Rule Syntax overhaul

[keddelzz]

• Allow ^^ in Rule1[_] on token (see here) so that I can write NUMBER ^^ (_.toInt).
• Allow a single token as right hand side of a rule (see here), so that I don't have to explicitly call singletonRule.
• Infer better type for regular operations on Rule1[_]
  • optional (?) (see here), so that I get the type def Optional: Rule[Option[String] :: HNil] = NUMBER ?.
  • positive closure (+) (see here), so that I get the type def Optional: Rule[(String :: HNil) :: Option[String :: HNil] :: HNil] = NUMBER +.
  • kleene (see here), so that I get the type def Optional: Rule[List[String :: HNil] :: HNil] = NUMBER *.
• Infer scanner- and parser-rule-types to make the definition of a grammar easier

  @grammar
  class TestParser {
    import scala.language.postfixOps
  
    val digit : RegExp = regex("[0-9]")
    val number: RegExp = digit *
  
    val NUMBER: Token = token(number)
  
    @start def S = NUMBER
  
  }

The example above triggers the error: 'TestParser.this.NUMBER' is not a token!, because the implicit conversion from Token -> Rule1[String] doesn't trigger and thus the macro doesn't accept the parse-rule definition. Explicitly annotating the rule S with the type Rule1[String] will get rid of the error. A smarter grammar analysis could trigger the implicit conversion manually.