[2.3.x] Only use new string dtype repr for the new (NaN-based) string dtype

Author: jorisvandenbossche

pandas/core/arrays/string_.py

@@ -119,10 +119,10 @@ class StringDtype(StorageExtensionDtype):
     Examples
     --------
     >>> pd.StringDtype()
-    <StringDtype(storage='python', na_value=<NA>)>
+    string[python]
 
     >>> pd.StringDtype(storage="pyarrow")
-    <StringDtype(na_value=<NA>)>
+    string[pyarrow]
     """
 
     @property
@@ -194,8 +194,11 @@ def __init__(
         self._na_value = na_value
 
     def __repr__(self) -> str:
-        storage = "" if self.storage == "pyarrow" else "storage='python', "
-        return f"<StringDtype({storage}na_value={self._na_value})>"
+        if self._na_value is libmissing.NA:
+            return f"{self.name}[{self.storage}]"
+        else:
+            storage = "" if self.storage == "pyarrow" else "storage='python', "
+            return f"<StringDtype({storage}na_value={self._na_value})>"
 
     def __eq__(self, other: object) -> bool:
         # we need to override the base class __eq__ because na_value (NA or NaN)

pandas/core/generic.py

@@ -7021,13 +7021,12 @@ def convert_dtypes(
         2  3  z   <NA>  <NA>    20  200.0
 
         >>> dfn.dtypes
-        a      Int32
-        b     string
-        c    boolean
-        d     string
-        e      Int64
-        f    Float64
-        dtype: object
+        a             Int32
+        b    string[python]
+        c           boolean
+        d    string[python]
+        e             Int64
+        f           Float64
 
         Start with a Series of strings and missing data represented by ``np.nan``.
 

pandas/io/formats/format.py

@@ -66,6 +66,7 @@
     ExtensionArray,
     TimedeltaArray,
 )
+from pandas.core.arrays.string_ import StringDtype
 from pandas.core.base import PandasObject
 import pandas.core.common as com
 from pandas.core.indexes.api import (
@@ -1231,6 +1232,8 @@ def _format(x):
                 return self.na_rep
             elif isinstance(x, PandasObject):
                 return str(x)
+            elif isinstance(x, StringDtype) and x.na_value is NA:
+                return repr(x)
             else:
                 # object dtype
                 return str(formatter(x))

pandas/tests/arrays/string_/test_string.py

@@ -126,11 +126,11 @@ def test_repr(dtype):
 def test_dtype_repr(dtype):
     if dtype.storage == "pyarrow":
         if dtype.na_value is pd.NA:
-            assert repr(dtype) == "<StringDtype(na_value=<NA>)>"
+            assert repr(dtype) == "string[pyarrow]"
         else:
             assert repr(dtype) == "<StringDtype(na_value=nan)>"
     elif dtype.na_value is pd.NA:
-        assert repr(dtype) == "<StringDtype(storage='python', na_value=<NA>)>"
+        assert repr(dtype) == "string[python]"
     else:
         assert repr(dtype) == "<StringDtype(storage='python', na_value=nan)>"
 

pandas/tests/io/formats/test_to_string.py

@@ -758,9 +758,9 @@ def test_to_string_string_dtype(self):
         result = df.dtypes.to_string()
         expected = dedent(
             """\
-            x            string
-            y            string
-            z    int64[pyarrow]"""
+            x    string[pyarrow]
+            y     string[python]
+            z     int64[pyarrow]"""
         )
         assert result == expected