| author | Christian Urban <urbanc@in.tum.de> | 
| Sat, 07 Jan 2017 14:52:26 +0000 | |
| changeset 471 | e5df48ff7033 | 
| parent 454 | 010237a7dae7 | 
| child 477 | 28e872e7efb3 | 
| permissions | -rw-r--r-- | 
| 471 | 1 | // version with explicit an n-times regular expression | 
| 2 | // this keeps the regular expression small | |
| 422 | 3 | |
| 120 
3e71efb25ce9
added
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
93diff
changeset | 4 | abstract class Rexp | 
| 424 
1129024b26d5
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
422diff
changeset | 5 | case object ZERO extends Rexp | 
| 
1129024b26d5
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
422diff
changeset | 6 | case object ONE extends Rexp | 
| 49 | 7 | case class CHAR(c: Char) extends Rexp | 
| 120 
3e71efb25ce9
added
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
93diff
changeset | 8 | case class ALT(r1: Rexp, r2: Rexp) extends Rexp | 
| 
3e71efb25ce9
added
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
93diff
changeset | 9 | case class SEQ(r1: Rexp, r2: Rexp) extends Rexp | 
| 
3e71efb25ce9
added
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
93diff
changeset | 10 | case class STAR(r: Rexp) extends Rexp | 
| 258 
1e4da6d2490c
updated programs
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
121diff
changeset | 11 | case class NTIMES(r: Rexp, n: Int) extends Rexp //explicit constructor | 
| 49 | 12 | |
| 13 | def nullable (r: Rexp) : Boolean = r match {
 | |
| 424 
1129024b26d5
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
422diff
changeset | 14 | case ZERO => false | 
| 
1129024b26d5
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
422diff
changeset | 15 | case ONE => true | 
| 49 | 16 | case CHAR(_) => false | 
| 17 | case ALT(r1, r2) => nullable(r1) || nullable(r2) | |
| 18 | case SEQ(r1, r2) => nullable(r1) && nullable(r2) | |
| 19 | case STAR(_) => true | |
| 120 
3e71efb25ce9
added
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
93diff
changeset | 20 | case NTIMES(r, i) => if (i == 0) true else nullable(r) | 
| 49 | 21 | } | 
| 22 | ||
| 23 | def der (c: Char, r: Rexp) : Rexp = r match {
 | |
| 424 
1129024b26d5
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
422diff
changeset | 24 | case ZERO => ZERO | 
| 
1129024b26d5
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
422diff
changeset | 25 | case ONE => ZERO | 
| 
1129024b26d5
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
422diff
changeset | 26 | case CHAR(d) => if (c == d) ONE else ZERO | 
| 49 | 27 | case ALT(r1, r2) => ALT(der(c, r1), der(c, r2)) | 
| 28 | case SEQ(r1, r2) => | |
| 29 | if (nullable(r1)) ALT(SEQ(der(c, r1), r2), der(c, r2)) | |
| 30 | else SEQ(der(c, r1), r2) | |
| 440 
e14cd32ad497
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
434diff
changeset | 31 | case STAR(r1) => SEQ(der(c, r1), STAR(r1)) | 
| 
e14cd32ad497
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
434diff
changeset | 32 | case NTIMES(r1, i) => | 
| 453 
a5525c780b96
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
441diff
changeset | 33 | if (i == 0) ZERO else der(c, SEQ(r1, NTIMES(r1, i - 1))) | 
| 49 | 34 | } | 
| 35 | ||
| 36 | def ders (s: List[Char], r: Rexp) : Rexp = s match {
 | |
| 37 | case Nil => r | |
| 120 
3e71efb25ce9
added
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
93diff
changeset | 38 | case c::s => ders(s, der(c, r)) | 
| 49 | 39 | } | 
| 40 | ||
| 261 
24531cfaa36a
updated handouts
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
258diff
changeset | 41 | def matches(r: Rexp, s: String) : Boolean = nullable(ders(s.toList, r)) | 
| 49 | 42 | |
| 43 | ||
| 258 
1e4da6d2490c
updated programs
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
121diff
changeset | 44 | //optional: one or zero times | 
| 424 
1129024b26d5
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
422diff
changeset | 45 | def OPT(r: Rexp) = ALT(r, ONE) | 
| 49 | 46 | |
| 422 | 47 | //evil regular expressions | 
| 414 | 48 | def EVIL1(n: Int) = SEQ(NTIMES(OPT(CHAR('a')), n), NTIMES(CHAR('a'), n))
 | 
| 440 
e14cd32ad497
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
434diff
changeset | 49 | val EVIL2 = SEQ(STAR(STAR(CHAR('a'))), CHAR('b'))
 | 
| 49 | 50 | |
| 51 | def time_needed[T](i: Int, code: => T) = {
 | |
| 52 | val start = System.nanoTime() | |
| 53 | for (j <- 1 to i) code | |
| 54 | val end = System.nanoTime() | |
| 55 | (end - start)/(i * 1.0e9) | |
| 56 | } | |
| 57 | ||
| 58 | ||
| 434 
8664ff87cd77
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
424diff
changeset | 59 | //test: (a?{n}) (a{n})
 | 
| 454 
010237a7dae7
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
453diff
changeset | 60 | for (i <- 1 to 1001 by 10) {
 | 
| 415 | 61 | println(i + " " + "%.5f".format(time_needed(2, matches(EVIL1(i), "a" * i)))) | 
| 120 
3e71efb25ce9
added
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
93diff
changeset | 62 | } | 
| 
3e71efb25ce9
added
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
93diff
changeset | 63 | |
| 454 
010237a7dae7
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
453diff
changeset | 64 | for (i <- 1 to 1001 by 10) {
 | 
| 434 
8664ff87cd77
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
424diff
changeset | 65 | println(i + " " + "%.5f".format(time_needed(2, matches(EVIL1(i), "a" * i)))) | 
| 
8664ff87cd77
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
424diff
changeset | 66 | } | 
| 
8664ff87cd77
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
424diff
changeset | 67 | |
| 441 | 68 | |
| 434 
8664ff87cd77
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
424diff
changeset | 69 | //test: (a*)* b | 
| 440 
e14cd32ad497
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
434diff
changeset | 70 | for (i <- 1 to 20) {
 | 
| 434 
8664ff87cd77
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
424diff
changeset | 71 | println(i + " " + "%.5f".format(time_needed(2, matches(EVIL2, "a" * i)))) | 
| 
8664ff87cd77
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
424diff
changeset | 72 | } | 
| 120 
3e71efb25ce9
added
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
93diff
changeset | 73 | |
| 440 
e14cd32ad497
updated
 Christian Urban <christian dot urban at kcl dot ac dot uk> parents: 
434diff
changeset | 74 | for (i <- 1 to 20) {
 | 
| 415 | 75 | println(i + " " + "%.5f".format(time_needed(2, matches(EVIL2, "a" * i)))) | 
| 414 | 76 | } | 
| 77 | ||
| 78 |