BUG: fix to_json() with JSON Table Schema work correctly with string dtype (#61900)

Co-authored-by: Khemkaran <[email protected]>
Co-authored-by: Joris Van den Bossche <[email protected]>

Author: 3 people

doc/source/whatsnew/v2.3.2.rst

@@ -22,7 +22,10 @@ become the default string dtype in pandas 3.0. See
 
 Bug fixes
 ^^^^^^^^^
--
+- Fix :meth:`~DataFrame.to_json` with ``orient="table"`` to correctly use the
+  "string" type in the JSON Table Schema for :class:`StringDtype` columns
+  (:issue:`61889`)
+
 
 .. ---------------------------------------------------------------------------
 .. _whatsnew_232.contributors:

pandas/io/json/_table_schema.py

@@ -90,8 +90,6 @@ def as_json_table_type(x: DtypeObj) -> str:
         return "datetime"
     elif lib.is_np_dtype(x, "m"):
         return "duration"
-    elif isinstance(x, ExtensionDtype):
-        return "any"
     elif is_string_dtype(x):
         return "string"
     else:
@@ -197,7 +195,7 @@ def convert_json_field_to_pandas_type(field) -> str | CategoricalDtype:
     """
     typ = field["type"]
     if typ == "string":
-        return "object"
+        return field.get("extDtype", None)
     elif typ == "integer":
         return field.get("extDtype", "int64")
     elif typ == "number":

pandas/tests/io/json/test_json_table_schema.py

@@ -70,7 +70,7 @@ def test_build_table_schema(self, df_schema, using_infer_string):
             "primaryKey": ["idx"],
         }
         if using_infer_string:
-            expected["fields"][2] = {"name": "B", "type": "any", "extDtype": "str"}
+            expected["fields"][2] = {"name": "B", "type": "string", "extDtype": "str"}
         assert result == expected
         result = build_table_schema(df_schema)
         assert "pandas_version" in result
@@ -120,10 +120,10 @@ def test_multiindex(self, df_schema, using_infer_string):
         if using_infer_string:
             expected["fields"][0] = {
                 "name": "level_0",
-                "type": "any",
+                "type": "string",
                 "extDtype": "str",
             }
-            expected["fields"][3] = {"name": "B", "type": "any", "extDtype": "str"}
+            expected["fields"][3] = {"name": "B", "type": "string", "extDtype": "str"}
         assert result == expected
 
         df.index.names = ["idx0", None]
@@ -303,7 +303,7 @@ def test_to_json(self, df_table, using_infer_string):
         ]
 
         if using_infer_string:
-            fields[2] = {"name": "B", "type": "any", "extDtype": "str"}
+            fields[2] = {"name": "B", "type": "string", "extDtype": "str"}
 
         schema = {"fields": fields, "primaryKey": ["idx"]}
         data = [
@@ -547,7 +547,7 @@ def test_convert_pandas_type_to_json_field_categorical(self, kind, ordered):
                 },
                 CategoricalDtype(categories=["a", "b", "c"], ordered=True),
             ),
-            ({"type": "string"}, "object"),
+            ({"type": "string"}, None),
         ],
     )
     def test_convert_json_field_to_pandas_type(self, inp, exp):

pandas/tests/io/json/test_json_table_schema_ext_dtype.py

@@ -50,7 +50,7 @@ def test_build_table_schema(self):
                 {"name": "index", "type": "integer"},
                 {"name": "A", "type": "any", "extDtype": "DateDtype"},
                 {"name": "B", "type": "number", "extDtype": "decimal"},
-                {"name": "C", "type": "any", "extDtype": "string"},
+                {"name": "C", "type": "string", "extDtype": "string"},
                 {"name": "D", "type": "integer", "extDtype": "Int64"},
             ],
             "primaryKey": ["index"],
@@ -80,10 +80,10 @@ def test_as_json_table_type_ext_decimal_dtype(self):
     @pytest.mark.parametrize("box", [lambda x: x, Series])
     def test_as_json_table_type_ext_string_array_dtype(self, box):
         string_data = box(array(["pandas"], dtype="string"))
-        assert as_json_table_type(string_data.dtype) == "any"
+        assert as_json_table_type(string_data.dtype) == "string"
 
     def test_as_json_table_type_ext_string_dtype(self):
-        assert as_json_table_type(StringDtype()) == "any"
+        assert as_json_table_type(StringDtype()) == "string"
 
     @pytest.mark.parametrize("box", [lambda x: x, Series])
     def test_as_json_table_type_ext_integer_array_dtype(self, box):
@@ -176,7 +176,7 @@ def test_build_string_series(self, sa):
 
         fields = [
             {"name": "id", "type": "integer"},
-            {"name": "a", "type": "any", "extDtype": "string"},
+            {"name": "a", "type": "string", "extDtype": "string"},
         ]
 
         schema = {"fields": fields, "primaryKey": ["id"]}
@@ -235,7 +235,7 @@ def test_to_json(self, da, dc, sa, ia):
             OrderedDict({"name": "idx", "type": "integer"}),
             OrderedDict({"name": "A", "type": "any", "extDtype": "DateDtype"}),
             OrderedDict({"name": "B", "type": "number", "extDtype": "decimal"}),
-            OrderedDict({"name": "C", "type": "any", "extDtype": "string"}),
+            OrderedDict({"name": "C", "type": "string", "extDtype": "string"}),
             OrderedDict({"name": "D", "type": "integer", "extDtype": "Int64"}),
         ]