windlathe

windlathe provides experimental, pre-alpha utilities for defining and validating metadata for multidimensional scientific model grids.

The current release validates versioned JSON manifests and provides structured validation errors. The API and file format may change before version 1.0. Python 3.10 or newer is supported.

Manifest format

Schema version 1 describes ordered grid axes and fields stored on the complete grid. A minimal two-dimensional manifest looks like this:

{
  "schema_version": 1,
  "axes": [
    {
      "name": "parameter_a",
      "units": "1",
      "coordinates": [0, 1]
    },
    {
      "name": "parameter_b",
      "units": "1",
      "coordinates": [10, 20, 30]
    }
  ],
  "fields": [
    {
      "name": "field_1",
      "units": "1",
      "shape": [2, 3],
      "data_reference": "field_1.json"
    }
  ],
  "provenance": ["generated example"]
}

Axis and field names must be unique within their lists. Coordinates must be finite numbers, and each field shape must match the coordinate counts in axis order. The optional provenance value is an array of non-empty strings. Unknown keys are rejected so misspellings do not silently pass validation.

Load a manifest from a path or an open text or binary file:

from windlathe import load_manifest

manifest = load_manifest("manifest.json")
print(manifest.axes[0].coordinates)
print(manifest.fingerprint())

load_manifest returns an immutable Manifest containing typed Axis and Field objects. Invalid JSON raises ManifestDecodeError; valid JSON that does not match the schema raises ManifestValidationError. A validation error provides path and message attributes in addition to its readable string.

Development environment

Create the Conda environment and activate it:

conda env create -f environment.yml
conda activate windlathe

Run the tests with:

python -m unittest discover -s test -v