3 ms·

My recommendation is to avoid using an ORM (SQLAlchemy or any other) to define the business logic models (which is the approach used in Django). That works well

My recommendation is to avoid using an ORM (SQLAlchemy or any other) to define the business logic models (which is the approach used in Django). That works well only if your business models and your database entities align very closely, but in many large systems this is not the case, end you end up running around in circles trying to align the two. Instead, use either an ORM, or even better the SQLALchemy core, to manage and abstract out the database; and use something like dataclasses, attrs or Pydantic to define your business logic models. In a network-oriented application, i.e. an API, I would recommend to use a separate tool to model the I/O entities, again unless they correspond very closely to the business ones.

Spivak 4y ago
So build your own application-specific bespoke ORM from dataclasses rather than using the one sitting right there?
- travisjungroth 4y ago
  When that pattern is done well, and it can be, it’s that you have a layer that doesn’t care that stuff is stored in a database. It’s lots of logic. That layer uses tools like dataclasses. Then you have some layer that stores things in the database. It’s like a record keeper. It uses an ORM or not. One benefit of this is ORMs are giant, complex objects by necessity. It’s good to only use them where they’re needed, at the database. Then you don’t have all their magic floating through your business code. They’re also rather opinionated. I, at least, experience whay more freedom when making a class from scratch than using an ORM. This comes at a cost of maintaining two versions of lots of things and a translation layer in between. It is overkill for super CRUDy stuff (most apps).