__slots__魔法

大家好,上一期我重点总结了有关类的基本知识,现在简单回顾一下,顺便加上一个创建类时常用的东西:__slots__

首先创建一个名人类:Celebrity

 class Celebrity:
     # 限定 Celebrity对象只能绑定name, age,domain属性,加速
    __slots__ = ['name','age',"domain"]
    # Class Attribute
    species = 'human'
    
    # Initializer / Instance Attributes
    def __init__(self, name, age, domain):
        self.name = name
        self.age = age
        self.domain = domain

可以发现用slots绑定了三个属性给Celebrity,slots的作用主要有两个:

  • 如果需要限定自定义类型的对象只能绑定某些属性,可以通过在类中定义__slots__变量来进行限定。需要注意的是__slots__的限定只对当前类的对象生效,对子类并不起任何作用。
  • 加速

现在可以做个实验,首先我们把slots绑定的domian属性去掉:

class Celebrity:
    
    # Class Attribute
    species = 'human'
    __slots__ = ['name', 'age']
    # Initializer / Instance Attributes
    def __init__(self, name, age,domain):
        self.name = name
        self.age = age
        self.domain = domain
        
female_leader = Celebrity("Miss Dong",65,"electrical appliance")

# Access the instance attributes
print("{} is {}.".format(
    female_leader.name, female_leader.age))

Out:AttributeError: 'Celebrity' object has no attribute 'domain'

会发现报错了,即便我们在init方法中有domain属性,但是由于slots中没有,所以Celebrity类下创建的实例都不能有domain

接下来让我们简单回顾一下如何调用类变量:

female_leader = Celebrity("Miss Dong", 65,"electrical appliance")
male_leader = Celebrity("Jack Ma", 55,"internet")

# Access the instance attributes
print("{} is {} and {} is {}.".format(
    female_leader.name, female_leader.age, male_leader.name, male_leader.age))

# Is male_leader a human?
if male_leader.species == "human":
    print("{0} is a {1}!".format(male_leader.name, male_leader.species))  

Out:
Miss Dong is 65 and Jack Ma is 55.
Jack Ma is a human!

*args

其实args应该和 kargs一起来说,但是今天先重点看一下它在对象中的应用,我们现在给Celebrity类新建3个实例,并且我们想知道年龄最大的是谁 ?

这种情况下*args很好用:

a = Celebrity("Miss Dong",65,"electrical appliance")
b = Celebrity("Jack Ma", 55,"internet")
c = Celebrity("Lei Jun", 50,"mobile")

def get_oldest(*args):
    return max(args)

print("The big brother is {} years old.".format(get_oldest(a.age, b.age, c.age)))

Out:
The big brother is 65 years old.

当然,其他的应用场景还有很多,不多列举了

类的继承

首先,我们在Celebrity类中新增两个方法:

  • description:对生成的大佬简单描述
  • speak: 大佬发言

完成后的结果如下:

class Celebrity:
     

    __slots__ = ['name', 'age',"domain"]
    species = 'human'

    def __init__(self, name, age, domain):
        self.name = name
        self.age = age
        self.domain = domain
        
     # instance method
    def description(self):
        return "{} is {} years old, working in the {} industry".format(self.name, self.age,self.domain)

    # instance method
    def speak(self, sound):
        return "{} says {}".format(self.name, sound)

现在新建两个类InternetBoss,MobileBoss,全部继承于Celebrity类:

    # Child class (inherits from Dog() class)
    class InternetBoss(Celebrity):
        pass

    # Child class (inherits from Dog() class)
    class MobileBoss(Celebrity):
        pass

如果我们什么都不做,会自动继承父类的 description和speak方法,做个实验,新建li作为InternetBoss的实例:

li = InternetBoss("Robbin",50,"advertisement")

调用description和speak方法:

li.description()
li.speak("What's your problem ?")

Out:
Robbin is 50 years old, working in the advertisement industry
Robbin says: What's your problem ?

再尝试一个MobileBoss的对象:

lei = MobileBoss("leijun", 50,"mobile")
lei.speak("Are you ok ?")

Out:
leijun says: Are you ok ?

可以发现都是一样的

类的多态, 复写父类的方法

对于类的多态,各种教程说的都太专业了,我的理解仅仅是:

  • 对父类现有方法A,当新增新的子类时,可以根据需要重写A。

在我们现在的例子中,可以复写description方法:

    class InternetBoss(Celebrity):
        def description(self):
            print("I'm Internet Boss !")
         
    class MobileBoss(Celebrity):
        def description(self):
            print("I'm Mobile phone Boss !")

这样InternetBoss类和MobileBoss类生成实例后,会调用自己的description方法:

li = InternetBoss("Robbin",50,"advertisement")
lei = MobileBoss("leijun", 50,"mobile")

li.description()
lei.description()

Out:
I'm Internet Boss !
I'm Mobile phone Boss !

isinstance() 和 issubclass()

Python 有两个判断继承的函数:

  • isinstance() 用于检查实例类型
  • issubclass() 用于检查类继承

现在还用我们的例子说明一下,首先,这是我们现有的三个类:

class Celebrity:
    __slots__ = ['name', 'age',"domain"]
    species = 'human'

    def __init__(self, name, age, domain):
        self.name = name
        self.age = age
        self.domain = domain
        
    def description(self):
        print( "{} is {} years old, working in the {} industry".format(self.name, self.age,self.domain))

    def speak(self, sound):
        print("{} says: {}".format(self.name, sound))
        
        
        
class InternetBoss(Celebrity):
    def description(self):
        print("I'm Internet Boss !")
         
class MobileBoss(Celebrity):
    def description(self):
        print("I'm Mobile phone Boss !")

然后我们分别用不同的类创建三个实例:

mingzhu = Celebrity("Miss Dong",65,"electrical appliance")
ma= InternetBoss("Pony", 48,"internet")
lei = MobileBoss("leijun", 50,"mobile")

现在使用issubclass()判断InternetBoss和MobileBoss是否继承自Celebrity:

# True
issubclass(InternetBoss,Celebrity) 

# True
issubclass(MobileBoss,Celebrity)

使用isinstance()查看mingzhu到底是谁的实例:

# True
isinstance(mingzhu,Celebrity)

# False
isinstance(mingzhu,InternetBoss)

# False
isinstance(mingzhu,MobileBoss)

同理查看ma到底是哪个类的实例:

# True
isinstance(ma,Celebrity)

# True
isinstance(ma,InternetBoss)

# False
isinstance(ma,MobileBoss)

因为InternetBoss是Celebrity子类,所以ma同时是Celebrity和InternetBoss的实例。

如果我们混用了issubclass和isinstance,会报错:

issubclass(ma,InternetBoss)

Out:
TypeError: issubclass() arg 1 must be a class

子类重写构造函数

刚才提到了,如果子类没有写构造函数init(),会自动继承父类的init,但我们通常需要子类有不同的初始函数,这样我们就需要自己复写一下,这里以InternetBoss为例:

class InternetBoss(Celebrity):
    def __init__(self,name, age, domain,hometown):
        super().__init__(name, age, domain)
        self.hometown = hometown
        
    
    def description(self):
        print("I'm Internet Boss !")
        
    def __repr__(self):
        return f"This is {self.name} speaking !"

使用了super()会保留需要的父类初始化参数,再添加自己的就行了,这里的repr我会下次总结,现在再新建实例:

总结

这次我记录了slots用法,*args 的一个使用场景,类的继承,复写父类方法,构造函数等基本概念,剩下的慢慢来,我会一点点补充。。。

Ps: 本文的实例名称均为杜撰,请不要对号入座...

我的其他文章已经放到了Github上,如果感兴趣的朋友可以去看看,链接如下:


alpha94511
549 声望996 粉丝

Python爱好者, 前端开发厌恶者