Coordinate workers across two exclusive targets

You are given two exclusive resources (targets) labeled "A" and "B".

You need to start N = 10 parallel workers (e.g., processes or threads), each of which performs:

1. setup() (may take variable time)
2. run(target) where target must be either "A" or "B"

The functions below must be treated as black boxes and must not be modified:

def setup():
    # DO NOT TOUCH
    ...

def run(target):
    # DO NOT TOUCH
    ...

Requirements

• At any point in time, no more than one worker may be executing run("A") .
• At any point in time, no more than one worker may be executing run("B") .
• You should maximize utilization of both targets: if there are workers ready to call run(...) , and a target is free, some worker should run on that target (i.e., don’t unnecessarily serialize all work on a single target).
• Your solution should be safe even if run(target) raises or the worker exits unexpectedly (i.e., resources must not be permanently lost).

Task

Implement the worker logic and any required shared synchronization so that starting 10 parallel copies of main() will satisfy the requirements. You may assume Python and typical concurrency primitives (e.g., locks, semaphores, queues), and you may structure the driver code that spawns the 10 workers.