fix: supports_multivalues_insert dialect option was mispelled (#278)

Author: Jim Fulton

sqlalchemy_bigquery/base.py

@@ -643,7 +643,7 @@ class BigQueryDialect(DefaultDialect):
     supports_pk_autoincrement = False
     supports_default_values = False
     supports_empty_insert = False
-    supports_multiline_insert = True
+    supports_multivalues_insert = True
     supports_unicode_statements = True
     supports_unicode_binds = True
     supports_native_decimal = True

tests/unit/test_sqlalchemy_bigquery.py

@@ -13,6 +13,8 @@
 import pytest
 import sqlalchemy
 
+from conftest import setup_table
+
 
 @pytest.fixture
 def mock_bigquery_client():
@@ -158,3 +160,14 @@ def test__remove_type_from_empty_in(inp, outp):
 
     r = BigQueryExecutionContext._BigQueryExecutionContext__remove_type_from_empty_in
     assert r(None, inp) == outp
+
+
+def test_multi_value_insert(faux_conn, last_query):
+    table = setup_table(faux_conn, "t", sqlalchemy.Column("id", sqlalchemy.Integer))
+    faux_conn.execute(table.insert().values([dict(id=i) for i in range(3)]))
+
+    last_query(
+        "INSERT INTO `t` (`id`) VALUES"
+        " (%(id_m0:INT64)s), (%(id_m1:INT64)s), (%(id_m2:INT64)s)",
+        {"id_m0": 0, "id_m1": 1, "id_m2": 2},
+    )