How does object-oriented programming work in Python?

A class is a callable object that creates instances. __init__ initialises a new instance after creation; it is not the constructor itself, which is __new__. Attributes can live on the instance or on the class, and lookup checks the instance first, then the class and its bases.

class Account:
    kind = "checking"

    def __init__(self, owner, balance=0):
        self.owner = owner
        self.balance = balance

    @property
    def is_overdrawn(self):
        return self.balance < 0

    @classmethod
    def empty(cls, owner):
        return cls(owner=owner, balance=0)

    @staticmethod
    def valid_owner(name):
        return bool(name and name.strip())

account = Account.empty("Asha")
print(account.is_overdrawn, Account.valid_owner("Asha"))

Encapsulation in Python

• A single underscore means internal by convention, not enforced
• A double underscore triggers name mangling to avoid accidental overrides
• Properties let you add validation without breaking attribute access
• There is no private keyword; discipline and clear APIs do the work

Prefer composition over deep class hierarchies. A class should have one reason to change, and dependencies should be passed in rather than constructed internally. That makes testing straightforward because a fake can be injected.