Versioning

asdf-pydantic provides a few tools and recommendations to help propagate versions of your models to ASDF extensions, tags, schemas, etc. following the guidelines in the ASDF documentation on versioning extensions.

Use the package version

asdf-pydantic recommends maintainers to use the versioning scheme of their Python package for all versioned resources in ASDF.

__version__ = ...

class MyModel(AsdfPydanticModel):
    _tag = f"asdf://asdf-pydantic/examples/tags/mymodel-{__version__}"


class MyModel(AsdfPydanticModel):
    _tag = TagDefinition(
        tag=f"asdf://asdf-pydantic/examples/tags/mymodel-{__version__}",
        schema_uri=f"asdf://asdf-pydantic/examples/schemas/mymodel-{__version__}",
    )


class MyExtension(Extension):
    extension_uri = f"asdf://asdf-pydantic/examples/extensions/test-extension-{__version__}"

Ways to construct __version__

Version from a version module

A version module is a de-facto standard for Python packages to define their version in a single location. This can be done by creating a version.py file in your package that contains the version in the variable, __version__.

Certain tools like setuptools-scm or hatch-vcs can automatically generate this module.

# mypackage/model.py
from mypackage.version import __version__

Version from the package metadata

Using importlib the version of the distribution package can be obtained. Here we use the magic variable __package__ inside its package module to get the package name and then retrieve the version from the package metadata.

# mypackage/model.py
import importlib

__version__ = importlib.import_module(__package__).__version__