Filter nodes by source Node gremlin

Question

So let's say I have data like so.

(id:1)->(id:2)->(id:3)->(id:4)->(id:5)->(id:6)->(id:9)->(id:10)
(id:5)->(id:7)->(id:8)->(id:6)

To be clear, Id 5 is the same node with 2 edges.

Here is a code sample:

g.addV('person').property('id',1).as('1').
  addV('person').property('id',2).as('2').
  addV('person').property('id',3).as('3').
  addV('person').property('id',4).as('4').
  addV('person').property('id',5).as('5').
  addV('person').property('id',6).as('6').
  addV('person').property('id',7).as('7').
  addV('person').property('id',8).as('8').
  addV('person').property('id',9).as('9').
  addV('person').property('id',10).as('10').
  addE('connection').from('1').to('2').
  addE('connection').from('2').to('3').
  addE('connection').from('3').to('4').
  addE('connection').from('4').to('5').
  addE('connection').from('5').to('6').
  addE('connection').from('6').to('9').
  addE('connection').from('9').to('10').
  addE('connection').from('5').to('7').
  addE('connection').from('7').to('8').
  addE('connection').from('8').to('6').iterate()

I need to traverse the graph, and exclude any node where 6 has a connection, in any direction, to 5. So I would get back:

(id:1)->(id:2)->(id:3)->(id:4)->(id:5)->(id:6)
(id:1)->(id:2)->(id:3)->(id:4)->(id:5)->(id:7)->(id:8)->(id:6)

instead of presenting the graph data that way, i'd suggest that you write you sample data as a gremlin script that can be run in the Gremlin Console. much easier to convey your example that way and saves the people who might answer that extra task allowing them to focus on your solution. — stephen mallette, Feb 15 '18 at 18:51

score 2 · Accepted Answer · answered Feb 16 '18 at 12:19

2

I'm not sure that I completely understand your question, but it sounds like you want to stop traversing as soon as you see a pattern where you encounter vertex 6 and vertex 6 has an edge to vertex 5 - if so then here's one way to do this:

gremlin> g.V().has('id',1).
......1>   repeat(both().simplePath()).
......2>     until(and(has('id',6),
......3>               both().has('id',5))).
......4>   path().by('id')
==>[1,2,3,4,5,6]
==>[1,2,3,4,5,7,8,6]

Note that I don't exactly match the output you described in your answer as the traversal starts from vertex 1, so the path taken will include that portion of the path in both cases.

answered Feb 16 '18 at 12:19

stephen mallette

45,298
5
67
135

When I modify the query to work with my actual data, it gave me this `The repeat()-traversal was not defined: RepeatStep(until([AndStep([[HasStep([id.eq(6)]), VertexStep(BOTH,vertex), HasStep([G3E_FNO.eq(5)])]])]),emit(false))` – FARSOS BULSARA Feb 16 '18 at 21:42
1

you must have a syntax error in your definition of repeat() somehow. maybe some bad parenthesis? – stephen mallette Feb 16 '18 at 22:22
You were right, though now it returns nothing. Would it matter if I was comparing strings instead of numbers? – FARSOS BULSARA Feb 16 '18 at 22:45
if the value in stored in the database are strings, then you need to use strings for your arguments in the traversal. – stephen mallette Feb 16 '18 at 23:26

Filter nodes by source Node gremlin

1 Answers1