diff -r 698ed1c96cd0 -r 4ce07c4abdb4 progs/compile.scala --- a/progs/compile.scala Thu Apr 09 07:42:23 2015 +0100 +++ b/progs/compile.scala Fri Apr 10 18:02:04 2015 +0100 @@ -1,134 +1,33 @@ -// A Compiler for the WHILE language +// A Small Compiler for the WHILE Language // -import matcher._ -import parser._ - -// 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[(Rexp, List[Char] => 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] + +// statements 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 Read(s: String) extends Stmt +// arithmetic expressions 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 +// boolean expressions 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 () - } -} +case class Bop(o: String, a1: AExp, a2: AExp) extends BExp -// 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 +// compiler headers needed for the JVM +// (contains an init method, as well as methods for read and write) val beginning = """ .class public XXX.XXX .super java/lang/Object @@ -149,6 +48,39 @@ return .end method +.method public static read()I + .limit locals 10 + .limit stack 10 + + ldc 0 + istore 1 ; this will hold our final integer +Label1: + getstatic java/lang/System/in Ljava/io/InputStream; + invokevirtual java/io/InputStream/read()I + istore 2 + iload 2 + ldc 10 ; the newline delimiter + isub + ifeq Label2 + iload 2 + ldc 32 ; the space delimiter + isub + ifeq Label2 + + iload 2 + ldc 48 ; we have our digit in ASCII, have to subtract it from 48 + isub + ldc 10 + iload 1 + imul + iadd + istore 1 + goto Label1 +Label2: + ;when we come here we have our integer computed in local variable 1 + iload 1 + ireturn +.end method .method public static main([Ljava/lang/String;)V .limit locals 200 @@ -171,9 +103,11 @@ x ++ "_" ++ counter.toString() } +// environments and instructions type Env = Map[String, String] type Instrs = List[String] +// arithmetic expression compilation 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") @@ -182,18 +116,19 @@ case Aop("*", a1, a2) => compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("imul\n") } +// boolean expression compilation 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) => + case Bop("=", a1, a2) => compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("if_icmpne " + jmp + "\n") - case Relop("!=", a1, a2) => + case Bop("!=", a1, a2) => compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("if_icmpeq " + jmp + "\n") - case Relop("<", a1, a2) => + case Bop("<", a1, a2) => compile_aexp(a1, env) ++ compile_aexp(a2, env) ++ List("if_icmpge " + jmp + "\n") } - +// statement compilation def compile_stmt(s: Stmt, env: Env) : (Instrs, Env) = s match { case Skip => (Nil, env) case Assign(x, a) => { @@ -204,8 +139,8 @@ 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) + val (instrs1, env1) = compile_block(bl1, env) + val (instrs2, env2) = compile_block(bl2, env1) (compile_bexp(b, env, if_else) ++ instrs1 ++ List("goto " + if_end + "\n") ++ @@ -216,7 +151,7 @@ case While(b, bl) => { val loop_begin = Fresh("Loop_begin") val loop_end = Fresh("Loop_end") - val (instrs1, env1) = compile_bl(bl, env) + val (instrs1, env1) = compile_block(bl, env) (List("\n" + loop_begin + ":\n\n") ++ compile_bexp(b, env, loop_end) ++ instrs1 ++ @@ -225,102 +160,58 @@ } case Write(x) => (List("iload " + env(x) + "\n" + "invokestatic XXX/XXX/write(I)V\n"), env) + case Read(x) => { + val index = if (env.isDefinedAt(x)) env(x) else env.keys.size.toString + (List("invokestatic XXX/XXX/read()I\n" + + "istore " + index + "\n"), env + (x -> index)) + } } -def compile_bl(bl: Block, env: Env) : (Instrs, Env) = bl match { +// compilation of a block (i.e. list of instructions) +def compile_block(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) + val (instrs2, env2) = compile_block(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 +// main compilation function for blocks +def compile(bl: Block, class_name: String) : String = { + val instructions = compile_block(bl, 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) +// Fibonacci numbers as a test-case +val fib_test = + List(Read("n"), // read n; + Assign("minus1",Num(0)), // minus1 := 0; + Assign("minus2",Num(1)), // minus2 := 1; + Assign("temp",Num(0)), // temp := 0; + While(Bop("<",Num(0),Var("n")), // while n > 0 do { + List(Assign("temp",Var("minus2")), // temp := minus2; + Assign("minus2",Aop("+",Var("minus1"),Var("minus2"))), // minus2 := minus1 + minus2; + Assign("minus1",Var("temp")), // minus1 := temp; + Assign("n",Aop("-",Var("n"),Num(1))))), // n := n - 1 }; + Write("minus1")) // write minus1 -//examples - -compile_to("loops.while", "loops.j") -//compile_to("fib.while", "fib.j") - - -// testing cases for time measurements +// prints out the JVM-assembly program -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._ +println(compile(fib_test, "fib")) -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 +// can be assembled with +// +// java -jar jvm/jasmin-2.4/jasmin.jar fib.j // -// java -jar jvm/jasmin-2.4/jasmin.jar loops.j +// and started with +// +// java fib/fib -