implement batched INSERT..VALUES () () for executemany

the feature is enabled for all built in backends
when RETURNING is used,
except for Oracle that doesn't need it,  and on
psycopg2 and mssql+pyodbc it is used for all INSERT statements,
not just those that use RETURNING.

third party dialects would need to opt in to the new feature
by setting use_insertmanyvalues to True.

Also adds dialect-level guards against using returning
with executemany where we dont have an implementation to
suit it.   execute single w/ returning still defers to the
server without us checking.

Fixes: #6047
Fixes: #7907
Change-Id: I3936d3c00003f02e322f2e43fb949d0e6e568304

1 files changed, 18 insertions, 0 deletions

diff --git a/lib/sqlalchemy/dialects/postgresql/provision.py b/lib/sqlalchemy/dialects/postgresql/provision.py
index 0d17f28e0..8dd8a4995 100644
--- a/lib/sqlalchemy/dialects/postgresql/provision.py
+++ b/lib/sqlalchemy/dialects/postgresql/provision.py
@@ -14,6 +14,7 @@ from ...testing.provision import log
 from ...testing.provision import prepare_for_drop_tables
 from ...testing.provision import set_default_schema_on_connection
 from ...testing.provision import temp_table_keyword_args
+from ...testing.provision import upsert
 
 
 @create_db.for_db("postgresql")
@@ -125,3 +126,20 @@ def prepare_for_drop_tables(config, connection):
             "idle in transaction: %s"
             % ("; ".join(row._mapping["query"] for row in rows))
         )
+
+
+@upsert.for_db("postgresql")
+def _upsert(cfg, table, returning, set_lambda=None):
+    from sqlalchemy.dialects.postgresql import insert
+
+    stmt = insert(table)
+
+    if set_lambda:
+        stmt = stmt.on_conflict_do_update(
+            index_elements=table.primary_key, set_=set_lambda(stmt.excluded)
+        )
+    else:
+        stmt = stmt.on_conflict_do_nothing()
+
+    stmt = stmt.returning(*returning)
+    return stmt