def test_wrong_ints(self, no_header_schema, test_input):
with pytest.raises(ParseErrors) as e:
parse_csv([[test_input, "Foo", "true"]],
json.loads(no_header_schema))
assert e.errors == [{
"row":
0,
"column":
"0",
"message":
f"invalid literal for int() with base 10: '{test_input}'",
}]
def test_wrong_floats(self, boligpriser_schema, boligpriser_header,
test_input):
with pytest.raises(ParseErrors) as e:
parse_csv(
[["0001", "Trøndelag", test_input]],
json.loads(boligpriser_schema),
header=boligpriser_header,
)
assert e.errors == [{
"row":
0,
"column":
"pris",
"message":
f"could not convert string to float: '{test_input}'",
}]
def test_incorrect_date_colum(self, dates_header, dates_schema):
csv_data = parse_csv(
[["1", "2020", "garbish data", "2020-01-01T12:01:01"]],
dates_schema,
header=dates_header,
)
validation_errors = JsonSchemaValidator(dates_schema).validate(
csv_data)
assert len(validation_errors) == 1
def test_incorrect_year_colum(self, dates_header, dates_schema):
invalid_years = ["abc", ""]
for invalid_year in invalid_years:
csv_data = parse_csv(
[["1", invalid_year, "2020-12-30", "2020-12-01T12:01:01"]],
dates_schema,
header=dates_header,
)
validation_errors = JsonSchemaValidator(dates_schema).validate(
csv_data)
assert len(validation_errors) == 1
def test_valid_year_colum(self, dates_header, dates_schema):
valid_years = ["2020", "-100", "9999"]
for valid_year in valid_years:
csv_data = parse_csv(
[["1", valid_year, "2020-12-30", "2020-12-01T12:01:01"]],
dates_schema,
header=dates_header,
)
validation_errors = JsonSchemaValidator(dates_schema).validate(
csv_data)
assert len(validation_errors) == 0
def validate_csv(event, context):
config = Config.from_lambda_event(event)
step_config = StepConfig.from_task_config(config.task_config)
s3_prefix = config.payload.output_dataset.s3_prefix
log_add(
header_row=step_config.header_row,
delimiter=step_config.delimiter,
quote=step_config.quote,
schema=step_config.schema,
output_prefix=s3_prefix,
)
if not step_config.schema:
log_add(notice="No Schema provided for validation")
config.payload.step_data.status = Status.VALIDATION_SUCCESS.value
# 2020.06: Validation done optionally - we now return ok if we don't supply a
# schema for the validation step
return asdict(config.payload.step_data)
input_prefix = next(
iter(config.payload.step_data.s3_input_prefixes.values()))
log_add(s3_input_prefix=input_prefix)
objects = s3.list_objects_v2(Bucket=BUCKET, Prefix=input_prefix)
s3_path = next(iter(objects["Contents"]))["Key"]
log_add(s3_input_path=s3_path)
response = s3.get_object(Bucket=BUCKET, Key=s3_path)
reader = csv.reader(
string_reader.from_response(response),
dialect="unix",
delimiter=step_config.delimiter,
quotechar=step_config.quote,
)
header = None
if step_config.header_row:
header = next(reader)
try:
csv_data = parse_csv(reader, step_config.schema, header)
except ParseErrors as p:
return _with_error(config, p.errors)
validation_errors = JsonSchemaValidator(
step_config.schema).validate(csv_data)
if validation_errors:
return _with_error(config, errors=validation_errors)
config.payload.step_data.status = Status.VALIDATION_SUCCESS.value
return asdict(config.payload.step_data)
def test_correct_dates(self, dates_header, dates_schema):
csv_data = parse_csv(
[
["1", "2020", "2020-01-01", "2020-01-01T12:01:01"],
["1", "2020", "2020-01-01", "2020-01-01 12:01:01"],
["1", "2020", "2020-01-01", "2020-12-01T12-01"],
],
dates_schema,
header=dates_header,
)
validation_errors = JsonSchemaValidator(dates_schema).validate(
csv_data)
assert len(validation_errors) == 0
def test_valid_date_time_colum(self, dates_header, dates_schema):
valid_dates = [
"0009-12-01T12:01:01",
"2020-12-12T12:01:01",
]
for valid_date in valid_dates:
csv_data = parse_csv(
[["1", "2020", "2020-12-30", valid_date]],
dates_schema,
header=dates_header,
)
validation_errors = JsonSchemaValidator(dates_schema).validate(
csv_data)
assert len(validation_errors) == 0
def test_incorrect_date_time_colum(self, dates_header, dates_schema):
invalid_dates = [
"2020-13-01T12:01:01",
"2020 12 32T12:01:01",
"garbish data",
]
for invalid_date in invalid_dates:
csv_data = parse_csv(
[["1", "2020", "2020-12-30", invalid_date]],
dates_schema,
header=dates_header,
)
validation_errors = JsonSchemaValidator(dates_schema).validate(
csv_data)
assert len(validation_errors) == 1
def test_parse_no_headers(self, no_header_schema):
data = parse_csv(
[["120", "Foo", "true"], ["999199", "Bar", "false"]],
json.loads(no_header_schema),
)
assert data == [
{
"0": 120,
"1": "Foo",
"2": True
},
{
"0": 999_199,
"1": "Bar",
"2": False
},
]
def test_parse_with_headers(self, boligpriser_schema, boligpriser_header):
data = parse_csv(
[
["001", "Østre byflak", "1010.01", "true"],
["002", "Hønse-Lovisaløkka", "5001,10", "false"],
],
json.loads(boligpriser_schema),
header=boligpriser_header,
)
assert data == [
{
"delbydel_id": "001",
"navn": "Østre byflak",
"pris": 1010.01,
"til_salg": True,
},
{
"delbydel_id": "002",
"navn": "Hønse-Lovisaløkka",
"pris": 5001.10,
"til_salg": False,
},
]
def test_empty_schema():
data = parse_csv([["1", "foo"], ["2", "bar"]], {})
assert data == [["1", "foo"], ["2", "bar"]]
def test_simple_array():
data = parse_csv([["1", "foo"], ["2", "bar"]], {"type": "array"})
assert data == [["1", "foo"], ["2", "bar"]]
def test_parse_empty_values(self, no_header_schema):
data = parse_csv([["55", "", "true"]], json.loads(no_header_schema))
assert data == [{"0": 55, "2": True}]