Adding fields/methods to existing classes.
A Python approach to method/field extensions like in C# and Kotlin, probably the closest.
pip install extclass
Suppose you have a Person
class in an already installed package:
from dataclasses import dataclass
@dataclass
class Person:
first_name: str
last_name: str
Then, you could add an extra method outside the source package somewhere in your project.
from extclass import extension
@extension
class PersonExt(Person):
age: int = 10
def full_name(self):
return f'{self.first_name} {self.last_name}'
With the help of the @extension
decorator, new field age
and full_name
are added to the origin Person
class. You can just call them as usual.
if __name__ == '__main__':
person = Person(first_name='rick', last_name='sanchez')
print(person.full_name())
print(person.age)
which print as
rick sanchez
10
However, the IDE cannot know the dynamic informations of the added attributes, so there would be a warning message:
Unresolved attribute reference 'full_name' for class 'Person'
And no autocomplete for the attributes in the IDE.
To handle the warnings also enabling autocomplete, you can simply typing.cast
the origin object to the new extended type.
from typing import cast
person = cast(PersonExt, person)
In this case, person
is still the old object but only marked as the extended type in order to cheat the IDE for better programming experiences.
Same method/field name in the extending class will not affect the origin class. So this package is very safe in any use-cases.
Classes of pydantic.BaseModel
can only attach new methods. Dynamic fields are internally forbidden by the pydantic package. So please do NOT extend fields for pydantic models, or an AttributeError
would be raised during runtime.
See test.py