I am currently using the PPrint library to generate case classes to paste into my regression test Scala code, based on the values my existing code returns:

case class MyCaseClass(
    byte: Option[Byte],
    int: Option[Int],
    long: Option[Long],
    short: Option[Short],
    double: Option[Double],
    float: Option[Float],
    boolean: Option[Boolean],
    string: Option[String],
    char: Option[Char]
)
val myValue = MyCaseClass(
  byte = Some(1),
  int = Some(1),
  long = Some(1),
  short = Some(1),
  double = Some(1.0),
  float = Some(1.0F),
  boolean = Some(true),
  string = Some("1"),
  char = Some('1')
)
pprint.PPrinter.BlackWhite.pprintln(myValue)

This returns valid generated code that I can use:

MyCaseClass(
  byte = Some(value = 1),
  int = Some(value = 1),
  long = Some(value = 1L),
  short = Some(value = 1),
  double = Some(value = 1.0),
  float = Some(value = 1.0F),
  boolean = Some(value = true),
  string = Some(value = "1"),
  char = Some(value = '1')
)

However, this is slightly more verbose than I want. Namely, I would pretty much never expliclty write a Some with an explicit value parameter (e.g. I would write Some(1) and not Some(value = 1)).

PPrint provides the defaultShowFieldNames property to disable parameter names, but this also disables the parameter names in my other case classes, which I do want generated.

pprint.PPrinter.BlackWhite.copy(defaultShowFieldNames = false).pprintln(myValue)

=>

MyCaseClass(
  Some(1),
  Some(1),
  Some(1L),
  Some(1),
  Some(1.0),
  Some(1.0F),
  Some(true),
  Some("1"),
  Some('1')
)

I know I can specify additional handlers to PPrint to specially handle Some values, but as far as I can tell, I need to handle each possible case separately in order to guarantee valid code regardless of the generic type of the Some:

pprint.PPrinter.BlackWhite
  .copy(additionalHandlers = {
    case Some(value: Byte)    => pprint.Tree.Literal(s"Some($value)")
    case Some(value: Int)     => pprint.Tree.Literal(s"Some($value)")
    case Some(value: Long)    => pprint.Tree.Literal(s"Some($value)")
    case Some(value: Short)   => pprint.Tree.Literal(s"Some($value)")
    case Some(value: Double)  => pprint.Tree.Literal(s"Some($value)")
    case Some(value: Float)   => pprint.Tree.Literal(s"Some(${value}F)")
    case Some(value: Boolean) => pprint.Tree.Literal(s"Some($value)")
    case Some(value: String)  => pprint.Tree.Literal(s"Some(\"$value\")")
    case Some(value: Char)    => pprint.Tree.Literal(s"Some('$value')")
  })
  .pprintln(myValue)

=>

MyCaseClass(
  byte = Some(1),
  int = Some(1),
  long = Some(1),
  short = Some(1),
  double = Some(1.0),
  float = Some(1.0F),
  boolean = Some(true),
  string = Some("1"),
  char = Some('1')
)

This approach doesn't handle all possible input values. For instance, if a value was another case class, it would fall back to the default Some(value = ...) handling.

How can I get PPrint to generate values in my desired format (parameter names for case classes other than Some, and no parameter names for Some values)?