是我在此螺纹Python脚本中使用全局变量的好实践

Question

我正在研究使用线程的任务，并且我使用全局变量来照顾我所遇到的问题（例如在线程之间共享变量）。我知道通常不建议使用全局变量，因此我想询问这是否适当使用全局变量。

这是代码。您会在function1（）和function2（）中找到全局变量。

from threading import Thread
from time import sleep
import random
import Queue

def getNextPrime(num):
    flag = False
    while(flag == False):
        num = num + 1
        flag = True
        for i in range(2, num):
            if num % i == 0:
                flag = False
                break
    # if we get here num should equal our next prime
    return num

def function1(self):
    i = 100
    global output_buffer
    output_buffer = Queue.Queue()
    while True:
        output_buffer.put("Thread 1: " + str(i))
        i -= 1
        sleep(1)

def function2(self):
    while True:
        global rand_num
        rand_num = random.randrange(4, 99999)
        output_buffer.put("Thread 2: " + str(rand_num))
        sleep(1)

def function3(self):
    while True:
        output_buffer.put("Thread 3: " + str(rand_num / 2.5))
        sleep(1)

def function4(self):
    prime_num = 1
    for i in range(0, 20):
        output_buffer.put("Thread 4: " + str(prime_num))
        prime_num = getNextPrime(prime_num)
        sleep(1)
# if I don't handle output like this I get weird behavior like two threads printing on the same line
def buffer_dump(self):
    while True:
        while not output_buffer.empty():
            print output_buffer.get()
        sleep(1)
if __name__ == "__main__":
    random.seed()
    thread1 = Thread(target=function1, args=(1, ))
    thread2 = Thread(target=function2, args=(1, ))
    thread3 = Thread(target=function3, args=(1, ))
    thread4 = Thread(target=function4, args=(1, ))
    output_thread = Thread(target=buffer_dump, args=(1, ))
    thread1.start()
    output_thread.start()
    sleep(2)
    thread2.start()
    sleep(2)
    thread3.start()
    sleep(2)
    thread4.start()

Answer 1

您很好地使用了全局变量。但是，如果将两个全局变量移到功能之外，则您的代码甚至可以更易读。它不会影响他们的范围。您也可以在此处阅读有关全局变量的常见问题解答，

https://docs.python.org/2/faq/programming.html#what-are-are-the-the-rules-for-local-local-and-global-variables-in-python

Answer 2

ah。我看到了问题。Python不允许在不同行上进行可变声明和iNitialisation。声明并初始化这样的同一行上的所有变量，

rand_num = random.randrange(4, 99999)

不喜欢这样，

global rand_num
    rand_num = random.randrange(4, 99999)

从所有变量中删除全局关键字。然后将所有变量放在所有功能定义之外的全局关键字