[ccplugin] provide a --chunksize option to index-in-es

related to #17090555

cubicweb_elasticsearch/ccplugin.py

@@ -52,6 +52,10 @@
          {'type': 'string', 'default': '',
           'help': 'all indexable types except given etypes'
           '[default: []]'}),
+        ('chunksize',
+         {'type': 'int', 'default': 100000,
+          'help': 'max number of entities to fetch at once (deafult: 100000)',
+          }),
     ]
 
     def run(self, args):
@@ -94,7 +98,9 @@
         if index_name is None:
             index_name = cnx.vreg.config['index-name']
         for etype in etypes:
-            for idx, entity in enumerate(indexable_entities(cnx, etype), 1):
+            for idx, entity in enumerate(
+                    indexable_entities(
+                        cnx, etype, chunksize=self.config.chunksize), 1):
                 try:
                     serializer = entity.cw_adapt_to('IFullTextIndexSerializable')
                     json = serializer.serialize(complete=False)