CODEWITHSUNDEEP
pythonpython-dataclasses
Hugo Trentesaux
Hugo Trentesaux

Reputation: 1999

python dataclass asdict ignores attributes without type annotation

Python documentation explains how to use dataclass asdict but it does not tell that attributes without type annotations are ignored:

from dataclasses import dataclass, asdict

@dataclass
class C:
  a : int
  b : int = 3
  c : str = "yes"
  d = "nope"

c = C(5)
asdict(c)
# this returns
# {'a': 5, 'b': 3, 'c': 'yes'}
# note that d is ignored

How can I make d attribute appear in the returned dict without implementing the function myself?

Upvotes: 0

Views: 942

Answers (2)

Wizard.Ritvik
Wizard.Ritvik

Reputation: 11662

A cheesy and somewhat unprofessional way is to annotate it with any random string value.

So, for example:


from dataclasses import dataclass, asdict

@dataclass
class C:
  a : int
  b : int = 3
  c : str = "yes"
  d: 'help,, i dunno how [a]ny of this w@rks!!' = "nope"

c = C(5)
print(asdict(c))

# this returns
# {'a': 5, 'b': 3, 'c': 'yes', 'd': 'nope'}

With some thought, this can actually be simplified down to:

@dataclass
class C:
  a : int
  b : int = 3
  c : str = "yes"
  d: ... = "nope"

Upvotes: 0

Alireza Jafari
Alireza Jafari

Reputation: 79

You can use Any as type annotation. For example:

from typing import Any
from dataclasses import dataclass, asdict

@dataclass
class C:
  a : int
  b : int = 3
  c : str = "yes"
  d : Any = "nope"

c = C(5)
asdict(c)
# this returns
# {'a': 5, 'b': 3, 'c': 'yes', 'd': 'nope'}
# Now, d is included as well!

Upvotes: 1

Related Questions