Added fix for signature of inherited dataclass

[howsunjow]

Change Summary

Added fix for signature of inherited dataclass when using the pydantic dataclass decorator.

There is a bug in the following code

from pydantic.dataclasses import dataclass

@dataclass
class A:
    a: int

@dataclass
class B(A):
    b: int

where the signature of the class B does not include the subclass (B) fields (see issue #7906 )

• Used generate_model_signature to create a temporary signature before running generate_dataclass_signature
• Added unit test for fix

Related issue number

fix #7906

Checklist

• The pull request title is a good summary of the changes - it will be used in the changelog
• Unit tests for the changes exist
• Tests pass on CI
• Documentation reflects the changes where applicable
• My PR is ready to review, please add a comment including the phrase "please review" to assign reviewers

please review

[sydney-runkle]

Hi @howsunjow,

Thanks so much for your contribution. The test that you've written looks great.

I left one comment regarding changes that we will likely want to make to the generate_dataclass_signature function. I think this would be a cleaner approach. I certainly wouldn't be opposed to centralizing some of the shared logic between the generate_dataclass_signature function and the generate_model_signature function once we figure out the commonalities that they share.

[sydney-runkle]

@howsunjow,

I left some very minor nit picks.

This looks much better - thanks for the change! I'd still like to reduce some of the redundancy that we have across the generate_model_signature and generate_dataclass_signature functions. Specifically, I think we can centralize some of the shared logic + use that in both cases.

Would you like to take a stab at that? If not, no worries, I'd be happy to take that on.

Great work 🌟, and thanks again for your contribution!

[howsunjow]

@howsunjow,

I left some very minor nit picks.

This looks much better - thanks for the change! I'd still like to reduce some of the redundancy that we have across the generate_model_signature and generate_dataclass_signature functions. Specifically, I think we can centralize some of the shared logic + use that in both cases.

Would you like to take a stab at that? If not, no worries, I'd be happy to take that on.

Great work 🌟, and thanks again for your contribution!

I've made some additional changes to centralize the the shared signature generation logic.

[sydney-runkle]

@howsunjow Nice! Thanks for refactoring to take advantage of shared code.

Merging now 😄. Thanks for your contribution ⭐!