fix: ensure SchemaField.field_dtype returns a string

Author: chelsea-lin

google/cloud/bigquery/_pandas_helpers.py

@@ -110,6 +110,7 @@ def _to_wkb(v):
 
 _PANDAS_DTYPE_TO_BQ = {
     "bool": "BOOLEAN",
+    "boolean": "BOOLEAN",
     "datetime64[ns, UTC]": "TIMESTAMP",
     "datetime64[ns]": "DATETIME",
     "float32": "FLOAT",
@@ -122,6 +123,7 @@ def _to_wkb(v):
     "uint16": "INTEGER",
     "uint32": "INTEGER",
     "geometry": "GEOGRAPHY",
+    "string": "STRING",
     date_dtype_name: "DATE",
     time_dtype_name: "TIME",
 }
@@ -520,7 +522,7 @@ def dataframe_to_bq_schema(dataframe, bq_schema):
 
         # Otherwise, try to automatically determine the type based on the
         # pandas dtype.
-        bq_type = _PANDAS_DTYPE_TO_BQ.get(dtype.name)
+        bq_type = _PANDAS_DTYPE_TO_BQ.get(dtype.name.lower())
         if bq_type is None:
             sample_data = _first_valid(dataframe.reset_index()[column])
             if (

google/cloud/bigquery/schema.py

@@ -292,7 +292,7 @@ def field_type(self):
         """
         type_ = self._properties.get("type")
         if type_ is None:  # Shouldn't happen, but some unit tests do this.
-            return None
+            return ""
         return cast(str, type_).upper()
 
     @property

tests/unit/test__pandas_helpers.py

@@ -1298,14 +1298,15 @@ def test_dataframe_to_bq_schema_returns_schema_with_pandas_gbq(
 def test_dataframe_to_bq_schema_w_named_index(module_under_test, monkeypatch):
     monkeypatch.setattr(module_under_test, "pandas_gbq", None)
 
-    df_data = collections.OrderedDict(
-        [
-            ("str_column", ["hello", "world"]),
-            ("int_column", [42, 8]),
-            ("bool_column", [True, False]),
-        ]
-    )
-    index = pandas.Index(["a", "b"], name="str_index")
+    str_column = pandas.Series(["hello", "world"], dtype="string")
+    int_column = pandas.Series([42, 8], dtype="Int64")
+    bool_column = pandas.Series([True, False], dtype="boolean")
+    df_data = {
+        "str_column": str_column,
+        "int_column": int_column,
+        "bool_column": bool_column,
+    }
+    index = pandas.Index(["a", "b"], name="str_index", dtype="string")
     dataframe = pandas.DataFrame(df_data, index=index)
 
     with pytest.warns(FutureWarning, match="pandas-gbq"):
@@ -1324,20 +1325,24 @@ def test_dataframe_to_bq_schema_w_named_index(module_under_test, monkeypatch):
 def test_dataframe_to_bq_schema_w_multiindex(module_under_test, monkeypatch):
     monkeypatch.setattr(module_under_test, "pandas_gbq", None)
 
-    df_data = collections.OrderedDict(
-        [
-            ("str_column", ["hello", "world"]),
-            ("int_column", [42, 8]),
-            ("bool_column", [True, False]),
-        ]
-    )
+    str_column = pandas.Series(["hello", "world"], dtype="string")
+    int_column = pandas.Series([42, 8], dtype="Int64")
+    bool_column = pandas.Series([True, False], dtype="boolean")
+    df_data = {
+        "str_column": str_column,
+        "int_column": int_column,
+        "bool_column": bool_column,
+    }
     index = pandas.MultiIndex.from_tuples(
         [
             ("a", 0, datetime.datetime(1999, 12, 31, 23, 59, 59, 999999)),
             ("a", 0, datetime.datetime(2000, 1, 1, 0, 0, 0)),
         ],
         names=["str_index", "int_index", "dt_index"],
     )
+    index = index.set_levels(
+        [index.levels[0].astype("string"), index.levels[1], index.levels[2]]
+    )
     dataframe = pandas.DataFrame(df_data, index=index)
 
     with pytest.warns(FutureWarning, match="pandas-gbq"):