awaitable-tools 0.0.1

Tools to bridge coroutines and non-coroutine awaitables in Python

Awaitable Tools

This package provides utilities for working with awaitable objects in Python.

Coroutines vs Awaitables

Coroutines

Objects returned from the function declared with the async def syntax. It is the standard way to create asynchronous programs.

async def coroutine():
    pass

Awaitable

Any object that implements __await__() for use in an await expression.

class Awaitable:
    def __await__(self):
        yield

Installation

pip install awaitable-tools

While most users work with coroutines, some advanced scenarios (like protocol implementations or framework development) require working directly with awaitables. This package fills that gap.

Technical Background

The Awaitable Protocol

An object is awaitable if its __await__() method returns an generator that:

1. Only yields None (any other value raises RuntimeError)
2. Final value uses return (not yield)
3. May use yield from with other compliant generators

class ValidAwaitable:
    def __await__(self):
        yield None  # Valid suspension
        # coroutine also implements __await__()
        yield from asyncio.sleep(1).__await__()
        return True

class InvalidAwaitable:
    def __await__(self):
        # Only valid if marked as @types.coroutine
        yield from asyncio.sleep(1)  # TypeError!
        yield "foo"  # RuntimeError!

About async/await

April 18, 2020, From yield to async/await