Misleading Dgraph tour example

Issues Dgraph
1

In this tour example Query Variables in another query block I | Blocksvars | Dgraph Tour

the first query consists of

  PJ as var(func:allofterms(name@en, "Peter Jackson")) @normalize @cascade {
    F as director.film
  }

However, removing both @cascade and @normalize return the same response. In fact from the tests I’ve been making @cascade does not really work for defining variables.

Am I missing something here?

2

There are differences, but sure, the example doesn’t make it clear.

without cascade

"num_uids": {
        "": 14,
        "_total": 88,
        "actor.film": 5,
        "director.film": 5,
        "name": 30,
        "performance.character": 17,
        "performance.film": 17
      }

With cascade

"num_uids": {
        "": 13,
        "_total": 76,
        "actor.film": 1,
        "director.film": 5,
        "name": 25,
        "performance.character": 16,
        "performance.film": 16
      }

The objective of cascade is to remove nodes that don’t have the predicates specified on the query body.

e.g:

{
  PJ as var(func:allofterms(name@en, "Peter Jackson")) @normalize @cascade {
    F as director.film
  }

  peterJ(func: uid(PJ)) @normalize @cascade {
    name : name@en
    predicate.test
    actor.film {
      performance.film @filter(uid(F)) {
        film_name: name@en
        predicate.test
      }
      performance.character {
        character: name@en
      }
    }
  }
}

This query above will return empty. Cuz none of the nodes has the predicate predicate.test.

The normalize directive just “flats” the response.

3

This looks like something the docs could use to fix.