BUG (string dtype): fix handling of string dtype in interchange protocol (pandas-dev#60333)

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

Author: WillAyd

pandas/core/interchange/from_dataframe.py

@@ -9,6 +9,8 @@
 
 import numpy as np
 
+from pandas._config import using_string_dtype
+
 from pandas.compat._optional import import_optional_dependency
 
 import pandas as pd
@@ -147,8 +149,6 @@ def protocol_df_chunk_to_pandas(df: DataFrameXchg) -> pd.DataFrame:
     -------
     pd.DataFrame
     """
-    # We need a dict of columns here, with each column being a NumPy array (at
-    # least for now, deal with non-NumPy dtypes later).
     columns: dict[str, Any] = {}
     buffers = []  # hold on to buffers, keeps memory alive
     for name in df.column_names():
@@ -347,8 +347,12 @@ def string_column_to_ndarray(col: Column) -> tuple[np.ndarray, Any]:
         # Add to our list of strings
         str_list[i] = string
 
-    # Convert the string list to a NumPy array
-    return np.asarray(str_list, dtype="object"), buffers
+    if using_string_dtype():
+        res = pd.Series(str_list, dtype="str")
+    else:
+        res = np.asarray(str_list, dtype="object")  # type: ignore[assignment]
+
+    return res, buffers  # type: ignore[return-value]
 
 
 def parse_datetime_format_str(format_str, data) -> pd.Series | np.ndarray:

pandas/tests/interchange/test_impl.py

@@ -6,8 +6,6 @@
 import numpy as np
 import pytest
 
-from pandas._config import using_string_dtype
-
 from pandas._libs.tslibs import iNaT
 from pandas.compat import (
     is_ci_environment,
@@ -401,7 +399,6 @@ def test_interchange_from_corrected_buffer_dtypes(monkeypatch) -> None:
     pd.api.interchange.from_dataframe(df)
 
 
-@pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)")
 def test_empty_string_column():
     # https://github.com/pandas-dev/pandas/issues/56703
     df = pd.DataFrame({"a": []}, dtype=str)
@@ -410,13 +407,12 @@ def test_empty_string_column():
     tm.assert_frame_equal(df, result)
 
 
-@pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)")
 def test_large_string():
     # GH#56702
     pytest.importorskip("pyarrow")
     df = pd.DataFrame({"a": ["x"]}, dtype="large_string[pyarrow]")
     result = pd.api.interchange.from_dataframe(df.__dataframe__())
-    expected = pd.DataFrame({"a": ["x"]}, dtype="object")
+    expected = pd.DataFrame({"a": ["x"]}, dtype="str")
     tm.assert_frame_equal(result, expected)
 
 
@@ -427,7 +423,6 @@ def test_non_str_names():
     assert names == ["0"]
 
 
-@pytest.mark.xfail(using_string_dtype(), reason="TODO(infer_string)")
 def test_non_str_names_w_duplicates():
     # https://github.com/pandas-dev/pandas/issues/56701
     df = pd.DataFrame({"0": [1, 2, 3], 0: [4, 5, 6]})
@@ -438,7 +433,7 @@ def test_non_str_names_w_duplicates():
             "Expected a Series, got a DataFrame. This likely happened because you "
             "called __dataframe__ on a DataFrame which, after converting column "
             r"names to string, resulted in duplicated names: Index\(\['0', '0'\], "
-            r"dtype='object'\). Please rename these columns before using the "
+            r"dtype='(str|object)'\). Please rename these columns before using the "
             "interchange protocol."
         ),
     ):