solutions/cw4/parser.sc
changeset 894 02ef5c3abc51
parent 864 b5b1bc0a603b
child 920 7af2eea19646
equal deleted inserted replaced
893:54a483a33763 894:02ef5c3abc51
       
     1 // CW3
       
     2 
       
     3 import $file.lexer
       
     4 import lexer._ 
       
     5 
       
     6 
       
     7 case class ~[+A, +B](_1: A, _2: B)
       
     8 type IsSeq[A] = A => Seq[_]
       
     9 
       
    10 abstract class Parser[I : IsSeq, T] {
       
    11   def parse(ts: I): Set[(T, I)]
       
    12 
       
    13   def parse_all(ts: I) : Set[T] =
       
    14     for ((head, tail) <- parse(ts); if tail.isEmpty) yield head
       
    15 }
       
    16 
       
    17 class SeqParser[I : IsSeq, T, S](p: => Parser[I, T], q: => Parser[I, S]) extends Parser[I, ~[T, S]] {
       
    18   def parse(sb: I) = 
       
    19     for ((head1, tail1) <- p.parse(sb); 
       
    20          (head2, tail2) <- q.parse(tail1)) yield (new ~(head1, head2), tail2)
       
    21 }
       
    22 
       
    23 class AltParser[I : IsSeq, T](p: => Parser[I, T], q: => Parser[I, T]) extends Parser[I, T] {
       
    24   def parse(sb: I) = p.parse(sb) ++ q.parse(sb)   
       
    25 }
       
    26 
       
    27 class FunParser[I : IsSeq, T, S](p: => Parser[I, T], f: T => S) extends Parser[I, S] {
       
    28   def parse(sb: I) = 
       
    29     for ((head, tail) <- p.parse(sb)) yield (f(head), tail)
       
    30 }
       
    31 
       
    32 // New parser that takes as input a list of tokens
       
    33 case class TokenListParser(ts: List[Token]) extends Parser[List[Token], List[Token]] {
       
    34     def parse(tsb: List[Token]) = {
       
    35         val (prefix, suffix) = tsb.splitAt(ts.length)
       
    36         if (prefix == ts) Set((prefix, suffix)) else Set()
       
    37     }
       
    38 }
       
    39 
       
    40 // Implicit definitions to go from a token 
       
    41 // or a list of tokens to a TokenListParser
       
    42 implicit def token2parser(t: Token) = TokenListParser(List(t))
       
    43 implicit def tokenList2parser(ts: List[Token]) = TokenListParser(ts)
       
    44 
       
    45 implicit def ParserOps[I : IsSeq, T](p: Parser[I, T]) = new {
       
    46   def || (q : => Parser[I, T]) = new AltParser[I, T](p, q)
       
    47   def ==>[S] (f: => T => S) = new FunParser[I, T, S](p, f)
       
    48   def ~[S] (q : => Parser[I, S]) = new SeqParser[I, T, S](p, q)
       
    49 }
       
    50 
       
    51 implicit def TokenOps(t: Token) = new {
       
    52     def || (q : => Parser[List[Token], List[Token]]) = new AltParser[List[Token], List[Token]](List(t), q)
       
    53     def || (qs : List[Token]) = new AltParser[List[Token], List[Token]](List(t), qs)
       
    54     def ==>[S] (f: => List[Token] => S) = new FunParser[List[Token], List[Token], S](List(t), f)
       
    55     def ~[S](q : => Parser[List[Token], S]) =
       
    56         new SeqParser[List[Token], List[Token], S](List(t), q)
       
    57     def ~ (qs : List[Token]) =
       
    58         new SeqParser[List[Token], List[Token], List[Token]](List(t), qs)
       
    59 }
       
    60 
       
    61 implicit def TokenListOps(ts: List[Token]) = new {
       
    62     def || (q : => Parser[List[Token], List[Token]]) = new AltParser[List[Token], List[Token]](ts, q)
       
    63     def || (qs : List[Token]) = new AltParser[List[Token], List[Token]](ts, qs)
       
    64     def ==>[S] (f: => List[Token] => S) = new FunParser[List[Token], List[Token], S](ts, f)
       
    65     def ~[S](q : => Parser[List[Token], S]) =
       
    66         new SeqParser[List[Token], List[Token], S](ts, q)
       
    67     def ~ (qs : List[Token]) =
       
    68         new SeqParser[List[Token], List[Token], List[Token]](ts, qs)
       
    69 }
       
    70 
       
    71 // Abstract Syntax Trees
       
    72 abstract class Stmt
       
    73 abstract class AExp
       
    74 abstract class BExp
       
    75 
       
    76 type Block = List[Stmt]
       
    77 
       
    78 case object Skip extends Stmt
       
    79 case class If(a: BExp, bl1: Block, bl2: Block) extends Stmt
       
    80 case class While(b: BExp, bl: Block) extends Stmt
       
    81 case class Assign(s: String, a: AExp) extends Stmt
       
    82 case class Read(s: String) extends Stmt
       
    83 case class WriteId(s: String) extends Stmt  // for printing values of variables
       
    84 case class WriteString(s: String) extends Stmt  // for printing words
       
    85 case class For(counter: String, lower: AExp, upper: AExp, code: Block) extends Stmt
       
    86 
       
    87 
       
    88 case class Var(s: String) extends AExp
       
    89 case class Num(i: Int) extends AExp
       
    90 case class Aop(o: String, a1: AExp, a2: AExp) extends AExp
       
    91 
       
    92 case object True extends BExp
       
    93 case object False extends BExp
       
    94 case class Bop(o: String, a1: AExp, a2: AExp) extends BExp
       
    95 case class And(b1: BExp, b2: BExp) extends BExp
       
    96 case class Or(b1: BExp, b2: BExp) extends BExp
       
    97 
       
    98 case class IdParser() extends Parser[List[Token], String] {
       
    99     def parse(tsb: List[Token]) = tsb match {
       
   100         case T_ID(id) :: rest => Set((id, rest))
       
   101         case _ => Set()
       
   102     }
       
   103 }
       
   104 
       
   105 case class NumParser() extends Parser[List[Token], Int] {
       
   106     def parse(tsb: List[Token]) = tsb match {
       
   107         case T_NUM(n) :: rest => Set((n, rest))
       
   108         case _ => Set()
       
   109     }
       
   110 }
       
   111 
       
   112 case class StringParser() extends Parser[List[Token], String] {
       
   113     def parse(tsb: List[Token]) = tsb match {
       
   114         case T_STRING(s) :: rest => Set((s, rest))
       
   115         case _ => Set()
       
   116     }
       
   117 }
       
   118 
       
   119 // WHILE Language Parsing
       
   120 lazy val AExp: Parser[List[Token], AExp] = 
       
   121   (Te ~ T_OP("+") ~ AExp) ==> { case x ~ _ ~ z => Aop("+", x, z): AExp } ||
       
   122   (Te ~ T_OP("-") ~ AExp) ==> { case x ~ _ ~ z => Aop("-", x, z): AExp } || Te
       
   123 lazy val Te: Parser[List[Token], AExp] = 
       
   124   (Fa ~ T_OP("*") ~ Te) ==> { case x ~ _ ~ z => Aop("*", x, z): AExp } || 
       
   125   (Fa ~ T_OP("/") ~ Te) ==> { case x ~ _ ~ z => Aop("/", x, z): AExp } || 
       
   126   (Fa ~ T_OP("%") ~ Te) ==> { case x ~ _ ~ z => Aop("%", x, z): AExp } || Fa  
       
   127 lazy val Fa: Parser[List[Token], AExp] = 
       
   128    (T_PAREN("(") ~ AExp ~ T_PAREN(")")) ==> { case _ ~ y ~ _ => y } || 
       
   129    IdParser() ==> Var  || 
       
   130    NumParser() ==> Num
       
   131 
       
   132 lazy val BExp: Parser[List[Token], BExp] = 
       
   133    (AExp ~ T_OP("==") ~ AExp) ==> { case x ~ _ ~ z => Bop("==", x, z): BExp } || 
       
   134    (AExp ~ T_OP("!=") ~ AExp) ==> { case x ~ _ ~ z => Bop("!=", x, z): BExp } || 
       
   135    (AExp ~ T_OP("<") ~ AExp) ==> { case x ~ _ ~ z => Bop("<", x, z): BExp } || 
       
   136    (AExp ~ T_OP(">") ~ AExp) ==> { case x ~ _ ~ z => Bop(">", x, z): BExp } ||
       
   137    (T_PAREN("(") ~ BExp ~ List(T_PAREN(")"), T_OP("&&")) ~ BExp) ==> { case _ ~ y ~ _ ~ v => And(y, v): BExp } ||
       
   138    (T_PAREN("(") ~ BExp ~ List(T_PAREN(")"), T_OP("||")) ~ BExp) ==> { case _ ~ y ~ _ ~ v => Or(y, v): BExp } ||
       
   139    (T_KEYWORD("true") ==> (_ => True: BExp )) || 
       
   140    (T_KEYWORD("false") ==> (_ => False: BExp )) ||
       
   141    (T_PAREN("(") ~ BExp ~ T_PAREN(")")) ==> { case _ ~ x ~ _ => x }
       
   142 
       
   143 lazy val Stmt: Parser[List[Token], Stmt] =
       
   144     T_KEYWORD("skip") ==> (_ => Skip: Stmt) ||
       
   145     (IdParser() ~ T_OP(":=") ~ AExp) ==> { case id ~ _ ~ z => Assign(id, z): Stmt } ||
       
   146     (T_KEYWORD("if") ~ BExp ~ T_KEYWORD("then") ~ Block ~ T_KEYWORD("else") ~ Block) ==> { case _ ~ y ~ _ ~ u ~ _ ~ w => If(y, u, w): Stmt } ||
       
   147     (T_KEYWORD("while") ~ BExp ~ T_KEYWORD("do") ~ Block) ==> { case _ ~ y ~ _ ~ w => While(y, w) : Stmt } ||
       
   148     (T_KEYWORD("read") ~ IdParser()) ==> { case _ ~ id => Read(id): Stmt} ||
       
   149     (T_KEYWORD("write") ~ IdParser()) ==> { case _ ~ id => WriteId(id): Stmt} ||
       
   150     (T_KEYWORD("write") ~ StringParser()) ==> { case _ ~ s => WriteString(s): Stmt} ||
       
   151     (T_KEYWORD("for") ~ IdParser() ~ T_OP(":=") ~ AExp ~ T_KEYWORD("upto") ~ AExp ~ T_KEYWORD("do") ~ Block) ==> {
       
   152       case _ ~ id ~ _ ~ lower ~ _ ~ upper ~ _ ~ blck => For(id, lower, upper, blck): Stmt
       
   153     }
       
   154 
       
   155 lazy val Stmts: Parser[List[Token], Block] =
       
   156     (Stmt ~ T_SEMI ~ Stmts) ==> { case x ~ _ ~ z => x :: z : Block } ||
       
   157     (Stmt ==> (s => List(s) : Block))
       
   158 
       
   159 lazy val Block: Parser[List[Token], Block] =
       
   160     (T_PAREN("{") ~ Stmts ~ T_PAREN("}")) ==> { case x ~ y ~ z => y} ||
       
   161     (Stmt ==> (s => List(s)))
       
   162