BUG: Handle non-dict items in json_normalize with max_level

doc/source/whatsnew/v3.0.0.rst

@@ -1092,6 +1092,7 @@ I/O
 - Bug in :class:`DataFrame` and :class:`Series` ``repr`` of :py:class:`collections.abc.Mapping` elements. (:issue:`57915`)
 - Fix bug in ``on_bad_lines`` callable when returning too many fields: now emits
   ``ParserWarning`` and truncates extra fields regardless of ``index_col`` (:issue:`61837`)
+- Bug in :func:`pandas.json_normalize` inconsistently handling non-dict items in ``data`` when ``max_level`` was set. The function will now raise a ``TypeError`` if ``data`` is a list containing non-dict items (:issue:`62829`)
 - Bug in :meth:`.DataFrame.to_json` when ``"index"`` was a value in the :attr:`DataFrame.column` and :attr:`Index.name` was ``None``. Now, this will fail with a ``ValueError`` (:issue:`58925`)
 - Bug in :meth:`.io.common.is_fsspec_url` not recognizing chained fsspec URLs (:issue:`48978`)
 - Bug in :meth:`DataFrame._repr_html_` which ignored the ``"display.float_format"`` option (:issue:`59876`)

pandas/io/json/_normalize.py

@@ -501,6 +501,13 @@ def _pull_records(js: dict[str, Any], spec: list | str) -> list:
         # GH35923 Fix pd.json_normalize to not skip the first element of a
         # generator input
         data = list(data)
+        for item in data:
+            if not isinstance(item, dict):
+                msg = (
+                    "All items in data must be of type dict, "
+                    f"found {type(item).__name__}"
+                )
+                raise TypeError(msg)
     else:
         raise NotImplementedError
 

pandas/tests/io/json/test_normalize.py

@@ -511,6 +511,17 @@ def test_max_level_with_records_path(self, max_level, expected):
         expected_df = DataFrame(data=expected, columns=result.columns.values)
         tm.assert_equal(expected_df, result)
 
+    def test_json_normalize_non_dict_items(self):
+        # gh-62829
+        data_list = [np.nan, {"id": 12}, {"id": 13}]
+        msg = "All items in data must be of type dict, found float"
+
+        with pytest.raises(TypeError, match=msg):
+            json_normalize(data_list, max_level=0)
+
+        with pytest.raises(TypeError, match=msg):
+            json_normalize(data_list)
+
     def test_nested_flattening_consistent(self):
         # see gh-21537
         df1 = json_normalize([{"A": {"B": 1}}])