def test_unsupported_dtype_gives_error():
reader = sample_reader()
with pytest.raises(ValueError) as excinfo:
# Remember, it won't yield anything if we don't "next" it
next(read_csv_records(reader, dtypes={"something": "no-such-dtype"}))
assert "Unrecognized dtype 'no-such-dtype'" in str(excinfo.value)
def test_basic_read_csv_records():
"""The base test of read_csv_records - no customisation.
"""
reader = sample_reader()
result = list(read_csv_records(reader))
assert result == [
{
"time": 100,
"col1": 1,
"col2": 10,
"col3": 1.0,
"col4": "abcd",
"col5": True,
"col6": None,
},
{
"time": 200,
"col1": 2,
"col2": 20,
"col3": 2.0,
"col4": "efgh",
"col5": False,
"col6": None,
},
]
def test_dtypes_plus_converters_change_parsing():
reader = sample_reader()
result = list(
read_csv_records(
reader, dtypes={"col1": "str", "col6": "str",}, converters={"col2": float,}
)
)
assert result == [
{
"time": 100,
"col1": "0001",
"col2": 10.0,
"col3": 1.0,
"col4": "abcd",
"col5": True,
"col6": "none",
},
{
"time": 200,
"col1": "0002",
"col2": 20.0,
"col3": 2.0,
"col4": "efgh",
"col5": False,
"col6": "",
},
]
def _read_csv_file(self,
file_like,
dialect=csv.excel,
columns=None,
encoding="utf-8",
dtypes=None,
converters=None,
**kwargs):
if columns is None:
reader = csv_dict_record_reader(file_like, encoding, dialect)
else:
reader = csv_text_record_reader(file_like, encoding, dialect,
columns)
return read_csv_records(reader, dtypes, converters, **kwargs)
def test_guess_dtype_gives_default_result():
reader = sample_reader()
result = list(
read_csv_records(
reader,
dtypes={
"time": "guess",
"col1": "guess",
"col2": "guess",
"col3": "guess",
"col4": "guess",
"col5": "guess",
"col6": "guess",
},
)
)
assert result == [
{
"time": 100,
"col1": 1,
"col2": 10,
"col3": 1.0,
"col4": "abcd",
"col5": True,
"col6": None,
},
{
"time": 200,
"col1": 2,
"col2": 20,
"col3": 2.0,
"col4": "efgh",
"col5": False,
"col6": None,
},
]
def test_dtypes_overridden_by_converters():
reader = sample_reader()
result = list(
read_csv_records(
reader,
dtypes={
"time": "bool", # overridden by converters
"col1": "str",
"col2": "int", # overridden by converters
"col6": "str",
},
converters={"time": int, "col2": float, "col5": str,},
)
)
assert result == [
{
"time": 100,
"col1": "0001",
"col2": 10.0,
"col3": 1.0,
"col4": "abcd",
"col5": "true",
"col6": "none",
},
{
"time": 200,
"col1": "0002",
"col2": 20.0,
"col3": 2.0,
"col4": "efgh",
"col5": "false",
"col6": "",
},
]