Match query on trigram list not working as expected

[ChStark]

What version of Dgraph are you using?

1.1.1

Have you tried reproducing the issue with the latest release?

yes

What is the hardware spec (RAM, OS)?

RAM: 1GB
Distributor ID: Ubuntu
Description: Ubuntu 18.04.3 LTS
Release: 18.04
Codename: bionic

Steps to reproduce the issue (command/config used to run Dgraph).

Lets say I have the following schema:

<product.title>: string @index(fulltext, trigram) @lang .
<product.tags>: [string] @index(fulltext, trigram) .

And I have the following data:

_:zelda <product.title> "The Legend of Zelda Links Awakening - Nintendo Switch - Standard Edition" .
_:zelda <product.tags> "zelda" .
_:zelda <product.tags> "links" .
_:zelda <product.tags> "awakening" .
_:zelda <product.tags> "switch" .
_:pokemon <product.title> "Pokemon Sword - Nintendo Switch - Standard Edition" .
_:pokemon <product.tags> "pokemon" .
_:pokemon <product.tags> "sword" .
_:pokemon <product.tags> "switch" .

I got the result of the previous 2 nodes if I query

{
  node(func: match( product.tags , "switch" , 0 ) ) {
    uid
    product.title
    product.tags
  }
}

I got the result of the pokemon title if I query

{
  node(func: match( product.tags , "pokemon" , 0 ) ) {
    uid
    product.title
    product.tags
  }
}

So far so good since both terms are exact

Expected behavior and actual result.

But if I do query (note the term "pokenon" being 1 step away considering levenshtein distance )

{
  node(func: match( product.tags , "pokenon" , 1 ) ) {
    uid
    product.title
    product.tags
  }
}

I got 0 results when the expected result is that I should get the pokemon product node

Note that it works if I query

{
  node(func: match( product.tags , "pokeon" , 1 ) ) {
    uid
    product.title
    product.tags
  }
}

{
  node(func: match( product.tags , "pokemmon" , 1 ) ) {
    uid
    product.title
    product.tags
  }
}

It works when the 1 step is add/remove letter, but not replacement of letter

I hope you can fix this, since this is the only think is keeping me of using this solution instead elasticsearch for a project