObject Utility Chrome Extension

1. Download or clone the source. The extension source code is included under /browser-plugins directory.
2. Open Chrome.
3. Open the Chrome's Preferences menu, and select "Extensions"
4. Enable the "Developer mode" checkbox.
5. Click on the "Load unpack extension..." button.
6. Then select the "chrome" folder under the 'wtframework/browser-plugins/' directory.
7. At this point the plugin should be installed. You can test out the plugin by going to another webpage, clicking on the WTF extension button in your chrome toolbar, this should open the WTF PageObject Utility window.

WTF Framework Features

Configurable Tests

Being able to run tests across different environments and settings is a powerful tool. WTF has a powerful tool for working with configurations called WTF_CONFIG_READER. By default, it'll look at the default.yaml file in the /configs directory. But you may specify using other config files by setting the WTF_ENV variable. This is useful to have different config files for your different test environments. Then in your CI system, you can just specify which config file to use.

In your tests, you can pull the values you have stored in your config file using the WTF_CONFIG_READER like this:

base_url = WTF_CONFIG_READER.get_value("baseurl")
webdriver.get( base_url + "/somelocation" )

This allows you to make your test environment agnostic, runnable across multiple configurations with just a switch of an environment variable. This is good for storing environment settings and locations, account information (like DB login), connection strings, etc...

You can also specify your selenium settings to which webdriver to use. Under selenium settings, you can configure whether you want a local or remote/grid, and which browser to use.

# Settings for Selenium WebDriver used for browser testing.
selenium:
  # Set type to LOCAL for running locally, and to REMOTE, to run on a remote 
  # webdriver.  Default is LOCAL
  type: LOCAL
  #type: REMOTE

  # Set to true, to reuse the same browser.  Set to false, to use a fresh browser 
  # instance each time.  If set to a number, this  would determine whether the browser 
  # session should be discarded after a certain time peroid.  
  # Default is 'true'
  reusebrowser: true

  # Terminate Selenium after all tests have run. Disabling this can be helpful
  # during debugging.  In operation you normally want to keep this to clean up 
  # after tests.
  shutdown_hook: true

  # Take screenshot of browser on error.
  take_screenshot: true
  # Take reference screenshot upon encountering a new page.
  take_reference_screenshot: true

  # remote_url is required if type=REMOTE.  Set this to point at the Remote Webdriver 
  # connection string.
  #remote_url: http://url.to.seleniumgrid:4444/wd/hub

  # Browser can be the following options.
  # ANDROID, CHROME, FIREFOX, HTMLUNIT, HTMLUNITWITHJS, 
  # INTERNETEXPLORER, IPAD, IPHONE, OPERA, SAFARI
  browser: FIREFOX
...

Then when you want to get an instance of webdriver, use the WTF_WEBDRIVER_MANAGER to get an instance of webdriver. This allows your tests to be agnostic of which webdriver it's instantiating. You can use a different config file for running it on CI, vs. running locally on your machine

# Get an instance of webdriver.
driver = WTF_WEBDRIVER_MANAGER.new_driver()

WTFBaseTest

WTF framework adds some functionality to Python's unittest that are helpful for more end to end level tests. In order to leverage this functionality, your tests should extend the WTFBaseTest base class.

WTFBaseTest comes with a ScreenCaptureTestWatcher. You may also implement your own test watcher by extending TestWatcher class, and overriding it's methods. This is useful for creating your own base test with your own actions such as recording results to Test Case Management upon test completion. If you like to do without the added functionality of WTFBaseTest, you can use WatchedTestCase and extend it.

Data Driven Testing

WTF framework provides a easy way of doing Data-Driven-Tests using CSV files. Data files are stored in the data/ folder, and can be easily accessed using the utility class WTF_DATA_MANAGER.get_data_file("nameOfCsvFile"). You can iterate a single test over those CSV row values by using the @ddt and @csvdata decorators.

You can have a csv file with first row the column headers like, `data/animals.csv'

Animal,Type,Size
Dog,Mammal,3.0
Cat,Mammal,1.5
Lizzard,Reptile,2.0

Then in your_data_driven_test.py, you can reference these values as follows:

# Use @ddt decorator at the class level.
@ddt
class TestCsvDataDrivenTest(TestCase):

	# Then use the @csvdata decorator to flag a test method data driven.
	@csvdata("testdata.csv")
    def test_csv_datadriven(self, parameter_dic):
    	#Then in your test, you can use the parameter passed into your test
    	# as a dictionary with key corresponding to your CSV headers.
    	animal = parameter_dic['Animal']
    	type = parameter_dic['Type']
    	size = parameter_dict['Size']
    	...

PageObjects & Chrome Extension

PageObjects is a common strategy for Selenium Webdriver programmers to create self contained PageObjects to encapsulate the low level UI details from their high level tests. This allows changes in pages to be maintained in their separate page objects so tests that use the page, need not worry about the details.

WTF provides handy chrome plugin to help you create page objects. See Chrome plugin installation instructions above. The chrome plugin will help you quickly generate the boiler-plate code, some simple validate methods, and help with the tedious task of mapping the WebElements on this page.

To Create a page object, do the following:

1. Go to your target page. Then click the WTF toolbar button, then select "Scan Page"
2. A popup window will open. You'll be presented with a form that'll include fields for naming your PageObject, setting the page verification method, and a button to map your page elements.
3. Fill in an appropriate name for your page object.
4. Adjust the page verification characteristics. You may want to change it from a string compare to a regular expression and replace variable parameters with wildcards.
5. Map the elements you want to include. To map an element, first click on the button labled Map New Element. You'll be taken back to your page window, dismiss the pop-up dialog, then click on the element you wish to map. After clicking on the element, you'll notice a new entry will be created in your PageObject Utility window. Fill in the fields and adjust the object identification properties. If the identification properties in the fields do not match an element on the page, the fields will turn red. Fix these before moving on.
6. A code preview will be displayed at the bottom part of this window. Review what's there before downloading.
7. Click on the "Download" link that's right above the Preview area. This will allow you to download this file. Save this file to your YourProject/tests/pages directory.
8. Edit this file in a code editor and add whatever high level method calls you want to expose. Then you'll have a fully functioning page object. For example, if you have a login form, you might want to create a method called login(username, password), and abstract away the act of logging to the high level tests. Then leave the details of filling in and submitting the form inside the page object method body.

You can now use this page object you created like this:

from wtframework.wtf.web.pages import PageFactory
from pages.homepage import HomePage
...
homepage = PageFactory.create_page(HomePage)
homepage.login(username, password) # you will need to implement this part.

Alternatively, you can use the PageUtils to wait for the page to load. This will allow you to specify a timeout period (in seconds) to wait for this page to finish loading.

from wtframework.wtf.web.pages import PageUtils
...
slow_loading_page = PageUtils.wait_until_page_loaded(YourPageClass, timeout=60)

Note: This will use the PageObject's _validate_page() to check if the page is matching the expected page. It's good to not use a web element in addition to URL or title validation, that way the page validation does not happen until page content appears on the screen.

Once you have created a PageOjbect, you'll want to go in and edit the file and make any changes to the mappings and page verification routines. As a good practice, it's good to write methods to expose your transactional logic as a higher level method call to avoid cluttering your high level tests and test flows with low level UI logic.

See: http://engineeringquality.blogspot.com/2012/12/python-quick-and-dirty-pageobject.html

Getting Help

• WTFramework Wiki - https://github.com/wiredrive/wtframework/wiki
• Documentation https://wtframework.readthedocs.org/en/latest/
• email - David Lai david@wiredrive.com

Misc

License

This framework is free and open source. Licensed under GPLv3. See 'LICENSE.TXT' for license details.

How to Contribute

You can fork this repository. To get the unit tests not marked as skipped running, you'll need to edit or supply your own config file with values for the selenium settings.

Development on this project is currently done using NVIE branching model.
http://nvie.com/posts/a-successful-git-branching-model/

CI Scripts are running on Travis, https://travis-ci.org/wiredrive/wtframework

To submit code changes, please submit pull requests against the development branch.

See here for other ways you can help. [https://github.com/wiredrive/wtframework/wiki/How-you-can-help.]

Credits

David Lai david@wiredrive.com