I have a dataclass which its signature gets updated based on the active configs before it gets initiated and may be modified after initiation. Accordingly, I have something like this:
from dataclasses import dataclass, field
@dataclass(slots=True, frozen=False)
class MyClass:
att_1: str = field(init=False)
att_2: str = field(init=False)
@classmethod
def update_class(cls, first, second):
cls.att_1 = first
cls.att_2 = second
My problem is that although the dataclass is not frozen, when I initiate the class I cannot change the attributes anymore. This is a sample of what I do.
MyClass.update_class("11", "12")
print(MyClass.att_1)
a = MyClass()
print(a)
print(MyClass.att_2)
print(MyClass.att_1)
a.att_2 = "002"
print(a)
Running this I get what is shown bellow:
11
MyClass(att_1='11', att_2='12')
12
11
Traceback (most recent call last):
File "/b.py", line 25, in <module>
a.att_2 = "002"
AttributeError: 'MyClass' object attribute 'att_2' is read-only
Can someone tell me why is att_2
read-only and how can I fix this?
So, the problem is that you are using __slots__
and then removing the __slots__
descriptors, breaking your class. So, consider (without the dataclass
involved):
>>> class Foo:
... __slots__ = "x", "y"
...
>>> foo = Foo()
>>> foo.x = 1
>>> foo.x
1
>>> Foo.x
<member 'x' of 'Foo' objects>
>>> Foo.x = None
>>> foo.x
>>> foo.x = 1
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
AttributeError: 'Foo' object attribute 'x' is read-only
So, one simple solution is to not use __slots__
, but it really isn't clear to me how this is all supposed to work or why you have an update_class
method (and what it's supposed to accomplish).