您的位置：首页 > 编程语言 > Python开发

[转]理解Python装饰器

2016-05-21 14:56 1066 查看

作者：xlzd
链接：http://www.zhihu.com/question/26930016/answer/81263287
来源：知乎
著作权归作者所有。商业转载请联系作者获得授权，非商业转载请注明出处。简单来讲，可以不严谨地把Python的装饰器看做一个包装函数的函数。比如，有一个函数：

C++

def func():
print 'func() run.'
if '__main__' == __name__:
func()

1
2
3
4
5
6

def func():

print 'func() run.'

if '__main__' == __name__:

func()

运行后将输出:

C++

func() run.

1 2	<span class="go">func() run.</span>

现在需要在函数运行前后打印一条日志, 但是又不希望或者没有权限修改函数内部的结构, 就可以用到装饰器(decorator):

C++

def log(function):
def wrapper(*args, **kwargs):
print 'before function [%s()] run.' % function.__name__
rst = function(*args, **kwargs)
print 'after function [%s()] run.' % function.__name__
return rst
return wrapper
@log
def func():
print 'func() run.'
if '__main__' == __name__:
func()

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

def log(function):

 def wrapper(*args, **kwargs):

 print 'before function [%s()] run.' % function.__name__

 rst = function(*args, **kwargs)

 print 'after function [%s()] run.' % function.__name__

 return rst

 return wrapper

@log

def func():

 print 'func() run.'

if '__main__' == __name__:

 func()

对于原来的函数”func()”并没有做修改,而是给其使用了装饰器log,运行后的输出为:

C++

before function [func()] run.
func() run.
after function [func()] run.

1
2
3
4

before function [func()] run.

func() run.

after function [func()] run.

把”@log”放到func()函数定义的地方,相当于执行了如下语句:

C++

func = log(func)

1 2	<span class="n">func</span> <span class="o">=</span> <span class="n">log</span><span class="p">(</span><span class="n">func</span><span class="p">)</span>

因为log()返回了一个函数, 所以原本的func指向了log()返回的函数wrapper。wrapper的参数列表为(*args, **kwargs), 所以其可以接受所有的参数调用, 在wrapper中,先打印了一行

‘before function [%s()] run.’ % function.__name__

(在Python中函数也是对象,函数的__name__是它的名字),然后执行了原来的函数并记录了返回值,在输出

‘after function [%s()] run.’ % function.__name__

后返回了函数的执行结果。

如果decorator本身需要传入参数，那就需要编写一个返回decorator的decorator。比如在Flask中:

C++

@app.route('/')
def index():
return 'hello, world!'

1
2
3
4

@app.route('/')

def index():

return 'hello, world!'

实现如下:

C++

import functools
def log(text=''):
def decorator(function):
@functools.wraps(function)
def wrapper(*args, **kwargs):
print 'before function [%s()] run, text: [%s].' % (function.__name__, text)
rst = function(*args, **kwargs)
print 'after function [%s()] run, text: [%s].' % (function.__name__, text)
return rst
return wrapper
return decorator
@log('log text')
def func():
print 'func() run.'
if '__main__' == __name__:
func()

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

import functools

def log(text=''):

 def decorator(function):

 @functools.wraps(function)

 def wrapper(*args, **kwargs):

 print 'before function [%s()] run, text: [%s].' % (function.__name__, text)

 rst = function(*args, **kwargs)

 print 'after function [%s()] run, text: [%s].' % (function.__name__, text)

 return rst

 return wrapper

 return decorator

@log('log text')

def func():

 print 'func() run.'

if '__main__' == __name__:

 func()

输出如下:

C++

before function [func()] run, text: [log text].
func() run.
after function [func()] run, text: [log text].

1
2
3
4

before function [func()] run, text: [log text].

func() run.

after function [func()] run, text: [log text].

最后脑洞小开一下, 有没有办法实现既支持不带参数(如log), 又支持带参数(如log(‘text’))的decorator吗?

C++

import functools
def log(argument):
if not callable(argument):
def decorator(function):
@functools.wraps(function)
def wrapper(*args, **kwargs):
print 'before function [%s()] run, text: [%s].' % (function.__name__, text)
rst = function(*args, **kwargs)
print 'after function [%s()] run, text: [%s].' % (function.__name__, text)
return rst
return wrapper
return decorator
def wrapper(*args, **kwargs):
print 'before function [%s()] run.' % function.__name__
rst = function(*args, **kwargs)
print 'after function [%s()] run.' % function.__name__
return rst
return wrapper

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

import functools

def log(argument):

 if not callable(argument):

 def decorator(function):

 @functools.wraps(function)

 def wrapper(*args, **kwargs):

 print 'before function [%s()] run, text: [%s].' % (function.__name__, text)

 rst = function(*args, **kwargs)

 print 'after function [%s()] run, text: [%s].' % (function.__name__, text)

 return rst

 return wrapper

 return decorator

 def wrapper(*args, **kwargs):

 print 'before function [%s()] run.' % function.__name__

 rst = function(*args, **kwargs)

 print 'after function [%s()] run.' % function.__name__

 return rst

 return wrapper

如上~~~

内容来自用户分享和网络整理，不保证内容的准确性，如有侵权内容，可联系管理员处理

标签： Decarator Partten python 装饰器

相关文章推荐

新的分享

章节导航