[Fix #111] Adding API for filtering object while indexing

django_elasticsearch_dsl/documents.py

@@ -159,8 +159,8 @@ def parallel_bulk(self, actions, **kwargs):
     @classmethod
     def generate_id(cls, object_instance):
         """
-        The default behavior is to use the Django object's pk (id) as the 
-        elasticseach index id (_id). If needed, this method can be overloaded 
+        The default behavior is to use the Django object's pk (id) as the
+        elasticseach index id (_id). If needed, this method can be overloaded
         to change this default behavior.
         """
         return object_instance.pk
@@ -177,7 +177,8 @@ def _prepare_action(self, object_instance, action):
 
     def _get_actions(self, object_list, action):
         for object_instance in object_list:
-            yield self._prepare_action(object_instance, action)
+            if self.should_index_object(object_instance):
+                yield self._prepare_action(object_instance, action)
 
     def _bulk(self, *args, **kwargs):
         """Helper for switching between normal and parallel bulk operation"""
@@ -187,6 +188,13 @@ def _bulk(self, *args, **kwargs):
         else:
             return self.bulk(*args, **kwargs)
 
+    def should_index_object(self, obj):
+        """
+        Overwriting this method and returning a boolean value 
+        should determine whether the object should be indexed.
+        """
+        return True
+
     def update(self, thing, refresh=None, action='index', parallel=False, **kwargs):
         """
         Update each document in ES for a model, iterable of models or queryset