DEPR: Deprecate use of un-supported numpy dt64/td64 dtype for pandas.array

doc/source/whatsnew/v2.1.0.rst

@@ -275,6 +275,8 @@ Deprecations
 - Deprecated behavior of :func:`assert_series_equal` and :func:`assert_frame_equal` considering NA-like values (e.g. ``NaN`` vs ``None`` as equivalent) (:issue:`52081`)
 - Deprecated constructing :class:`SparseArray` from scalar data, pass a sequence instead (:issue:`53039`)
 - Deprecated positional indexing on :class:`Series` with :meth:`Series.__getitem__` and :meth:`Series.__setitem__`, in a future version ``ser[item]`` will *always* interpret ``item`` as a label, not a position (:issue:`50617`)
+- Deprecated the use of dt64/td64 dtypes with :func:`pandas.array` (:issue:`53439`)
+-
 
 .. ---------------------------------------------------------------------------
 .. _whatsnew_210.performance:

pandas/core/construction.py

@@ -41,6 +41,7 @@
     maybe_convert_platform,
     maybe_infer_to_datetimelike,
     maybe_promote,
+    _ensure_nanosecond_dtype,
 )
 from pandas.core.dtypes.common import (
     is_list_like,
@@ -109,8 +110,6 @@ def array(
         ============================== =======================================
         :class:`pandas.Interval`       :class:`pandas.arrays.IntervalArray`
         :class:`pandas.Period`         :class:`pandas.arrays.PeriodArray`
-        :class:`datetime.datetime`     :class:`pandas.arrays.DatetimeArray`
-        :class:`datetime.timedelta`    :class:`pandas.arrays.TimedeltaArray`
         :class:`int`                   :class:`pandas.arrays.IntegerArray`
         :class:`float`                 :class:`pandas.arrays.FloatingArray`
         :class:`str`                   :class:`pandas.arrays.StringArray` or
@@ -184,26 +183,6 @@ def array(
     ['a', 'b']
     Length: 2, dtype: str32
 
-    Finally, Pandas has arrays that mostly overlap with NumPy
-
-      * :class:`arrays.DatetimeArray`
-      * :class:`arrays.TimedeltaArray`
-
-    When data with a ``datetime64[ns]`` or ``timedelta64[ns]`` dtype is
-    passed, pandas will always return a ``DatetimeArray`` or ``TimedeltaArray``
-    rather than a ``PandasArray``. This is for symmetry with the case of
-    timezone-aware data, which NumPy does not natively support.
-
-    >>> pd.array(['2015', '2016'], dtype='datetime64[ns]')
-    <DatetimeArray>
-    ['2015-01-01 00:00:00', '2016-01-01 00:00:00']
-    Length: 2, dtype: datetime64[ns]
-
-    >>> pd.array(["1H", "2H"], dtype='timedelta64[ns]')
-    <TimedeltaArray>
-    ['0 days 01:00:00', '0 days 02:00:00']
-    Length: 2, dtype: timedelta64[ns]
-
     Examples
     --------
     If a dtype is not specified, pandas will infer the best dtype from the values.
@@ -295,6 +274,9 @@ def array(
     )
     from pandas.core.arrays.string_ import StringDtype
 
+    if isinstance(dtype, np.dtype) and dtype.kind in "mM":
+        _ensure_nanosecond_dtype(dtype)
+
     if lib.is_scalar(data):
         msg = f"Cannot pass scalar '{data}' to 'pandas.array'."
         raise ValueError(msg)

pandas/tests/arrays/test_array.py

@@ -1,5 +1,6 @@
 import datetime
 import decimal
+import re
 
 import numpy as np
 import pytest
@@ -28,6 +29,20 @@
 )
 
 
+def test_dt64_array():
+    # PR 53439
+    dtype_unit_lst = ["M8[h]", "M8[m]", "m8[h]", "M8[m]"]
+
+    for unit in dtype_unit_lst:
+        dtype_var = np.dtype(unit)
+        msg = (
+            "dtype={} is not supported. Supported resolutions are 's', 'ms', "
+            "'us', and 'ns'".format(dtype_var)
+        )
+        with pytest.raises(TypeError, match=re.escape(msg)):
+            pd.array([], dtype=dtype_var)
+
+
 @pytest.mark.parametrize(
     "data, dtype, expected",
     [
@@ -203,6 +218,8 @@
     ],
 )
 def test_array(data, dtype, expected):
+    with open("/home/richard/Desktop/file.txt", "a+") as fil:
+        fil.write(f"{dtype}\n")
     result = pd.array(data, dtype=dtype)
     tm.assert_equal(result, expected)