diff -r 3020f8c76baa -r 9e7897f25e13 testing4/re.scala --- a/testing4/re.scala Wed Nov 28 23:26:47 2018 +0000 +++ b/testing4/re.scala Thu Nov 29 17:15:11 2018 +0000 @@ -1,8 +1,9 @@ // Part 1 about Regular Expression Matching //========================================== -object CW8a { +//object CW9a { +// Regular Expressions abstract class Rexp case object ZERO extends Rexp case object ONE extends Rexp @@ -38,10 +39,11 @@ def ~ (r: String) = SEQ(s, r) } -// (1a) Complete the function nullable according to +// (1) Complete the function nullable according to // the definition given in the coursework; this // function checks whether a regular expression -// can match the empty string +// can match the empty string and Returns a boolean +// accordingly. def nullable (r: Rexp) : Boolean = r match { case ZERO => false @@ -52,10 +54,10 @@ case STAR(_) => true } -// (1b) Complete the function der according to +// (2) Complete the function der according to // the definition given in the coursework; this // function calculates the derivative of a -// regular expression w.r.t. a character +// regular expression w.r.t. a character. def der (c: Char, r: Rexp) : Rexp = r match { case ZERO => ZERO @@ -68,11 +70,12 @@ case STAR(r1) => SEQ(der(c, r1), STAR(r1)) } -// (1c) Complete the function der according to +// (3) Complete the simp function according to // the specification given in the coursework; this -// function simplifies a regular expression; -// however it does not simplify inside STAR-regular -// expressions +// function simplifies a regular expression from +// the inside out, like you would simplify arithmetic +// expressions; however it does not simplify inside +// STAR-regular expressions. def simp(r: Rexp) : Rexp = r match { case ALT(r1, r2) => (simp(r1), simp(r2)) match { @@ -90,11 +93,12 @@ case r => r } -// (1d) Complete the two functions below; the first + +// (4) Complete the two functions below; the first // calculates the derivative w.r.t. a string; the second // is the regular expression matcher taking a regular // expression and a string and checks whether the -// string matches the regular expression +// string matches the regular expression. def ders (s: List[Char], r: Rexp) : Rexp = s match { case Nil => r @@ -102,12 +106,13 @@ } // main matcher function -def matcher(r: Rexp, s: String): Boolean = nullable(ders(s.toList, r)) +def matcher(r: Rexp, s: String) = nullable(ders(s.toList, r)) -// (1e) Complete the size function for regular -// expressions according to the specification +// (5) Complete the size function for regular +// expressions according to the specification // given in the coursework. + def size(r: Rexp): Int = r match { case ZERO => 1 case ONE => 1 @@ -138,11 +143,13 @@ size(simp(der('a', der('a', EVIL)))) // => 8 size(simp(der('a', der('a', der('a', EVIL))))) // => 8 -// Java needs around 30 seconds for matching 28 a's with EVIL. +// Python needs around 30 seconds for matching 28 a's with EVIL. +// Java 9 and later increase this to an "astonishing" 40000 a's in +// around 30 seconds. // // Lets see how long it takes to match strings with -// 0.5 Million a's...it should be in the range of some -// seconds. +// 5 Million a's...it should be in the range of a +// couple of seconds. def time_needed[T](i: Int, code: => T) = { val start = System.nanoTime() @@ -154,6 +161,16 @@ for (i <- 0 to 5000000 by 500000) { println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL, "a" * i)))) } + +// another "power" test case +simp(Iterator.iterate(ONE:Rexp)(r => SEQ(r, ONE | ONE)).drop(100).next) == ONE + +// the Iterator produces the rexp +// +// SEQ(SEQ(SEQ(..., ONE | ONE) , ONE | ONE), ONE | ONE) +// +// where SEQ is nested 100 times. + */ -} +//}