uparallel

μparallel - oversimplified helper for easy parallel functions execution

Yes, it's oversimplified:

from uparallel import uparallel
import requests

@uparallel(3)
def wikipedia(word):
    return requests.get(f"https://www.wikiwand.com/en/{word.capitalize()}")

results = [wikipedia(word) for word in 'fork spoon spork knife cup home'.split()]
for line in results:
    print(line.url, line.status_code)

So, unparallel(num) is a wrapper for any function to execute num of them in parallel. There's three limitations:

1. It will wait until last of calls finish their execution
2. It will break on any execption, so catch it inside the function
3. Returned result is a thin wrapper, not result of function itself.

Last one need more explanation. Instead of result of function you will got a special proxy object. It will behave more or less like your normal object (like in this example with requests). If it's not, call result.wait() or even just result() instead of just result.

@uparallel(2)
def add2(x):
    return x+2
	
results = [add2(x) for x in range(10)]
for y in results:
	print(y()-2, y)

And don't afraid to read the code, it's less than 100 lines long.