naistangz / UnitTestingPython

Unit Testing in Python using Pytest

Using pytest

Why?

• To run standalone test functions
• Ability to use standard assert methods
• Unlike unittest, where you place tests into classes and handle class inheritance.
• Easier to define tests, unittest has over 10 different assertion methods
• Pytest uses fixtures to automate test setup, teardown and common test scenarios for use in varying scopes
• Pytest is an open source project

Types of test

• Unit
• Integration
• Linting
• Smoke
• Functional Tests

Objectives

• Pytest framework's core features: assertions, exceptions, fixtures, parametrisation
• Modularising common test utility functions with conftest.py
• Using TDD to add new functionality

Project

• Adding destinations onto a map

Testing methods

• Happy Path Testing

  • Organising positive cases into their own functions

• Sad Path Testing

  • Negative test cases to validate against
  • Halting the flow of the code execution or throw an exception
  • Log that an action could not be taken and continue code execution.

Fixtures

• Automate test setup and teardown
• Comes in handy when tests use operations, e.g database or API contention
• Can be scoped to execute at different times in our set suite

Scoping Fixtures

@pytest.fixture(scope = "session")
    - Once per session 

@pytest.fixture(scope = "module")
    - Once per module 

@pytest.fixture(scope = "class")
    - Once per class of tests

@pytest.fixture(scope = "function")
    - Once per test

• When using fixtures we can scope them according to the following four scopes from highest (test session) to lowest scope (function level).
• Using scoping minimises the number of active fixtures during test runs.
• Scoping may help you avoid adding operations that slow down entire test suite.
• This happens because they're involved for the entire session rather than for each function.
• E.g. happy path test, transforms data to include population counts for each row of the country data, it appends a new column indicating if the row was updated or not.
  • In clean_map.csv file, Andorra in line 2 is indicated in the population dictionary, it would gain an additional population field with the value from the dictionary
  • After it has been transformed it will also receive a field updated, which would evaluate to true since it has been updated.
  • The country Angola in line 9 is not indicated in our population dictionary.
  • Population field would be a none value.
  • In sad_path testing, we make sure that we cannot attempt to transform the data twice on lines 58 through 60 - we define the inability to transform data twice in the following script in our source code.

Using conftest.py for common functions

• Stores common utility test fixtures and extension code often referred to as 'hooks'
• Pytest collects the fixtures in the file so they are globally accessible within the testing directory
• It must be placed under the /tests folder

---

Understanding TDD

• Test-driven development is a software development process to illustrate what features the system will carry out.
• The developer writes a test that will surely fail without code to implement its behaviour (aka red failure state)
• Developer must make sure it eventually passes by writing the minimum implementation code needed to get this test passing.
• TDD allows developers to focus on functionality in small pieces on the forefront.
• Facilitates code reuse

---

Managing data in Docker with volumes

• Volumes is a way to persist data generated by and used by Docker containers
• Decouples container from storage meaning storage is separated from the container and even if you delete or remove the container, the storage is still available
• Volumes used to share volume or data among different containers