无法腌制生成器对象

如何解决无法腌制生成器对象

import multiprocessing as mp
import time
import datetime

def yielder(IMapIterator):
    while True:
        try:
            yield IMapIterator.next()
        except:
            return

def download_one(x):
    time.sleep(1)
    return x + '_/'

def test_one(x):
    time.sleep(0.3)
    return x + '\\'

pool_download = mp.Pool(2)
iterator_download = pool_download.imap(download_one,['a','b','c','d'])
generator_download = yielder(iterator_download)

pool_test = mp.Pool(2)
iterator_test = pool_test.imap(test_one,generator_download)
generator_test = yielder(iterator_test)

for i in range(20):
    print(str(datetime.datetime.utcNow()),generator_test.next())
    time.sleep(0.5)

其中 generator_download 是生成器。也有以下非工作代码，标题错误，其中 generator_download 也是一个生成器：

import multiprocessing as mp

def yielder(queue):
    while True:
        element = queue.get()
        if element is None:
            break
        else:
            yield element

manager = mp.Manager()
queue_download = manager.Queue()
pool_downloader = manager.Pool(2)
list_urls = ['a','d']

def downloader(url):
    time.sleep(np.random.random())
    return a + '_'

generator_download = yielder(queue_download)
r = pool_downloader.imap(downloader,generator_download)

有人能解释一下区别吗？在全球范围内，我希望有 3 个队列和 2 个池，它们始终如一地持续下载和测试一些对象，直到最后一个队列为空。

解决方法

通过替换解决

pool_downloader = manager.Pool(2)

到

pool_downloader = mp.Pool(2)

我不知道我理解的对不对，不过好像经理的副本试图将自己转移到自己身上，导致错误的原因