Backport PR pandas-dev#47763: BUG: fix regression in Series[string] setitem setting a scalar with a mask

Author: jorisvandenbossche

doc/source/whatsnew/v1.4.4.rst

@@ -15,6 +15,7 @@ including other versions of pandas.
 Fixed regressions
 ~~~~~~~~~~~~~~~~~
 - Fixed regression in :func:`concat` materializing :class:`Index` during sorting even if :class:`Index` was already sorted (:issue:`47501`)
+- Fixed regression in setting ``None`` or non-string value into a ``string``-dtype Series using a mask (:issue:`47628`)
 -
 
 .. ---------------------------------------------------------------------------

pandas/core/arrays/string_.py

@@ -17,6 +17,7 @@
 from pandas._typing import (
     Dtype,
     Scalar,
+    npt,
     type_t,
 )
 from pandas.compat import pa_version_under1p01
@@ -413,6 +414,12 @@ def __setitem__(self, key, value):
 
         super().__setitem__(key, value)
 
+    def _putmask(self, mask: npt.NDArray[np.bool_], value) -> None:
+        # the super() method NDArrayBackedExtensionArray._putmask uses
+        # np.putmask which doesn't properly handle None/pd.NA, so using the
+        # base class implementation that uses __setitem__
+        ExtensionArray._putmask(self, mask, value)
+
     def astype(self, dtype, copy: bool = True):
         dtype = pandas_dtype(dtype)
 

pandas/tests/arrays/string_/test_string.py

@@ -553,3 +553,23 @@ def test_isin(dtype, request, fixed_now_ts):
     result = s.isin(["a", fixed_now_ts])
     expected = pd.Series([True, False, False])
     tm.assert_series_equal(result, expected)
+
+
+def test_setitem_scalar_with_mask_validation(dtype):
+    # https://github.com/pandas-dev/pandas/issues/47628
+    # setting None with a boolean mask (through _putmaks) should still result
+    # in pd.NA values in the underlying array
+    ser = pd.Series(["a", "b", "c"], dtype=dtype)
+    mask = np.array([False, True, False])
+
+    ser[mask] = None
+    assert ser.array[1] is pd.NA
+
+    # for other non-string we should also raise an error
+    ser = pd.Series(["a", "b", "c"], dtype=dtype)
+    if type(ser.array) is pd.arrays.StringArray:
+        msg = "Cannot set non-string value"
+    else:
+        msg = "Scalar must be NA or str"
+    with pytest.raises(ValueError, match=msg):
+        ser[mask] = 1