Activity

1. gkorland commented on Jul 11, 2020

   

   gkorland

   on Jul 11, 2020

   Last edited by gkorland

   Contributor

   More actions

   First, thanks for the suggestion to contribute the code!
   Second, you're right pipeline can improve performance dramatically and we already added this support in other clients.
   But, it can't replace the sync API but added as alternative API.
2. boris-42 commented on Jul 11, 2020

   

   boris-42

   on Jul 11, 2020

   ContributorAuthor

   More actions

   @gkorland First of all great that this is something on mind ;).

   So you suggest to implement something like bulk_queries(queries) ?

   Another thing is this place of code (as well please take a look at #81 )

       def commit(self):
           """
           Create entire graph.
           """
           if len(self.nodes) == 0 and len(self.edges) == 0:
               return None
   
           query = 'CREATE '
           for _, node in self.nodes.items():
               query += str(node) + ','
   
           query += ','.join([str(edge) for edge in self.edges])
   
           # Discard leading comma.
           if query[-1] is ',':
               query = query[:-1]
   
           return self.query(query)
   

   The problem is that it only works for creation of things but not updates.
   As soon as we add functionality to update/delete we would switch to multiple Graph commands e.g.:
   MATCH + DELETE, MERGE + SET in such situation we would need to use multiple queries (that was sort of initial point, sorry for being unclear it was a bit late)