Python学习笔记：装饰器

Python 装饰器的基本概念和应用

• 代码编写要遵循开放封闭原则，虽然在这个原则是用的面向对象开发，但是也适用于函数式编程，简单来说，它规定已经实现的功能代码不允许被修改，但可以被扩展，即：

1. 封闭：已实现的功能代码块
2. 开放：对扩展开发

装饰器是 Python 高阶函数的语法糖，可以为已经存在的对象添加额外的功能，比如：

1. 引入日志
2. 函数执行时间统计
3. 执行函数前预备处理
4. 执行函数后清理功能
5. 权限校验等场景
6. 缓存

Python 装饰器的基本实现

装饰器的例程：

 #!/usr/bin/env python3
 #-*- coding:utf-8 -*-
 #File Name:04_decorator_simple.py
 #Created Time:2019-01-09 15:24:38

 import time

 def timefun(func):
     # wrapped_fun 即为闭包, func 为要装饰的函数
     def wrapped_fun():
         start_time = time.time()
         func()
         end_time = time.time()
         print("%s\t运行用时 %f s" % (func.__name__, end_time - start_time))
     return wrapped_fun # 返回内部函数的引用

 def foo1():
     for i in range(10000):
         pass

 @timefun # 相当于 foo = timefun(foo)
 def foo():
     for i in range(100000):
         pass

 if __name__ == "__main__":
     foo1 = timefun(foo1)
     foo1()
     foo()

装饰器例程

foo1    运行用时 0.000491 s
foo     运行用时 0.002976 s

装饰器例程的运行结果

运行过程分析见下图：

Python 装饰器实现的基础为闭包，其会在闭包中调用目标函数，关于闭馆的详细内容，请参考 Python 学习笔记：闭包

Python 装饰器的使用

多个装饰器

 #!/usr/bin/env python3
 #-*- coding:utf-8 -*-
 #File Name:05_decorator_multi.py
 #Created Time:2019-01-09 15:55:31

 def make_body(func):
     """添加 body"""
     def wrapped():
         return "<b>" + func() + "</b>"
     return wrapped

 def make_italic(func):
     """设置为斜体"""
     def wrapped():
         return "<i>" + func() + "</i>"
     return wrapped

 # 相当于 make_body(make_italic(foo))
 @make_body
 @make_italic
 def foo():
     return "Hello World"

 def foo1():
     return "Hello Python"

 if __name__ == "__main__":
     print(foo())
     foo1 = make_body(make_italic(foo1))
     print(foo1())

01 多个装饰器

<b><i>Hello World</i></b>
<b><i>Hello Python</i></b>

01 多个装饰器——运行结果

多个装饰器的运行过程分析：

含有不定长参数

 #!/usr/bin/env python3
 #-*- coding:utf-8 -*-
 #File Name:06_decorator_multi_var.py
 #Created Time:2019-01-09 16:20:30

 from time import ctime

 def timefun(func):
     def wrapped_func(*args, **kwargs):
         print("%s called at %s" % (func.__name__, ctime()))
         print("wrapped_func: ", end="")
         print(args, kwargs)
         func(*args, **kwargs)  # 拆包
     return wrapped_func

 @timefun
 def foo(a,b,*args, **kwargs):
     print("foo: %s, %s, %s, %s" %(a, b, args, kwargs))

 if __name__ == "__main__":
     foo(1,2,3,4,5, tmp=2)

02 含有不定长参数

foo called at Wed Jan  9 16:32:48 2019
wrapped_func: (1, 2, 3, 4, 5) {'tmp': 2}
foo: 1, 2, (3, 4, 5), {'tmp': 2}

02 含有不定长参数——运行结果

带有返回值

 #!/usr/bin/env python3
 #-*- coding:utf-8 -*-
 #File Name: 07_decorator_multi_retrun.py
 #Created Time:2019-01-09 16:20:30

 from time import ctime

 def timefun(func):
     def wrapped_func(*args, **kwargs):
         print("%s called at %s" % (func.__name__, ctime()))
         print("wrapped_func: ", end="")
         print(args, kwargs)
         return func(*args, **kwargs)  # 此处如何没有 return，则26行会输出 None
     return wrapped_func

 @timefun
 def foo(a,b,*args, **kwargs):
     print("foo: %s, %s, %s, %s" %(a, b, args, kwargs))
     return "Hello world!"

 if __name__ == "__main__":
     print(foo(1,2,3,4,5, tmp=2))

03带有返回值

foo called at Wed Jan  9 16:38:05 2019
wrapped_func: (1, 2, 3, 4, 5) {'tmp': 2}
foo: 1, 2, (3, 4, 5), {'tmp': 2}
Hello world!

03带有返回值——运行结果

装饰器中设置外部变量

 #!/usr/bin/env python3
 #-*- coding:utf-8 -*-
 #File Name: 08_decorator_outside_var.py
 #Created Time:2019-01-09 16:20:30

 def timefun_arg(outside_var="Hello"):
     def timefun(func):
         def wrapped_func(*args, **kwargs):
             print("wrapped_func: %s" % outside_var)
             return func(*args, **kwargs)
         return wrapped_func
     return timefun

 @timefun_arg()
 def foo(a,b,*args, **kwargs):
     return "foo: Hello world!"

 @timefun_arg("Python")  # 相当于 foo1 = timefun_arg("Python")(foo1)
 def foo1(a,b,*args, **kwargs):
     return "foo1: Hello world!"

 if __name__ == "__main__":
     print(foo(1,2,3,4,5, tmp=2))
     print(foo1(1,2,3,4,5, tmp=2))

04 装饰器中设置外部变量

定义多层函数，相当于双层的装饰器。

wrapped_func: Hello
foo: Hello world!
wrapped_func: Python
foo1: Hello world!

04装饰器中设置外部变量——运行结果

类装饰器

 #!/usr/bin/env python3
 #-*- coding:utf-8 -*-
 #File Name:09_decorator_class.py
 #Created Time:2019-01-09 16:53:33

 class Test(object):
     def __init__(self, func):
         print("******初始化******")
         print("Called by %s." % func.__name__)
         self.func = func

     def __call__(self):
         return self.func()

 @Test  # 相当于 foo = Test(foo)
 def foo():
     return "Hello world!"

 def foo1():
     return "Hello python!"

 if __name__ == "__main__":
     print(foo())
     foo1 = Test(foo1)
     print(foo1())

05 类装饰器

类对象借助 __call__()  魔术方法，即可实现相应的类装饰器

******初始化******
Called by foo.
Hello world!
******初始化******
Called by foo1.
Hello python!

05 类装饰器运行结果