BUG: Fixed encoding of pd.NA with to_json (pandas-dev#31748)

Author: ArtificialQualia

doc/source/whatsnew/v1.0.2.rst

@@ -25,8 +25,9 @@ Fixed regressions
 Bug fixes
 ~~~~~~~~~
 
--
--
+**I/O**
+
+- Using ``pd.NA`` with :meth:`DataFrame.to_json` now correctly outputs a null value instead of an empty object (:issue:`31615`)
 
 .. ---------------------------------------------------------------------------
 

pandas/_libs/src/ujson/python/objToJSON.c

@@ -54,6 +54,7 @@ static PyTypeObject *cls_dataframe;
 static PyTypeObject *cls_series;
 static PyTypeObject *cls_index;
 static PyTypeObject *cls_nat;
+static PyTypeObject *cls_na;
 PyObject *cls_timedelta;
 
 npy_int64 get_nat(void) { return NPY_MIN_INT64; }
@@ -151,6 +152,7 @@ int PdBlock_iterNext(JSOBJ, JSONTypeContext *);
 void *initObjToJSON(void) {
     PyObject *mod_pandas;
     PyObject *mod_nattype;
+    PyObject *mod_natype;
     PyObject *mod_decimal = PyImport_ImportModule("decimal");
     type_decimal =
         (PyTypeObject *)PyObject_GetAttrString(mod_decimal, "Decimal");
@@ -176,6 +178,12 @@ void *initObjToJSON(void) {
         Py_DECREF(mod_nattype);
     }
 
+    mod_natype = PyImport_ImportModule("pandas._libs.missing");
+    if (mod_natype) {
+        cls_na = (PyTypeObject *)PyObject_GetAttrString(mod_natype, "NAType");
+        Py_DECREF(mod_natype);
+    }
+
     /* Initialise numpy API */
     import_array();
     // GH 31463
@@ -1909,6 +1917,10 @@ void Object_beginTypeContext(JSOBJ _obj, JSONTypeContext *tc) {
                      "%R (0d array) is not JSON serializable at the moment",
                      obj);
         goto INVALID;
+    } else if (PyObject_TypeCheck(obj, cls_na)) {
+        PRINTMARK();
+        tc->type = JT_NULL;
+        return;
     }
 
 ISITERABLE:

pandas/tests/io/json/test_pandas.py

@@ -1640,3 +1640,44 @@ def test_deprecate_numpy_argument_read_json(self):
         with tm.assert_produces_warning(FutureWarning):
             result = read_json(expected.to_json(), numpy=True)
             tm.assert_frame_equal(result, expected)
+
+    def test_frame_int_overflow(self):
+        # GH 30320
+        encoded_json = json.dumps([{"col": "31900441201190696999"}, {"col": "Text"}])
+        expected = DataFrame({"col": ["31900441201190696999", "Text"]})
+        result = read_json(encoded_json)
+        tm.assert_frame_equal(result, expected)
+
+    @pytest.mark.parametrize(
+        "dataframe,expected",
+        [
+            (
+                pd.DataFrame({"x": [1, 2, 3], "y": ["a", "b", "c"]}),
+                '{"(0, \'x\')":1,"(0, \'y\')":"a","(1, \'x\')":2,'
+                '"(1, \'y\')":"b","(2, \'x\')":3,"(2, \'y\')":"c"}',
+            )
+        ],
+    )
+    def test_json_multiindex(self, dataframe, expected):
+        series = dataframe.stack()
+        result = series.to_json(orient="index")
+        assert result == expected
+
+    def test_to_s3(self, s3_resource):
+        # GH 28375
+        mock_bucket_name, target_file = "pandas-test", "test.json"
+        df = DataFrame({"x": [1, 2, 3], "y": [2, 4, 6]})
+        df.to_json(f"s3://{mock_bucket_name}/{target_file}")
+        assert target_file in (
+            obj.key for obj in s3_resource.Bucket("pandas-test").objects.all()
+        )
+
+    def test_json_pandas_na(self):
+        # GH 31615
+        result = pd.DataFrame([[pd.NA]]).to_json()
+        assert result == '{"0":{"0":null}}'
+
+    def test_json_pandas_nulls(self, nulls_fixture):
+        # GH 31615
+        result = pd.DataFrame([[nulls_fixture]]).to_json()
+        assert result == '{"0":{"0":null}}'