Functions in Python

Functions are first‑class objects defined with def; they can be passed, returned, and stored in data structures.

Defining functions

Parameters are positional by default; use annotations for documentation and tooling.

def add(a: int, b: int) -> int:
    return a + b

Docstrings document behavior.

def greet(name: str) -> str:
    """Return a friendly greeting for the given name."""
    return f"Hello, {name}!"

Default and keyword arguments

Defaults are evaluated at definition time; use None sentinels for mutable defaults.

def power(x, exp=2):
    return x ** exp

from typing import Optional, List

def append_safe(item: int, bucket: Optional[List[int]] = None) -> List[int]:
    bucket = [] if bucket is None else bucket
    bucket.append(item)
    return bucket

*args and **kwargs

Collect extra positional and keyword arguments.

def log(event, *args, **kwargs):
    print(event, args, kwargs)

Keyword‑only and positional‑only

Force keywords or positional with * and /.

def config(host, /, *, port=80):
    return host, port

Returning multiple values

Python returns tuples; unpack on the caller side.

def divmod_(a, b):
    return a // b, a % b
q, r = divmod_(7, 3)

Lambdas and function objects

Lambdas create small anonymous functions; functions are objects with attributes like __name__.

square = lambda n: n*n

Prefer def for readability, especially with complex logic.

Closures

Inner functions capture names from enclosing scopes.

def make_counter():
    n = 0
    def inc():
        nonlocal n
        n += 1
        return n
    return inc

Decorators (preview)

Wrap functions to add behavior (logging, caching). See the Decorators section for details.

Summary

• Use annotations and docstrings to clarify intent
• Avoid mutable default arguments; prefer None sentinels
• Master *args/**kwargs, keyword‑only, and positional‑only for ergonomic APIs