Let's understand this concept with the given example,
class Parent:
def __init__(self):
self.name = "ayush"
self.__age = 21
def get_name(self):
return self.name
def __get_age(self):
return self.__age
obj = Parent()
print(obj.get_name())
print(obj.__get_age())
print(obj.__age)
The above code seems to be syntactically correct, but on executing the following error will occur.
AttributeError: 'Parent' object has no attribute '__get_age'
So, why the above code is giving AttributeError?
The answer is simple, this is because of name mangling in python.
The answer to this problem is in the problem itself. Since AttributeError is Raised when an attribute reference or assignment fails.
Let's explore the problem a little bit more, by analyzing the list of names in the Parent class scope. for that, we can use dir()
print(dir(obj))
['_Parent__age', '_Parent__get_age', '__class__', '__delattr__', '__dict__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__getattribute__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__le__', '__lt__', '__module__', '__ne__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', '__weakref__', 'get_name', 'name']
Here we have two attributes ( 'Parentage', '_Parent_get_age' ), whose name are implicitly merged with the class name in other words, we can say the attribute name is mangled with the class name.
Since, there is no references for the __age and __get_age identifiers in the Parent class scope, hence causes AttributeError.
What is Mangling?
Name Mangling is nothing but the mechanism, by which any identifiers with 2 leading underscores ( like __age) are textually replaced with classname_age.
Till now, we have considered this mechanism as a problem, which restricts the naming of identifiers. However, in the actual scenario, it is helpful in some situation.
Let's discuss some cases where the name mangling is used.
- Sometimes, mangling can be considered a way to implement private members in python. In the above example, we have seen that AttributeError is raised, when we are trying to access the mangled variables outside the class however inside the class the variables are referenced with the same name. But there is a catch, we can access these private member by referencing the attributes with the mangled name.
print(obj._Parent__age)
So, no perfect privacy in python ... 😩😩😩
- Overriding: Overriding is the ability of OOPs that allows the subclass to override the method of a parent class. The method in the subclass is of the same name and signature as the super class method.
class Parent:
def __init__(self):
self.name = "ayush"
self.age = 21
def get_name(self):
return self.name
def get_age(self):
return f"Parent class - {self.age}"
__get_age = get_age # private copy of original get_age method
class Child(Parent):
def __init__(self):
self.name = "abc"
self.age = 23
def get_name(self):
return self.name
def get_age(self):
return f"Child class - {self.age}"
child = Child()
print(child.get_age()) # Child class - 23
print(child._Parent__get_age()) # Parent class - 23
As it is clear from the above example, Name mangling is helpful for letting subclasses override methods without breaking intraclass method calls.
Conclusion
The conclusion from the above discussion is that implicit conversion of a variable (like __var ) name into a class mangled name is the inbuilt feature of python. Mangled variables can be used in various situations as discussed above.
Top comments (0)