updated json normalize and added tests

Author: arthurlw

pandas/io/json/_normalize.py

@@ -267,7 +267,7 @@ def _simple_json_normalize(
 
 
 def json_normalize(
-    data: dict | list[dict] | Series,
+    data: dict | list[dict] | Series | str | bytes,
     record_path: str | list | None = None,
     meta: str | list[str | list[str]] | None = None,
     meta_prefix: str | None = None,
@@ -285,8 +285,8 @@ def json_normalize(
 
     Parameters
     ----------
-    data : dict, list of dicts, or Series of dicts
-        Unserialized JSON objects.
+    data : dict, list of dicts, Series of dicts/JSON strings/bytes, or JSON string/bytes
+        Unserialized JSON objects or JSON strings/bytes.
     record_path : str or list of str, default None
         Path in each object to list of records. If not passed, data will be
         assumed to be an array of records.
@@ -434,7 +434,28 @@ def json_normalize(
     1          2
 
     Returns normalized data with columns prefixed with the given string.
+
+    >>> # JSON string input
+    >>> json_str = '{"id": 1, "name": {"first": "John", "last": "Doe"}}'
+    >>> pd.json_normalize(json_str)
+       id name.first name.last
+    0   1       John      Doe
     """
+    if isinstance(data, (str, bytes)):
+        import json
+        data = json.loads(data)
+    
+    if isinstance(data, Series):
+        if data.empty:
+            return DataFrame()
+        
+        sample = data.iloc[0]
+        if isinstance(sample, (str, bytes)):
+            import json
+            data = data.apply(json.loads)
+        index = data.index
+    else:
+        index = None
 
     def _pull_field(
         js: dict[str, Any], spec: list | str, extract_record: bool = False
@@ -485,11 +506,6 @@ def _pull_records(js: dict[str, Any], spec: list | str) -> list:
                 )
         return result
 
-    if isinstance(data, Series):
-        index = data.index
-    else:
-        index = None
-
     if isinstance(data, list) and not data:
         return DataFrame()
     elif isinstance(data, dict):

pandas/tests/io/json/test_normalize.py

@@ -569,6 +569,76 @@ def test_series_index(self, state_data):
         result = json_normalize(series, "counties")
         tm.assert_index_equal(result.index, idx.repeat([3, 2]))
 
+    def test_json_string_input(self):
+        # GH61006: Accept JSON as str input
+        json_str = '{"id": 1, "name": {"first": "John", "last": "Doe"}}'
+        result = json_normalize(json_str)
+        expected = DataFrame({
+            "id": [1],
+            "name.first": ["John"],
+            "name.last": ["Doe"]
+        })
+        tm.assert_frame_equal(result, expected)
+
+        json_array_str = '''[
+            {"id": 1, "name": {"first": "John", "last": "Doe"}},
+            {"id": 2, "name": {"first": "Jane", "last": "Smith"}}
+        ]'''
+        result = json_normalize(json_array_str)
+        expected = DataFrame({
+            "id": [1, 2],
+            "name.first": ["John", "Jane"],
+            "name.last": ["Doe", "Smith"]
+        })
+        tm.assert_frame_equal(result, expected)
+
+    def test_json_bytes_input(self):
+        # GH61006: Accept JSON as bytes input
+        json_bytes = b'{"id": 1, "name": {"first": "John", "last": "Doe"}}'
+        result = json_normalize(json_bytes)
+        expected = DataFrame({
+            "id": [1],
+            "name.first": ["John"],
+            "name.last": ["Doe"]
+        })
+        tm.assert_frame_equal(result, expected)
+
+    def test_series_json_string(self):
+        # GH61006:
+        s = Series([
+            '{"value": 0.0}',
+            '{"value": 0.5}',
+            '{"value": 1.0}'
+        ])
+        result = json_normalize(s)
+        expected = DataFrame({
+            "value": [0.0, 0.5, 1.0]
+        })
+        tm.assert_frame_equal(result, expected)
+
+    def test_series_json_string_with_index(self):
+        # GH61006:
+        s = Series(
+            ['{"value": 0.0}', '{"value": 0.5}'],
+            index=['a', 'b']
+        )
+        result = json_normalize(s)
+        expected = DataFrame(
+            {"value": [0.0, 0.5]},
+            index=['a', 'b']
+        )
+        tm.assert_frame_equal(result, expected)
+
+    def test_invalid_json_string(self):
+        invalid_json = '{"id": 1, "name": {"first": "John", "last": "Doe"'
+        with pytest.raises(json.JSONDecodeError):
+            json_normalize(invalid_json)
+
+    def test_non_json_string(self):
+        non_json = "Hello World"
+        with pytest.raises(json.JSONDecodeError):
+            json_normalize(non_json)
+
 
 class TestNestedToRecord:
     def test_flat_stays_flat(self):