Python decorator是什么？（20180517）

最近在跟着学廖雪峰老师的教程。在项目实例阶段，直接上来就是decorator()的写法，很是摸不到头脑，所以在csdn博客圈里补下这块的知识。

原来decorator是为了少写代码啊！ 做几个试验

定义一个简单的功能函数

>>> def do_something():
	for i in range(1000):
		pass
	print('test')
	
>>> do_something()
test

函数嵌套调用

现在输出功能函数的执行时间，首先使用函数嵌套调用的方式（不是装饰器写法），定义一个performance函数，以函数为入参，输出runtime，作为fun函数的性能情况

>>> import time
>>> def performance(fun):
	start = time.time()
	fun()
	runtime = time.time() - start
	print(runtime)

>>> def do_something():
	for i in range(1000):
		pass
	print('test')	

>>> performance(do_something)test0.03552699089050293>>>

最简单的装饰器

用decorator的方式改写上述写法，采用python的@写法，在do_something()函数调用前，使用performance装饰器

>>> import time
>>> def performance(fun):
	start = time.time()
	fun()
	runtime = time.time() - start
	print(runtime)

>>> @performance
def do_something():
for i in range(10000):
pass
print('test')



test
0.08151555061340332
>>> 

带单参函数的装饰器

假设对do_something()函数做改造，允许传入一个参数，再次改写装饰器

>>> import time
>>> def performance(fun):
	def wrapper(name):
		start = time.time()
		fun(name)
		runtime = time.time() - start
		print(runtime)
	return wrapper

>>> @performance
def do_something(name):
	for i in range(10000):
		pass
	print("test",name)

	
>>> do_something('haha')
test haha
0.055966854095458984
>>> 

带多参函数的装饰器

再次对do_something函数做改造，允许传入多个参数，在此情况下，对装饰器做改写

>>> import time
>>> def performance(fun):
	def wrapper(*args, **kw):
		start = time.time()
		fun(*args, **kw)
		runtime = time.time() - start
		print(runtime)
	return wrapper

>>> @performance
def do_something(name):
	for i in range(1000):
		pass
	print('test',name)

	
>>> @performance
def do_something2(name, user):
	for i in range(1000):
		pass
	print(user,' test ',name)

	
>>> do_something('Jack')
test Jack
0.06430554389953613
>>> do_something2('haha','chen')
chen  test  haha
0.11893939971923828
>>> 

装饰器带入参

再次改造装饰器，装饰器有入参调用次数，默认为1

>>> import time
>>> def performance(runtimes=1):
	def _performance(run):
		def wrapper(*args, **kw):
			start = time.time()
			for i in range(runtimes):
				run(*args, **kw)
			runtime = time.time() - start
			print(runtime)
		return wrapper
	return _performance

>>> @performance(2)
def do_something(*args, **kw):
	for i in range(10000):
		pass
	print(args)

	
>>> do_something('chen','it works')
('chen', 'it works')
('chen', 'it works')
0.053436994552612305
>>> 

总结：

装饰器写法，是为了减少代码量，尽量做到代码复用。引用官方手册的说明，装饰器写法（写法二），等同于写法一。

def f(...):
    ...

#写法一：
f = staticmethod(f)
#写法二：
@staticmethod
def f(...):
    ...