Python

Magic Attributes

Intermediate

Magic attributes, also known as dunder attributes, are special variables that Python automatically creates and attaches to modules, classes, functions, and objects. They provide important information such as the name of a class, its documentation, parent classes, method resolution order, and more. Understanding these attributes helps you inspect, debug, and work more effectively with Python objects.

Magic Attributes (Dunder Attributes)

While magic methods (dunder methods) define behavior, magic attributes (dunder attributes) provide metadata and internal information about objects, classes, functions, and modules. Python has 50+ built-in dunder attributes. You usually don’t implement most of them yourself — Python sets them automatically. However, understanding them is essential for introspection, debugging, and metaprogramming.

Most Important Magic Attributes

Attribute	Belongs To	Purpose
__name__	Modules, Classes, Functions	Name of the module, class, or function
__doc__	Modules, Classes, Functions, Methods	Docstring (documentation string)
__module__	Classes, Functions	Name of the module where it was defined
__class__	Instances	The class of the object ( type(obj) )
__dict__	Instances, Classes, Modules	Dictionary containing writable attributes
__bases__	Classes	Tuple of base (parent) classes
__mro__	Classes	Method Resolution Order (tuple)
__qualname__	Classes, Functions	Qualified name (e.g. Class.method )
__annotations__	Functions, Classes	Dictionary of type hints
__slots__	Classes	Memory optimization by restricting attributes
__file__	Modules	Full path to the module’s source file
__package__	Modules	Package name (for packages and sub-packages)
__spec__	Modules	Module spec used by the import system

class MagicDemo:
    """This is a demo class to showcase magic attributes."""
    
    class_var = "Hello"

    def __init__(self, value):
        self.instance_var = value

    def method(self):
        """A sample method."""
        pass


# Usage
obj = MagicDemo(42)

print("=== Object ===")
print(obj.__class__)           # <class '__main__.MagicDemo'>
print(obj.__dict__)            # {'instance_var': 42}

print("\n=== Class ===")
print(MagicDemo.__name__)      # MagicDemo
print(MagicDemo.__module__)    # __main__
print(MagicDemo.__qualname__)  # MagicDemo
print(MagicDemo.__doc__)       # This is a demo class...
print(MagicDemo.__bases__)     # (<class 'object'>,)
print(MagicDemo.__mro__)       # (<class '__main__.MagicDemo'>, <class 'object'>)

print("\n=== Function ===")
print(MagicDemo.method.__name__)       # method
print(MagicDemo.method.__qualname__)   # MagicDemo.method
print(MagicDemo.method.__doc__)        # A sample method.

print("\n=== Module ===")
import os
print(os.__name__)      # os
print(os.__file__)      # /path/to/os.py
print(os.__package__)   # ''

Magic attributes give you direct access to metadata about your code. The most commonly used ones include __name__, __doc__, __class__, __dict__, __bases__, __mro__, and __module__. While you rarely need to create your own dunder attributes, knowing how to read and use them is an essential skill for writing clean, introspective, and professional Python code. Mastering both magic methods and magic attributes will significantly improve how you design and understand Python classes.

Classes Magic Attributes OOP Objects Python

← All training