def test_validity():
input_schema = catalog.get_schema(
URI('https://odp.saeon.ac.za/schema/metadata/saeon/iso19115'))
input_json = catalog.load_json(
URI('https://odp.saeon.ac.za/schema/metadata/saeon/iso19115-example'))
output_schema = catalog.get_schema(
URI('https://odp.saeon.ac.za/schema/metadata/saeon/datacite4'))
output_json = catalog.load_json(
URI('https://odp.saeon.ac.za/schema/metadata/saeon/datacite4-example-translated'
))
assert input_schema.validate().valid
assert input_schema.evaluate(JSON(input_json)).valid
assert output_schema.validate().valid
assert output_schema.evaluate(JSON(output_json)).valid
def test_translate_iso19115_to_datacite():
input_schema = catalog.get_schema(
URI('https://odp.saeon.ac.za/schema/metadata/saeon/iso19115'))
input_json = catalog.load_json(
URI('https://odp.saeon.ac.za/schema/metadata/saeon/iso19115-example.json'
))
output_json = catalog.load_json(
URI('https://odp.saeon.ac.za/schema/metadata/saeon/datacite4-example-translated.json'
))
result = input_schema.evaluate(JSON(input_json))
patch = result.output('patch', scheme='saeon/datacite4')
translation = result.output('translation', scheme='saeon/datacite4')
assert JSONPatch(*patch).evaluate(None) == translation
# work in progress
# assert translation == output_json
assert translation.keys() == output_json.keys()
for k in translation:
if k == 'contributors':
# todo: resolve leftover empty arrays/objects when there are
# no source values to fill them
continue
assert translation[k] == output_json[k]
async def list_catalogs(paginator: Paginator = Depends(), ):
return paginator.paginate(
select(Catalog), lambda row: CatalogModel(
id=row.Catalog.id,
schema_id=row.Catalog.schema_id,
schema_uri=row.Catalog.schema.uri,
schema_=schema_catalog.get_schema(URI(row.Catalog.schema.uri)).
value,
))
async def list_tags(paginator: Paginator = Depends(), ):
return paginator.paginate(
select(Tag), lambda row: TagModel(
id=row.Tag.id,
cardinality=row.Tag.cardinality,
public=row.Tag.public,
scope_id=row.Tag.scope_id,
schema_id=row.Tag.schema_id,
schema_uri=row.Tag.schema.uri,
schema_=schema_catalog.get_schema(URI(row.Tag.schema.uri)).value,
))
async def get_tag(tag_id: str, ):
tag = Session.execute(
select(Tag).where(Tag.id == tag_id)).scalar_one_or_none()
if not tag:
raise HTTPException(HTTP_404_NOT_FOUND)
return TagModel(
id=tag.id,
cardinality=tag.cardinality,
public=tag.public,
scope_id=tag.scope_id,
schema_id=tag.schema_id,
schema_uri=tag.schema.uri,
schema_=schema_catalog.get_schema(URI(tag.schema.uri)).value,
)
async def get_schema(
schema_id: str,
):
schema = Session.execute(
select(Schema).
where(Schema.id == schema_id)
).scalar_one_or_none()
if not schema:
raise HTTPException(HTTP_404_NOT_FOUND)
return SchemaModel(
id=schema.id,
type=schema.type,
uri=schema.uri,
schema_=schema_catalog.get_schema(URI(schema.uri)).value,
)
async def list_schemas(
schema_type: SchemaType = None,
paginator: Paginator = Depends(),
):
stmt = select(Schema)
if schema_type:
stmt = stmt.where(Schema.type == schema_type)
return paginator.paginate(
stmt,
lambda row: SchemaModel(
id=row.Schema.id,
type=row.Schema.type,
uri=row.Schema.uri,
schema_=schema_catalog.get_schema(URI(row.Schema.uri)).value,
)
)
def _evaluate_record(self, record_id: str, timestamp: datetime) -> bool:
"""Evaluate a record model (API) against the publication schema for
a catalog, and commit the result to the catalog_record table.
The catalog_record entry is stamped with the `timestamp` of the latest
contributing change (from catalog/record/record_tag/collection_tag).
"""
catalog = Session.get(Catalog, self.catalog_id)
record = Session.get(Record, record_id)
catalog_record = (Session.get(CatalogRecord, (self.catalog_id, record_id)) or
CatalogRecord(catalog_id=self.catalog_id, record_id=record_id))
record_model = output_record_model(record)
record_json = JSON(record_model.dict())
publication_schema = schema_catalog.get_schema(URI(catalog.schema.uri))
if (result := publication_schema.evaluate(record_json)).valid:
catalog_record.validity = result.output('flag')
catalog_record.published = True
catalog_record.published_record = self._create_published_record(record_model).dict()
self._save_published_doi(record_model)
from fastapi import HTTPException
from jschon import JSONSchema, URI
from sqlalchemy import select
from starlette.status import HTTP_422_UNPROCESSABLE_ENTITY
from odp.api.models import RecordModelIn, TagInstanceModelIn
from odp.db import Session
from odp.db.models import Schema, SchemaType, Tag
from odp.lib.schema import schema_catalog
async def get_tag_schema(tag_instance_in: TagInstanceModelIn) -> JSONSchema:
if not (tag := Session.execute(
select(Tag).where(
Tag.id == tag_instance_in.tag_id)).scalar_one_or_none()):
raise HTTPException(HTTP_422_UNPROCESSABLE_ENTITY, 'Invalid tag id')
schema = Session.get(Schema, (tag.schema_id, SchemaType.tag))
return schema_catalog.get_schema(URI(schema.uri))
async def get_metadata_schema(record_in: RecordModelIn) -> JSONSchema:
if not (schema := Session.get(Schema,
(record_in.schema_id, SchemaType.metadata))):
raise HTTPException(HTTP_422_UNPROCESSABLE_ENTITY, 'Invalid schema id')
return schema_catalog.get_schema(URI(schema.uri))
@router.get(
'/{catalog_id}',
response_model=CatalogModel,
dependencies=[Depends(Authorize(ODPScope.CATALOG_READ))],
)
async def get_catalog(catalog_id: str, ):
if not (catalog := Session.get(Catalog, catalog_id)):
raise HTTPException(HTTP_404_NOT_FOUND)
return CatalogModel(
id=catalog.id,
schema_id=catalog.schema_id,
schema_uri=catalog.schema.uri,
schema_=schema_catalog.get_schema(URI(catalog.schema.uri)).value,
)
@router.get(
'/{catalog_id}/records',
response_model=Page[PublishedRecordModel],
dependencies=[Depends(Authorize(ODPScope.CATALOG_READ))],
)
async def list_published_records(
catalog_id: str,
paginator: Paginator = Depends(),
):
if not Session.get(Catalog, catalog_id):
raise HTTPException(HTTP_404_NOT_FOUND)
