Skip to content
← All tracks

Data Modelling and Invariants

Python

Dataclasses, `TypedDict`, `NamedTuple`, enums and validation at the boundary — making illegal states unrepresentable.

This track is written for Python, which isn't the mode you're browsing in.

0 / 20 solved · 5 articles
Continue →
  1. 1. __slots__, part 1: what it actually saves (with numbers) Read
  2. 2. Choosing a record type: dataclass, NamedTuple, TypedDict, attrs, pydantic, msgspec Read
  3. 3. Descriptor-typed fields: the third way to compute a field Read
  4. 4. make_dataclass and building record types at runtime Read
  5. 5. Migration table: dataclass behaviour from 3.10 to 3.15 Read
  6. 6. Not solved yet. Dataclasses: slots, frozen, order
  7. 7. Not solved yet. asdict() and astuple(): the deep-copy tax
  8. 8. Not solved yet. Serializing dataclasses properly: round-trips, paths and schema versions
  9. 9. Not solved yet. Converters and validate-on-assignment: what attrs buys you
  10. 10. Not solved yet. dataclass_transform (PEP 681): making your own decorator type-checkable
  11. 11. Not solved yet. @dataclass is a code generator: what it emits
  12. 12. Not solved yet. eq, hash, and the eq/frozen/unsafe_hash truth table
  13. 13. Not solved yet. replace(), copy.replace(), and the mypy gap between them
  14. 14. Not solved yet. field(): the seven knobs that change generated behaviour
  15. 15. Not solved yet. frozen=True: what it buys, what it costs, and what it does not do
  16. 16. Not solved yet. Writing generic code over dataclasses under --strict
  17. 17. Not solved yet. kw_only, KW_ONLY and the inheritance ordering wall
  18. 18. Not solved yet. The mutable-default trap — and why the guard is weaker than you think
  19. 19. Not solved yet. NamedTuple: when a record should be a tuple
  20. 20. Not solved yet. Parse at the boundary, dataclasses in the core
  21. 21. Not solved yet. __post_init__ and InitVar: parameters that are consumed, not stored
  22. 22. Not solved yet. Runtime introspection after PEP 649: Field.type is not a type
  23. 23. Not solved yet. Generic dataclasses, Self, and the covariance question
  24. 24. Not solved yet. order=True, the 3.13 __eq__ change, and orderings that are not tuples
  25. 25. Not solved yet. __slots__, part 2: the eight things it breaks

Check yourself

4 questions · one attempt each

These do not count toward finishing the track. They are here to catch the things that are easy to read past.

0 / 4

Why does the first dataclass raise, and what does default_factory fix? 

@dataclass
class Bad:
    xs: list = []

@dataclass
class Good:
    xs: list = field(default_factory=list)
Question 1 of 4