62
|
1 |
|
|
2 |
// regular expressions including NOT
|
|
3 |
abstract class Rexp
|
|
4 |
|
|
5 |
case object NULL extends Rexp
|
|
6 |
case object EMPTY extends Rexp
|
|
7 |
case class CHAR(c: Char) extends Rexp
|
|
8 |
case class ALT(r1: Rexp, r2: Rexp) extends Rexp
|
|
9 |
case class SEQ(r1: Rexp, r2: Rexp) extends Rexp
|
|
10 |
case class STAR(r: Rexp) extends Rexp
|
|
11 |
case class NOT(r: Rexp) extends Rexp
|
|
12 |
|
|
13 |
|
|
14 |
// some convenience for typing in regular expressions
|
|
15 |
def charlist2rexp(s : List[Char]) : Rexp = s match {
|
|
16 |
case Nil => EMPTY
|
|
17 |
case c::Nil => CHAR(c)
|
|
18 |
case c::s => SEQ(CHAR(c), charlist2rexp(s))
|
|
19 |
}
|
|
20 |
implicit def string2rexp(s : String) : Rexp = charlist2rexp(s.toList)
|
|
21 |
|
|
22 |
|
|
23 |
// nullable function: tests whether the regular
|
|
24 |
// expression can recognise the empty string
|
|
25 |
def nullable (r: Rexp) : Boolean = r match {
|
|
26 |
case NULL => false
|
|
27 |
case EMPTY => true
|
|
28 |
case CHAR(_) => false
|
|
29 |
case ALT(r1, r2) => nullable(r1) || nullable(r2)
|
|
30 |
case SEQ(r1, r2) => nullable(r1) && nullable(r2)
|
|
31 |
case STAR(_) => true
|
|
32 |
case NOT(r) => !(nullable(r))
|
|
33 |
}
|
|
34 |
|
|
35 |
// tests whether a regular expression
|
|
36 |
// cannot recognise more
|
|
37 |
def no_more (r: Rexp) : Boolean = r match {
|
|
38 |
case NULL => true
|
|
39 |
case EMPTY => false
|
|
40 |
case CHAR(_) => false
|
|
41 |
case ALT(r1, r2) => no_more(r1) && no_more(r2)
|
|
42 |
case SEQ(r1, r2) => if (nullable(r1)) (no_more(r1) && no_more(r2)) else no_more(r1)
|
|
43 |
case STAR(_) => false
|
|
44 |
case NOT(r) => !(no_more(r))
|
|
45 |
}
|
|
46 |
|
|
47 |
|
|
48 |
// derivative of a regular expression w.r.t. a character
|
|
49 |
def der (c: Char, r: Rexp) : Rexp = r match {
|
|
50 |
case NULL => NULL
|
|
51 |
case EMPTY => NULL case CHAR(d) => if (c == d) EMPTY else NULL
|
|
52 |
case ALT(r1, r2) => ALT(der(c, r1), der(c, r2))
|
|
53 |
case SEQ(r1, r2) =>
|
|
54 |
if (nullable(r1)) ALT(SEQ(der(c, r1), r2), der(c, r2))
|
|
55 |
else SEQ(der(c, r1), r2)
|
|
56 |
case STAR(r) => SEQ(der(c, r), STAR(r))
|
|
57 |
case NOT(r) => NOT(der (c, r))
|
|
58 |
}
|
|
59 |
|
|
60 |
// regular expression for specifying
|
|
61 |
// ranges of characters
|
|
62 |
def RANGE(s : List[Char]) : Rexp = s match {
|
|
63 |
case Nil => NULL
|
|
64 |
case c::Nil => CHAR(c)
|
|
65 |
case c::s => ALT(CHAR(c), RANGE(s))
|
|
66 |
}
|
|
67 |
|
|
68 |
// one or more
|
|
69 |
def PLUS(r: Rexp) = SEQ(r, STAR(r))
|
|
70 |
|
|
71 |
|
|
72 |
type Rule[T] = (Rexp, List[Char] => T)
|
|
73 |
|
|
74 |
def error (s: String) = throw new IllegalArgumentException ("Cannot tokenize: " + s)
|
|
75 |
|
|
76 |
def munch[T](r: Rexp, action: List[Char] => T, s: List[Char], t: List[Char]) : Option[(List[Char], T)] =
|
|
77 |
s match {
|
|
78 |
case Nil if (nullable(r)) => Some(Nil, action(t))
|
|
79 |
case Nil => None
|
|
80 |
case c::s if (no_more(der (c, r)) && nullable(r)) => Some(c::s, action(t))
|
|
81 |
case c::s if (no_more(der (c, r))) => None
|
|
82 |
case c::s => munch(der (c, r), action, s, t ::: List(c))
|
|
83 |
}
|
|
84 |
|
|
85 |
def one_token[T](rs: List[Rule[T]], s: List[Char]) : (List[Char], T) = {
|
|
86 |
val somes = rs.map { (r) => munch(r._1, r._2, s, Nil) } .flatten
|
|
87 |
if (somes == Nil) error(s.mkString) else (somes sortBy (_._1.length) head)
|
|
88 |
}
|
|
89 |
|
|
90 |
def tokenize[T](rs: List[Rule[T]], s: List[Char]) : List[T] = s match {
|
|
91 |
case Nil => Nil
|
|
92 |
case _ => one_token(rs, s) match {
|
|
93 |
case (rest, token) => token :: tokenize(rs, rest)
|
|
94 |
}
|
|
95 |
}
|
|
96 |
|
|
97 |
|
|
98 |
|
|
99 |
|
|
100 |
|