README.md

Metaclasses

Metaclasses are classes that create classes — they control what happens when Python sees class Foo:.

Files

File	Description
example.py	Singleton metaclass, validation, type() demo

---

Descriptive Example

Scenario

Ensure a Database class has only one instance across the entire application.

class SingletonMeta(type):
    _instances = {}

    def __call__(cls, *args, **kwargs):
        if cls not in cls._instances:
            cls._instances[cls] = super().__call__(*args, **kwargs)
        return cls._instances[cls]

class Database(metaclass=SingletonMeta):
    def __init__(self, url):
        self.url = url

db1 = Database("postgres://localhost")
db2 = Database("mysql://other")
print(db1 is db2)   # True — same instance
print(db1.url)      # postgres://localhost (first call wins)

Creating a class at runtime

Dynamic = type("Dynamic", (), {"x": 42, "greet": lambda self: "hi"})
print(Dynamic().x)   # 42

---

Interview Q&A

Q1: What is a metaclass?
A: The class of a class. MyClass.__class__ is usually type. Metaclasses customize class creation via __new__ and __init__.

Q2: What happens when Python executes class Foo:?
A: Body collected into namespace → metaclass __new__ creates class object → metaclass __init__ initializes it → name Foo bound to class.

Q3: When use metaclass vs class decorator?
A: Metaclass: control creation of entire inheritance hierarchies. Class decorator: transform a single class. Decorator is simpler for most cases.

Q4: What is the metaclass of type?
A: type itself — it's its own metaclass.

Q5: Implement Singleton without metaclass?
A: Override __new__ on the class itself, or use a module-level instance, or @functools.lru_cache on factory function.

Q6: Real-world metaclass usage?
A: Django ORM models, dataclass generation, API registration, enforcing interfaces on subclasses. Prefer __init_subclass__ or decorators when possible.

---

Run

python3 example.py