3. Graph Relationships

In this step you connect the content you created earlier with typed relations and query the resulting graph. You will link people to the articles they wrote, let people follow each other, and then ask the graph who is most followed and what is reachable from a given person.

All statements go through the SQL endpoint used in the previous steps:

curl -s -X POST localhost:8090/api/sql/docs-graph \
  -H "Authorization: Bearer $TOKEN" -H 'content-type: application/json' \
  -d '{"sql": "…"}'

1. Create the nodes

Use a workspace that allows the two node types. The social:Person and social:Article types below have a name/city and a title property.

INSERT INTO 'social' (path, node_type, name, properties) VALUES
  ('/alice', 'social:Person', 'alice', '{"name":"Alice","city":"Zurich"}'::jsonb),
  ('/bob',   'social:Person', 'bob',   '{"name":"Bob","city":"Bern"}'::jsonb),
  ('/carol', 'social:Person', 'carol', '{"name":"Carol","city":"Zurich"}'::jsonb),
  ('/dave',  'social:Person', 'dave',  '{"name":"Dave","city":"Basel"}'::jsonb);

INSERT INTO 'social' (path, node_type, name, properties) VALUES
  ('/graph-intro',     'social:Article', 'graph-intro',     '{"title":"Graph Intro"}'::jsonb),
  ('/advanced-graphs', 'social:Article', 'advanced-graphs', '{"title":"Advanced Graphs"}'::jsonb);

2. Relate them

RELATE creates one directed edge. Both endpoints name their workspace.

RELATE FROM path='/bob'   IN WORKSPACE 'social' TO path='/alice' IN WORKSPACE 'social' TYPE 'follows';
RELATE FROM path='/carol' IN WORKSPACE 'social' TO path='/alice' IN WORKSPACE 'social' TYPE 'follows';
RELATE FROM path='/dave'  IN WORKSPACE 'social' TO path='/alice' IN WORKSPACE 'social' TYPE 'follows';
RELATE FROM path='/dave'  IN WORKSPACE 'social' TO path='/bob'   IN WORKSPACE 'social' TYPE 'follows' WEIGHT 2.5;

RELATE FROM path='/alice' IN WORKSPACE 'social' TO path='/graph-intro' IN WORKSPACE 'social' TYPE 'authored';
RELATE FROM path='/advanced-graphs' IN WORKSPACE 'social'
       TO   path='/graph-intro'     IN WORKSPACE 'social' TYPE 'references' WEIGHT 0.9;

Each statement answers {"affected_rows": 1}.

3. Query the graph

Who follows whom:

SELECT * FROM GRAPH_TABLE(
  MATCH (a:Person)-[r:follows]->(b:Person)
  COLUMNS (a.name AS follower, b.name AS followed, r.weight AS weight)
);

{"rows":[{"follower":"bob","followed":"alice","weight":null},
         {"follower":"carol","followed":"alice","weight":null},
         {"follower":"dave","followed":"bob","weight":2.5},
         {"follower":"dave","followed":"alice","weight":null}]}

Articles and their authors, reading the title property straight off the node variable:

SELECT * FROM GRAPH_TABLE(
  MATCH (p:Person)-[:authored]->(a:Article)
  COLUMNS (p.name AS author, a.title AS title)
);
-- alice, "Graph Intro"

Everything dave can reach in up to three hops:

SELECT DISTINCT reached FROM GRAPH_TABLE(
  MATCH (a:Person)-[:follows]->{1,3}(b:Person)
  WHERE a.path = '/dave'
  COLUMNS (b.name AS reached)
) AS g;
-- bob, alice

4. Rank people

Graph algorithm functions run inside COLUMNS. Use an outer ORDER BY to sort the result:

SELECT * FROM GRAPH_TABLE(
  MATCH (n:Person)
  COLUMNS (n.name AS name, pageRank(n) AS rank, in_degree(n) AS followers)
) ORDER BY rank DESC;

{"rows":[{"name":"alice","rank":0.274,"followers":3},
         {"name":"bob","rank":0.112,"followers":1},
         {"name":"dave","rank":0.079,"followers":0},
         {"name":"carol","rank":0.079,"followers":0}]}

5. Combine with document queries

GRAPH_TABLE is a table, so join it back to the workspace for any column the pattern did not carry:

SELECT g.follower, s.properties->>'city' AS city
FROM GRAPH_TABLE(
  MATCH (a:Person)-[:follows]->(b:Person) WHERE b.path = '/alice'
  COLUMNS (a.name AS follower, a.id AS follower_id)
) AS g
JOIN 'social' s ON s.id = g.follower_id;

6. Remove an edge

UNRELATE FROM path='/dave' IN WORKSPACE 'social'
         TO   path='/bob'  IN WORKSPACE 'social' TYPE 'follows';

What you learned

• RELATE … TYPE … [WEIGHT …] creates a directed edge; UNRELATE removes it.
• GRAPH_TABLE(MATCH … WHERE … COLUMNS (…)) turns a pattern into rows.
• Node variables expose properties by name; edge variables expose relation_type and weight.
• Algorithms such as pageRank(n) run inside COLUMNS.

Next: Branching workflow. For the full pattern grammar see the GRAPH_TABLE reference.