Rework roundtripping logic

Author: rhshadrach

doc/source/whatsnew/v2.3.0.rst

@@ -35,7 +35,7 @@ Other enhancements
 - The semantics for the ``copy`` keyword in ``__array__`` methods (i.e. called
   when using ``np.array()`` or ``np.asarray()`` on pandas objects) has been
   updated to work correctly with NumPy >= 2 (:issue:`57739`)
-- :meth:`~Series.to_hdf` and :meth:`~DataFrame.to_hdf` now round-trip with ``StringDtype`` (:issue:`60663`)
+- :meth:`~Series.to_hdf` and :meth:`~DataFrame.to_hdf` now round-trip with ``StringDtype`` preserving the ``na_value`` but not necessarily the storage (:issue:`60663`)
 - The :meth:`~Series.cumsum`, :meth:`~Series.cummin`, and :meth:`~Series.cummax` reductions are now implemented for ``StringDtype`` columns when backed by PyArrow (:issue:`60633`)
 - The :meth:`~Series.sum` reduction is now implemented for ``StringDtype`` columns (:issue:`59853`)
 

pandas/io/pytables.py

@@ -38,7 +38,6 @@
     writers as libwriters,
 )
 from pandas._libs.lib import is_string_array
-from pandas._libs.missing import NA
 from pandas._libs.tslibs import timezones
 from pandas.compat._optional import import_optional_dependency
 from pandas.compat.pickle_compat import patch_pickle
@@ -3030,15 +3029,6 @@ def read_array(self, key: str, start: int | None = None, stop: int | None = None
             ret = node[0][start:stop]
             dtype = getattr(attrs, "value_type", None)
             if dtype is not None:
-                if dtype == "str[python]":
-                    dtype = StringDtype("python", np.nan)
-                elif dtype == "string[python]":
-                    dtype = StringDtype("python", NA)
-                elif dtype == "str[pyarrow]":
-                    dtype = StringDtype("pyarrow", np.nan)
-                else:
-                    assert dtype == "string[pyarrow]"
-                    dtype = StringDtype("pyarrow", NA)
                 ret = pd_array(ret, dtype=dtype)
         else:
             dtype = getattr(attrs, "value_type", None)
@@ -3283,15 +3273,7 @@ def write_array(
             vlarr = self._handle.create_vlarray(self.group, key, _tables().ObjectAtom())
             vlarr.append(value.to_numpy())
             node = getattr(self.group, key)
-            if value.dtype == StringDtype("python", np.nan):
-                node._v_attrs.value_type = "str[python]"
-            elif value.dtype == StringDtype("python", NA):
-                node._v_attrs.value_type = "string[python]"
-            elif value.dtype == StringDtype("pyarrow", np.nan):
-                node._v_attrs.value_type = "str[pyarrow]"
-            else:
-                assert value.dtype == StringDtype("pyarrow", NA)
-                node._v_attrs.value_type = "string[pyarrow]"
+            node._v_attrs.value_type = str(value.dtype)
         elif empty_array:
             self.write_array_empty(key, value)
         else:

pandas/tests/io/pytables/test_put.py

@@ -243,7 +243,8 @@ def test_put_str_frame(setup_path, performance_warning, string_dtype_arguments):
         _maybe_remove(store, "df")
 
         store.put("df", df)
-        expected = df
+        expected_dtype = "str" if dtype.na_value is np.nan else "string"
+        expected = df.astype(expected_dtype)
         result = store.get("df")
         tm.assert_frame_equal(result, expected)
 
@@ -256,7 +257,8 @@ def test_put_str_series(setup_path, performance_warning, string_dtype_arguments)
         _maybe_remove(store, "df")
 
         store.put("ser", ser)
-        expected = ser
+        expected_dtype = "str" if dtype.na_value is np.nan else "string"
+        expected = ser.astype(expected_dtype)
         result = store.get("ser")
         tm.assert_series_equal(result, expected)