📜  在Python中扩展类方法

📅  最后修改于: 2022-05-13 01:54:18.850000             🧑  作者: Mango

在Python中扩展类方法

在Python中,当一个子类定义一个已经存在于其超类中的函数以便以自己的方式添加一些其他功能时,子类中的函数被称为扩展方法,该机制称为扩展。这是Python显示多态性的一种方式。这类似于Java中的覆盖和 C++ 中的虚拟方法。来自子类实例的调用执行子类的函数版本。要调用超类的版本,使用super()方法。

为什么要扩展函数?
在Python中扩展类方法的目标可以通过继承来实现。扩展方法提供的主要优点之一是它使代码可重用。此外,多个子类可以共享相同的代码,甚至可以根据需要对其进行更新。

我们将在这里处理这两种情况

案例1:没有扩展类方法

# definition of superclass "Triangles"
class Triangles(object):
      
    count = 0
      
    # Calling Constructor
    def __init__(self, name, s1, s2, s3):
        self.name = name
        self.s1 = s1
        self.s2 = s2
        self.s3 = s3
        Triangles.count+= 1
  
    def setName(self, name):
        self.name = name
  
    def setdim(self, s1, s2, s3):
        self.s1 = s1
        self.s2 = s2
        self.s3 = s3
  
    def getcount(self):
        return Triangles.count
       
    def __str__(self):
        return 'Name: '+self.name+'\nDimensions: '+str(self.s1)+','+str(self.s2)+','+str(self.s3)
      
      
# describing a subclass 
# inherits from Triangles
class Peri(Triangles):
      
    # function to calculate the area     
    def calculate(self):
        self.pm = 0
        self.pm = self.s1 + self.s2 + self.s3
         
    # function to display just the area 
    # because it is not extended
    def display(self):
        return self.pm
          
      
def main():
      
    # instance of the subclass
    p = Peri('PQR', 2, 3, 4)
    # call to calculate()
    p.calculate()
    # explicit call to __str__()
    print(p.__str__())
    # call to display()
    print(p.display())
      
main()

输出:

Name: PQR
Dimensions: 2, 3, 4
9

案例2:扩展类方法

# definition of superclass "Triangles"
class Triangles(object):
      
    count = 0
      
    def __init__(self, name, s1, s2, s3):
        self.name = name
        self.s1 = s1
        self.s2 = s2
        self.s3 = s3
        Triangles.count+= 1
  
    def setName(self, name):
        self.name = name
  
    def setdim(self, s1, s2, s3):
        self.s1 = s1
        self.s2 = s2
        self.s3 = s3
  
    def getcount(self):
        return Triangles.count
  
    # superclass's version of display()
    def display(self):
        return 'Name: '+self.name+'\nDimensions: '+str(self.s1)+', '+str(self.s2)+', '+str(self.s3)
  
# definition of the subclass
# inherits from "Triangles"
class Peri(Triangles):
          
    def calculate(self):
        self.pm = 0
        self.pm = self.s1 + self.s2 + self.s3
         
    # extended method 
    def display(self):
          
        # calls display() of superclass 
        print (super(Peri, self).display())
          
        # adding its own properties 
        return self.pm
          
      
def main():
      
    # instance of the subclass
    p = Peri('PQR', 2, 3, 4)
      
    # call to calculate
    p.calculate()
      
    # one call is enough 
    print(p.display())
      
main()

输出:

Name: PQR
Dimensions: 2, 3, 4
9

在这两种情况下产生的输出是相同的,唯一的区别是第二种情况使其他子类更容易继承方法并根据需要“扩展”它们,如上所述增加了代码的可重用性。