| author | Christian Urban <urbanc@in.tum.de> | 
| Sat, 22 Oct 2016 13:11:33 +0100 | |
| changeset 458 | d01568431081 | 
| parent 455 | 192f4c59633e | 
| child 467 | 3fc9b036321d | 
| permissions | -rw-r--r-- | 
| 
121
 
43c116860e47
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
93 
diff
changeset
 | 
1  | 
|
| 422 | 2  | 
abstract class Rexp  | 
| 
407
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
3  | 
case object ZERO extends Rexp  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
4  | 
case object ONE extends Rexp  | 
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
5  | 
case class CHAR(c: Char) extends Rexp  | 
| 422 | 6  | 
case class ALT(r1: Rexp, r2: Rexp) extends Rexp  | 
7  | 
case class SEQ(r1: Rexp, r2: Rexp) extends Rexp  | 
|
8  | 
case class STAR(r: Rexp) extends Rexp  | 
|
9  | 
case class NTIMES(r: Rexp, n: Int) extends Rexp  | 
|
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
10  | 
|
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
11  | 
// nullable function: tests whether the regular  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
12  | 
// expression can recognise the empty string  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
13  | 
def nullable (r: Rexp) : Boolean = r match {
 | 
| 
407
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
14  | 
case ZERO => false  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
15  | 
case ONE => true  | 
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
16  | 
case CHAR(_) => false  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
17  | 
case ALT(r1, r2) => nullable(r1) || nullable(r2)  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
18  | 
case SEQ(r1, r2) => nullable(r1) && nullable(r2)  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
19  | 
case STAR(_) => true  | 
| 
121
 
43c116860e47
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
93 
diff
changeset
 | 
20  | 
case NTIMES(r, i) => if (i == 0) true else nullable(r)  | 
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
21  | 
}  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
22  | 
|
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
23  | 
// derivative of a regular expression w.r.t. a character  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
24  | 
def der (c: Char, r: Rexp) : Rexp = r match {
 | 
| 
407
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
25  | 
case ZERO => ZERO  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
26  | 
case ONE => ZERO  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
27  | 
case CHAR(d) => if (c == d) ONE else ZERO  | 
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
28  | 
case ALT(r1, r2) => ALT(der(c, r1), der(c, r2))  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
29  | 
case SEQ(r1, r2) =>  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
30  | 
if (nullable(r1)) ALT(SEQ(der(c, r1), r2), der(c, r2))  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
31  | 
else SEQ(der(c, r1), r2)  | 
| 441 | 32  | 
case STAR(r1) => SEQ(der(c, r1), STAR(r1))  | 
| 
455
 
192f4c59633e
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
441 
diff
changeset
 | 
33  | 
case NTIMES(r1, i) =>  | 
| 
 
192f4c59633e
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
441 
diff
changeset
 | 
34  | 
if (i == 0) ZERO else der(c, SEQ(r1, NTIMES(r1, i - 1)))  | 
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
35  | 
}  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
36  | 
|
| 422 | 37  | 
def simp(r: Rexp) : Rexp = r match {
 | 
38  | 
  case ALT(r1, r2) => (simp(r1), simp(r2)) match {
 | 
|
39  | 
case (ZERO, r2s) => r2s  | 
|
40  | 
case (r1s, ZERO) => r1s  | 
|
41  | 
case (r1s, r2s) => if (r1s == r2s) r1s else ALT (r1s, r2s)  | 
|
42  | 
}  | 
|
43  | 
  case SEQ(r1, r2) =>  (simp(r1), simp(r2)) match {
 | 
|
44  | 
case (ZERO, _) => ZERO  | 
|
45  | 
case (_, ZERO) => ZERO  | 
|
46  | 
case (ONE, r2s) => r2s  | 
|
47  | 
case (r1s, ONE) => r1s  | 
|
48  | 
case (r1s, r2s) => SEQ(r1s, r2s)  | 
|
49  | 
}  | 
|
50  | 
case r => r  | 
|
51  | 
}  | 
|
52  | 
||
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
53  | 
// derivative w.r.t. a string (iterates der)  | 
| 414 | 54  | 
def ders2(s: List[Char], r: Rexp) : Rexp = (s, r) match {
 | 
| 
407
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
55  | 
case (Nil, r) => r  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
56  | 
case (s, ZERO) => ZERO  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
57  | 
case (s, ONE) => if (s == Nil) ONE else ZERO  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
58  | 
case (s, CHAR(c)) => if (s == List(c)) ONE else  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
59  | 
if (s == Nil) CHAR(c) else ZERO  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
60  | 
case (s, ALT(r1, r2)) => ALT(ders2(s, r2), ders2(s, r2))  | 
| 422 | 61  | 
case (c::s, r) => ders2(s, simp(der(c, r)))  | 
| 
407
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
62  | 
}  | 
| 
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
63  | 
|
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
64  | 
// main matcher function  | 
| 
407
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
65  | 
def matcher(r: Rexp, s: String) : Boolean = nullable(ders2(s.toList, r))  | 
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
66  | 
|
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
67  | 
|
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
68  | 
//one or zero  | 
| 
407
 
4b454a6d1814
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
121 
diff
changeset
 | 
69  | 
def OPT(r: Rexp) = ALT(r, ONE)  | 
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
70  | 
|
| 422 | 71  | 
def EVIL1(n: Int) = SEQ(NTIMES(OPT(CHAR('a')), n), NTIMES(CHAR('a'), n))
 | 
| 441 | 72  | 
val EVIL2 = SEQ(STAR(STAR(CHAR('a'))), CHAR('b'))
 | 
| 422 | 73  | 
|
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
74  | 
|
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
75  | 
def time_needed[T](i: Int, code: => T) = {
 | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
76  | 
val start = System.nanoTime()  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
77  | 
for (j <- 1 to i) code  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
78  | 
val end = System.nanoTime()  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
79  | 
(end - start)/(i * 1.0e9)  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
80  | 
}  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
81  | 
|
| 
425
 
d14a9bdac26f
updated
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents: 
422 
diff
changeset
 | 
82  | 
// warmup  | 
| 441 | 83  | 
//test: (a?{n}) (a{n})
 | 
84  | 
for (i <- 1 to 7000001 by 500000) {
 | 
|
85  | 
println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL1(i), "a" * i))))  | 
|
86  | 
}  | 
|
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
87  | 
|
| 441 | 88  | 
for (i <- 1 to 7000001 by 500000) {
 | 
| 414 | 89  | 
println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL1(i), "a" * i))))  | 
| 
92
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
90  | 
}  | 
| 
 
e85600529ca5
moved scala files
 
Christian Urban <christian dot urban at kcl dot ac dot uk> 
parents:  
diff
changeset
 | 
91  | 
|
| 441 | 92  | 
//test: (a*)* b  | 
93  | 
for (i <- 1 to 7000001 by 500000) {
 | 
|
| 414 | 94  | 
println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL2, "a" * i))))  | 
95  | 
}  | 
|
96  | 
||
| 441 | 97  | 
for (i <- 1 to 7000001 by 500000) {
 | 
98  | 
println(i + " " + "%.5f".format(time_needed(2, matcher(EVIL2, "a" * i))))  | 
|
99  | 
}  | 
|
100  | 
||
101  | 
||
102  |