import scala.language.implicitConversions
import scala.language.reflectiveCalls
/* Note, in the lectures I did not show the type consraint
* I <% Seq[_] , which means that the input type I can be
* treated, or seen, as a sequence. */
abstract class Parser[I <% Seq[_], T] {
def parse(ts: I): Set[(T, I)]
def parse_all(ts: I) : Set[T] =
for ((head, tail) <- parse(ts);
if (tail.isEmpty)) yield head
}
class SeqParser[I <% Seq[_], T, S](p: => Parser[I, T],
q: => Parser[I, S]) extends Parser[I, (T, S)] {
def parse(sb: I) =
for ((head1, tail1) <- p.parse(sb);
(head2, tail2) <- q.parse(tail1)) yield ((head1, head2), tail2)
}
class AltParser[I <% Seq[_], T](p: => Parser[I, T],
q: => Parser[I, T]) extends Parser[I, T] {
def parse(sb: I) = p.parse(sb) ++ q.parse(sb)
}
class FunParser[I <% Seq[_], T, S](p: => Parser[I, T],
f: T => S) extends Parser[I, S] {
def parse(sb: I) =
for ((head, tail) <- p.parse(sb)) yield (f(head), tail)
}
// atomic parsers
case class CharParser(c: Char) extends Parser[String, Char] {
def parse(sb: String) =
if (sb.head == c) Set((c, sb.tail)) else Set()
}
case class StringParser(s: String) extends Parser[String, String] {
def parse(sb: String) = {
val (prefix, suffix) = sb.splitAt(s.length)
if (prefix == s) Set((prefix, suffix)) else Set()
}
}
case object NumParser extends Parser[String, String] {
val reg = "[0-9]+".r
def parse(sb: String) = reg.findPrefixOf(sb) match {
case None => Set()
case Some(s) => Set(sb.splitAt(s.length))
}
}
// convenience
implicit def string2parser(s : String) = StringParser(s)
implicit def ParserOps[I<% Seq[_], T](p: Parser[I, T]) = new {
def || (q : => Parser[I, T]) = new AltParser[I, T](p, q)
def ==>[S] (f: => T => S) = new FunParser[I, T, S](p, f)
def ~[S] (q : => Parser[I, S]) = new SeqParser[I, T, S](p, q)
}
implicit def StringOps(s: String) = new {
def || (q : => Parser[String, String]) = new AltParser[String, String](s, q)
def || (r: String) = new AltParser[String, String](s, r)
def ==>[S] (f: => String => S) = new FunParser[String, String, S](s, f)
def ~[S] (q : => Parser[String, S]) =
new SeqParser[String, String, S](s, q)
def ~ (r: String) =
new SeqParser[String, String, String](s, r)
}
// a parse palindromes
lazy val Pal : Parser[String, String] =
(("a" ~ Pal ~ "a") ==> { case ((x, y), z) => x + y + z } ||
("b" ~ Pal ~ "b") ==> { case ((x, y), z) => x + y + z } || "")
println("Palindrom: " + Pal.parse_all("ababbaba"))
// well-nested parenthesis parser
lazy val P : Parser[String, String] =
"(" ~ P ~ ")" ~ P ==> { case (((u, x), y), z) => "{" + x + "}" + z } || ""
P.parse_all("(((()()))())")
P.parse_all("(((()()))()))")
P.parse_all(")(")
P.parse_all("()")
// arithmetic expressions
lazy val E: Parser[String, String] =
(F ~ "*" ~ T) ==> { case ((x, y), z) => x + y + z } || F
lazy val F: Parser[String, String] =
((T ~ "+" ~ T) ==> { case ((x, y), z) => x + y + z } ||
(T ~ "-" ~ T) ==> { case ((x, y), z) => x + y + z } || T)
lazy val T: Parser[String, String] =
("(" ~ E ~ ")") ==> { case ((x, y), z) => x + y + z } || NumParser
println(E.parse_all("1*2+3"))
println(E.parse_all("1+2*3"))
println(E.parse_all("(1+2)+3"))
println(E.parse_all("1+2+3")) // this is not parsed, because of
// how the grammar is set up
// non-ambiguous vs ambiguous grammars
lazy val S : Parser[String, String] =
("1" ~ S ~ S) ==> { case ((x, y), z) => x + y + z } || ""
S.parse_all("1" * 15)
lazy val U : Parser[String, String] =
("1" ~ U) ==> { case (x, y) => x + y } || ""
U.parse("11")
U.parse("11111")
U.parse("11011")
U.parse_all("1" * 100 + "0")
lazy val UCount : Parser[String, Int] =
("1" ~ UCount) ==> { case (x, y) => y + 1 } ||
"" ==> { (x) => 0 }
UCount.parse("11111")
UCount.parse_all("11111")
// Single Character parser
lazy val One : Parser[String, String] = "1"
lazy val Two : Parser[String, String] = "2"
One.parse("1")
One.parse("111")
(One ~ One).parse("111")
(One ~ One ~ One).parse("111")
(One ~ One ~ One ~ One).parse("1111")
(One || Two).parse("111")