Detailed explanation of the function and principle of python's super()

The definition of object methods in Python is very weird. The first parameter is usually named self (equivalent to this in other languages), which is used to pass the object itself. It does not need to be passed explicitly when calling, the system will automatically pass it.

The protagonist we introduce today is super(). Super() is very commonly used in class inheritance. It solves some problems of subclasses calling parent class methods. When the parent class is called multiple times, it is only executed once, which optimizes execution. Logic, let’s look at it in detail below.

Cite an example:

class Foo:
  def bar(self, message):
    print(message)

>>> Foo().bar("Hello, Python.")
Hello, Python.

When there is an inheritance relationship, sometimes it is necessary to call the method of the parent class in the subclass. In this case, the simplest way is to convert the object call When calling into a class, it should be noted that the self parameter needs to be passed explicitly at this time, for example:

class FooParent:
  def bar(self, message):
    print(message)
class FooChild(FooParent):
  def bar(self, message):
    FooParent.bar(self, message)

>>> FooChild().bar("Hello, Python.")
Hello, Python.

There are some disadvantages in doing this, for example, if the parent class name is modified, then in the child There will be many modifications involved in the class. In addition, Python is a language that allows multiple inheritance. The method shown above needs to be written multiple times when there is multiple inheritance, which is cumbersome. In order to solve these problems, Python introduced the super() mechanism. The example code is as follows:

class FooParent:
  def bar(self, message):
    print(message)
class FooChild(FooParent):
  def bar(self, message):
    super(FooChild, self).bar(message)

On the surface, it looks like the super(FooChild, self).bar(message) method and FooParent. The results of the bar(self, message) method are consistent. In fact, the internal processing mechanisms of the two methods are very different. When multiple inheritance is involved, there will be obvious differences. Let’s give an example directly:

Code 1:

>>> FooChild().bar("Hello, Python.")
Hello, Python.

Result:

Enter E

Enter B

Enter A

Leave A

Leave B

Enter C

Enter A

Leave A

Leave C

Enter D

Enter A

Leave A

Leave D

Leave E

The execution sequence is easy to understand. The only thing to note is that the public parent class A is executed multiple times .

Code 2:

class A:
  def __init__(self):
    print("Enter A")
    print("Leave A")
class B(A):
  def __init__(self):
    print("Enter B")
    A.__init__(self)
    print("Leave B")
class C(A):
  def __init__(self):
    print("Enter C")
    A.__init__(self)
    print("Leave C")
class D(A):
  def __init__(self):
    print("Enter D")
    A.__init__(self)
    print("Leave D")
class E(B, C, D):
  def __init__(self):
    print("Enter E")
    B.__init__(self)
    C.__init__(self)
    D.__init__(self)
    print("Leave E")
E()

Result:

Enter E

Enter B

Enter C

D

Enter A

Leave A

Leave D

Leave C

Leave B

Leave E

In the super mechanism, it can be guaranteed that the public parent class is only executed once. As for the order of execution, it is in accordance with MRO (Method Resolution Order): method resolution order. of. This MRO mechanism will be introduced in detail later.