Use Executor for getting responses

Use an Executor of the same type as configured to get backend responses.This leverages the batching functionality, allowing batched requests to be processed concurrently.

Author: lucasrcosta

graphql_server/__init__.py

@@ -5,6 +5,8 @@
 from graphql import get_default_backend
 from graphql.error import format_error as default_format_error
 from graphql.execution import ExecutionResult
+from graphql.execution.executors.sync import SyncExecutor
+from promise import Promise, is_thenable
 
 from .error import HttpQueryError
 
@@ -67,11 +69,25 @@ def run_http_query(
 
     all_params = [get_graphql_params(entry, extra_data) for entry in data]
 
-    responses = [
-        get_response(schema, params, catch_exc, allow_only_query, **execute_options)
+    executor = execute_options.get('executor')
+    if executor is None:
+        response_executor = SyncExecutor()
+    else:
+        response_executor = executor.__class__()
+
+    response_promises = [
+        response_executor.execute(
+            get_response, schema, params, catch_exc, allow_only_query, **execute_options)
         for params in all_params
     ]
 
+    response_executor.wait_until_finished()
+
+    responses = map(
+        lambda result: result.get() if is_thenable(result) else result,
+        response_promises
+    )
+
     return responses, all_params