| author | Christian Urban <christian.urban@kcl.ac.uk> | 
| Fri, 17 Oct 2025 11:20:49 +0100 | |
| changeset 1009 | 7fd1997bd14c | 
| parent 742 | 155426396b5f | 
| permissions | -rw-r--r-- | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 1 | |
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 2 | abstract class Rexp | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 3 | case object ZERO extends Rexp | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 4 | case object ONE extends Rexp | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 5 | case class CHAR(c: Char) extends Rexp | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 6 | case class ALT(r1: Rexp, r2: Rexp) extends Rexp | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 7 | case class SEQ(r1: Rexp, r2: Rexp) extends Rexp | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 8 | case class STAR(r: Rexp) extends Rexp | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 9 | case class NTIMES(r: Rexp, n: Int) extends Rexp | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 10 | case class UPNTIMES(r: Rexp, n: Int) extends Rexp | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 11 | |
| 623 | 12 | // the nullable function: tests whether the regular | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 13 | // expression can recognise the empty string | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 14 | def nullable (r: Rexp) : Boolean = r match {
 | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 15 | case ZERO => false | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 16 | case ONE => true | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 17 | case CHAR(_) => false | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 18 | case ALT(r1, r2) => nullable(r1) || nullable(r2) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 19 | case SEQ(r1, r2) => nullable(r1) && nullable(r2) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 20 | case STAR(_) => true | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 21 | case NTIMES(r, i) => if (i == 0) true else nullable(r) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 22 | case UPNTIMES(r: Rexp, n: Int) => true | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 23 | } | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 24 | |
| 623 | 25 | // the derivative of a regular expression w.r.t. a character | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 26 | def der (c: Char, r: Rexp) : Rexp = r match {
 | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 27 | case ZERO => ZERO | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 28 | case ONE => ZERO | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 29 | case CHAR(d) => if (c == d) ONE else ZERO | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 30 | case ALT(r1, r2) => ALT(der(c, r1), der(c, r2)) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 31 | case SEQ(r1, r2) => | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 32 | if (nullable(r1)) ALT(SEQ(der(c, r1), r2), der(c, r2)) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 33 | else SEQ(der(c, r1), r2) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 34 | case STAR(r1) => SEQ(der(c, r1), STAR(r1)) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 35 | case NTIMES(r1, i) => | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 36 | if (i == 0) ZERO else | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 37 | if (nullable(r1)) SEQ(der(c, r1), UPNTIMES(r1, i - 1)) | 
| 457 | 38 | else SEQ(der(c, r1), NTIMES(r1, i - 1)) | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 39 | case UPNTIMES(r1, i) => | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 40 | if (i == 0) ZERO | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 41 | else SEQ(der(c, r1), UPNTIMES(r1, i - 1)) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 42 | } | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 43 | |
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 44 | def simp(r: Rexp) : Rexp = r match {
 | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 45 |   case ALT(r1, r2) => (simp(r1), simp(r2)) match {
 | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 46 | case (ZERO, r2s) => r2s | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 47 | case (r1s, ZERO) => r1s | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 48 | case (r1s, r2s) => if (r1s == r2s) r1s else ALT (r1s, r2s) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 49 | } | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 50 |   case SEQ(r1, r2) =>  (simp(r1), simp(r2)) match {
 | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 51 | case (ZERO, _) => ZERO | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 52 | case (_, ZERO) => ZERO | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 53 | case (ONE, r2s) => r2s | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 54 | case (r1s, ONE) => r1s | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 55 | case (r1s, r2s) => SEQ(r1s, r2s) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 56 | } | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 57 | case r => r | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 58 | } | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 59 | |
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 60 | |
| 623 | 61 | // the derivative w.r.t. a string (iterates der) | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 62 | def ders (s: List[Char], r: Rexp) : Rexp = s match {
 | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 63 | case Nil => r | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 64 | case c::s => ders(s, simp(der(c, r))) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 65 | } | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 66 | |
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 67 | |
| 623 | 68 | // the main matcher function | 
| 69 | def matches(r: Rexp, s: String) : Boolean = nullable(ders(s.toList, r)) | |
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 70 | |
| 623 | 71 | // one or zero | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 72 | def OPT(r: Rexp) = ALT(r, ONE) | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 73 | |
| 623 | 74 | // evil regular expressions | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 75 | def EVIL1(n: Int) = SEQ(NTIMES(OPT(CHAR('a')), n), NTIMES(CHAR('a'), n))
 | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 76 | val EVIL2 = SEQ(STAR(STAR(CHAR('a'))), CHAR('b'))
 | 
| 471 | 77 | val EVIL3 = SEQ(STAR(ALT(CHAR('a'), SEQ(CHAR('a'),CHAR('a')))), CHAR('b'))
 | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 78 | |
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 79 | def time_needed[T](i: Int, code: => T) = {
 | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 80 | val start = System.nanoTime() | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 81 | for (j <- 1 to i) code | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 82 | val end = System.nanoTime() | 
| 623 | 83 | "%.5f".format((end - start) / (i * 1.0e9)) | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 84 | } | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 85 | |
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 86 | |
| 623 | 87 | // test: (a?{n}) (a{n})
 | 
| 88 | for (i <- 0 to 8000 by 1000) {
 | |
| 89 |   println(s"$i: ${time_needed(2, matches(EVIL1(i), "a" * i))}")
 | |
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 90 | } | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 91 | |
| 623 | 92 | // test: (a*)* b | 
| 93 | for (i <- 0 to 6000000 by 500000) {
 | |
| 94 |   println(s"$i: ${time_needed(2, matches(EVIL2, "a" * i))}")
 | |
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 95 | } | 
| 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 96 | |
| 471 | 97 | |
| 98 | val r0 = simp(der('a', EVIL3))
 | |
| 99 | val r1 = simp(der('a', r0))
 | |
| 100 | val r2 = simp(der('a', r1))
 | |
| 101 | val r3 = simp(der('a', r2))
 | |
| 102 | val r4 = simp(der('a', r3))
 | |
| 103 | val r5 = simp(der('a', r4))
 | |
| 104 | val r6 = simp(der('a', r5))
 | |
| 105 | ||
| 106 | //test: (a|aa)* b | |
| 519 | 107 | /* | 
| 623 | 108 | for (i <- 0 to 100 by 10) {
 | 
| 109 |   println(s"$i: ${time_needed(2, matches(EVIL3, "a" * i ++ "c"))}")
 | |
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 110 | } | 
| 519 | 111 | */ | 
| 455 
192f4c59633e
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: diff
changeset | 112 |