make-it-sync

This very simple library helps provide a synchronous interface for your python async functions and class methods.

Introduction

from asyncio import sleep
from make_it_sync import make_sync

async def simple_func(a: int) -> int:
    'Simple sleeper function to test calling mechanics'
    await sleep(0.01)
    return a + 1


t_wrap = make_sync(simple_func)
print (t_wrap(4))

# Prints out 5

It works by running an async event loop and executing the function. It will use the current thread if no loop is running, otherwise it will create a new thread and run the function there.

Features:

1. Will wrap a stand-alone function
2. All arguments (keyword and positional) are passed
3. Instance methods on classes may be wrapped
4. Abstract methods are correctly handled.

Usage

To install pip install make-it-async.

The make_sync function creates a new function that will call the function you pass to it:

async def simple_func(a: int) -> int:
    'Simple sleeper function to test calling mechanics'
    await sleep(0.01)
    return a + 1


t_wrap = make_sync(simple_func)

You may treat t_wrap as a python function. You could use make_sync as a function decorator, though that isn't the normal usage as that would hid the async version of the function. Normally, make_sync is used to provide a non-async, alternate, version of the function.

It is also possible to use make_sync with abstract functions:

class abc_base(ABC):
    @abstractmethod
    async def doit_async(self):
        raise NotImplementedError()

    doit = make_sync(doit_async)

class abc_derived(abc_base):
    async def doit_async(self):
        return 42

a = abs_derived()
print(a.doit())
# Will print out 42

The abstract dispatch will occur at runtime and the call to doit_async will be pulled from the the sub-class. This allows you to define the asynchronous API in an ABC, and build a common set of synchronous methods.