def from_string(s, all_types):
"""
Parse string as condition object.
Parameters
----------
s: str
String to parse.
all_types: Dict[str, pyarrow.DataType]
Mapping from all known columns to pyarrow types.
Returns
-------
condition: Condition
Parsed condition.
Raises
------
ValueError: If condition cannot be parsed.
"""
m = re.match(
pattern=r"""
^ # anchor
\s* # optional space
\(? # optional open bracket
\s* # optional space
([^!<>=\s]+) # column name
\s* # optional space
(==|=|<=|<|>=|>|!=) # operator
([^)=]+) # value
\)? # optional closing bracket
\s* # optional space
$ # anchor
""",
string=s,
flags=re.VERBOSE,
)
if not m:
raise ValueError('Cannot parse condition "{s}"'.format(s=s))
col, op, var = m.groups()
col_obj = C(col)
pa_type = all_types.get(col)
if pa_type is None:
raise ValueError(
'Unknown column "{col}" in condition "{s}"'.format(col=col,
s=s))
var_f = get_str_to_python_converter(pa_type)
var_obj = var_f(var.strip())
if (op == "==") or (op == "="):
return col_obj == var_obj
elif op == "<=":
return col_obj <= var_obj
elif op == "<":
return col_obj < var_obj
elif op == ">=":
return col_obj >= var_obj
elif op == ">":
return col_obj > var_obj
elif op == "!=":
return col_obj != var_obj
else:
raise RuntimeError("unreachable")
def test_get_str_to_python_converter_unknown_type():
pa_type = pa.struct([("f1", pa.int32()), ("f2", pa.bool_())])
with pytest.raises(ValueError) as exc:
get_str_to_python_converter(pa_type)
assert str(exc.value) == "Cannot handle type struct<f1: int32, f2: bool>"
def test_get_str_to_python_converter_failes(s, pa_type):
converter = get_str_to_python_converter(pa_type)
with pytest.raises(ValueError):
converter(s)
def test_get_str_to_python_converter_ok(s, pa_type, expected):
converter = get_str_to_python_converter(pa_type)
actual = converter(s)
assert actual == expected
assert isinstance(actual, type(expected))