From Linkedin Python post "Python parallelism in one line - A Better Model for Day to Day Threading Tasks" Does anyone know of similar "compact" examples of handling parallelism for the case where the workers may produce tasks that are put into the FIFO queue (maybe with `while` instead of map) ?

concurrent futures with FIFO

# requires python 3.2.3 or higher
import concurrent.futures

def f(x):
    print ("working on task: %s" %{x})
    if x-1 > 0 :
        return x-1 # I want this task back in the queue.

def main():
    Queue = [x for x in range(4)]
    with concurrent.futures.ThreadPoolExecutor(max_workers=3) as executor:
        while len(Queue)>0:
            task = Queue.pop(0)
            future = executor.submit(f, task)
            result = future.result()
            if result is not None:
                Queue.append(result) # here the f may return a task is to be added to the queue.
        print ("processing completed.")

if __name__ == "__main__":
    main()

    # expected output:
    # "working on task: 1"
    # "working on task: 2" (returns 1 to the Queue)
    # "working on task: 3" (returns 2 to the Queue)
    # "working on task: 4" (returns 3 to the Queue)
    # "working on task: 1" 
    # "working on task: 2" (returns 1 to the Queue)
    # "working on task: 3" (returns 2 to the Queue)
    # "working on task: 1"
    # "working on task: 2" (returns 1 to the Queue)
    # "working on task: 1"