For this project, I'm using poetry. Poetry is a dependency management tool and I use it to manage all the dependencies that I'll be using for this project.
To install poetry navigate to the following link https://python-poetry.org/ Click on documentation Under installation, click "With the official install"
Alternatively, you can do
pip install poetryHowever, this way of installing it is not recommended as you'd have to install it separately for every version of Python you have installed on your system.
To start the project first type poetry init to create a
pyproject.toml file that all dependencies meta data about
the project will be stored. Then you should see a pyproject.toml
file created
Type poetry shell to activate the new virtual environment
Note: if it says that poetry is already activated some- where else, go to folder folder it mentions and type exit. Exit will deactivate your current virutal environment. For example if you get: Virtual environment already activated: /home//.cache/ pypoetry/virtualenvs/my-project-0wt3KWFj-py3.7
Do the following
cd /home/<user>/.cache/pypoetry/virtualenvs/
exit
To add a dependency, use poetry add --dev <package_name>@latest
for example:
poetry add --dev selenium@latest
poetry add --dev pytest@latest
Alternatively, you can install multiple dependencies in one line by writing them one after another like so:
poetry add --dev black flake8 pre-commit tox python-dotenv
Selenium is used for UI automation, pytest and tox for testing our code for different versions of Python, black and flake8 for PEP-8 formatting, and pre-commit for auto-linting during each commit. Finially, I'm using python-dotenv to store environment variables and load them securely.
Note: some packages may require a higher version of Python to work. So you can jut go in the pyproject.toml and update the version (e.g. python_version >= 3.9.)
After all dependencies have been install, a poetry.lock
file should have been generatered.
You can use the command poetry show --tree to see dep-
endencies in tree hierarchy format. This will give
you a better sense of what packages belong to what group
of dependencies.
The first thing I did when creating this project was to create my fixtures. Fixtures give your test contextual information that they can use to set up your project.
In conftest.py I have 2 main fixtures. The first is
driver, which is the Chrome driver that is used
to spawn a new test window and gives other test
context as to what it should do.
The second fixture is called envars which is used for tests that need to access sensitive information such as username and password.
test_suite.py contains all the tests that
pytest will run.
The first, test will test that login is successful. The second tests that we are able to add new customers with no problems. The new customer's credentials are saved in a file and accessed in the last test. This test tries to log in as the new customer and verifies that the login was successful.
I have a folder called pom which stands for page object model. First I defined a base page that describes the behavior of a web element we want to use such as find, wait for, and find all. It also takes a driver object upon initialization. wait_for will wait for an element to be loaded in the page before it attempts to access it. This helps to account for race conditions. I use find and wait interchangeably. However, I would rather use wait_for if I'm changing pages and I want to make sure all the elements have been loaded. For the most part, I don't use find_all unless the only selector I can find targets a group of elements. In that case I would just refer to the element that I need by index.
All subsequent classes inherit from this base class. I tried my best to implement a separation of concerns in how I defined my classes. I have all my selectors defined in these classes. I try to make each method pertain to the class it belongs in. If I feel like there's a piece of functionality that doesn't belong to that class, I'll define it in a helper_functions folder and import it as needed.
To get my selectors, I would inspect each element, copy the selector, and test that the selector works by typing the following into the console:
document.querySelector(<CSS_SELECTOR>)
To run the tests, simply type:
pytest test_file.py
Make sure you are in the folder pom before you do so.
The .pre-commit-config.yaml file defines how
I want my commit hook to work
tox.ini contains which PEP-8 rules I want to ignore
when linting. It also defines tests that I want to
run with different version of Python. This will be
triggered when you commit a change to git.