是否有与 Java 的 FixedThreadPool 等效的 Python 库？

Question

Python 中是否有一个简单的 ThreadPool 库，它具有例如 pool.execute(function, args) 方法，当池已满时应该阻塞（直到其中一个线程空闲）？

我曾尝试使用多处理包中的 ThreadPool，但它的 pool.apply_async() 函数在池已满时不会阻塞。实际上，我根本不明白它的行为。

Answer 1

这个ActiveState Code Recipes page 有一个基于Python queue 的实现来进行阻塞。在 execute 的位置使用 add_task。

## {{{ http://code.activestate.com/recipes/577187/ (r9)
from Queue import Queue
from threading import Thread

class Worker(Thread):
    """Thread executing tasks from a given tasks queue"""
    def __init__(self, tasks):
        Thread.__init__(self)
        self.tasks = tasks
        self.daemon = True
        self.start()

    def run(self):
        while True:
            func, args, kargs = self.tasks.get()
            try: func(*args, **kargs)
            except Exception, e: print e
            self.tasks.task_done()

class ThreadPool:
    """Pool of threads consuming tasks from a queue"""
    def __init__(self, num_threads):
        self.tasks = Queue(num_threads)
        for _ in range(num_threads): Worker(self.tasks)

    def add_task(self, func, *args, **kargs):
        """Add a task to the queue"""
        self.tasks.put((func, args, kargs))

    def wait_completion(self):
        """Wait for completion of all the tasks in the queue"""
        self.tasks.join()

if __name__ == '__main__':
    from random import randrange
    delays = [randrange(1, 10) for i in range(100)]

    from time import sleep
    def wait_delay(d):
        print 'sleeping for (%d)sec' % d
        sleep(d)

    # 1) Init a Thread pool with the desired number of threads
    pool = ThreadPool(20)

    for i, d in enumerate(delays):
        # print the percentage of tasks placed in the queue
        print '%.2f%c' % ((float(i)/float(len(delays)))*100.0,'%')

        # 2) Add the task to the queue
        pool.add_task(wait_delay, d)

    # 3) Wait for completion
    pool.wait_completion()
## end of http://code.activestate.com/recipes/577187/ }}}

Answer 2

我修改了@ckhan 的答案并添加了回调功能。

#Custom ThreadPool implementation from http://code.activestate.com/recipes/577187/
# Usage:
#       def worker_method(data):
#           ''' do processing '''
#           return data
#           
#       def on_complete_callback(data, isfailed):
#           print ('on complete %s' % data)
#           
#       pool = ThreadPool(5)
#       for data in range(1,10)
#           pool.add_task(worker_method, on_complete_callback, data)    
#       pool.wait_completion()

from Queue import Queue
from threading import Thread
import thread

class Worker(Thread):
    """Thread executing tasks from a given tasks queue"""
    def __init__(self, tasks, thread_id):
        Thread.__init__(self)
        self.tasks = tasks
        self.daemon = True
        self.start()
        self.thread_id = thread_id

    def run(self):
        while True:
            func, callback, args, kargs = self.tasks.get()
            try:
                data = func(*args, **kargs)
                callback(data, False)
            except Exception, e:                
                callback(e, True)
            self.tasks.task_done()

class ThreadPool:
    """Pool of threads consuming tasks from a queue"""
    def __init__(self, num_threads):
        self.tasks = Queue(num_threads)
        for i in range(num_threads): Worker(self.tasks, i)

    def add_task(self, func, callback, *args, **kargs):
        """Add a task to the queue"""
        self.tasks.put((func, callback, args, kargs))

    def wait_completion(self):
        """Wait for completion of all the tasks in the queue"""
        self.tasks.join()

Answer 3

from multiprocessing.pool import ThreadPool