Python学习之——装饰器
- 参考
- 基础
- 闭包概念
- 装饰器
- 系统自带的装饰器
- @property
- @staticmethod
- @classmethod
- 自定义装饰器
- 函数的装饰器
- 无参数
- 有参数
- 类的装饰器
- 无参数
- 有参数
- @functools.wraps
- 装饰器类
- 装饰器实现单例模式
参考
python装饰器的4种类型:函数装饰函数、函数装饰类、类装饰函数、类装饰类
9.4 定义一个带参数的装饰器
基础
闭包概念
Python闭包(Closure)详解
闭包概念:在一个内部函数中,对外部作用域的变量进行引用,并且外部函数的返回值为内部函数,那么内部函数就叫做闭包。
示例:
def outer_func(year):def inner_func(month):print(f'year:{year}, month:{month}')return inner_funcclosure_func = outer_func('2023')
closure_func (12)
调用func时,产生了闭包inner_func,其持有外层函数的自由变量year,当函数func的生命周期结束之后,year这个变量依然存在,因为它被闭包引用了,不会被回收。
闭包的特点:
内部函数可以读取外层函数内的局部变量,并让其常驻内存,不会被回收,所以注意内存泄漏
装饰器
一文搞懂什么是Python装饰器
装饰器是用来增强函数/类功能的一个函数
import loggingdef no_arg_decorator(func):call_count = 0def inner_func(*arg, **kwargs):# do somethings, examplenonlocal call_countcall_count += 1print(f'call {func.__name__} func, total call count:{call_count }')func(*arg, **kwargs)return inner_funcdef arg_decorator(*outer_arg, **outer_kwargs):def inner_func(func):def inner_func2(*arg, **kwargs):# do somethings use outer_arg, outer_kwargs, examplefile_path = outer_kwargs.get('log_file_path')if file_path:logging.basicConfig(filename=file_path,format='%(asctime)s - %(name)s - %(levelname)s - %(message)s-%(funcName)s',level=logging.DEBUG)logging.debug(f'debug, call {func.__name__}, arg:{arg}, kwargs:{kwargs}')func(*arg, **kwargs)return inner_func2return inner_funcdef test_func1(a=1):pass# 1.不用语法糖@符号
# 1.1 无参装饰器
wraped_func1 = no_arg_decorator(test_func1)# 1.2 有参装饰器
wraped_func2 = arg_decorator(log_file_path='log.txt')(test_func1)# 2.采用语法糖@符号
# 2.1 无参装饰器
@no_arg_decorator
def test_func2(a=1):pass# 2.2 有参装饰器
@arg_decorator(log_file_path='log.txt')
def test_func3(a=1):passif __name__ == '__main__':print('=========================')wraped_func1()wraped_func1()wraped_func2(a=11)test_func2()test_func3(a=111)# 结果
=========================
call test_func1 func, total call count:1
call test_func1 func, total call count:2
call test_func2 func, total call count:1
系统自带的装饰器
@property
Python 中 property() 函数及 @property 装饰器
class Food(object):def __init__(self):self._price = 0@propertydef price(self):return self._price @price.setterdef price(self, value):if value < 0:raise ValueError('price must large than 0!')self._price = value@price.deleterdef price(self):del self._price
@staticmethod
staticmethod修饰过的方法叫静态方法,可以直接通过类调用方法,这样做的好处是执行效率比较高,也可以通过实例调用该方法。
from datetime import datetimeTIME_FORMAT_STR = "%Y/%m/%d/ %H:%M:%S"class TimeUtil():@staticmethoddef timestamp_to_utc_str(timestamp: float, format_str=TIME_FORMAT_STR) -> str:"""时间戳转utc-0时区的时间"""datetime_obj: datetime = datetime.utcfromtimestamp(timestamp)return datetime_obj.strftime(format_str)@staticmethoddef timestamp_to_local_str(timestamp: float, format_str=TIME_FORMAT_STR) -> str:"""时间戳转当前本地时区的时间"""datetime_obj: datetime = datetime.fromtimestamp(timestamp)return datetime_obj.strftime(format_str)
@classmethod
classmethod修饰过的方法叫类方法,可以直接通过类或者实例调用方法
利用@classmethod实现单例模式
from datetime import datetimeclass SingletonBase(object):__instance = None@classmethoddef get_instance(cls, *arg, **kwargs):if cls.__instance is None:cls.__instance = cls()cls.__instance.init(*arg, **kwargs)return cls.__instancedef init(self, *arg, **kwargs):passclass TestMgr(SingletonBase):def init(self, *arg, **kwargs):print(f'I am TestMgr Singleton, arg:{arg}, kwargs:{kwargs}')if __name__ == '__main__':test_mgr = TestMgr.get_instance('hello', 'world', time=datetime.now().strftime("%Y/%m/%d/ %H:%M:%S"))
自定义装饰器
python装饰器2:类装饰器
python装饰器的4种类型:函数装饰函数、函数装饰类、类装饰函数、类装饰类
函数的装饰器
无参数
import timeall_func = {}# 用于函数注册的装饰器
def register_func(func):all_func[func.__name__] = funcreturn func# 用于统计函数耗时的装饰器
def calc_time(func):def inner_func():start = time.time()func()end = time.time()print(f'call {func.__name__} func, cost time:{int(end - start)} second')return inner_func@calc_time
def test_func4():time.sleep(2)if __name__ == '__main__':print('=========================')test_func4()# 结果
=========================
call test_func4 func, cost time:2 second
有参数
# -*- coding:utf-8 -*-
from functools import wrapstest_dict = {}def Decorator(arg1=0, arg2=1):def inner_func(func):funcName = func.__name__test_dict[funcName] = func# 没有return func,则该装饰器默认返回None,会导致被Decorator装饰后的函数无法被调用,# 有return func,则被Decorator装饰后的函数还是被赋值为原函数,后续还可以调用# return funcreturn inner_funcdef decorator(*arg):return Decorator(*arg)def Decorator2(arg1=0, arg2=1):def inner_func(func):@wraps(func)def wrapper(*args, **kwargs):# do somethingreturn func(*args, **kwargs)funcName = func.__name__test_dict[funcName] = func# 有return wrapper,则被Decorator2装饰后的函数还是被赋值为原函数,后续还可以调用return wrapperreturn inner_funcdef decorator2(*arg):return Decorator2(*arg)@Decorator()
def test_func_a():pass@decorator()
def test_func_b():pass@Decorator2()
def test_func_c():pass@decorator2()
def test_func_d():passif __name__ == "__main__":print(test_dict)# 结果
{'test_func_a': <function test_func_a at 0x000001F7B6851D00>, 'test_func_b': <function test_func_b at 0x000001F7B6851DA0>, 'test_func_c': <function test_func_c at 0x000001F7B6851E40>, 'test_func_d': <function test_func_d at 0x000001F7B6851F80>}
类的装饰器
无参数
# decorator1(cls)返回一个函数inner_func的写法
def decorator1(cls):def inner_func(a):print('class name:', cls.__name__)return cls(a)return inner_func# decorator2(cls)返回一个类inner_cls的写法
def decorator2(cls):class inner_cls:def __init__(self, *args, **kwargs):self.wrapped_obj = cls(*args, **kwargs)def __getattr__(self, name):return getattr(self.wrapped_obj, name)return inner_cls@decorator1
class FooClass1():def __init__(self, a):self.a = adef fun(self):print('self.a =', self.a)@decorator2
class FooClass2():def __init__(self, a):self.a = adef fun(self):print('self.a =', self.a)foo1 = FooClass1('Foo1_a')
foo1.fun()foo2 = FooClass2('Foo2_a')
foo2.fun()
有参数
1.通过装饰器实现子类的自动注册
all_subcls = {}def register_subcls(base_cls_name):def decorate(sub_cls):if base_cls_name not in all_subcls :all_subcls[base_cls_name] = {}all_subcls[base_cls_name][sub_cls.__name__] = sub_clsreturn decoratedef get_subcls(base_cls_name, sub_cls_name):if base_cls_name not in all_subcls:return Noneelse:return all_subcls[base_cls_name].get(sub_cls_name)# 使用
class TestBase1(object):@staticmethoddef create_instance(sub_cls_name, *args, **kwargs):sub_cls = get_subcls('TestBase1', sub_cls_name)return sub_cls(*args, **kwargs) if sub_cls else Nonedef __init__(self, *args, **kwargs):self.args = argsself.kwargs = kwargsdef print_self(self):print(f'TestBase1, args:{args}, kwargs:{kwargs}')@register_subcls('TestBase1')
class TestA(TestBase1):def print_self(self):print(f'TestA, args:{args}, kwargs:{kwargs}')@register_subcls('TestBase1')
class TestB(TestBase1):def print_self(self):print(f'TestB, args:{args}, kwargs:{kwargs}')
- 通过装饰器给类增加静态函数/类函数/成员函数
# 实例方法
def add_instance_method(cls):def inner_func(func):setattr(cls, func.__name__, func)return inner_func# 类方法
def add_class_method(cls):def inner_func(func):setattr(cls, func.__name__, classmethod(func))return inner_func# 静态方法
def add_static_method(cls):def inner_func(func):setattr(cls, func.__name__, staticmethod(func))return inner_funcdef add_func_decorator(cls):@add_instance_method(cls)def test_instance_func(self):print('test_instance_func')@add_class_method(cls)def test_class_func(cls):print('test_class_func')@add_static_method(cls)def test_static_func():print('test_static_func')return cls@add_func_decorator
class TestClass():def __init__(self):passif __name__ == '__main__':TestClass.test_class_func()TestClass.test_static_func()test_obj = TestClass()test_obj.test_instance_func()# 结果
test_class_func
test_static_func
test_instance_func
@functools.wraps
functools.wraps对我们的装饰器函数进行了装饰之后,消除了装饰器对原函数的影响。
具体Python装饰器深度解析
装饰器类
之前都是用函数来构造装饰器,其实类也可以用来构建装饰器
装饰器类
先熟悉一下Python __call__()方法
Python 中,凡是可以将 () 直接应用到自身并执行,都称为可调用对象。可调用对象包括自定义的函数、Python 内置函数以及本节所讲的类实例对象。
__ call__()方法的功能类似于在类中重载 () 运算符,使得类实例对象可以像调用普通函数那样,以“对象名()”的形式使用。
from functools import wrapsclass logit(object):def __init__(self, logfile='out.log'):self.logfile = logfiledef __call__(self, func):@wraps(func)def wrapped_function(*args, **kwargs):log_string = func.__name__ + " was called"print(log_string)with open(self.logfile, 'a') as fp:fp.write(log_string + '\n')# 发送一个通知self.notify()return func(*args, **kwargs)return wrapped_functiondef notify(self):pass# 使用
@logit()
def myfunc1():pass
装饰器实现单例模式
Python单例模式(Singleton)的N种实现
1.使用函数装饰器实现单例
def Singleton(cls):_instance = {}def inner():if cls not in _instance:_instance[cls] = cls()return _instance[cls]return inner@Singleton
class Cls(object):def __init__(self):passcls1 = Cls()
cls2 = Cls()
print(id(cls1) == id(cls2))
2.使用类装饰器实现单例
class Singleton(object):def __init__(self, cls):self._cls = clsself._instance = {}def __call__(self):if self._cls not in self._instance:self._instance[self._cls] = self._cls()return self._instance[self._cls]@Singleton
class Cls2(object):def __init__(self):passcls1 = Cls2()
cls2 = Cls2()
print(id(cls1) == id(cls2))# 同时,由于是面对对象的,这里还可以这么用
class Cls3():passCls3 = Singleton(Cls3)
cls3 = Cls3()
cls4 = Cls3()
print(id(cls3) == id(cls4))