solution/cw5/fun_parser.sc
author Christian Urban <christian.urban@kcl.ac.uk>
Mon, 29 Aug 2022 01:16:32 +0200
changeset 873 a25da86f7c8c
parent 868 8fb3b6d3be70
permissions -rw-r--r--
updated

// A parser for the Fun language
//================================
//
// call with 
//
//     amm fun_parser.sc fact.fun
//
//     amm fun_parser.sc defs.fun
//
// this will generate a parse-tree from a list
// of tokens

import scala.language.implicitConversions    
import scala.language.reflectiveCalls

import $file.fun_tokens, fun_tokens._ 


// Parser combinators
//    type parameter I needs to be of Seq-type
//
abstract class Parser[I, T](implicit ev: I => Seq[_]) {
  def parse(ts: I): Set[(T, I)]

  def parse_single(ts: I) : T = 
    parse(ts).partition(_._2.isEmpty) match {
      case (good, _) if !good.isEmpty => good.head._1
      case (good, err) if err.isEmpty => {
        println (s"Parse Error\n $good \n $err") ; sys.exit(-1) }
      case (_, err) => { 
	println (s"Parse Error\n${err.minBy(_._2.length)}") ; sys.exit(-1) }
    }
}

// convenience for writing grammar rules
case class ~[+A, +B](_1: A, _2: B)

class SeqParser[I, T, S](p: => Parser[I, T], 
                         q: => Parser[I, S])(implicit ev: I => Seq[_]) extends Parser[I, ~[T, S]] {
  def parse(sb: I) = 
    for ((head1, tail1) <- p.parse(sb); 
         (head2, tail2) <- q.parse(tail1)) yield (new ~(head1, head2), tail2)
}

class AltParser[I, T](p: => Parser[I, T], 
                      q: => Parser[I, T])(implicit ev: I => Seq[_]) extends Parser[I, T] {
  def parse(sb: I) = p.parse(sb) ++ q.parse(sb)   
}

class FunParser[I, T, S](p: => Parser[I, T], 
                         f: T => S)(implicit ev: I => Seq[_]) extends Parser[I, S] {
  def parse(sb: I) = 
    for ((head, tail) <- p.parse(sb)) yield (f(head), tail)
}

// convenient combinators
implicit def ParserOps[I, T](p: Parser[I, T])(implicit ev: I => Seq[_]) = 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)
}

def ListParser[I, T, S](p: => Parser[I, T], 
                        q: => Parser[I, S])(implicit ev: I => Seq[_]): Parser[I, List[T]] = {
  (p ==> ((s) => List(s))) ||
  (p ~ q ~ ListParser(p, q)) ==> { case x ~ _ ~ z => x :: z : List[T] }
}

case class TokParser(tok: Token) extends Parser[List[Token], Token] {
  def parse(ts: List[Token]) = ts match {
    case t::ts if (t == tok) => Set((t, ts)) 
    case _ => Set()
  }
}

implicit def token2tparser(t: Token) = TokParser(t)

implicit def TokOps(t: Token) = new {
  def || (q : => Parser[List[Token], Token]) = new AltParser[List[Token], Token](t, q)
  def ==>[S] (f: => Token => S) = new FunParser[List[Token], Token, S](t, f)
  def ~[S](q : => Parser[List[Token], S]) = new SeqParser[List[Token], Token, S](t, q)
}

case object EmptyParser extends Parser[List[Token], String] {
  def parse(ts: List[Token]) = Set(("", ts))
}

case object NumParser extends Parser[List[Token], Int] {
  def parse(ts: List[Token]) = ts match {
    case T_NUM(n)::ts => Set((n, ts)) 
    case _ => Set ()
  }
}

case object FNumParser extends Parser[List[Token], Double] {
  def parse(ts: List[Token]) = ts match {
    case T_FNUM(x)::ts => Set((x, ts)) 
    case _ => Set()
  }
}

case object IdParser extends Parser[List[Token], String] {
  def parse(ts: List[Token]) = ts match {
    case T_ID(s)::ts => Set((s, ts)) 
    case _ => Set ()
  }
}

case object CharConstParser extends Parser[List[Token], Int] {
  def parse(ts: List[Token]) = ts match {
    case T_CHR(c)::ts => Set((c, ts)) 
    case _ => Set ()
  }
}

case object TyParser extends Parser[List[Token], String] {
  def parse(ts: List[Token]) = ts match {
    case T_TY(s)::ts => Set((s, ts)) 
    case _ => Set ()
  }
}


// Abstract syntax trees for the Fun language
abstract class Exp 
abstract class BExp 
abstract class Decl 

case class Def(name: String, args: List[(String, String)], ty: String, body: Exp) extends Decl
case class Main(e: Exp) extends Decl
case class Const(name: String, v: Int) extends Decl
case class FConst(name: String, x: Double) extends Decl

case class Call(name: String, args: List[Exp]) extends Exp
case class If(a: BExp, e1: Exp, e2: Exp) extends Exp
case class Var(s: String) extends Exp
case class Num(i: Int) extends Exp     // integer numbers
case class FNum(i: Double) extends Exp  // floating numbers
case class ChConst(c: Int) extends Exp // char constant
case class Aop(o: String, a1: Exp, a2: Exp) extends Exp
case class Sequence(e1: Exp, e2: Exp) extends Exp
case class Bop(o: String, a1: Exp, a2: Exp) extends BExp


// arithmetic expressions (there needs to be an F in the SEMICOLON case)
lazy val Exp: Parser[List[Token], Exp] = 
  (T_KWD("if") ~ BExp ~ T_KWD("then") ~ Exp ~ T_KWD("else") ~ Exp) ==>
    { case _ ~ x ~ _ ~ y ~ _ ~ z => If(x, y, z): Exp } ||
  (F ~ T_SEMI ~ Exp) ==> { case x ~ _ ~ y => Sequence(x, y): Exp } || L
lazy val L: Parser[List[Token], Exp] = 
  (T ~ T_OP("+") ~ Exp) ==> { case x ~ _ ~ z => Aop("+", x, z): Exp } ||
  (T ~ T_OP("-") ~ Exp) ==> { case x ~ _ ~ z => Aop("-", x, z): Exp } || T  
lazy val T: Parser[List[Token], Exp] = 
  (F ~ T_OP("*") ~ T) ==> { case x ~ _ ~ z => Aop("*", x, z): Exp } || 
  (F ~ T_OP("/") ~ T) ==> { case x ~ _ ~ z => Aop("/", x, z): Exp } || 
  (F ~ T_OP("%") ~ T) ==> { case x ~ _ ~ z => Aop("%", x, z): Exp } || F
lazy val F: Parser[List[Token], Exp] = 
  (IdParser ~ T_LPAREN ~ T_RPAREN) ==> { case x ~ _ ~ _ => Call(x, Nil): Exp } ||
  (IdParser ~ T_LPAREN ~ ListParser(Exp, T_COMMA) ~ T_RPAREN) ==> { case x ~ _ ~ z ~ _ => Call(x, z): Exp } ||
  (T_LPAREN ~ Exp ~ T_RPAREN) ==> { case _ ~ y ~ _ => y: Exp } || 
  IdParser ==> { case x => Var(x): Exp } || 
  NumParser ==> { case x => Num(x): Exp } ||
  CharConstParser ==> { case x => ChConst(x): Exp } ||
  FNumParser ==> { case x => FNum(x): Exp }

// boolean expressions
lazy val BExp: Parser[List[Token], BExp] = 
  (Exp ~ T_OP("==") ~ Exp) ==> { case x ~ _ ~ z => Bop("==", x, z): BExp } || 
  (Exp ~ T_OP("!=") ~ Exp) ==> { case x ~ _ ~ z => Bop("!=", x, z): BExp } || 
  (Exp ~ T_OP("<") ~ Exp)  ==> { case x ~ _ ~ z => Bop("<",  x, z): BExp } || 
  (Exp ~ T_OP(">") ~ Exp)  ==> { case x ~ _ ~ z => Bop("<",  z, x): BExp } || 
  (Exp ~ T_OP("<=") ~ Exp) ==> { case x ~ _ ~ z => Bop("<=", x, z): BExp } || 
  (Exp ~ T_OP("=>") ~ Exp) ==> { case x ~ _ ~ z => Bop("<=", z, x): BExp } ||
  (T_LPAREN ~ BExp ~ T_RPAREN) ==> { case _ ~ b ~ _ => b : BExp } 

lazy val Arg : Parser[List[Token], (String, String)] = 
  (IdParser ~ T_COLON ~ TyParser) ==> { case x ~ _ ~ ty => (x, ty) }  

lazy val Defn: Parser[List[Token], Decl] = {
   (T_KWD("def") ~ IdParser ~ T_LPAREN ~ T_RPAREN ~ T_COLON ~ TyParser ~ T_OP("=") ~ Exp) ==>
     { case _ ~ y ~ _ ~ _ ~ _~ ty ~ _ ~ r => Def(y, Nil, ty, r): Decl } ||
   (T_KWD("def") ~ IdParser ~ T_LPAREN ~ ListParser(Arg, T_COMMA) ~ T_RPAREN ~ T_COLON ~ TyParser ~ T_OP("=") ~ Exp) ==>
     { case _ ~ y ~ _ ~ w ~ _ ~ _~ ty ~ _ ~ r => Def(y, w, ty, r): Decl }
}

lazy val Const_decl: Parser[List[Token], Decl] =
   (T_KWD("val") ~ Arg ~ T_OP("=") ~ NumParser) ==>
     { case _ ~ x ~ _ ~ v => Const(x._1, v): Decl } ||
   (T_KWD("val") ~ Arg ~ T_OP("=") ~ FNumParser) ==>
     { case _ ~ x ~ _ ~ v => FConst(x._1, v): Decl } 

lazy val Prog: Parser[List[Token], List[Decl]] =
  (Defn ~ T_SEMI ~ Prog) ==> { case x ~ _ ~ z => x :: z : List[Decl] } ||
  (Const_decl ~ T_SEMI ~ Prog) ==> { case x ~ _ ~ z => x :: z : List[Decl] } ||
  (Exp ==> ((s) => List(Main(s)) : List[Decl]))



// Reading tokens and Writing parse trees

//import ammonite.ops._

def parse_tks(tks: List[Token]) : List[Decl] = {
  //println(Prog.parse(tks))
  Prog.parse_single(tks)
}

//@doc("Parses a file.")
@main
def main(fname: String) : Unit = {
  val tks = tokenise(os.read(os.pwd / fname))
  println(parse_tks(tks))
}