Excluding type tag from json

[arshadalis]

trafficstars

Is there a way to exclude the discriminator type tag from the output json. For example

sealed trait Foo
case class Bar(s :String, w: Long) extends Foo
case class Baz(q : String, p: Int) extends Foo

the output json should be lets say for Bar

{
"s": "someabc"
"w": 12
}

i.e. It should not include "_type": "Bar" in the json.

I found following are the ways to customise the type tag

TypeTagSetting.ShortClassName
TypeTagSetting.FullClassName
TypeTagSetting.UserDefinedName

Is there a way to disable it ?

[julienrf]

How would you deserialize such JSON blobs? I mean, how would you detect which concrete class, Bar or Baz you should construct?

[julienrf]

One way to disable it is to have distinct formats, one format for Bar, and one format for Baz (but no format for Foo)

[vangogiel]

I can relate to this issue. You could have another TypeTagOWrites function, e.g.

    def default(): TypeTagOWrites =
      new TypeTagOWrites {
        def owrites[A](typeName: String, owrites: OWrites[A]): OWrites[A] =
          OWrites[A](a => owrites.writes(a))
      }

In my case, I want to have all my case classes with different parameters written with one implicit. I do not want to have multiple implicits. However, I want to also add other fields from my sealed trait. I want that to be done in a specific order and different than provided with the "flat" implicit.

I want other fields between my "typeName" and the owrites.writes(a)).

I tested it locally and it works. I can provide the code.

[vangogiel]

Unless there is a different solution provided with your library that I am unable to see.