strtobool support for Python 3.12 (#880)

* pyiceberg strtobool function to replace distutil strtobool function

* unit_test for strtobool

* Update pyiceberg/types.py

Co-authored-by: Fokko Driesprong <[email protected]>

* revamp unit_test

* revamp unit_test

---------

Co-authored-by: Fokko Driesprong <[email protected]>

Author: MehulBatra

pyiceberg/types.py

@@ -92,6 +92,22 @@ def _parse_fixed_type(fixed: Any) -> int:
         return fixed
 
 
+def strtobool(val: str) -> bool:
+    """Convert a string representation of truth to true (1) or false (0).
+
+    True values are 'y', 'yes', 't', 'true', 'on', and '1'; false values
+    are 'n', 'no', 'f', 'false', 'off', and '0'.  Raises ValueError if
+    'val' is anything else.
+    """
+    val = val.lower()
+    if val in ("y", "yes", "t", "true", "on", "1"):
+        return True
+    elif val in ("n", "no", "f", "false", "off", "0"):
+        return False
+    else:
+        raise ValueError(f"Invalid truth value: {val!r}")
+
+
 class IcebergType(IcebergBaseModel):
     """Base type for all Iceberg Types.
 

pyiceberg/utils/config.py

@@ -16,12 +16,12 @@
 # under the License.
 import logging
 import os
-from distutils.util import strtobool
 from typing import List, Optional
 
 import strictyaml
 
 from pyiceberg.typedef import UTF8, FrozenDict, RecursiveDict
+from pyiceberg.types import strtobool
 
 PYICEBERG = "pyiceberg_"
 DEFAULT = "default"

tests/test_types.py

@@ -43,6 +43,7 @@
     TimestamptzType,
     TimeType,
     UUIDType,
+    strtobool,
 )
 
 non_parameterized_types = [
@@ -630,3 +631,21 @@ def test_deepcopy_of_singleton_fixed_type() -> None:
 
     for lhs, rhs in zip(list_of_fixed_types, copied_list):
         assert id(lhs) == id(rhs)
+
+
+def test_strtobool() -> None:
+    # Values that should return True
+    true_values = ["y", "yes", "t", "true", "on", "1"]
+    for val in true_values:
+        assert strtobool(val) is True, f"Expected True for value: {val}"
+
+    # Values that should return False
+    false_values = ["n", "no", "f", "false", "off", "0"]
+    for val in false_values:
+        assert strtobool(val) is False, f"Expected False for value: {val}"
+
+    # Values that should raise ValueError
+    invalid_values = ["maybe", "2", "trueish", "falseish", "", " "]
+    for val in invalid_values:
+        with pytest.raises(ValueError, match=f"Invalid truth value: {val!r}"):
+            strtobool(val)