Python并发ThreadPoolExecutor - 如果满足条件则停止执行

Question

我一直在尝试使用 ThreadPoolExecutor 进行多线程，如果其中一个进程满足某个条件，则可以安全地停止执行，这样剩余的线程就不必最终确定。

我有以下概念，但它不起作用，它一直在运行：

import time, sys
from concurrent.futures import ThreadPoolExecutor

data = {"future1": 'blank', "future2": 'blank', "future3": 'blank'}

def function1(n):
   global data #not sure if this is necessary, as it seems to be able to access this anyway?
   print(n)
   time.sleep(1)
   data['future1'] = n

def function2(n):
   global data
   print(n)
   time.sleep(2)
   data['future2'] = n

def function3(n):
   global data
   print(n)
   time.sleep(3)
   data['future3'] = n

with ThreadPoolExecutor(max_workers=4) as executor:
  while True:
    future1=executor.submit(function1, 'test1')
    future2=executor.submit(function2, 'test2')
    future3=executor.submit(function3, 'test3')

    if data['future2']!='blank':
      executor.shutdown(wait=False)
      sys.exit()

不知道我在这里做错了什么，任何帮助将不胜感激。

Answer 1

这是完整的答案，事实证明 executor.shutdown(wait=False) 不是 Sachin 提到的要走的路。完全归功于https://gist.github.com/clchiou/f2608cbe54403edb0b13

import time, sys
from concurrent.futures import ThreadPoolExecutor
import concurrent.futures.thread

data = {"future1": None, "future2": None, "future3": None}

def function1(n):
   time.sleep(1)
   data['future1'] = n
   print(n)

def function2(n):
   time.sleep(2)
   data['future2'] = n
   print(n)

def function3(n):
   time.sleep(3)
   data['future3'] = n
   print(n)

with ThreadPoolExecutor(max_workers=4) as executor:
  executor.submit(function1, 'test1')
  executor.submit(function2, 'test2')
  executor.submit(function3, 'test3')

  while True:
    if any(v is not None for v in data.values()):
      executor._threads.clear()
      concurrent.futures.thread._threads_queues.clear()
      break

print(data)

Answer 2

您正在while True: 循环中运行线程，该循环一次又一次地启动调用。

with ThreadPoolExecutor(max_workers=4) as executor:

    future1=executor.submit(function1, 'test1')
    future2=executor.submit(function2, 'test2')
    future3=executor.submit(function3, 'test3')

    while True:
        if data['future2']!='blank':
            executor.shutdown(wait=False)
            sys.exit()

executor.shutdown(wait=False) 仍然会等待其子进程完成