feat: support date data type for to_datetime() (#1902)

* feat: support date data type for to_datetime()

* add test preconditions

* fix test

* fix lint

* fix tests

* fix lint

* fix test

* update docstring

* update docstring

Author: sycai

bigframes/core/tools/datetimes.py

@@ -13,7 +13,7 @@
 # limitations under the License.
 
 from collections.abc import Mapping
-from datetime import datetime
+from datetime import date, datetime
 from typing import Optional, Union
 
 import bigframes_vendored.constants as constants
@@ -28,7 +28,7 @@
 
 def to_datetime(
     arg: Union[
-        Union[int, float, str, datetime],
+        Union[int, float, str, datetime, date],
         vendored_pandas_datetimes.local_iterables,
         bigframes.series.Series,
         bigframes.dataframe.DataFrame,
@@ -38,7 +38,7 @@ def to_datetime(
     format: Optional[str] = None,
     unit: Optional[str] = None,
 ) -> Union[pd.Timestamp, datetime, bigframes.series.Series]:
-    if isinstance(arg, (int, float, str, datetime)):
+    if isinstance(arg, (int, float, str, datetime, date)):
         return pd.to_datetime(
             arg,
             utc=utc,
@@ -62,7 +62,11 @@ def to_datetime(
             f"Unit parameter is not supported for non-numerical input types. {constants.FEEDBACK_LINK}"
         )
 
-    if arg.dtype in (bigframes.dtypes.TIMESTAMP_DTYPE, bigframes.dtypes.DATETIME_DTYPE):
+    if arg.dtype in (
+        bigframes.dtypes.TIMESTAMP_DTYPE,
+        bigframes.dtypes.DATETIME_DTYPE,
+        bigframes.dtypes.DATE_DTYPE,
+    ):
         to_type = (
             bigframes.dtypes.TIMESTAMP_DTYPE if utc else bigframes.dtypes.DATETIME_DTYPE
         )

bigframes/operations/datetime_ops.py

@@ -50,6 +50,7 @@ def output_type(self, *input_types: dtypes.ExpressionType) -> dtypes.ExpressionT
             dtypes.FLOAT_DTYPE,
             dtypes.INT_DTYPE,
             dtypes.STRING_DTYPE,
+            dtypes.DATE_DTYPE,
         ):
             raise TypeError("expected string or numeric input")
         return pd.ArrowDtype(pa.timestamp("us", tz=None))
@@ -67,6 +68,7 @@ def output_type(self, *input_types: dtypes.ExpressionType) -> dtypes.ExpressionT
             dtypes.FLOAT_DTYPE,
             dtypes.INT_DTYPE,
             dtypes.STRING_DTYPE,
+            dtypes.DATE_DTYPE,
         ):
             raise TypeError("expected string or numeric input")
         return pd.ArrowDtype(pa.timestamp("us", tz="UTC"))

tests/system/small/operations/test_datetimes.py

@@ -13,12 +13,15 @@
 # limitations under the License.
 
 import datetime
+import typing
 
 import numpy
+from packaging import version
 from pandas import testing
 import pandas as pd
 import pytest
 
+import bigframes.pandas as bpd
 import bigframes.series
 from bigframes.testing.utils import assert_series_equal
 
@@ -548,3 +551,23 @@ def test_timedelta_dt_accessors_on_wrong_type_raise_exception(scalars_dfs, acces
 
     with pytest.raises(TypeError):
         access(bf_df["timestamp_col"])
+
+
+@pytest.mark.parametrize(
+    "col",
+    # TODO(b/431276706) test timestamp_col too.
+    ["date_col", "datetime_col"],
+)
+def test_to_datetime(scalars_dfs, col):
+    if version.Version(pd.__version__) <= version.Version("2.1.0"):
+        pytest.skip("timezone conversion bug")
+    bf_df, pd_df = scalars_dfs
+
+    actual_result = typing.cast(
+        bigframes.series.Series, bpd.to_datetime(bf_df[col])
+    ).to_pandas()
+
+    expected_result = pd.Series(pd.to_datetime(pd_df[col]))
+    testing.assert_series_equal(
+        actual_result, expected_result, check_dtype=False, check_index_type=False
+    )

third_party/bigframes_vendored/pandas/core/tools/datetimes.py

@@ -1,17 +1,22 @@
 # Contains code from https://github.com/pandas-dev/pandas/blob/main/pandas/core/tools/datetimes.py
 
-from datetime import datetime
+from datetime import date, datetime
 from typing import List, Mapping, Tuple, Union
 
 import pandas as pd
 
-from bigframes import constants, series
+from bigframes import constants, dataframe, series
 
 local_iterables = Union[List, Tuple, pd.Series, pd.DataFrame, Mapping]
 
 
 def to_datetime(
-    arg,
+    arg: Union[
+        Union[int, float, str, datetime, date],
+        local_iterables,
+        series.Series,
+        dataframe.DataFrame,
+    ],
     *,
     utc=False,
     format=None,
@@ -58,7 +63,7 @@ def to_datetime(
         dtype: timestamp[us, tz=UTC][pyarrow]
 
     Args:
-        arg (int, float, str, datetime, list, tuple, 1-d array, Series):
+        arg (int, float, str, datetime, date, list, tuple, 1-d array, Series):
             The object to convert to a datetime.
         utc (bool, default False):
             Control timezone-related parsing, localization and conversion. If True, the