后浪小萌新Python --- Lock

一、数据冲突

一般发生在多个线程对同一个数据进行操作的时候

注意: Queue对象本身就是线程安全的，不会出现数据冲突

举例：

from threading import Thread
data = 0


def up_data():
    global data
    for _ in range(100008):
        data += 1


def down_data():
    global data
    for _ in range(100000):
        data -= 1


if __name__ == '__main__':
    # 情况1: 在一个线程中对同一个数据进行加和减的操作
    # up_data()
    # down_data()
    # print(data)
    # 情况2：在两个线程中进行加和减操作
    t1 = Thread(target=up_data)
    t2 = Thread(target=down_data)
    t1.start()
    t2.start()
    t1.join()
    t2.join()
    print(data)

执行结果:

8

二、手动Lock

Lock - 锁类， 传统锁，重量级锁（加锁和解锁需要手动进行）

举例：

from threading import Thread, Lock

data = 0
# 1.创建锁对象(有多少个数据就创建多个锁)
data_lock = Lock()


def up_data():
    global data
    for _ in range(100008):
        # 2.加锁
        data_lock.acquire()
        data += 1
        # 3.解锁
        data_lock.release()


def down_data():
    global data
    for _ in range(100000):
        data_lock.acquire()
        data -= 1
        data_lock.release()


if __name__ == '__main__':
    t1 = Thread(target=up_data)
    t2 = Thread(target=down_data)
    t1.start()
    t2.start()
    t1.join()
    t2.join()
    print(data)

执行结果:

8

三、自动锁RLock

举例：

from threading import Thread, RLock

data = 0
data_lock = RLock()


def up_data():
    global data
    for _ in range(100008):
        # 使用锁（会自动加锁和解锁）
        with data_lock:
            data += 1


def down_data():
    global data
    for _ in range(100000):
        with data_lock:
            data -= 1


if __name__ == '__main__':
    t1 = Thread(target=up_data)
    t2 = Thread(target=down_data)
    t1.start()
    t2.start()
    t1.join()
    t2.join()
    print(data)

执行结果:

8