I have two questions. First off, this is my reflection code for instantiating a class constructor via reflection:
case class RowToScalaConverter[T: TypeTag]() {
private val tpe: _root_.scala.reflect.runtime.universe.Type = typeOf[T]
private val symbol: _root_.scala.reflect.runtime.universe.ClassSymbol = tpe.typeSymbol.asClass
private val classLoaderMirror = runtimeMirror(getClass.getClassLoader)
private val classMirror: _root_.scala.reflect.runtime.universe.ClassMirror = classLoaderMirror.reflectClass(symbol)
private val constructorSymbol: _root_.scala.reflect.runtime.universe.Symbol = tpe.decl(termNames.CONSTRUCTOR)
private val constructorMethod: _root_.scala.reflect.runtime.universe.MethodMirror = classMirror.reflectConstructor(constructorSymbol.asMethod)
private def build(args: Seq[_]): T = constructorMethod(args: _ *).asInstanceOf[T]
def rowToClass(row: Row): T = {
val args = getRowArgs(row)
print(args)
build(args)
}
}
(Ps, I'm quite aware this is not type safe and that's ok with me, but it's the best I can do for what I want)
First question: Is there a cleaner way to do this? It seems like a lot of boilerplate. I'm not too worried about thread safety, As I plan to define a fixed number of these converters for a certain class [T] at runtime. It just looks kind of ugly.
Second: Would this cause any sort of memory leaks in regards to case class instantiation?
