Check one JSONSchema is a subset of another. Plays nicely with Pydantic.
pip install isjsonschemasubsetCurrently in Alpha.
from pathlib import Path
from isjsonschemasubset import dump, issubset, load
path_to_a = Path(__file__).parent / "a.json"
path_to_b = Path(__file__).parent / "b.json"
dump(A, path_to_a) # where A is a `type[pydantic.BaseModel]`
dump(B, path_to_b) # where B is a `type[pydantic.BaseModel]`
json_schema_a = load(path_to_a)
json_schema_b = load(path_to_b)
for error in issubset(json_schema_a, json_schema_b):
...Yields errors if type A is not a subset of B.
In this context, "is a subset of" means that if we were to do:
B(**a.model_dump(mode="json"))We would not get an error. This is very important when deserializing data from JSON columns in the database and for writing backwards compatible APIs.
See tests/test_schema_versions.py for further example usage. In this case, every time we change Foo (or anything that Foo refers to recursively), we dump a new schema in tests/schemas/Foo and check all of the schemas are backwards compatible.
git clone [email protected]:leontrolski/isjsonschemasubset.git
pip install -r requirements-dev.txt
pytest
mypy src tests --strict# bump version
python -m pip install build twine
python -m build
twine check dist/*
twine upload dist/*