Primitive tokenizer in Scala

lex.scala

package eu.wausoft.lex
import java.lang.{String => JString} // renames the built-in string to JString to avoid token clashing

trait Token
case class  String(value: JString)  extends Token
case class  Number(value: JString)  extends Token
case class  Atom(value: JString)    extends Token
case class  Keyword(value: JString) extends Token
case object Unknown extends Token

object Lexer {
  // triple-strings """ are "block-strings", they also ignore escape characters like @"" in C#
  private val string = """(".*")""".r        // matches strings with any character between the ""
  private val number = """(\d*.\d+|\d*)""".r // matches the numbers in the following formats: 123 123.456 .123
  private val atom   = """('[\w\-]+)""".r    // matches expressions like 'hello 'foo-bar 'baz_quux
  private val keywd  = """(:[\w\-]+)""".r    // matches expressions like :hello :foo-bar :baz_quux
	
  // (\s+(?![^"]*")) matches only on whitespace outside of a "" pair, allowing strings with spaces inside
  def tokenize(str: JString): List[Token] = str.split("""(\s+(?![^"]*"))""").map {
    case string(v) => String(v)
    case number(v) => Number(v)
    case atom(v)   => Atom(v)
    case keywd(v)  => Keyword(v)
    case _ => Unknown
  }.toList
}