fix(python): Ensure the cs.temporal() selector uses wildcard time zone matching for Datetime

py-polars/polars/datatypes/classes.py

@@ -422,8 +422,8 @@ class Datetime(TemporalType):
     time_zone
         Time zone string, as defined in zoneinfo (to see valid strings run
         `import zoneinfo; zoneinfo.available_timezones()` for a full list).
-        When using to match dtypes, can use "*" to check for Datetime columns
-        that have any timezone.
+        When used to match dtypes, can set this to "*" to check for Datetime
+        columns that have any (non-null) timezone.
 
     Notes
     -----

py-polars/polars/datatypes/constants.py

@@ -36,6 +36,9 @@
         Datetime("ms"),
         Datetime("us"),
         Datetime("ns"),
+        Datetime("ms", "*"),
+        Datetime("us", "*"),
+        Datetime("ns", "*"),
     ]
 )
 DURATION_DTYPES: frozenset[PolarsDataType] = DataTypeGroup(

py-polars/tests/unit/series/buffers/test_from_buffers.py

@@ -9,6 +9,17 @@
 from polars.testing import assert_series_equal
 from polars.testing.parametric import series
 
+# TODO: Define data type groups centrally somewhere in the test suite
+DATETIME_DTYPES: set[pl.PolarsDataType] = {
+    pl.Datetime,
+    pl.Datetime("ms"),
+    pl.Datetime("us"),
+    pl.Datetime("ns"),
+}
+TEMPORAL_DTYPES: set[pl.PolarsDataType] = (
+    {pl.Date, pl.Time} | pl.DURATION_DTYPES | DATETIME_DTYPES
+)
+
 
 @given(
     s=series(
@@ -34,7 +45,7 @@ def test_series_from_buffers_numeric(s: pl.Series) -> None:
     assert_series_equal(s, result)
 
 
-@given(s=series(allowed_dtypes=pl.TEMPORAL_DTYPES, chunked=False))
+@given(s=series(allowed_dtypes=TEMPORAL_DTYPES, chunked=False))
 def test_series_from_buffers_temporal_with_validity(s: pl.Series) -> None:
     validity = s.is_not_null()
     physical = pl.Int32 if s.dtype == pl.Date else pl.Int64

py-polars/tests/unit/test_selectors.py

@@ -1,12 +1,22 @@
+import sys
+from datetime import datetime
 from typing import Any
 
 import pytest
 
 import polars as pl
 import polars.selectors as cs
+from polars.dependencies import _ZONEINFO_AVAILABLE
 from polars.selectors import expand_selector, is_selector
 from polars.testing import assert_frame_equal
 
+if sys.version_info >= (3, 9):
+    from zoneinfo import ZoneInfo
+elif _ZONEINFO_AVAILABLE:
+    # Import from submodule due to typing issue with backports.zoneinfo package:
+    # https://github.com/pganssle/zoneinfo/issues/125
+    from backports.zoneinfo._zoneinfo import ZoneInfo
+
 
 def assert_repr_equals(item: Any, expected: str) -> None:
     """Assert that the repr of an item matches the expected string."""
@@ -356,6 +366,33 @@ def test_selector_temporal(df: pl.DataFrame) -> None:
     assert df.select(cs.date() | cs.time()).schema == {"ghi": pl.Time, "JJK": pl.Date}
 
 
+def test_selector_temporal_13665() -> None:
+    df = pl.DataFrame(
+        data={"utc": [datetime(1950, 7, 5), datetime(2099, 12, 31)]},
+        schema={"utc": pl.Datetime(time_zone="UTC")},
+    ).with_columns(
+        idx=pl.int_range(0, 2),
+        utc=pl.col("utc").dt.replace_time_zone(None),
+        tokyo=pl.col("utc").dt.convert_time_zone("Asia/Tokyo"),
+        hawaii=pl.col("utc").dt.convert_time_zone("US/Hawaii"),
+    )
+    for selector in (cs.datetime(), cs.datetime("us"), cs.temporal()):
+        assert df.select(selector).to_dict(as_series=False) == {
+            "utc": [
+                datetime(1950, 7, 5, 0, 0),
+                datetime(2099, 12, 31, 0, 0),
+            ],
+            "tokyo": [
+                datetime(1950, 7, 5, 10, 0, tzinfo=ZoneInfo(key="Asia/Tokyo")),
+                datetime(2099, 12, 31, 9, 0, tzinfo=ZoneInfo(key="Asia/Tokyo")),
+            ],
+            "hawaii": [
+                datetime(1950, 7, 4, 14, 0, tzinfo=ZoneInfo(key="US/Hawaii")),
+                datetime(2099, 12, 30, 14, 0, tzinfo=ZoneInfo(key="US/Hawaii")),
+            ],
+        }
+
+
 def test_selector_expansion() -> None:
     df = pl.DataFrame({name: [] for name in "abcde"})