dataclass

Python Dataclasses

---

Python’s dataclasses module, introduced in Python 3.7, provides a decorator and functions for automatically adding special methods to user-defined classes. This is particularly useful for classes that are primarily used to store data, as it reduces boilerplate code.

from dataclasses import dataclass, field, asdict
import json

@dataclass
class MyDataClass:
    x: int
    y: int = 0 # default value, optional

The dataclass will provide __init__, __repr__, and __eq__ methods automatically based on the class attributes.

With field, it allows for more customization of individual fields, such as setting default values, excluding fields from certain methods, and specifying metadata.
Additional field options

• init=False: Exclude the field from the generated __init__ method

@dataclass
class Item:
    name: str
    price: float
    uuid: str = field(default_factory=lambda: str(uuid.uuid4()), repr=False)
 
# default factory provide a function to generate default values
# repr = False means it will be excluded when printing the object

Making a list with dataclass is different

@dataclass
class Objects:
    items: list[Item] = field(default_factory=list)

Serialize and Deserialize with dataclass

# serialize
@dataclass
class Object:
    items: list[Item] = field(default_factory=list)
 
myobj = Object(items=[Item(name="item1", price=10.0), Item(name="item2", price=20.0)])
mydict = asdict(myobj)
json.dumps(mydict, indent=4)

# deserialize
json_str = '''
'''
data = json.loads(json_str)
myobj = Object(**data)