// Author: Zhuo Ying Jiang Li

// Starting code by Dr Christian Urban

// lexer

// Use this command to print the list of tokens:

// amm fun_token.sc <name>.fun

type Token = (String, String)

type Tokens = List[Token]

// regular expressions including records

abstract class Rexp

case class RANGE(chars: List[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 OPTIONAL(r: Rexp) extends Rexp

case class PLUS(r: Rexp) extends Rexp

case class NTIMES(r: Rexp, n: Int) extends Rexp

case class RECD(x: String, r: Rexp) extends Rexp  // records for extracting strings or tokens

// values

case class Opt(v: Val) extends Val

case class Pls(vs: List[Val]) extends Val

case class Nt(vs: List[Val]) extends Val

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

implicit def string2rexp(s : String) : Rexp =

  def ? = OPTIONAL(r)

  def + = PLUS(r)

  def ^ (n: Int) = NTIMES(r, n)

  def ? = OPTIONAL(s)

  def + = PLUS(s)

  def ^ (n: Int) = NTIMES(s, n)

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

  case RANGE(_) => false

  case OPTIONAL(r1) => true

  case PLUS(r1) => nullable(r1)

  case NTIMES(r1, n) => if (n == 0) true else nullable(r1)

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

  case RANGE(chars) => if (chars.contains(c)) ONE else ZERO

  case SEQ(r1, r2) =>

  case OPTIONAL(r) => der(c, r)

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

  case NTIMES(r1, n) => if (n == 0) ZERO else SEQ(der(c, r1), NTIMES(r1, n - 1))

// extracts a string from a value

  case Sequ(v1, v2) => flatten(v1) ++ flatten(v2)

  case Opt(v) => flatten(v)

  case Pls(vs) => vs.map(flatten).mkString

  case Nt(vs) => vs.map(flatten).mkString

// used for tokenising a string

def env(v: Val) : Tokens = v match {

  case Opt(v) => env(v)

  case Pls(vs) => vs.flatMap(env)

  case Nt(vs) => vs.flatMap(env)

// The injection and mkeps part of the lexer

//===========================================

  case RANGE(chars) => throw new Exception("lexing error")  // this will never be called but the coursework asks for it so...

  case ALT(r1, r2) =>

  case OPTIONAL(r) => Opt(Empty)

  case PLUS(r) => Pls(List(mkeps(r))) // scala define a list with one element

  case NTIMES(r, n) => if (n == 0) Nt(Nil) else Nt(List.fill(n)(mkeps(r))) // wrong

  case _ => throw new Exception("lexing error")

  case (CHAR(d), Empty) => Chr(c)

  case (RANGE(chars), Empty) => Chr(c)

  case (OPTIONAL(r1), v) => Opt(inj(r1, c, v))

  case (PLUS(r1), Sequ(v1, Stars(vs))) => Pls(inj(r1, c, v1)::vs)

  case (NTIMES(r1, n), Sequ(v1, Nt(vs))) => Nt(inj(r1, c, v1)::vs)

def F_SEQ_Empty1(f1: Val => Val, f2: Val => Val) =

def F_SEQ_Empty2(f1: Val => Val, f2: Val => Val) =

// simplification

                else (ALT (r1s, r2s), F_ALT(f1s, f2s))

  case Nil => if (nullable(r)) mkeps(r) else

    { throw new Exception("lexing error") }

def lexing_simp(r: Rexp, s: String) =

  env(lex_simp(r, s.toList))

// FUN language lexer

val DIGIT = RANGE("0123456789".toList)

val LOWERCASE = RANGE("abcdefghijklmnopqrstuvwxyz".toList)

val UPPERCASE = RANGE("ABCDEFGHIJKLMNOPQRSTUVWXYZ".toList)

val SYM = RANGE("!\"#$%&'()*+,-./:;<>=?`@[]\\^_{}|~".toList)  // I referenced the CPP ASCII table https://en.cppreference.com/w/cpp/language/ascii

val KEYWORD : Rexp = "val" | "if" | "then" | "else" | "def" | "skip" // "skip" is hardcoded because hanoi.fun calls skip() without parentheses

val TYPE : Rexp = "Int" | "Double" | "Void"

val GLOBAL_ID : Rexp = UPPERCASE ~ ("_" | LOWERCASE | DIGIT | UPPERCASE).% // start with capital letter and followed by any case

val ID : Rexp = LOWERCASE ~ ("_" | UPPERCASE | LOWERCASE | DIGIT).% // start with lowercase 

val SEMI : Rexp = ";"

val COLON : Rexp = ":"

val OP : Rexp = "=" | "==" | "-" | "+" | "*" | "!=" | "<" | ">" | "<=" | ">=" | "%" | "/" // no && and || operators

val INT : Rexp = DIGIT.+

val DOUBLE : Rexp = DIGIT.+ ~ "." ~ DIGIT.+  // negative numbers sign is lexed as operator, but the parser will identify negative numbers

val COMMA : Rexp = "," 

val WHITESPACES: Rexp = (" " | "\n" | "\t" | "\r").+ // whitespaces are either " " or \n or \t or \r

val LPAREN : Rexp = RANGE("({".toList)

val RPAREN : Rexp = RANGE(")}".toList)

val CH : Rexp = "'" ~ (LOWERCASE | UPPERCASE | DIGIT | SYM | " " | "\\n" | "\\t" | "\\r") ~ "'"  // \n, \t and \r should also be tokenized, any character should be, whitespaces too

val COMMENT : Rexp = ("//" ~ (LOWERCASE | UPPERCASE | SYM | DIGIT | RANGE(" \t\r".toList)).% ~ "\n") | ("/*" ~ (LOWERCASE | UPPERCASE | SYM | DIGIT | RANGE(" \n\t\r".toList)).% ~ "*/")

val FUN_REGS = (("keyword" $ KEYWORD) |

                ("type" $ TYPE) |

                ("global" $ GLOBAL_ID) |

                ("id" $ ID) |

                ("op" $ OP) |

                ("double" $ DOUBLE) |

                ("int" $ INT) |

                ("semi" $ SEMI) |

                ("colon" $ COLON) |

                ("comma" $ COMMA) |

                ("ch" $ CH) |

                ("par" $ (LPAREN | RPAREN)) |

                COMMENT | WHITESPACES).%

def fun_lex(program: String) : Tokens = {

  lexing_simp(FUN_REGS, program)

def tokenise(program: String) : Tokens = {

  lexing_simp(FUN_REGS, program)

import scala.io.Source._

def lex(filename: String) = {

  // read file

  val fun_code = fromFile(filename).getLines.mkString("\n")

  // print tokens to screen

  println(fun_lex(fun_code).mkString("\n"))