Find Related and Connection

Where we left off

Every traversal function so far started from a known seed_id. In practice, you often start from a search term instead: "show me everything related to whoever matches 'Alice'." graph.find_related() and graph.connection() are the two functions built for that.

`graph.find_related()`: search, then expand

SELECT node_table_name, node_id FROM graph.find_related(
  property_key := 'name', property_value := 'Alice',
  source_table := 'public.people'::regclass,
  max_depth := 2,
  target_table := 'public.people'::regclass
);

This is graph.find() and graph.expand() from Lesson 14, fused: it searches for the starting node by property, then expands from whatever it finds, in one call. Useful whenever the caller has a name, not an ID, which in practice is most of the time.

`graph.connection()`: two searches, one path between them

SELECT source_table_name, target_table_name, hop_count, readable_path
FROM graph.connection(
  source_key := 'name', source_value := 'Dan',
  target_key := 'name', target_value := 'Alice',
  source_table := 'public.people'::regclass,
  target_table := 'public.people'::regclass,
  max_depth := 4
);

Where find_related searches from one side, connection searches *both* ends by property and finds the path between whatever it finds. It returns the same step-by-step shape as graph.shortest_path(), plus a readable_path column that renders the whole thing as one string:

people:p4 --works_at--> companies:c1 | companies:c1 --works_at--> people:p1

That's graph.format_path()'s formatting (Lesson 16 doesn't cover it directly, but this is where you'd reach for it) applied automatically, handy for logging or showing a path to an end user without writing your own string-building loop.