Robust Python

Chapter 10. User-Defined Types: Classes

Classes will be the final user-defined type that I’ll cover in this book. Many developers learn classes early, and this is both a boon and a bane. Classes are used in many frameworks and codebases, so it pays off to be fluent in class design. However, when developers learn classes too early, they miss the nuance of when and, more importantly, when not to use them.

Think back to your use of classes. Could you represent that data as a dataclass instead? What about a set of free functions? I’ve seen too many codebases that use classes everywhere when they really shouldn’t, and maintainability suffers because of it.

However, I’ve also come across codebases that swing the pendulum the other way: using no classes at all. This also affects maintainability; it is easy to break assumptions and have inconsistent data throughout. In Python, you should strive for a balance. Classes have a place in your codebase, but it is important to recognize their strengths and weaknesses. It’s time to really dig deep, cast aside your preconceptions, and learn how classes help you make more robust code.

Class Anatomy

Classes are intended to be another way of grouping related data together. They have decades of history in the object-oriented paradigm and, at first glance, don’t differ that much from what you learned about data classes. In fact, you can write a class just like you wrote a dataclass:

class Person:
    name: str = ""
    years_experience: int