// regular expressions including NOT

abstract class Rexp

case object NULL extends Rexp

case object EMPTY extends Rexp

case object ALLC extends Rexp            // recognises any character

case class CHAR(c: Char) extends Rexp

case class ALT(r1: Rexp, r2: Rexp) extends Rexp

case class SEQ(r1: Rexp, r2: Rexp) extends Rexp

case class STAR(r: Rexp) extends Rexp

case class NOT(r: Rexp) extends Rexp     // negation of a regular expression

// nullable function: tests whether the regular 

// expression can recognise the empty string

def nullable (r: Rexp) : Boolean = r match {

  case NULL => false

  case EMPTY => true

  case ALLC => false

  case CHAR(_) => false

  case ALT(r1, r2) => nullable(r1) || nullable(r2)

  case SEQ(r1, r2) => nullable(r1) && nullable(r2)

  case STAR(_) => true

  case NOT(r) => !(nullable(r))

}

// tests whether a regular expression 

// cannot recognise more

def no_more (r: Rexp) : Boolean = r match {

  case NULL => true

  case EMPTY => false

  case ALLC => false

  case CHAR(_) => false

  case ALT(r1, r2) => no_more(r1) && no_more(r2)

  case SEQ(r1, r2) => if (nullable(r1)) (no_more(r1) && no_more(r2)) else no_more(r1)

  case STAR(_) => false

  case NOT(r) => !(no_more(r))

}

// derivative of a regular expression w.r.t. a character

def der (c: Char, r: Rexp) : Rexp = r match {

  case NULL => NULL

  case EMPTY => NULL  

  case ALLC => EMPTY

  case CHAR(d) => if (c == d) EMPTY else NULL

  case ALT(r1, r2) => ALT(der(c, r1), der(c, r2))

  case SEQ(r1, r2) => 

    if (nullable(r1)) ALT(SEQ(der(c, r1), r2), der(c, r2))

    else SEQ(der(c, r1), r2)

  case STAR(r) => SEQ(der(c, r), STAR(r))

  case NOT(r) => NOT(der (c, r))

}

// regular expression for specifying 

// ranges of characters

def Range(s : List[Char]) : Rexp = s match {

  case Nil => NULL

  case c::Nil => CHAR(c)

  case c::s => ALT(CHAR(c), Range(s))

}

def RANGE(s: String) = Range(s.toList)

// one or more

def PLUS(r: Rexp) = SEQ(r, STAR(r))

// many alternatives

def Alts(rs: List[Rexp]) : Rexp = rs match {

  case Nil => NULL

  case r::Nil => r

  case r::rs => ALT(r, Alts(rs))

}

def ALTS(rs: Rexp*) = Alts(rs.toList)

// repetitions

def Seqs(rs: List[Rexp]) : Rexp = rs match {

  case Nil => NULL

  case r::Nil => r

  case r::rs => SEQ(r, Seqs(rs))

}

def SEQS(rs: Rexp*) = Seqs(rs.toList)

// some convenience for typing in regular expressions

def charlist2rexp(s : List[Char]) : Rexp = s match {

  case Nil => EMPTY

  case c::Nil => CHAR(c)

  case c::s => SEQ(CHAR(c), charlist2rexp(s))

}

implicit def string2rexp(s : String) : Rexp = charlist2rexp(s.toList)

type Rule[T] = (Rexp, List[Char] => T)

case class Tokenizer[T](rules: List[Rule[T]], excl: List[T] = Nil) {

  def munch(r: Rexp, action: List[Char] => T, s: List[Char], t: List[Char]) : Option[(List[Char], T)] = 

    s match {

      case Nil if (nullable(r)) => Some(Nil, action(t))

      case Nil => None

      case c::s if (no_more(der (c, r)) && nullable(r)) => Some(c::s, action(t))

      case c::s if (no_more(der (c, r))) => None

      case c::s => munch(der (c, r), action, s, t ::: List(c))

    }

  def one_token(s: List[Char]) : Either[(List[Char], T), String] = {

    val somes = rules.map { (r) => munch(r._1, r._2, s, Nil) }.flatten

    if (somes == Nil) Right(s.mkString) 

    else Left(somes sortBy (_._1.length) head)

  }

  def tokenize(cs: List[Char]) : List[T] = cs match {

    case Nil => Nil

    case _ => one_token(cs) match {

      case Left((rest, token)) => token :: tokenize(rest)

      case Right(s) => { println("Cannot tokenize: \"" + s + "\""); Nil } 

    }

  }

  def fromString(s: String) : List[T] = 

    tokenize(s.toList).filterNot(excl.contains(_))

  def fromFile(name: String) : List[T] = 

    fromString(io.Source.fromFile(name).mkString)

}

// parser combinators with input type I and return type T

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

  def parse_single(ts: I) : T = parse_all(ts).toList match {

    case t::Nil => t

    case _ => { println ("Parse Error") ; sys.exit(-1) }

  }

  def || (right : => Parser[I, T]) : Parser[I, T] = new AltParser(this, right)

  def ==>[S] (f: => T => S) : Parser [I, S] = new FunParser(this, f)

  def ~[S] (right : => Parser[I, S]) : Parser[I, (T, S)] = new SeqParser(this, right)

  def ~>[S] (right : => Parser[I, S]) : Parser[I, S] = this ~ right ==> (_._2)

  def <~[S] (right : => Parser[I, S]) : Parser[I, T] = this ~ right ==> (_._1)

}

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)

}

// A parser and evaluator for teh while language

// 

//:load matcher.scala

//:load parser3.scala

// some regular expressions

val SYM = RANGE("ABCDEFGHIJKLMNOPQRSTUVXYZabcdefghijklmnopqrstuvwxyz_")

val DIGIT = RANGE("0123456789")

val ID = SEQ(SYM, STAR(ALT(SYM, DIGIT))) 

val NUM = PLUS(DIGIT)

val KEYWORD = ALTS("skip", "while", "do", "if", "then", "else", "true", "false", "write") 

val SEMI: Rexp = ";"

val OP: Rexp = ALTS(":=", "=", "-", "+", "*", "!=", "<", ">")

val WHITESPACE = PLUS(RANGE(" \n"))

val RPAREN: Rexp = ")"

val LPAREN: Rexp = "("

val BEGIN: Rexp = "{"

val END: Rexp = "}"

val COMMENT = SEQS("/*", NOT(SEQS(STAR(ALLC), "*/", STAR(ALLC))), "*/")

// tokens for classifying the strings that have been recognised

abstract class Token

case object T_WHITESPACE extends Token

case object T_COMMENT extends Token

case object T_SEMI extends Token

case object T_LPAREN extends Token

case object T_RPAREN extends Token

case object T_BEGIN extends Token

case object T_END extends Token

case class T_ID(s: String) extends Token

case class T_OP(s: String) extends Token

case class T_NUM(s: String) extends Token

case class T_KWD(s: String) extends Token

val lexing_rules: List[Rule[Token]] = 

  List((KEYWORD, (s) => T_KWD(s.mkString)),

       (ID, (s) => T_ID(s.mkString)),

       (OP, (s) => T_OP(s.mkString)),

       (NUM, (s) => T_NUM(s.mkString)),

       (SEMI, (s) => T_SEMI),

       (LPAREN, (s) => T_LPAREN),

       (RPAREN, (s) => T_RPAREN),

       (BEGIN, (s) => T_BEGIN),

       (END, (s) => T_END),

       (WHITESPACE, (s) => T_WHITESPACE),

       (COMMENT, (s) => T_COMMENT))

// the tokenizer

val Tok = Tokenizer(lexing_rules, List(T_WHITESPACE, T_COMMENT))

// the abstract syntax trees

abstract class Stmt

abstract class AExp

abstract class BExp 

type Block = List[Stmt]

case object Skip extends Stmt

case class If(a: BExp, bl1: Block, bl2: Block) extends Stmt

case class While(b: BExp, bl: Block) extends Stmt

case class Assign(s: String, a: AExp) extends Stmt

case class Write(s: String) extends Stmt

case class Var(s: String) extends AExp

case class Num(i: Int) extends AExp

case class Aop(o: String, a1: AExp, a2: AExp) extends AExp

case object True extends BExp

case object False extends BExp

case class Relop(o: String, a1: AExp, a2: AExp) extends BExp

// atomic parsers

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)

case object NumParser extends Parser[List[Token], Int] {

  def parse(ts: List[Token]) = ts match {

    case T_NUM(s)::ts => Set((s.toInt, 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 ()

  }

}

// arithmetic expressions

lazy val AExp: Parser[List[Token], AExp] = 

  (T ~ T_OP("+") ~ AExp) ==> { case ((x, y), z) => Aop("+", x, z): AExp } ||

  (T ~ T_OP("-") ~ AExp) ==> { case ((x, y), z) => Aop("-", x, z): AExp } || T  

lazy val T: Parser[List[Token], AExp] = 

  (F ~ T_OP("*") ~ T) ==> { case ((x, y), z) => Aop("*", x, z): AExp } || F

lazy val F: Parser[List[Token], AExp] = 

  (T_LPAREN ~> AExp <~ T_RPAREN) || 

  IdParser ==> Var || 

  NumParser ==> Num

// boolean expressions

lazy val BExp: Parser[List[Token], BExp] = 

  (T_KWD("true") ==> ((_) => True: BExp)) || 

  (T_KWD("false") ==> ((_) => False: BExp)) ||

  (T_LPAREN ~> BExp <~ T_RPAREN) ||

  (AExp ~ T_OP("=") ~ AExp) ==> { case ((x, y), z) => Relop("=", x, z): BExp } || 

  (AExp ~ T_OP("!=") ~ AExp) ==> { case ((x, y), z) => Relop("!=", x, z): BExp } || 

  (AExp ~ T_OP("<") ~ AExp) ==> { case ((x, y), z) => Relop("<", x, z): BExp } || 

  (AExp ~ T_OP(">") ~ AExp) ==> { case ((x, y), z) => Relop("<", z, x): BExp } 

lazy val Stmt: Parser[List[Token], Stmt] =

  (T_KWD("skip") ==> ((_) => Skip: Stmt)) ||

  (IdParser ~ T_OP(":=") ~ AExp) ==> { case ((x, y), z) => Assign(x, z): Stmt } ||

  (T_KWD("if") ~ BExp ~ T_KWD("then") ~ Block ~ T_KWD("else") ~ Block) ==>

    { case (((((x,y),z),u),v),w) => If(y, u, w): Stmt } ||

  (T_KWD("while") ~ BExp ~ T_KWD("do") ~ Block) ==> { case (((x, y), z), w) => While(y, w) } || 

  (T_KWD("write") ~ IdParser) ==> { case (x, y) => Write(y) } 

lazy val Stmts: Parser[List[Token], Block] =

  (Stmt ~ T_SEMI ~ Stmts) ==> { case ((x, y), z) => x :: z : Block } ||

  (Stmt ==> ((s) => List(s) : Block))

lazy val Block: Parser[List[Token], Block] =

  (T_BEGIN ~> Stmts <~ T_END) || 

  (Stmt ==> ((s) => List(s)))

// compiler

val beginning = """

.class public XXX.XXX

.super java/lang/Object

.method public <init>()V

   aload_0

   invokenonvirtual java/lang/Object/<init>()V

   return

.end method

.method public static write(I)V 

    .limit locals 5 

    .limit stack 5 

    iload 0 

    getstatic java/lang/System/out Ljava/io/PrintStream; 

    swap 

    invokevirtual java/io/PrintStream/println(I)V 

    return 

.end method

.method public static main([Ljava/lang/String;)V

   .limit locals 200

   .limit stack 200

"""

val ending = """

   return

.end method

"""

// for generating new labels

var counter = -1

def Fresh(x: String) = {

  counter += 1

  x ++ "_" ++ counter.toString()

}

type Env = Map[String, String]

type Instrs = List[String]

def compile_aexp(a: AExp, env : Env) : Instrs = a match {

  case Num(i) => List("ldc " + i.toString + "\n")

  case Var(s) => List("iload " + env(s) + "\n")

  case Aop("+", a1, a2) => compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("iadd\n")

  case Aop("-", a1, a2) => compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("isub\n")

  case Aop("*", a1, a2) => compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("imul\n")

}

def compile_bexp(b: BExp, env : Env, jmp: String) : Instrs = b match {

  case True => Nil

  case False => List("goto " + jmp + "\n")

  case Relop("=", a1, a2) => 

    compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("if_icmpne " + jmp + "\n")

  case Relop("!=", a1, a2) => 

    compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("if_icmpeq " + jmp + "\n")

  case Relop("<", a1, a2) => 

    compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("if_icmpge " + jmp + "\n")

}

def compile_stmt(s: Stmt, env: Env) : (Instrs, Env) = s match {

  case Skip => (Nil, env)

  case Assign(x, a) => {

    val index = if (env.isDefinedAt(x)) env(x) else env.keys.size.toString

    (compile_aexp(a, env) ++ 

     List("istore " + index + "\n"), env + (x -> index))

  } 

  case If(b, bl1, bl2) => {

    val if_else = Fresh("If_else")

    val if_end = Fresh("If_end")

    val (instrs1, env1) = compile_bl(bl1, env)

    val (instrs2, env2) = compile_bl(bl2, env1)

    (compile_bexp(b, env, if_else) ++

     instrs1 ++

     List("goto " + if_end + "\n") ++

     List("\n" + if_else + ":\n\n") ++

     instrs2 ++

     List("\n" + if_end + ":\n\n"), env2)

  }

  case While(b, bl) => {

    val loop_begin = Fresh("Loop_begin")

    val loop_end = Fresh("Loop_end")

    val (instrs1, env1) = compile_bl(bl, env)

    (List("\n" + loop_begin + ":\n\n") ++

     compile_bexp(b, env, loop_end) ++

     instrs1 ++

     List("goto " + loop_begin + "\n") ++

     List("\n" + loop_end + ":\n\n"), env1)

  }

  case Write(x) => 

    (List("iload " + env(x) + "\n" + "invokestatic XXX/XXX/write(I)V\n"), env)

}

def compile_bl(bl: Block, env: Env) : (Instrs, Env) = bl match {

  case Nil => (Nil, env)

  case s::bl => {

    val (instrs1, env1) = compile_stmt(s, env)

    val (instrs2, env2) = compile_bl(bl, env1)

    (instrs1 ++ instrs2, env2)

  }

}

def compile(input: String) : String = {

  val class_name = input.split('.')(0)

  val tks = Tok.fromFile(input)

  val ast = Stmts.parse_single(tks)

  val instructions = compile_bl(ast, Map.empty)._1

  (beginning ++ instructions.mkString ++ ending).replaceAllLiterally("XXX", class_name)

}

def compile_to(input: String, output: String) = {

  val fw = new java.io.FileWriter(output) 

  fw.write(compile(input)) 

  fw.close()

}

//

val tks = Tok.fromString("x := x + 1")

val ast = Stmt.parse_single(tks)

println(compile_stmt(ast, Map("x" -> "n"))._1.mkString)

//examples

compile_to("loops.while", "loops.j")

//compile_to("fib.while", "fib.j")

// testing cases for time measurements

/*

def time_needed[T](i: Int, code: => T) = {

  val start = System.nanoTime()

  for (j <- 1 to i) code

  val end = System.nanoTime()

  (end - start)/(i * 1.0e9)

}

// for testing

import scala.sys.process._

val test_prog = """

start := XXX;

x := start;

y := start;

z := start;

while 0 < x do {

 while 0 < y do {

  while 0 < z do {

    z := z - 1

  };

  z := start;

  y := y - 1

 };     

 y := start;

 x := x - 1

};

write x;

write y;

write z

"""

def compile_test(n: Int) : Unit = {

  val class_name = "LOOP"

  val tks = Tok.fromString(test_prog.replaceAllLiterally("XXX", n.toString))

  val ast = Stmts.parse_single(tks)

  val instructions = compile_bl(ast, Map.empty)._1

  val assembly = (beginning ++ instructions.mkString ++ ending).replaceAllLiterally("XXX", class_name)

  val fw = new java.io.FileWriter(class_name + ".j") 

  fw.write(assembly) 

  fw.close()

  val test = ("java -jar jvm/jasmin-2.4/jasmin.jar " + class_name + ".j").!!

  println(n + " " + time_needed(2, ("java " + class_name + "/" + class_name).!!))

}

List(1, 5000, 10000, 50000, 100000, 250000, 500000, 750000, 1000000).map(compile_test(_))

// javabyte code assmbler

//

// java -jar jvm/jasmin-2.4/jasmin.jar loops.j

*/