Bharat Kammakatla
Bharat Kammakatla's Blog ✍️


Bharat Kammakatla's Blog ✍️

✡️ Selenium Python Automation Framework - How to build ?

✡️ Selenium Python Automation Framework - How to build ?

Bharat Kammakatla's photo
Bharat Kammakatla
·Aug 22, 2020·

4 min read

Python is easy to learn and simpler to code when compared with other traditional languages. Nowadays, software development is making a transition towards python. Taking into consideration the holistic view of a software product, most of the companies are more inclined to maintain the same stack of technology in software development and testing life cycles. In this blog, I will put forward to you the steps and the process of how to build a Selenium Test Automation Framework in python. For a full working code, kindly refer to my Github repo.

  • Create a python project. You can choose an IDE of your own. However, I have preferred to use PyCharm as it is most industry wide used.
  • Create a new Virtual Environment.
  • Create a package named tests. Entire framework is be divided into divided into different packages. Other would be subsequently created in later steps.
  • Create a class for each testcase. Every testcase should be implemented as pytest method ex: def test_e2e(self) and every pytest method should be wrapped under a class.
  • Use fixtures for setup and teardown methods. Place the browser invocation and closure codes inside fixtures.
  • Generalise the fixture methods for all testcases by placing them in file.
  • Create a utilities package and place all the reusable classes, methods there. Doing so makes our code less reductant.
  • Create a Base class inside utilities package and call the fixture there @pytest.mark.usefixtures("setup"). Inherit Base class to all the testcases.
  • Create a package resources. Place browser drivers (ex: chromedriver.exe, geckodriver), input_data files, etc there.
  • Pass command line arguments to select browser at runtime. Means we should be able to select which browser we need to run our cases on. (ex: py.test --browser_name=chrome). Use pytest_addoption(parser) method and request.config.getoption("browser_name") function to achieve this.
  • Implement pageObject mechanism:
    • Create a package named pageObjects.
    • Create separate classes for each page (ex: HomePage, LoginPage, BookingPage, etc).
    • Define constructor which will initialize the driver which was passed form the testcases.
    • Define objects and their locators as class variables
    • Define getter methods for each class variable.
  • Create package actions. Define all the actions(like click, sendkeys, select, etc). Create and object for actions in each testcase and use actions object to call all the actions.
  • Also if there are any other reusable functions, you can place them in Base class inside utilities package.
  • Implement the logging feature. Use logging package to achieve this.
  • Implement HTML reports: Use pytest-html to achieve this. Run the cases with --html=report.html parameter which helps us to create html reports.
  • Also add screenshots for failed cases. To achieve this we should tweak few functions inside pytest-html package, you can find the tweaked code in the my framework code.
  • Parameterize data from external excel sheet using openpyxcl library.

📚 Framework code: Refer to my GitHub repository

💻 GitHub Profile:

👨‍💻 Website :

Dat's it guyz. Thank you 😊

Share this