python-patterns

Define a class inheriting from typing.Protocol with the required method signatures, then annotate function parameters with that protocol. Any class implementing those methods satisfies the type checker without explicit inheritance, enabling flexible, loosely coupled code.

Use dataclasses for data transfer objects and value objects that mainly store fields. They auto-generate __init__, __repr__, and __eq__, support frozen=True for immutability, and field(default_factory=...) for mutable defaults like lists.

Build a list of coroutines and pass them to asyncio.gather, which executes them concurrently and returns results in order. Entry points use asyncio.run to start the event loop for the top-level async function.

Generator expressions use parentheses and evaluate lazily, producing items one at a time without storing the full sequence in memory. List comprehensions eagerly build the entire list, which is wasteful for large or pipelined data processing.

Define a base exception class for your domain, then subclass it for specific errors like UserNotFoundError or ValidationError. Store contextual attributes such as user_id or field in __init__ and pass a formatted message to super().

Python 3.10+ supports match statements that combine with union type hints like str | int | None. You can match on types using case str(), case int(), or case None to branch logic declaratively.