eyconf.validation.backends.json_schema.JsonSchemaValidator¶

class eyconf.validation.backends.json_schema.JsonSchemaValidator(allow_additional: bool = False)¶

JSON Schema validation backend for dataclass schemas using Draft 2020-12.

Features full support for Annotated, Literal, NotRequired, Unions, nested dataclasses, aliases via metadata, and Sequence types. Cached schema generation.

Attributes

allow_additional

Methods

`__init__`([allow_additional])
`to_json_schema`(schema[, check_schema])	Convert TypedDict/dataclass to JSON Schema.
`validate`(data, schema)	Protocol match: Validate data against schema-derived JSON Schema.
`validate_and_construct`(data, schema)	Validate and construct dataclass instance.

__init__(allow_additional: bool = False) → None¶

to_json_schema(schema: type[D], check_schema: bool = True) → dict[str, Any]¶

Convert TypedDict/dataclass to JSON Schema.

Parameters:

schema (type[D]) – Dataclass/TypedDict type to convert.
check_schema (bool, default=True) – Validate schema with Draft202012Validator.
allow_additional (bool or None, default=None) – TODO: Allow extra properties; uses __allow_additional__ if None.

Returns:

Draft-2020-12 compliant schema.

Return type:

JsonSchema

Raises:

ValueError – Unsupported types (non-str dict keys, etc.).
jsonschema.SchemaError – Invalid generated schema.

Notes

Cached for repeated calls.

validate(data: D | dict[str, Any], schema: type[D]) → None¶: Protocol match: Validate data against schema-derived JSON Schema.

validate_and_construct(data: D | dict[str, Any], schema: type[D]) → D¶: Validate and construct dataclass instance.