Python 3 check if function is executing inside multiprocessing pool

Question

Is it possible to check in f if it's inside a Pool? Alternatively what is a good catch that only catches that specific error?

a = [[1, 2], [1, 2]]

def g(x):
    return x

def f(x):
    try:
        with Pool(2) as p:
            print(p.map(g, x))
    except Exception as e:
        print(type(e))
    return x

with Pool(2) as p:
    print(p.map(f, a))

Answer 1

Based on johndue's answer, but cleaner:

import multiprocessing
multiprocessing.current_process().daemon  # True if already a subprocess

To avoid potential code duplication, I recommend to use a ThreadPool as a drop-in replacement with the same interface:

import multiprocessing
import multiprocessing.pool

if multiprocessing.current_process().daemon:
    pool_cls = multiprocessing.pool.ThreadPool
else:
    pool_cls = multiprocessing.pool.Pool

with pool_cls() as pool:
    pool.map()

If you want to apply Thread/Process specific parameters, eg, number of workers, have a look at functools.partial() .

Answer 2

This works for me

from multiprocessing import current_process
if "daemon" in current_process()._config:
or
if "ForkPoolWorker" in str(current_process()):