Different python objects for dot notation lovers

Why would I want to be able to use dot notation?

For years I've been using dictionaries to store all kind of things, especially when I needed to generate report on large amount of data.
Recently I was working on developing a rest api at work and I wanted to be able to store small pieces of flat data and access them using dot notation as I find it more elegant and convenient than using the old ['key'] of a dictionary.
My criterias were the following:

• I didn't want to "hack" any data type by modifying their behaviors.
• I only want to use the standard library.

---

Solutions:

• SimpleNamespace (3.3+)

• namedtuple (2.6+)

• enum (3.4+)

• dataclass (3.7+)

---

Examples

SimpleNamespace :

It allows you to instantiate an object immediately with any attributes that you like and create/modify them as necessary (it is a subclass of Object):

from types import SimpleNamespace

dummy_object = SimpleNamespace()
#Creation of attributes
dummy_object.name = 'dummy_name'
dummy_object.age = 36
#Update of name attribute
dummy_object.name = 'dummy_name_test'
#Delete the name attribute
del dummy_object.name

namedtuple :

As stated by its name it allows you to create a tuple and access its data by name rather than index.
As for a tuple, you can't modify it.

from collections import namedtuple

person = namedtuple('Person', ['name', 'age'])
dummy_object = person('dummy_name', 36)

Enum :

An enum is a class where the attributes name are bound to constant values.
Attributes and values can't be modified.

from enum import Enum

class dummy_object(Enum):
    name = 'dummy_name'
    age = 36

---

dataclass :

It's just a simple decorator that add special methods to classes (such as __init__ and __repr__).
As for SimpleNamespace you can modify attributes and values.

from dataclasses import dataclass

#You can add attributes without default value with type hinting.
@dataclass
class Person:
    name: str
    age: int
dummy_person = Person()

#You can add attributes with default value like with regular class.
@dataclass
class Person:
    name = 'dummy_name'
    age = 36
dummy_person = Person()

---

As a summary here is a quick table to understand the behavior of each kind of object.

	SimpleNamespace	namedtupled	Enum	dataclass
Instantiation	Yes	Yes	No	Yes
Attribute mutability	Yes	No	No	Yes
Value mutability	Yes	No	No	Yes
Identity compare	Values	Values	ID	ID

Obviously you can always use regular classes or dictionnary but these objects might simplify your workflow and improve the readability of your code.

---

References
https://docs.python.org/3/library/types.html#types.SimpleNamespace
https://docs.python.org/3/library/collections.html#collections.namedtuple
https://docs.python.org/3/library/enum.html
https://docs.python.org/3/library/dataclasses.html