Python使用Condition对象实现多线程同步

使用Condition对象可以在某些事件触发后才处理数据或执行特定的功能代码，可以用于不同线程之间的通信或通知，以实现更高级别的同步。在内部实现上，Condition对象总是与某种锁对象相关联。

Condition对象除了具有acquire()和release()方法之外，还有wait()、wait_for()、notify()、notify_all()等方法：

• wait(timeout=None)方法会释放锁，并阻塞当前线程直到超时或其他线程针对同一个Condition对象调用了notify()/notify_all()方法，被唤醒之后当前线程会重新尝试获取锁并在成功获取锁之后结束wait()方法，然后继续执行；
• wait_for(predicate, timeout=None)方法阻塞当前线程直到超时或者指定条件得到满足；
• notify(n=1)唤醒等待该Condition对象的一个或多个线程，该方法并不负责释放锁；
• notify_all()方法会唤醒等待该Condition对象的所有线程。

本文代码模拟了经典的生产者-消费者问题，使用列表模拟物品池，生产者往里放物品，消费者从中获取物品，物品池满时生产者等待，空时消费者等待。

import threading
from random import randint
from time import sleep
#自定义生产者线程类
class Producer(threading.Thread):
    def __init__(self, threadname):
        threading.Thread.__init__(self,name=threadname)
        
    def run(self):
        global x
        
        while True:
            sleep(1)
            #获取锁
            con.acquire()
            #假设共享列表中最多能容纳5个元素
            if len(x) == 5:
                #如果共享列表已满，生产者等待
                print('Producer is waiting.....')
                con.wait()
            else:
                r = randint(1, 1000)
                print('Produced:', r)
                #产生新元素，添加至共享列表
                x.append(r)
                #唤醒等待条件的线程
                con.notify()
                
            #释放锁
            con.release()
#自定义消费者线程类
class Consumer(threading.Thread):
    def __init__(self, threadname):
        threading.Thread.__init__(self, name =threadname)
        
    def run(self):
        global x
        
        while True:
            sleep(2)
            #获取锁
            con.acquire()
            if not x:
                #等待
                print('Consumer is waiting.....')
                con.wait()
            else:
                print('Consumed:', x.pop(0))
                con.notify()
                
            con.release()
        
#创建Condition对象以及生产者线程和消费者线程
con = threading.Condition()
x = []
Producer('Producer').start()
Consumer('Consumer').start()

某次运行部分结果如下：

Produced: 696

Produced: 970

Consumed: 696

Produced: 546

Produced: 30

Consumed: 970

Produced: 824

Produced: 68

Consumed: 546

Produced: 409

Produced: 172

Consumed: 30

Produced: 820

Producer is waiting.....

Consumed: 824

Produced: 2

Producer is waiting.....

Consumed: 68

Produced: 473

Producer is waiting.....

Consumed: 409

Produced: 167

Consumed: 172

Produced: 192

Producer is waiting.....

Consumed: 820

Produced: 789

Consumed: 2

Produced: 855

Producer is waiting.....

Consumed: 473

Produced: 754

Producer is waiting.....

Consumed: 167

Produced: 549

Consumed: 192

Produced: 100

Producer is waiting.....

Consumed: 789

Produced: 862

Consumed: 855

Produced: 537

Producer is waiting.....

Consumed: 754

Produced: 715

Consumed: 549

Produced: 79

Producer is waiting.....

Consumed: 100

Produced: 445

Consumed: 862

Produced: 64

Producer is waiting.....

Consumed: 537

Produced: 755

Consumed: 715

Produced: 8

^C（按Crtl+Break退出）