dbml-builder

Generates Pydantic and SQLAlchemy from a DBML file.

This package is for users wanting to use their data model represented in DBML in production. dbml-builder accomplishes this by:

1. Generating Pydantic and SQLAlchemy code.
2. Verifying existing generated code to see if it matches the specified version and has not been changed since creation.

Currently, there doesn't seem to be a good solution for code generation with DBML in Python hence the creation of dbml-builder. Additionally, large software systems tend to break as Pydantic schemas are modified which is the reason why the package includes verification functionality.

dbml-builder is new and actively developed. If you have any feature requests or issues, please submit them here.

Installation

Install using pip:

pip install dbml_builder

Usage

Generate your ORM and schemas by running:

model-build generate ./project.dbml ./generated

or call generate_models directly in Python code.

You can check to if the model code is still valid by running:

model-build check v0.9.3 ./generated

or call verify directly in Python code.

Note that the version is what is specified in the note for a given project in DBML.

Example

Suppose we have a project:

>> ls
src/  LICENSE  poetry.lock  data-model.dbml  pyproject.toml

where src contains your code for your python project.

We can automatically generate code using:

pip install dbml_builder
model-build generate ./data-model.dbml ./src/generated

We can now submit src/generated to version control and use the generated code in a module:

from generated.schema import SOME_PYDANTIC_SCHEMA
from generated.orm import SOME_SQLALCHEMY_TABLE

We can also ensure the generated code is not changed by placing a check in our code:

# src/main.py
from dbml_builder import verify

verify('v0.1.0', '../data-model.dbml')