python装饰器

By | 2017年3月8日

这是在Python学习小组上介绍的内容,现学现卖、多练习是好的学习方式。

第一步:最简单的函数,准备附加额外功能

1
2
3
4
5
6
7
8
# -*- coding:gbk -*-
”’示例1: 最简单的函数,表示调用了两次”’
def myfunc():
    print(“myfunc() called.”)
myfunc()
myfunc()

 

第二步:使用装饰函数在函数执行前和执行后分别附加额外功能

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
# -*- coding:gbk -*-
”’示例2: 替换函数(装饰)
装饰函数的参数是被装饰的函数对象,返回原函数对象
装饰的实质语句: myfunc = deco(myfunc)”’
def deco(func):
    print(“before myfunc() called.”)
    func()
    print(”  after myfunc() called.”)
    return func
def myfunc():
    print(” myfunc() called.”)
myfunc = deco(myfunc)
myfunc()
myfunc()

第三步:使用语法糖@来装饰函数

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
# -*- coding:gbk -*-
”’示例3: 使用语法糖@来装饰函数,相当于“myfunc = deco(myfunc)”
但发现新函数只在第一次被调用,且原函数多调用了一次”’
def deco(func):
    print(“before myfunc() called.”)
    func()
    print(”  after myfunc() called.”)
    return func
@deco
def myfunc():
    print(” myfunc() called.”)
myfunc()
myfunc()

第四步:使用内嵌包装函数来确保每次新函数都被调用

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
# -*- coding:gbk -*-
”’示例4: 使用内嵌包装函数来确保每次新函数都被调用,
内嵌包装函数的形参和返回值与原函数相同,装饰函数返回内嵌包装函数对象”’
def deco(func):
    def _deco():
        print(“before myfunc() called.”)
        func()
        print(”  after myfunc() called.”)
        # 不需要返回func,实际上应返回原函数的返回值
    return _deco
@deco
def myfunc():
    print(” myfunc() called.”)
    return ‘ok’
myfunc()
myfunc()

第五步:对带参数的函数进行装饰

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
# -*- coding:gbk -*-
”’示例5: 对带参数的函数进行装饰,
内嵌包装函数的形参和返回值与原函数相同,装饰函数返回内嵌包装函数对象”’
def deco(func):
    def _deco(a, b):
        print(“before myfunc() called.”)
        ret = func(a, b)
        print(”  after myfunc() called. result: %s” % ret)
        return ret
    return _deco
@deco
def myfunc(a, b):
    print(” myfunc(%s,%s) called.” % (a, b))
    return a + b
myfunc(1, 2)
myfunc(3, 4)

第六步:对参数数量不确定的函数进行装饰

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
# -*- coding:gbk -*-
”’示例6: 对参数数量不确定的函数进行装饰,
参数用(*args, **kwargs),自动适应变参和命名参数”’
def deco(func):
    def _deco(*args, **kwargs):
        print(“before %s called.” % func.__name__)
        ret = func(*args, **kwargs)
        print(”  after %s called. result: %s” % (func.__name__, ret))
        return ret
    return _deco
@deco
def myfunc(a, b):
    print(” myfunc(%s,%s) called.” % (a, b))
    return a+b
@deco
def myfunc2(a, b, c):
    print(” myfunc2(%s,%s,%s) called.” % (a, b, c))
    return a+b+c
myfunc(1, 2)
myfunc(3, 4)
myfunc2(1, 2, 3)
myfunc2(3, 4, 5)

第七步:让装饰器带参数

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
# -*- coding:gbk -*-
”’示例7: 在示例4的基础上,让装饰器带参数,
和上一示例相比在外层多了一层包装。
装饰函数名实际上应更有意义些”’
def deco(arg):
    def _deco(func):
        def __deco():
            print(“before %s called [%s].” % (func.__name__, arg))
            func()
            print(”  after %s called [%s].” % (func.__name__, arg))
        return __deco
    return _deco
@deco(“mymodule”)
def myfunc():
    print(” myfunc() called.”)
@deco(“module2”)
def myfunc2():
    print(” myfunc2() called.”)
myfunc()
myfunc2()

第八步:让装饰器带 类 参数

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
# -*- coding:gbk -*-
”’示例8: 装饰器带类参数”’
class locker:
    def __init__(self):
        print(“locker.__init__() should be not called.”)
    @staticmethod
    def acquire():
        print(“locker.acquire() called.(这是静态方法)”)
    @staticmethod
    def release():
        print(”  locker.release() called.(不需要对象实例)”)
def deco(cls):
    ”’cls 必须实现acquire和release静态方法”’
    def _deco(func):
        def __deco():
            print(“before %s called [%s].” % (func.__name__, cls))
            cls.acquire()
            try:
                return func()
            finally:
                cls.release()
        return __deco
    return _deco
@deco(locker)
def myfunc():
    print(” myfunc() called.”)
myfunc()
myfunc()

第九步:装饰器带类参数,并分拆公共类到其他py文件中,同时演示了对一个函数应用多个装饰器

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
# -*- coding:gbk -*-
”’mylocker.py: 公共类 for 示例9.py”’
class mylocker:
    def __init__(self):
        print(“mylocker.__init__() called.”)
    @staticmethod
    def acquire():
        print(“mylocker.acquire() called.”)
    @staticmethod
    def unlock():
        print(”  mylocker.unlock() called.”)
class lockerex(mylocker):
    @staticmethod
    def acquire():
        print(“lockerex.acquire() called.”)
    @staticmethod
    def unlock():
        print(”  lockerex.unlock() called.”)
def lockhelper(cls):
    ”’cls 必须实现acquire和release静态方法”’
    def _deco(func):
        def __deco(*args, **kwargs):
            print(“before %s called.” % func.__name__)
            cls.acquire()
            try:
                return func(*args, **kwargs)
            finally:
                cls.unlock()
        return __deco
    return _deco
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
# -*- coding:gbk -*-
”’示例9: 装饰器带类参数,并分拆公共类到其他py文件中
同时演示了对一个函数应用多个装饰器”’
from mylocker import *
class example:
    @lockhelper(mylocker)
    def myfunc(self):
        print(” myfunc() called.”)
    @lockhelper(mylocker)
    @lockhelper(lockerex)
    def myfunc2(self, a, b):
        print(” myfunc2() called.”)
        return a + b
if __name__==”__main__”:
    a = example()
    a.myfunc()
    print(a.myfunc())
    print(a.myfunc2(1, 2))
    print(a.myfunc2(3, 4))

发表评论

您的电子邮箱地址不会被公开。 必填项已用*标注