Modifying a JSON array in Scala with circe

2019-09-24 01:25发布

I have a JSON string with array as the following:

{
"cars": {
    "Nissan": [
        {"model":"Sentra", "doors":4},
        {"model":"Maxima", "doors":4}
    ],
    "Ford": [
        {"model":"Taurus", "doors":4},
        {"model":"Escort", "doors":2}
    ]
}
}

I would like to edit a new card brand, using circe at scala.
Instead of

"Nissan": [
        {"model":"Sentra", "doors":4},
        {"model":"Maxima", "doors":4},
    ]

I would like to have as a result:

"Nissan": [
        {"model":"Sentra", "doors":1000},
    ],

Thanks.

标签: scala circe
1条回答
Deceive 欺骗
2楼-- · 2019-09-24 02:20

Made after reading the manual, please do the same next time!

Apart from your JSON being invalid (Trailing comma on object in array "Nissan"), it should look something like this:

import cats.syntax.either._
import io.circe._, io.circe.parser._

val json: String = """
{
"cars": {
    "Nissan": [
        {"model":"Sentra", "doors":4},
        {"model":"Maxima", "doors":4}
    ],
    "Ford": [
        {"model":"Taurus", "doors":4},
        {"model":"Escort", "doors":2}
    ]
}
}
"""

val newJson = parse(json).toOption
  .flatMap { doc =>
    doc.hcursor
      .downField("cars")
      .downField("Nissan")
      .withFocus(_ =>
          Json.arr(
            Json.fromFields(
              Seq(
                ("model", Json.fromString("Sentra")),
                ("doors", Json.fromInt(1000))
              )
            )
          )
      )
      .top
  }

newJson match {
  case Some(v) => println(v.toString)
  case None => println("Failure!")
}

Try it out! (Rerun to see correct indentation!)


newJson is actually an Option, so if parsing or modifying fails, you will get None.

Calling toOption on parse(json) converts the returned Either[Json] (parsing succeeded / failed) to an Option[Json].

We need to use an Option[Json] in the first place because .top returns a Option[Json] (modifying succeeded / failed), too.

This way we can flatMap and don't have to deal with any nested types (see here).

查看更多
登录 后发表回答