| author | Christian Urban <christian dot urban at kcl dot ac dot uk> | 
| Thu, 13 Oct 2016 13:13:27 +0100 | |
| changeset 450 | b93eaa833d31 | 
| parent 440 | e14cd32ad497 | 
| child 453 | a5525c780b96 | 
| permissions | -rw-r--r-- | 
| 54 | 1  | 
abstract class Rexp  | 
| 
424
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
2  | 
case object ZERO extends Rexp  | 
| 
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
3  | 
case object ONE extends Rexp  | 
| 49 | 4  | 
case class CHAR(c: Char) extends Rexp  | 
5  | 
case class ALT(r1: Rexp, r2: Rexp) extends Rexp  | 
|
6  | 
case class SEQ(r1: Rexp, r2: Rexp) extends Rexp  | 
|
7  | 
case class STAR(r: Rexp) extends Rexp  | 
|
8  | 
||
9  | 
// nullable function: tests whether the regular  | 
|
10  | 
// expression can recognise the empty string  | 
|
11  | 
def nullable (r: Rexp) : Boolean = r match {
 | 
|
| 
424
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
12  | 
case ZERO => false  | 
| 
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
13  | 
case ONE => true  | 
| 49 | 14  | 
case CHAR(_) => false  | 
15  | 
case ALT(r1, r2) => nullable(r1) || nullable(r2)  | 
|
16  | 
case SEQ(r1, r2) => nullable(r1) && nullable(r2)  | 
|
17  | 
case STAR(_) => true  | 
|
18  | 
}  | 
|
19  | 
||
20  | 
// derivative of a regular expression w.r.t. a character  | 
|
21  | 
def der (c: Char, r: Rexp) : Rexp = r match {
 | 
|
| 
424
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
22  | 
case ZERO => ZERO  | 
| 
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
23  | 
case ONE => ZERO  | 
| 
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
24  | 
case CHAR(d) => if (c == d) ONE else ZERO  | 
| 49 | 25  | 
case ALT(r1, r2) => ALT(der(c, r1), der(c, r2))  | 
26  | 
case SEQ(r1, r2) =>  | 
|
27  | 
if (nullable(r1)) ALT(SEQ(der(c, r1), r2), der(c, r2))  | 
|
28  | 
else SEQ(der(c, r1), r2)  | 
|
| 
440
 
e14cd32ad497
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
434 
diff
changeset
 | 
29  | 
case STAR(r1) => SEQ(der(c, r1), STAR(r1))  | 
| 49 | 30  | 
}  | 
31  | 
||
32  | 
// derivative w.r.t. a string (iterates der)  | 
|
33  | 
def ders (s: List[Char], r: Rexp) : Rexp = s match {
 | 
|
34  | 
case Nil => r  | 
|
| 54 | 35  | 
case c::s => ders(s, der(c, r))  | 
| 49 | 36  | 
}  | 
37  | 
||
38  | 
// main matcher function  | 
|
| 
261
 
24531cfaa36a
updated handouts
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
258 
diff
changeset
 | 
39  | 
def matches(r: Rexp, s: String) : Boolean = nullable(ders(s.toList, r))  | 
| 49 | 40  | 
|
| 
424
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
41  | 
//examples from the homework  | 
| 
343
 
539b2e88f5b9
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
261 
diff
changeset
 | 
42  | 
val r = STAR(ALT(SEQ(CHAR('a'), CHAR('b')), CHAR('b')))
 | 
| 
 
539b2e88f5b9
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
261 
diff
changeset
 | 
43  | 
der('a', r)
 | 
| 
 
539b2e88f5b9
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
261 
diff
changeset
 | 
44  | 
der('b', r)
 | 
| 
 
539b2e88f5b9
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
261 
diff
changeset
 | 
45  | 
der('c', r)
 | 
| 49 | 46  | 
|
| 
258
 
1e4da6d2490c
updated programs
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
119 
diff
changeset
 | 
47  | 
//optional: one or zero times  | 
| 
424
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
48  | 
def OPT(r: Rexp) = ALT(r, ONE)  | 
| 49 | 49  | 
|
50  | 
//n-times  | 
|
51  | 
def NTIMES(r: Rexp, n: Int) : Rexp = n match {
 | 
|
| 
424
 
1129024b26d5
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
52  | 
case 0 => ONE  | 
| 49 | 53  | 
case 1 => r  | 
54  | 
case n => SEQ(r, NTIMES(r, n - 1))  | 
|
55  | 
}  | 
|
56  | 
||
| 
258
 
1e4da6d2490c
updated programs
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
119 
diff
changeset
 | 
57  | 
// the evil regular expression  a?{n} a{n}
 | 
| 415 | 58  | 
def EVIL1(n: Int) = SEQ(NTIMES(OPT(CHAR('a')), n), NTIMES(CHAR('a'), n))
 | 
| 422 | 59  | 
|
60  | 
// the evil regular expression (a*)*b  | 
|
| 
440
 
e14cd32ad497
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
434 
diff
changeset
 | 
61  | 
val EVIL2 = SEQ(STAR(STAR(CHAR('a'))), CHAR('b'))
 | 
| 49 | 62  | 
|
| 
258
 
1e4da6d2490c
updated programs
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
119 
diff
changeset
 | 
63  | 
//for measuring time  | 
| 49 | 64  | 
def time_needed[T](i: Int, code: => T) = {
 | 
65  | 
val start = System.nanoTime()  | 
|
66  | 
for (j <- 1 to i) code  | 
|
67  | 
val end = System.nanoTime()  | 
|
68  | 
(end - start)/(i * 1.0e9)  | 
|
69  | 
}  | 
|
70  | 
||
| 
434
 
8664ff87cd77
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
433 
diff
changeset
 | 
71  | 
//test: (a?{n}) (a{n})
 | 
| 
433
 
c08290ee4f1f
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
424 
diff
changeset
 | 
72  | 
for (i <- 1 to 20) {
 | 
| 
 
c08290ee4f1f
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
424 
diff
changeset
 | 
73  | 
println(i + ": " + "%.5f".format(time_needed(2, matches(EVIL1(i), "a" * i))))  | 
| 
 
c08290ee4f1f
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
424 
diff
changeset
 | 
74  | 
}  | 
| 
 
c08290ee4f1f
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
424 
diff
changeset
 | 
75  | 
|
| 415 | 76  | 
for (i <- 1 to 20) {
 | 
77  | 
println(i + ": " + "%.5f".format(time_needed(2, matches(EVIL1(i), "a" * i))))  | 
|
| 49 | 78  | 
}  | 
| 
363
 
0d6deecdb2eb
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
343 
diff
changeset
 | 
79  | 
|
| 
434
 
8664ff87cd77
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
433 
diff
changeset
 | 
80  | 
//test: (a*)* b  | 
| 
440
 
e14cd32ad497
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
434 
diff
changeset
 | 
81  | 
for (i <- 1 to 20) {
 | 
| 
434
 
8664ff87cd77
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
433 
diff
changeset
 | 
82  | 
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: 
433 
diff
changeset
 | 
83  | 
}  | 
| 
 
8664ff87cd77
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
433 
diff
changeset
 | 
84  | 
|
| 
440
 
e14cd32ad497
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
434 
diff
changeset
 | 
85  | 
for (i <- 1 to 20) {
 | 
| 415 | 86  | 
println(i + " " + "%.5f".format(time_needed(2, matches(EVIL2, "a" * i))))  | 
87  | 
}  | 
|
| 
363
 
0d6deecdb2eb
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
343 
diff
changeset
 | 
88  |