Advanced Python Dataclasses

Also Known As

TL;DR

Explanation

Basic dataclasses auto-generate boilerplate. Advanced: frozen=True makes instances immutable (like PHP readonly), slots=True uses __slots__ for memory efficiency (Python 3.10+), field(default_factory=list) for mutable defaults, post_init for validation, KW_ONLY for keyword-only fields. @dataclass(order=True) generates comparison methods. Combine with __post_init__ for validation logic. For data validation with coercion, use Pydantic; for pure data containers, dataclasses are simpler.

Common Misconception

Why It Matters

Common Mistakes

• Mutable default argument: @dataclass class Foo: items: list = [] — use field(default_factory=list).
• Not using frozen=True for value objects — mutable 'value objects' are not true value objects.
• Comparing dataclasses without order=True — __lt__/__gt__ not generated by default.
• Using dataclasses where Pydantic is needed — dataclasses don't validate or coerce types at runtime.

Code Examples

# Mutable default — shared across all instances:
@dataclass
class Order:
    items: list = []  # BUG: all Order instances share the same list!

o1 = Order(); o2 = Order()
o1.items.append('widget')
print(o2.items)  # ['widget'] — shared reference!

from dataclasses import dataclass, field
from typing import ClassVar

@dataclass(frozen=True, slots=True)  # Immutable + memory-efficient
class Money:
    amount: int  # Cents
    currency: str
    CURRENCIES: ClassVar[set] = {'USD', 'EUR', 'GBP'}

    def __post_init__(self):
        if self.currency not in self.CURRENCIES:
            raise ValueError(f'Unknown currency: {self.currency}')
        if self.amount < 0:
            raise ValueError('Amount cannot be negative')

@dataclass
class Order:
    items: list = field(default_factory=list)  # Correct mutable default

References

• ↗ https://docs.python.org/3/library/dataclasses.html

Tags

Related Terms