progs/re3a.scala
author cu
Thu, 19 Oct 2017 11:04:43 +0100
changeset 524 e264779c3411
parent 519 3921e68f5cd1
child 623 8e63f9745f46
permissions -rw-r--r--
updated
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
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
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    12
// nullable function: tests whether the regular 
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
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    25
// derivative of a regular expression w.r.t. a character
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
85586bada20a updated
Christian Urban <urbanc@in.tum.de>
parents: 455
diff changeset
    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
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    61
// derivative w.r.t. a string (iterates der)
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
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    68
// main matcher function
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    69
def matcher(r: Rexp, s: String) : Boolean = nullable(ders(s.toList, r))
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    70
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    71
//one or zero
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
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    74
//evil regular expressions
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
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
    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()
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    83
  (end - start)/(i * 1.0e9)
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
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    87
//test: (a?{n}) (a{n})
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    88
for (i <- 1 to 8001 by 1000) {
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    89
  println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL1(i), "a" * i))))
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
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    92
for (i <- 1 to 8001 by 1000) {
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    93
  println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL1(i), "a" * i))))
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
    94
}
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
//test: (a*)* b
519
3921e68f5cd1 updated
cu
parents: 471
diff changeset
    97
for (i <- 1 to 6000001 by 500000) {
471
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
    98
  println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL2, "a" * i))))
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
    99
}
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   100
519
3921e68f5cd1 updated
cu
parents: 471
diff changeset
   101
for (i <- 1 to 6000001 by 500000) {
455
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
   102
  println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL2, "a" * i))))
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
   103
}
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
   104
471
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   105
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   106
val r0 = simp(der('a', EVIL3))
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   107
val r1 = simp(der('a', r0))
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   108
val r2 = simp(der('a', r1))
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   109
val r3 = simp(der('a', r2))
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   110
val r4 = simp(der('a', r3))
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   111
val r5 = simp(der('a', r4))
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   112
val r6 = simp(der('a', r5))
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   113
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   114
//test: (a|aa)* b
519
3921e68f5cd1 updated
cu
parents: 471
diff changeset
   115
/*
471
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   116
for (i <- 1 to 7001 by 500) {
e5df48ff7033 updated
Christian Urban <urbanc@in.tum.de>
parents: 457
diff changeset
   117
  println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL3, "a" * i ++ "c"))))
455
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
   118
}
519
3921e68f5cd1 updated
cu
parents: 471
diff changeset
   119
 */
455
192f4c59633e updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
diff changeset
   120