1. Janusgraph-Users
  2. Topics

Creating a gremlin pipeline from an arraylist

Raymond Canzanese <r...@...>
 

I have an arraylist a of edges that I want to make gremlin queries over.  In the old days, I would do: 

a._() 

And have a pipeline I could work with.  Now it seems I can do:

g.inject(a).unfold()

or

g.E(a)

Which of these techniques should I prefer?  Is one of them more efficient than the other?
Jason Plurad <plu...@...>
 

You probably should benchmark it, but I'd think that the injection would be faster since you already have the edges resolved. I think using the graph step g.E(a) would ultimately re-lookup the edges by id.


On Tuesday, August 8, 2017 at 12:23:50 PM UTC-4, Raymond Canzanese wrote:
I have an arraylist a of edges that I want to make gremlin queries over.  In the old days, I would do: 

a._() 

And have a pipeline I could work with.  Now it seems I can do:

g.inject(a).unfold()

or

g.E(a)

Which of these techniques should I prefer?  Is one of them more efficient than the other?

1 - 2 of 2