Backport PR #60134 on branch 2.3.x (TST (string dtype): remove xfails in extension tests + fix categorical/string dispatch) (#60178)

Backport PR #60134: TST (string dtype): remove xfails in extension tests + fix categorical/string dispatch

Co-authored-by: Joris Van den Bossche <jorisvandenbossche@gmail.com>

Author: meeseeksmachine

pandas/core/arrays/string_.py

@@ -915,7 +915,6 @@ def _cmp_method(self, other, op):
             if not is_array_like(other):
                 other = np.asarray(other)
             other = other[valid]
-            other = np.asarray(other)
 
         if op.__name__ in ops.ARITHMETIC_BINOPS:
             result = np.empty_like(self._ndarray, dtype="object")

pandas/tests/extension/base/ops.py

@@ -5,10 +5,6 @@
 import numpy as np
 import pytest
 
-from pandas._config import using_string_dtype
-
-from pandas.compat import HAS_PYARROW
-
 from pandas.core.dtypes.common import is_string_dtype
 
 import pandas as pd
@@ -134,12 +130,6 @@ class BaseArithmeticOpsTests(BaseOpsUtil):
     series_array_exc: type[Exception] | None = TypeError
     divmod_exc: type[Exception] | None = TypeError
 
-    # TODO(infer_string) need to remove import of pyarrow
-    @pytest.mark.xfail(
-        using_string_dtype() and not HAS_PYARROW,
-        reason="TODO(infer_string)",
-        strict=False,
-    )
     def test_arith_series_with_scalar(self, data, all_arithmetic_operators):
         # series & scalar
         if all_arithmetic_operators == "__rmod__" and is_string_dtype(data.dtype):
@@ -149,11 +139,6 @@ def test_arith_series_with_scalar(self, data, all_arithmetic_operators):
         ser = pd.Series(data)
         self.check_opname(ser, op_name, ser.iloc[0])
 
-    @pytest.mark.xfail(
-        using_string_dtype() and not HAS_PYARROW,
-        reason="TODO(infer_string)",
-        strict=False,
-    )
     def test_arith_frame_with_scalar(self, data, all_arithmetic_operators):
         # frame & scalar
         if all_arithmetic_operators == "__rmod__" and is_string_dtype(data.dtype):
@@ -163,22 +148,12 @@ def test_arith_frame_with_scalar(self, data, all_arithmetic_operators):
         df = pd.DataFrame({"A": data})
         self.check_opname(df, op_name, data[0])
 
-    @pytest.mark.xfail(
-        using_string_dtype() and not HAS_PYARROW,
-        reason="TODO(infer_string)",
-        strict=False,
-    )
     def test_arith_series_with_array(self, data, all_arithmetic_operators):
         # ndarray & other series
         op_name = all_arithmetic_operators
         ser = pd.Series(data)
         self.check_opname(ser, op_name, pd.Series([ser.iloc[0]] * len(ser)))
 
-    @pytest.mark.xfail(
-        using_string_dtype() and not HAS_PYARROW,
-        reason="TODO(infer_string)",
-        strict=False,
-    )
     def test_divmod(self, data):
         ser = pd.Series(data)
         self._check_divmod_op(ser, divmod, 1)
@@ -194,7 +169,6 @@ def test_divmod_series_array(self, data, data_for_twos):
         other = pd.Series(other)
         self._check_divmod_op(other, ops.rdivmod, ser)
 
-    @pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)", strict=False)
     def test_add_series_with_extension_array(self, data):
         # Check adding an ExtensionArray to a Series of the same dtype matches
         # the behavior of adding the arrays directly and then wrapping in a

pandas/tests/extension/test_categorical.py

@@ -148,7 +148,6 @@ def test_map(self, data, na_action):
         result = data.map(lambda x: x, na_action=na_action)
         tm.assert_extension_array_equal(result, data)
 
-    @pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)", strict=False)
     def test_arith_frame_with_scalar(self, data, all_arithmetic_operators, request):
         # frame & scalar
         op_name = all_arithmetic_operators
@@ -160,7 +159,6 @@ def test_arith_frame_with_scalar(self, data, all_arithmetic_operators, request):
             )
         super().test_arith_frame_with_scalar(data, op_name)
 
-    @pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)", strict=False)
     def test_arith_series_with_scalar(self, data, all_arithmetic_operators, request):
         op_name = all_arithmetic_operators
         if op_name == "__rmod__":

pandas/tests/extension/test_numpy.py

@@ -18,8 +18,6 @@
 import numpy as np
 import pytest
 
-from pandas._config import using_string_dtype
-
 from pandas.core.dtypes.dtypes import NumpyEADtype
 
 import pandas as pd
@@ -244,15 +242,13 @@ def test_insert_invalid(self, data, invalid_scalar):
     frame_scalar_exc = None
     series_array_exc = None
 
-    @pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)", strict=False)
     def test_divmod(self, data):
         divmod_exc = None
         if data.dtype.kind == "O":
             divmod_exc = TypeError
         self.divmod_exc = divmod_exc
         super().test_divmod(data)
 
-    @pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)", strict=False)
     def test_divmod_series_array(self, data):
         ser = pd.Series(data)
         exc = None
@@ -261,7 +257,6 @@ def test_divmod_series_array(self, data):
             self.divmod_exc = exc
         self._check_divmod_op(ser, divmod, data)
 
-    @pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)", strict=False)
     def test_arith_series_with_scalar(self, data, all_arithmetic_operators, request):
         opname = all_arithmetic_operators
         series_scalar_exc = None
@@ -275,7 +270,6 @@ def test_arith_series_with_scalar(self, data, all_arithmetic_operators, request)
         self.series_scalar_exc = series_scalar_exc
         super().test_arith_series_with_scalar(data, all_arithmetic_operators)
 
-    @pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)", strict=False)
     def test_arith_series_with_array(self, data, all_arithmetic_operators):
         opname = all_arithmetic_operators
         series_array_exc = None
@@ -284,7 +278,6 @@ def test_arith_series_with_array(self, data, all_arithmetic_operators):
         self.series_array_exc = series_array_exc
         super().test_arith_series_with_array(data, all_arithmetic_operators)
 
-    @pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)", strict=False)
     def test_arith_frame_with_scalar(self, data, all_arithmetic_operators, request):
         opname = all_arithmetic_operators
         frame_scalar_exc = None