I am writing a Customer Serializer. In that Serializer I would like to somehow say: "and this thing you already know how to serialize".
My current approach looks like that:
import org.json4s.native.Serialization._
import org.json4s.JsonDSL.WithBigDecimal._
object WindowSerializer extends CustomSerializer[Window](format =>
( [omitted],
{
case Window(frame, size) =>
( "size" -> size ) ~
( "frame" -> parse(write(frame)) )
}))
That parse(write(frame))
things is both ugly and inefficient. How to fix that?
You can call Extraction.decompose(a: Any)(implicit formats: Formats): JValue
which produces a JValue
from some value using runtime reflection.
import org.json4s._
import org.json4s.jackson.JsonMethods._
import org.json4s.JsonDSL._
import java.util.UUID
case class Thing(name: String)
case class Box(id: String, thing: Thing)
class BoxSerializer extends CustomSerializer[Box](format => ({
case jv: JValue =>
val id = (jv \ "id").extract[String]
val thing = (jv \ "thing").extract[Thing]
Box(id, thing)
}, {
case b: Box =>
("token" -> UUID.randomUUID().toString()) ~
("id" -> box.id) ~
("thing" -> Extraction.decompose(box.thing))
}))
implicit val formats = DefaultFormats + new BoxSerializer
val box = Box("1000", Thing("a thing"))
// decompose the value to JSON
val json = Extraction.decompose(box)
println(pretty(json))
// {
// "token" : "d9bd49dc-11b4-4380-ab10-f6df005a384c",
// "id" : "1000",
// "thing" : {
// "name" : "a thing"
// }
// }
// and read a value of type Box back from the JSON
println(json.extract[Box])
// Box(1000,Thing(a thing))
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With