def test_minimal_study_to_csv():
session = make_session()
patient_1 = Patient(DateOfBirth="1900-01-01", Sex="M")
patient_2 = Patient(DateOfBirth="1900-01-01", Sex="F")
session.add_all([patient_1, patient_2])
session.commit()
study = StudyDefinition(population=patients.all(), sex=patients.sex())
with tempfile.NamedTemporaryFile(mode="w+") as f:
study.to_csv(f.name)
results = list(csv.DictReader(f))
assert results == [
{"patient_id": str(patient_1.Patient_ID), "sex": "M"},
{"patient_id": str(patient_2.Patient_ID), "sex": "F"},
]
def test_sqlcmd_and_odbc_outputs_match():
session = make_session()
patient = Patient(DateOfBirth="1950-01-01")
patient.CodedEvents.append(
CodedEvent(CTV3Code="XYZ", NumericValue=50, ConsultationDate="2002-06-01")
)
session.add(patient)
session.commit()
study = StudyDefinition(
population=patients.with_these_clinical_events(codelist(["XYZ"], "ctv3"))
)
with tempfile.NamedTemporaryFile() as input_csv_odbc, tempfile.NamedTemporaryFile() as input_csv_sqlcmd:
# windows line endings
study.to_csv(input_csv_odbc.name, with_sqlcmd=False)
# unix line endings
study.to_csv(input_csv_sqlcmd.name, with_sqlcmd=True)
assert filecmp.cmp(input_csv_odbc.name, input_csv_sqlcmd.name, shallow=False)
def test_duplicate_id_checking():
study = StudyDefinition(population=patients.all())
# A bit of a hack: overwrite the queries we're going to run with a query which
# deliberately returns duplicate values
study.queries = [
(
"dummy_query",
"""
SELECT * FROM (
VALUES
(1,1),
(2,2),
(3,3),
(1,4)
) t (patient_id, foo)
""",
)
]
with pytest.raises(RuntimeError):
study.to_dicts()
with pytest.raises(RuntimeError):
with tempfile.NamedTemporaryFile(mode="w+") as f:
study.to_csv(f.name)
