OGC Environmental Data Retrieval (EDR) API Pydantic

This repository contains the edr-pydantic Python package. It provides Pydantic models for the OGC Environmental Data Retrieval (EDR) API. This can, for example, be used to help develop an EDR API using FastAPI.

Install

pip install edr-pydantic

Or to install from source:

pip install git+https://github.com/KNMI/edr-pydantic.git

Usage

from edr_pydantic.collections import Collection
from edr_pydantic.data_queries import EDRQuery, EDRQueryLink, DataQueries
from edr_pydantic.extent import Extent, Spatial
from edr_pydantic.link import Link
from edr_pydantic.observed_property import ObservedProperty
from edr_pydantic.parameter import Parameters, Parameter
from edr_pydantic.unit import Unit
from edr_pydantic.variables import Variables

c = Collection(
    id="hrly_obs",
    title="Hourly Site Specific observations",
    description="Observation data for UK observing sites",
    extent=Extent(
        spatial=Spatial(
            bbox=[
                [-15.0, 48.0, 5.0, 62.0]
            ],
            crs="WGS84"
        )
    ),
    links=[
        Link(
            href="https://example.org/uk-hourly-site-specific-observations",
            rel="service-doc"
        )
    ],
    data_queries=DataQueries(
        position=EDRQuery(
            link=EDRQueryLink(
                href="https://example.org/edr/collections/hrly_obs/position?coords={coords}",
                rel="data",
                variables=Variables(
                    query_type="position",
                    output_formats=[
                        "CoverageJSON"
                    ]
                )
            )
        )
    ),
    parameter_names=Parameters({
        "Wind Direction": Parameter(
            unit=Unit(
                label="degree true"
            ),
            observedProperty=ObservedProperty(
                id="https://codes.wmo.int/common/quantity-kind/_windDirection",
                label="Wind Direction"
            ),
            dataType="integer"
        )
    })
)

print(c.model_dump_json(indent=2, exclude_none=True, by_alias=True))

Will print

{
  "id": "hrly_obs",
  "title": "Hourly Site Specific observations",
  "description": "Observation data for UK observing sites",
  "links": [
    {
      "href": "https://example.org/uk-hourly-site-specific-observations",
      "rel": "service-doc"
    }
  ],
  "extent": {
    "spatial": {
      "bbox": [
        [
          -15.0,
          48.0,
          5.0,
          62.0
        ]
      ],
      "crs": "WGS84"
    }
  },
  "data_queries": {
    "position": {
      "link": {
        "href": "https://example.org/edr/collections/hrly_obs/position?coords={coords}",
        "rel": "data",
        "variables": {
          "query_type": "position",
          "output_formats": [
            "CoverageJSON"
          ]
        }
      }
    }
  },
  "parameter_names": {
    "Wind Direction": {
      "type": "Parameter",
      "data-type": "integer",
      "unit": {
        "label": "degree true"
      },
      "observedProperty": {
        "id": "https://codes.wmo.int/common/quantity-kind/_windDirection",
        "label": "Wind Direction"
      }
    }
  }
}

IMPORTANT: The arguments by_alias=True to model_dump_json() or model_dump() is required to get the output as shown above. Without by_alias=True the attribute data-type will be wrongly outputted as dataType. This is due an issue in the EDR spec and Pydantic.

Contributing

Make an editable install from within the repository root

pip install -e '.[test]'

Running tests

pytest tests/

Linting and typing

Linting and typing (mypy) is done using pre-commit hooks.

pip install pre-commit
pre-commit install
pre-commit run

Related packages

• CoverageJSON Pydantic
• geojson-pydantic

Real world usage

This library is used to build an OGC Environmental Data Retrieval (EDR) API, serving automatic weather data station data from The Royal Netherlands Meteorological Institute (KNMI). See the KNMI Data Platform EDR API.

TODOs

Help is wanted in the following areas to fully implement the EDR spec:

• See TODOs in code listing various small inconsistencies in the spec
• In various places there could be more validation on content

License

Apache License, Version 2.0