def test_metadata_schema(
vocabulary_clear, full_input_record, full_record,
minimal_input_record, minimal_record):
"""Test metadata schema."""
# Test full attributes
data = MetadataSchemaV1().load(full_input_record)
assert data == full_record
# Test minimal attributes
data = MetadataSchemaV1().load(minimal_input_record)
assert data == minimal_record
def test_publication_date(
vocabulary_clear, minimal_input_record, minimal_record):
def assert_publication_dates(data, expected):
assert data['publication_date'] == expected_record['publication_date']
assert (
data['_publication_date_search'] ==
expected_record['_publication_date_search']
)
expected_record = minimal_record
# Partial
minimal_input_record['publication_date'] = '2020-02'
expected_record['publication_date'] = '2020-02'
expected_record['_publication_date_search'] = '2020-02-01'
data = MetadataSchemaV1().load(minimal_input_record)
assert_publication_dates(data, expected_record)
# Interval (asymmetrical is allowed!)
minimal_input_record['publication_date'] = '2020-02-02/2025-12'
expected_record['publication_date'] = '2020-02-02/2025-12'
expected_record['_publication_date_search'] = '2020-02-02'
data = MetadataSchemaV1().load(minimal_input_record)
assert_publication_dates(data, expected_record)
# Invalid date
minimal_input_record['publication_date'] = 'invalid'
with pytest.raises(ValidationError):
data = MetadataSchemaV1().load(minimal_input_record)
# Invalid interval
minimal_input_record['publication_date'] = '2025-12/2020-02-02'
with pytest.raises(ValidationError):
data = MetadataSchemaV1().load(minimal_input_record)
def test_identifiers(minimal_input_record):
"""Test Identifiers field."""
# No 'identifiers' field at all is supported
data = MetadataSchemaV1().load(minimal_input_record)
assert data.get('identifiers') == minimal_input_record.get('identifiers')
# Empty dict
minimal_input_record['identifiers'] = {}
data = MetadataSchemaV1().load(minimal_input_record)
assert data['identifiers'] == minimal_input_record['identifiers']
# Minimal
minimal_input_record['identifiers'] = {
"DOI": "10.5281/zenodo.9999999",
}
data = MetadataSchemaV1().load(minimal_input_record)
assert data['identifiers'] == minimal_input_record['identifiers']
# Different schemes
minimal_input_record['identifiers'] = {
"DOI": "10.5281/zenodo.9999999",
"ARK": "ark:/123/456",
}
data = MetadataSchemaV1().load(minimal_input_record)
assert data['identifiers'] == minimal_input_record['identifiers']
# With duplicate schemes, only last one is picked
minimal_input_record['identifiers'] = {
"DOI": "10.5281/zenodo.9999999",
"DOI": "10.5281/zenodo.0000000",
}
data = MetadataSchemaV1().load(minimal_input_record)
assert data['identifiers'] == minimal_input_record['identifiers']
assert data['identifiers']['DOI'] == "10.5281/zenodo.0000000"
# Invalid: no identifier
minimal_input_record['identifiers'] = {
"DOI": ""
}
with pytest.raises(ValidationError):
data = MetadataSchemaV1().load(minimal_input_record)
# Invalid: no scheme
minimal_input_record['identifiers'] = {
"": "10.5281/zenodo.9999999"
}
with pytest.raises(ValidationError):
data = MetadataSchemaV1().load(minimal_input_record)
def test_extensions(db, minimal_record):
"""Test MetadataSchemaV1 dump() for 'extensions' field.
Right now we are going for a schema level test which turns out to still be
heavy, so something fundamental is lacking.
"""
data = {
'extensions': {
'dwc:family': 'Felidae',
'dwc:behavior': 'Plays with yarn, sleeps in cardboard box.',
'nubiomed:right_or_wrong': True,
}
}
minimal_record.update(data)
record = Record.create(minimal_record)
db.session.commit()
serialized_record = MetadataSchemaV1().dump(record) # returns MarshmalDict
assert serialized_record['extensions'] == data['extensions']
def test_dumping_empty_record():
empty_record = dump_empty(MetadataSchemaV1())
assert empty_record == {
'_access': {'files_restricted': None, 'metadata_restricted': None},
'_default_preview': None,
'_communities': None,
'_contact': None,
'_created_by': None,
'_embargo_date': None,
'_files': [
{
'bucket': None,
'checksum': None,
'key': None,
'links': None,
'size': None,
'type': None
}
],
'_internal_notes': [
{
'user': None,
'timestamp': None,
'note': None
}
],
'_owners': [None],
'access_right': None,
'contributors': [
{
'affiliations': [
{
'name': None,
'identifiers': None,
}
],
'family_name': None,
'given_name': None,
'identifiers': None,
'name': None,
'role': None,
'type': None,
}
],
'creators': [
{
'affiliations': [
{
'name': None,
'identifiers': None,
}
],
'family_name': None,
'given_name': None,
'identifiers': None,
'name': None,
'type': None,
}
],
'dates': [
{
'type': None,
'end': None,
'description': None,
'start': None
}
],
'extensions': None,
'descriptions': [
{
'type': None,
'lang': None,
'description': None
}
],
'language': None,
'locations': [
{
'description': None,
'point': {
'lon': None,
'lat': None
},
'place': None
}
],
'licenses': [
{
'identifier': None,
'scheme': None,
'uri': None,
'license': None
}
],
'version': None,
'publication_date': None,
'references': [
{
'scheme': None,
'reference_string': None,
'identifier': None
}
],
'related_identifiers': [
{
'resource_type': {
'subtype': None,
'type': None
},
'scheme': None,
'relation_type': None,
'identifier': None
}
],
'resource_type': {
'subtype': None,
'type': None
},
'subjects': [{'subject': None, 'identifier': None, 'scheme': None}],
'titles': [
{
'type': None,
'lang': None,
'title': None
}
],
# TODO: Investigate the impact of these 2 fields on
# frontend to backend to frontend flow
'identifiers': None,
'recid': None
}
def test_embargo_date(vocabulary_clear, minimal_input_record):
# Test embargo validation
minimal_input_record["embargo_date"] = "1000-01-01"
with pytest.raises(ValidationError):
data = MetadataSchemaV1().load(minimal_input_record)