@dataclass is a decorator which is part of the Python dataclasses module. When the @dataclass decorator is used, it automatically generates special methods such as: 

• _ _ init _ _.: Constructor to initialize fields
• _ _ repr _ _ : String representation of the object
• _ _ eq _ _  : Equality comparison between objects
• _ _ hash_ _ : Enables use as dictionary keys (if values are hashable)

 Along with the methods listed above, the @dataclass decorator has two important attributes.

• Order: If True, (the default is False), __lt__(), __le__(), __gt__(), and __ge__() methods will be generated; i.e., @dataclass (order = True).
• Immutability: Fields can be made immutable using the frozen=True parameter; i.e., @dataclass(frozen=True). 

 In a nutshell, the primary goal of the @dataclass decorator is to simplify the creation of classes. 

Advantages of the dataclass Decorator

Using the dataclass decorator has several advantages:

• Boilerplate reduction: It reduces the amount of boilerplate code needed for classes by automatically generating common special methods.
• Readability: It improves the readability of the code by making it more concise and focused on the data representation.
• Default values: You can provide default values for attributes directly in the class definition, reducing the need for explicit __init__() methods.
• Immutability:By combining @dataclass with the frozen=True option, you can create immutable data classes, ensuring that instances cannot be modified after creation.

Usage

from dataclasses import dataclass

@dataclass
class Person:
    name: str
    age: int

In this example, the Person class is annotated with @dataclass, and two fields (name and age), are declared. The __init__(), __repr__(), __eq__(), and __hash__() methods are automatically generated. Here's an explanation of how to use each generated method:

• __init__(self, ...): The __init__ method is automatically generated with parameters corresponding to the annotated attributes. You can create instances of the class by providing values for the attributes.

person = Person('Sam', 45)

• __repr__(self) -> str: The __repr__ method returns a string representation of the object, useful for debugging and logging. When you print an object or use it in an f-string, the __repr__ method is called.

person # Person(name='Sam', age=45)

• __eq__(self, other) -> bool: The __eq__ method checks for equality between two objects based on their attributes. It is used when you compare objects using the equality operator (==).

# Usage
person1 = Person('Sam', 45)
person1

person2 = Person('Sam', 46)
person2

print(person1 == person2) # False. 

• __hash__(self) -> int: The __hash__ method generates a hash value for the object, allowing instances to be used in sets and dictionaries. It is required when the class is used as a key in a dictionary or an element in a set.

Ordering

If you include the order=True option, additional ordering methods (__lt__, __le__, __gt__, and __ge__) are generated. These methods allow instances to be compared using less than, less than or equal, greater than, and greater than or equal operators. If you perform a comparison on the Person object without order, TypeError will be thrown.  

print(person1 < person2)
# // TypeError: '<' not supported between instances of 'Person' and 'Person'

After adding ordering, we can perform comparisons.

@dataclass(order=True)
class Person:
    name: str
    age: int


# Usage
person1 = Person('Sam', 45)
person1

person2 = Person('Sam', 46)
person2

print(person1 < person2) # False. 

order is False by default, meaning comparison methods are not generated unless explicitly enabled. Comparisons are based on field values, not object identities.

Immutability

@dataclass can be made immutable using the frozen=True attribute; the default is False.

@dataclass
class Person:
    name: str
    age: int

person = Person('Sam', 45)
person.name = 'Sam2'
person # Person(name='Sam2', age=45)

In the code above, we are able to reassign values to the Person name field. After adding frozen=True, the exception will be thrown and reassignment is not allowed. 

@dataclass(frozen=True)
class Person:
    name: str
    age: int

person = Person('Sam', 45)
person.name = 'Sam2'

# FrozenInstanceError: cannot assign to field 'name'

Be aware of performance implications: frozen=True adds a slight overhead because of additional checks for immutability.

Default Value

Using the dataclassesmodule, we can assign the default value to the fields in the class definition. 

from dataclasses import dataclass, field

@dataclass
class Person:
    name: str
    age: int = field(default=20)


# Usage
person = Person('Sam')
person # Person(name='Sam', age=20)

Default values can be of any data type, including other data classes or mutable objects. They are evaluated only once when the class is defined, not each time an instance is created.