转：如何更通俗地讲解Python的装饰器？

本文参考：

装饰器本质上是一个Python函数，它可以让其他函数在不需要做任何代码变动的前提下增加额外功能，装饰器的返回值也是一个函数对象。它经常用于插入日志、性能测试、事务处理、缓存、权限校验等场景。

示例1：简单装饰器

import logging
 
def use_logging(func):
  
    def wrapper():
        logging.warning('{} is running'.format(func.__name__))
        return func()   # 把foo当做参数传递进来时，执行func()就相当于执行foo()
    return wrapper
  
def foo():
    print('i am foo')
  
ff = use_logging(foo)  # 因为装饰器use_logging(foo)返回的是函数对象wrapper，这条语句相当于ff=wrapper
ff()                   # 执行ff()就相当于执行wrapper()

示例2：语法糖

def use_logging(func):
 
    def wrapper():
        logging.warn("%s is running" % func.__name__)
        return func()
    return wrapper
 
@use_logging
def foo():
    print("i am foo")
 
foo()

示例3：业务函数test带参数

def wrapper(name):
        logging.warn("%s is running" % func.__name__)
        return func(name)
    return wrapper
 
@use_logging
def foo(name):
    print("i am %s" % name)

def square_it(fn):
    def warp(*args):
        return fn(*args)**2
    return warp
 
@square_it
def test(x):
    return x + 2
 
print(test(2))

示例4：带参数的装饰器

def use_logging(level):
    def decorator(func):
        def wrapper(*args, **kwargs):
            if level == "warn":
                logging.warn("%s is running" % func.__name__)
            elif level == "info":
                logging.info("%s is running" % func.__name__)
            return func(*args)
        return wrapper
 
    return decorator
 
@use_logging(level="warn")
def foo(name='foo'):
    print("i am %s" % name)
 
foo()