Unlock Python’s Magic Methods: A Practical Guide to Special Attributes

Special Attributes

Attribute

Meaning

__name__

Name of class, function, or method

__module__

Module where the class is defined

__class__

Class to which an object or class belongs

__bases__

Tuple of base classes

__doc__

Docstring of class or function (None if not defined)

__mro__

Method resolution order (class.mro())

__dict__

Writable dictionary of attributes for a class or instance

__dir__

List of member names; called by dir()

Magic Methods

Categories

Creation, initialization and destruction (__init__, __del__)

Hashing (__hash__)

Boolean evaluation (__bool__)

Visualization (__repr__, __str__, __bytes__)

Operator overloading

Container size and length

Callable objects (__call__)

Context management (__enter__, __exit__)

Reflection (__getattr__, __getattribute__, etc.)

__hash__

The hash() function returns an integer; defining __hash__ makes instances hashable.

class A:
    def __init__(self, name, age=18):
        self.name = name
        self.age = age

    def __hash__(self):
        return 1

    def __repr__(self):
        return self.name

print(hash(A('tom')))
print((A('tom'), A('tom')))
print({A('tom'), A('tom')})
print({A('tom'): 1, A('tom'): 2})
print({A('tom'), A('tom')})
print({tuple('t'), tuple('t')})
print({('tom',), ('tom',)})
print({'tom', 'tom'})

This example shows that equal hash values do not guarantee object equality; hash collisions can occur. To remove duplicate keys from a set, you must also implement __eq__.

class A:
    def __init__(self, name, age=18):
        self.name = name
        self.age = age

    def __hash__(self):
        return 1

    def __eq__(self, other):
        return self.name == other.name

    def __repr__(self):
        return self.name

print({A('tom'), A('tom')})

__eq__

Corresponds to the == operator; returns a bool indicating whether two objects are equal.

Comparison with __hash__:

__hash__ only provides a hash value for use as a set or dict key; __eq__ is needed to decide actual equality.

Equal hash values may be a collision and do not imply equality.

Providing __hash__ alone is usually for using the object as a key.

Non‑hashable objects return False for isinstance(obj, collections.Hashable).

__bool__

The built‑in bool() calls __bool__; if not defined, Python falls back to __len__ (non‑zero is True). If __len__ is also missing, all instances are considered True.

Visualization Magic Methods

__repr__

repr() returns the official string representation of an object; if __repr__ is not defined, the default shows the object's memory address.

__str__

str() (and print/format) use __str__ if defined; otherwise they fall back to __repr__.

__bytes__

bytes() returns a bytes representation of the object.

class A:
    def __init__(self, name, age=18):
        self.name = name
        self.age = age

    def __repr__(self):
        return 'repr:{},{}'.format(self.name, self.age)

    def __str__(self):
        return 'str: {},{}'.format(self.name, self.age)

    def __bytes__(self):
        import json
        return json.dumps(self.__dict__).encode()

print(A('tom'))
print([A('tom')])
print([str(A('tom'))])

Operator Overloading

Operator

Special Method

Meaning

<, <=, ==, >, >=, !=

__lt__, __le__, __eq__, __gt__, __ge__, __ne__

Comparison operators

+, -, *, /, %, **, divmod

__add__, __sub__, __mul__, __truediv__, __mod__, __floordiv__, __pow__, __divmod__

Arithmetic operators (including shift and bitwise)

@functools.total_ordering decorator

Implements the full set of ordering methods from a minimal subset, but should be used sparingly.

Container‑Related Methods

Method

Meaning

__len__

len() returns the size; also used by bool() when __bool__ is missing.

__iter__

Returns an iterator for iteration.

__contains__

Implements the in operator; falls back to __iter__ if absent.

__getitem__

Enables self[key] access; raises KeyError if key not found.

__setitem__

Sets a value for self[key].

__missing__

Called by dict subclasses when a key is absent.

Callable Objects

In Python everything is an object, including functions; calling a function actually invokes its __call__ method.

Context Management

If a class implements both __enter__() and __exit__(), its instances can be used with the with statement.

class Point:
    def __init__(self):
        print('init')

    def __enter__(self):
        print('enter')
        return self

    def __exit__(self, exc_type, exc_val, exc_tb):
        print('exit')
        return True  # suppress exceptions

with Point() as p:
    print('do something')

__enter__ returns the object bound after as; __exit__ receives exception information and can suppress it by returning a true value.

Parameters

__enter__ takes no arguments.

__exit__ receives three arguments (exc_type, exc_val, exc_tb); they are None when no exception occurs.

Context Application Scenarios

Enhancing functionality before and after code execution (similar to decorators).

Resource management, e.g., closing files, network connections, database connections.

Permission checks performed in __enter__ before code runs.

contextlib.contextmanager

This decorator turns a generator function into a context manager, avoiding the need to implement __enter__ and __exit__ in a class.

Reflection‑Related Functions and Methods

Built‑in Reflection Functions

getattr(object, name[, default]) – returns attribute value or default.

setattr(object, name, value) – sets or creates an attribute.

hasattr(object, name) – checks if an attribute exists.

Reflection Magic Methods

__getattr__()

class Base:
    n = 0

class Point(Base):
    z = 6
    def __init__(self, x, y):
        self.x = x
        self.y = y
    def __getattr__(self, item):
        return f"missing {item}"

p = Point(4, 5)
print(p.x)   # 4
print(p.z)   # 6
print(p.n)   # 0
print(p.t)   # missing t

If an attribute is not found via normal lookup, __getattr__ is called; otherwise an AttributeError is raised.

setattr()

class Point(Base):
    def __setattr__(self, key, value):
        print(f"setattr {key}={value}")
        self.__dict__[key] = value

Intercepts attribute creation or modification; you must update __dict__ manually for the change to take effect.

__getattribute__

class Base:
    n = 0

class Point(Base):
    z = 6
    def __init__(self, x, y):
        self.x = x
        self.y = y
    def __getattribute__(self, item):
        return item  # simplistic example

__getattribute__ is invoked for *every* attribute access; it must delegate to the superclass to avoid infinite recursion and should return the computed value or raise an exception, after which __getattr__ may be called if the attribute is still missing.