Scala通过模式匹配识别字符串

Question

I wonder how to avoid casting Any to String but rather use pattern matching.

Having collected a dataframe from spark like df.select('column).first.toSeq.head Direct casting sort of is a workaround df.select('column).first.toSeq.head.asInstanceOf[String] , but I would prefer to use scala native pattern matching like

val collectedFromSpark: Any = "someString"
  val realString:String = collectedFromSpark match{
    case s:String => _
    case _ => throw new Exception("expected something else")
  }

However, realString:String only receives Any and not String .

How can I formulate this cast in a scala native way?

Answer 1

Change your case to this:

case s: String => s

Now it will know that s is a String .