SerialAlchemy might be sticking around

Oh hi. Do you actually use SerialAlchemy? Well, I have some news.

With the release of SQLAlchemy 2.0, it's possible to easily integrate dataclasses with models... which means it's easy to change models to dicts... WHICH MEANS it's possible to serialize a model into JSON without the use of SerialAlchemy.

However...

After recently trying to make some utilities to work with dataclass-based Models, I realized I was basically recreating this library. Since this library still works, what's the point of redoing it?

--

SerialAlchemy adds serialization to SQLAlchemy ORM objects via a mixin class. It is tightly coupled with SQLAlchemy, because it's the best thing ever invented.

About

Here's the elevator pitch: are you defining your models with SQLAlchemy and again with Pydantic? Do you have trouble keeping clear which models are SQLAlchemy models and which models are Pydantic models? Do you have a way to keep your models' models in sync with DB changes?

THAT'S TOO MANY MODELS for my simple brain and my lazy disposition, in any case.

I made SerialAlchemy mainly because I hate repeating myself.

MIT Licensed