feat: BatchEventProcessor as Default Event Processor (#378)

ForwardingEventProcessor sends calls in a synchronous manner so to reduce the time it is replaced with BatchEventProcessor which sends calls in an asynchronous manner.

Author: ozayr-zaviar

optimizely/optimizely.py

@@ -25,7 +25,7 @@
 from .decision_service import Decision
 from .error_handler import NoOpErrorHandler as noop_error_handler
 from .event import event_factory, user_event_factory
-from .event.event_processor import ForwardingEventProcessor
+from .event.event_processor import BatchEventProcessor
 from .event_dispatcher import EventDispatcher as default_event_dispatcher
 from .helpers import enums, validator
 from .helpers.enums import DecisionSources
@@ -50,7 +50,8 @@ def __init__(
             notification_center=None,
             event_processor=None,
             datafile_access_token=None,
-            default_decide_options=None
+            default_decide_options=None,
+            event_processor_options=None
     ):
         """ Optimizely init method for managing Custom projects.
 
@@ -78,6 +79,7 @@ def __init__(
                            optimizely.event.event_processor.BatchEventProcessor.
           datafile_access_token: Optional string used to fetch authenticated datafile for a secure project environment.
           default_decide_options: Optional list of decide options used with the decide APIs.
+          event_processor_options: Optional dict of options to be passed to the default batch event processor.
         """
         self.logger_name = '.'.join([__name__, self.__class__.__name__])
         self.is_valid = True
@@ -86,8 +88,19 @@ def __init__(
         self.error_handler = error_handler or noop_error_handler
         self.config_manager = config_manager
         self.notification_center = notification_center or NotificationCenter(self.logger)
-        self.event_processor = event_processor or ForwardingEventProcessor(
-            self.event_dispatcher, logger=self.logger, notification_center=self.notification_center,
+        event_processor_defaults = {
+            'batch_size': 1,
+            'flush_interval': 30,
+            'timeout_interval': 5,
+            'start_on_init': True
+        }
+        if event_processor_options:
+            event_processor_defaults.update(event_processor_options)
+        self.event_processor = event_processor or BatchEventProcessor(
+            self.event_dispatcher,
+            logger=self.logger,
+            notification_center=self.notification_center,
+            **event_processor_defaults
         )
 
         if default_decide_options is None:

tests/test_event_processor.py

@@ -116,7 +116,7 @@ class BatchEventProcessorTest(base.BaseTest):
     MAX_BATCH_SIZE = 10
     MAX_DURATION_SEC = 0.2
     MAX_TIMEOUT_INTERVAL_SEC = 0.1
-    TEST_TIMEOUT = 10
+    TEST_TIMEOUT = 15
 
     def setUp(self, *args, **kwargs):
         base.BaseTest.setUp(self, 'config_dict_with_multiple_experiments')