Adjust tests in base folder for arrow string option (#56124)

Author: phofl

pandas/tests/base/test_constructors.py

@@ -138,7 +138,9 @@ class TestConstruction:
             "object-string",
         ],
     )
-    def test_constructor_datetime_outofbound(self, a, constructor):
+    def test_constructor_datetime_outofbound(
+        self, a, constructor, request, using_infer_string
+    ):
         # GH-26853 (+ bug GH-26206 out of bound non-ns unit)
 
         # No dtype specified (dtype inference)
@@ -150,7 +152,10 @@ def test_constructor_datetime_outofbound(self, a, constructor):
             assert result.dtype == "M8[s]"
         else:
             result = constructor(a)
-            assert result.dtype == "object"
+            if using_infer_string and "object-string" in request.node.callspec.id:
+                assert result.dtype == "string"
+            else:
+                assert result.dtype == "object"
             tm.assert_numpy_array_equal(result.to_numpy(), a)
 
         # Explicit dtype specified

pandas/tests/base/test_conversion.py

@@ -20,6 +20,7 @@
     SparseArray,
     TimedeltaArray,
 )
+from pandas.core.arrays.string_arrow import ArrowStringArrayNumpySemantics
 
 
 class TestToIterable:
@@ -215,7 +216,9 @@ def test_iter_box_period(self):
         ),
     ],
 )
-def test_values_consistent(arr, expected_type, dtype):
+def test_values_consistent(arr, expected_type, dtype, using_infer_string):
+    if using_infer_string and dtype == "object":
+        expected_type = ArrowStringArrayNumpySemantics
     l_values = Series(arr)._values
     r_values = pd.Index(arr)._values
     assert type(l_values) is expected_type
@@ -358,17 +361,23 @@ def test_to_numpy(arr, expected, index_or_series_or_array, request):
 @pytest.mark.parametrize(
     "arr", [np.array([1, 2, 3], dtype="int64"), np.array(["a", "b", "c"], dtype=object)]
 )
-def test_to_numpy_copy(arr, as_series):
+def test_to_numpy_copy(arr, as_series, using_infer_string):
     obj = pd.Index(arr, copy=False)
     if as_series:
         obj = Series(obj.values, copy=False)
 
     # no copy by default
     result = obj.to_numpy()
-    assert np.shares_memory(arr, result) is True
+    if using_infer_string and arr.dtype == object:
+        assert np.shares_memory(arr, result) is False
+    else:
+        assert np.shares_memory(arr, result) is True
 
     result = obj.to_numpy(copy=False)
-    assert np.shares_memory(arr, result) is True
+    if using_infer_string and arr.dtype == object:
+        assert np.shares_memory(arr, result) is False
+    else:
+        assert np.shares_memory(arr, result) is True
 
     # copy=True
     result = obj.to_numpy(copy=True)

pandas/tests/base/test_misc.py

@@ -3,6 +3,8 @@
 import numpy as np
 import pytest
 
+from pandas._config import using_pyarrow_string_dtype
+
 from pandas.compat import PYPY
 
 from pandas.core.dtypes.common import (
@@ -80,7 +82,10 @@ def test_ndarray_compat_properties(index_or_series_obj):
     assert Series([1]).item() == 1
 
 
-@pytest.mark.skipif(PYPY, reason="not relevant for PyPy")
+@pytest.mark.skipif(
+    PYPY or using_pyarrow_string_dtype(),
+    reason="not relevant for PyPy doesn't work properly for arrow strings",
+)
 def test_memory_usage(index_or_series_memory_obj):
     obj = index_or_series_memory_obj
     # Clear index caches so that len(obj) == 0 report 0 memory usage
@@ -175,7 +180,9 @@ def test_access_by_position(index_flat):
     assert index[-1] == index[size - 1]
 
     msg = f"index {size} is out of bounds for axis 0 with size {size}"
-    if is_dtype_equal(index.dtype, "string[pyarrow]"):
+    if is_dtype_equal(index.dtype, "string[pyarrow]") or is_dtype_equal(
+        index.dtype, "string[pyarrow_numpy]"
+    ):
         msg = "index out of bounds"
     with pytest.raises(IndexError, match=msg):
         index[size]

pandas/tests/base/test_unique.py

@@ -1,6 +1,8 @@
 import numpy as np
 import pytest
 
+from pandas._config import using_pyarrow_string_dtype
+
 import pandas as pd
 import pandas._testing as tm
 from pandas.tests.base.common import allow_na_ops
@@ -98,6 +100,7 @@ def test_nunique_null(null_obj, index_or_series_obj):
 
 
 @pytest.mark.single_cpu
+@pytest.mark.xfail(using_pyarrow_string_dtype(), reason="decoding fails")
 def test_unique_bad_unicode(index_or_series):
     # regression test for #34550
     uval = "\ud83d"  # smiley emoji

pandas/tests/base/test_value_counts.py

@@ -14,6 +14,7 @@
     Series,
     Timedelta,
     TimedeltaIndex,
+    array,
 )
 import pandas._testing as tm
 from pandas.tests.base.common import allow_na_ops
@@ -113,7 +114,7 @@ def test_value_counts_null(null_obj, index_or_series_obj):
     tm.assert_series_equal(result, expected)
 
 
-def test_value_counts_inferred(index_or_series):
+def test_value_counts_inferred(index_or_series, using_infer_string):
     klass = index_or_series
     s_values = ["a", "b", "b", "b", "b", "c", "d", "d", "a", "a"]
     s = klass(s_values)
@@ -125,7 +126,9 @@ def test_value_counts_inferred(index_or_series):
         tm.assert_index_equal(s.unique(), exp)
     else:
         exp = np.unique(np.array(s_values, dtype=np.object_))
-        tm.assert_numpy_array_equal(s.unique(), exp)
+        if using_infer_string:
+            exp = array(exp)
+        tm.assert_equal(s.unique(), exp)
 
     assert s.nunique() == 4
     # don't sort, have to sort after the fact as not sorting is
@@ -147,7 +150,7 @@ def test_value_counts_inferred(index_or_series):
     tm.assert_series_equal(hist, expected)
 
 
-def test_value_counts_bins(index_or_series):
+def test_value_counts_bins(index_or_series, using_infer_string):
     klass = index_or_series
     s_values = ["a", "b", "b", "b", "b", "c", "d", "d", "a", "a"]
     s = klass(s_values)
@@ -201,7 +204,9 @@ def test_value_counts_bins(index_or_series):
         tm.assert_index_equal(s.unique(), exp)
     else:
         exp = np.array(["a", "b", np.nan, "d"], dtype=object)
-        tm.assert_numpy_array_equal(s.unique(), exp)
+        if using_infer_string:
+            exp = array(exp)
+        tm.assert_equal(s.unique(), exp)
     assert s.nunique() == 3
 
     s = klass({}) if klass is dict else klass({}, dtype=object)
@@ -246,7 +251,7 @@ def test_value_counts_datetime64(index_or_series, unit):
     expected_s = Series([3, 2, 1], index=idx, name="count")
     tm.assert_series_equal(s.value_counts(), expected_s)
 
-    expected = pd.array(
+    expected = array(
         np.array(
             ["2010-01-01 00:00:00", "2009-01-01 00:00:00", "2008-09-09 00:00:00"],
             dtype=f"datetime64[{unit}]",