Validating Metadata

In the Data Package Library metadata validation is separated from Metadata Modeling. If you have a metadata descriptor or you exported it from a model, you can validate it against JSONSchema based on the Data Package Standard validation rules. This validation is stricter than models validation and it also supports custom profiles defined as per the Data Package Standard.

Note

Most examples here use Schema metadata class but it is also applicable to other classes

Validating from a Path

It can be validated from local or remote descriptor:

from dplib.actions.schema.check import check_schema

errors = check_schema('data/schema.json')
# if there are errors it will contain `dplib.errors.MetadataError`

Validating from a Dict

A Python dictionary can be passed to the validator

from dplib.actions.schema.check import check_schema

errors = check_schema({"missingValues": "-"})
# it will contain `dplib.errors.MetadataError`

Validating a Model

It's also possible to validate a model:

from dplib.models import Schema, Field
from dplib.actions.schema.check import check_schema

schema = Schema()
schema.add_field(Field(name='string', type='string'))
schema.profile = '<my-custom-profile-url>'
errors = check_schema(schema)
# if there are errors it will contain `dplib.errors.MetadataError`