代码之家 › 专栏 › 技术社区 › abhishek_rathaur

蟒蛇装饰器

python-decorators python

-1

abhishek_rathaur · 技术社区 · 6 年前

from random import randint
import  time
state = 0 #close

open_time = 0
failure_count = 0
count = 0
status = {0 : "closed" , 2 : " open" , 1 : "half closed"}
def circuitbreaker(func):
    global count
    global open_time , state
    print("circuit status "+ status.get(state))
    if state ==0: #close state
        try:
            func()
        except Exception as ex:
            print(ex)
            count+=1
            if count>2:
                print("opening circuit breaker")
                state = 2
                open_time = int(time.time())
    elif (state == 2): #open state
        if( time.time() - open_time > 5) :
            state = 1
        else:
            print("circuit opened")
    else:
        try:
            func()
            count = 0
            open_time = 0
            print("closing circuit breaker")
            state = 0
        except Exception as ex:
            state = 2
            open_time = int(time.time())
            print("opening circuit breaker")

@circuitbreaker
def generic_func():
    a = randint(0,9)
    print("hello")
    print("random number = "+str(a))
    if a>4:
        return a
    else:
        raise Exception('Yalla!')

if __name__=="__main__":

    # while(True):
    #     generic_func()
        time.sleep(1)

我有这个密码。我有几个问题: 1)为什么调用泛型函数,即使我在main中对其进行了注释。

2)在主函数中取消注释注释部分时。我得到以下错误。如何正确调用此泛型函数。我的动机是实现一个断路器,当调用函数中出现某种错误或异常时,断路器将闭合。我可以直接使用: 断路器(调用函数),但我想使用decorators

Traceback (most recent call last):
circuit status closed
hello
  File "/Users/abhishekkumar/PycharmProjects/RateLimiter/circuitbreaker.py", line 53, in <module>
random number = 1
Yalla!
    generic_func()
TypeError: 'NoneType' object is not callable

Process finished with exit code 1

问题是decorator应该返回一个函数对象,并且应该在函数中包含相关逻辑,然后返回该函数,否则它将不返回任何对象。

3 回复 | 直到 6 年前

Dan D. 6 年前

问题1的答案是:这是因为断路器decorator,因为它的逻辑是在模块导入期间执行的,并且它调用了decorated函数。检查下面几行

...
try:
    func() # <-- here
except Exception as ex:
    print(ex)
    ...

实现decorators的方法是返回一个包装函数,它包含业务逻辑:

from functools import wraps

def decorator(f):
    @wraps(f)
    def wrapper(*args, **kwargs):
        ... your logic here ...
    return wrapper

问题2的答案来自上一个问题。

Patrick Artner 6 年前

Answers to almost all questions you never asked about decorators

装饰某物的函数应该返回它自己所做的功能-而不是做所有的东西-你返回的不是一个“函数指针”,而是一个隐式的 None 从你的除臭剂(含蓄地说,你什么也不回)。这个 没有 然后叫…

如何修复:

def circuitbreaker(func):
    def the_works():
        global count
        global open_time , state
        print("circuit status "+ status.get(state))
        # .. all your other code ...
    return the_works

for _ in range(5): 
    generic_func()

固定输出 :

circuit status closed
hello
random number = 3
Yalla!
circuit status closed
hello
random number = 3
Yalla!
circuit status closed
hello
random number = 0
Yalla!
opening circuit breaker
circuit status  open
circuit opened
circuit status  open
circuit opened

hellokangning 6 年前

decorators在decorated函数定义之后立即运行,这通常是在导入时。在装潢师 circuitbreaker 你打电话 generic_func 已经。

Here 是fluent python的一个例子:

registry = []

def register(func):
    print('running register(%s)' % func)
    registry.append(func)
    return func

@register
def f1():
    print('running f1()')

@register
def f2():
    print('running f2()')

def f3():
    print('running f3()')

def main():
    print('registry ->', registry)
    f1()
    f2()
    f3()

if __name__ == '__main__':
    main()

输出是

running register(<function f1 at 0x1055ae378>)
running register(<function f2 at 0x1055ae400>)
registry -> [<function f1 at 0x1055ae378>, <function f2 at 0x1055ae400>]
running f1()
running f2()

这是你想要的吗: How do I catch an exception in a decorator but allow the caller to catch it as well?