BUG: to_sql fails for pyarrow date dtype (#53854)

doc/source/whatsnew/v2.1.0.rst

@@ -364,12 +364,14 @@ Datetimelike
 - :meth:`DatetimeIndex.map` with ``na_action="ignore"`` now works as expected. (:issue:`51644`)
 - Bug in :class:`DateOffset` which had inconsistent behavior when multiplying a :class:`DateOffset` object by a constant (:issue:`47953`)
 - Bug in :func:`date_range` when ``freq`` was a :class:`DateOffset` with ``nanoseconds`` (:issue:`46877`)
+- Bug in :meth:`DataFrame.to_sql` raising ``ValueError`` for pyarrow-backed date like dtypes (:issue:`53854`)
 - Bug in :meth:`Timestamp.date`, :meth:`Timestamp.isocalendar`, :meth:`Timestamp.timetuple`, and :meth:`Timestamp.toordinal` were returning incorrect results for inputs outside those supported by the Python standard library's datetime module (:issue:`53668`)
 - Bug in :meth:`Timestamp.round` with values close to the implementation bounds returning incorrect results instead of raising ``OutOfBoundsDatetime`` (:issue:`51494`)
 - Bug in :meth:`arrays.DatetimeArray.map` and :meth:`DatetimeIndex.map`, where the supplied callable operated array-wise instead of element-wise (:issue:`51977`)
 - Bug in constructing a :class:`Series` or :class:`DataFrame` from a datetime or timedelta scalar always inferring nanosecond resolution instead of inferring from the input (:issue:`52212`)
 - Bug in parsing datetime strings with weekday but no day e.g. "2023 Sept Thu" incorrectly raising ``AttributeError`` instead of ``ValueError`` (:issue:`52659`)
 
+
 Timedelta
 ^^^^^^^^^
 - :meth:`TimedeltaIndex.map` with ``na_action="ignore"`` now works as expected (:issue:`51644`)

pandas/io/sql.py

@@ -974,10 +974,16 @@ def insert_data(self) -> tuple[list[str], list[np.ndarray]]:
         for i, (_, ser) in enumerate(temp.items()):
             if ser.dtype.kind == "M":
                 if isinstance(ser._values, ArrowExtensionArray):
-                    with warnings.catch_warnings():
-                        warnings.filterwarnings("ignore", category=FutureWarning)
-                        # GH#52459 to_pydatetime will return Index[object]
-                        d = np.asarray(ser.dt.to_pydatetime(), dtype=object)
+                    import pyarrow as pa
+
+                    if pa.types.is_date(ser.dtype.pyarrow_dtype):
+                        # GH#53854 to_pydatetime not supported for pyarrow date dtypes
+                        d = ser._values.to_numpy(dtype=object)
+                    else:
+                        with warnings.catch_warnings():
+                            warnings.filterwarnings("ignore", category=FutureWarning)
+                            # GH#52459 to_pydatetime will return Index[object]
+                            d = np.asarray(ser.dt.to_pydatetime(), dtype=object)
                 else:
                     d = ser._values.to_pydatetime()
             elif ser.dtype.kind == "m":

pandas/tests/io/test_sql.py

@@ -561,6 +561,7 @@ def test_dataframe_to_sql_arrow_dtypes(conn, request):
             "datetime": pd.array(
                 [datetime(2023, 1, 1)], dtype="timestamp[ns][pyarrow]"
             ),
+            "date": pd.array([date(2023, 1, 1)], dtype="date32[day][pyarrow]"),
             "timedelta": pd.array([timedelta(1)], dtype="duration[ns][pyarrow]"),
             "string": pd.array(["a"], dtype="string[pyarrow]"),
         }