line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
$VERSION = '0.01'; |
2
|
|
|
|
|
|
|
|
3
|
1
|
|
|
1
|
|
9031
|
use strict; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
703
|
|
4
|
|
|
|
|
|
|
|
5
|
|
|
|
|
|
|
package Data::Trie; |
6
|
|
|
|
|
|
|
|
7
|
|
|
|
|
|
|
#creates a new Trie node and initializes its value and daughters to zero |
8
|
|
|
|
|
|
|
sub new { |
9
|
23
|
|
|
23
|
0
|
57
|
my $self = {}; |
10
|
23
|
|
|
|
|
46
|
my $class = shift; |
11
|
|
|
|
|
|
|
#does this node terminate a word? |
12
|
23
|
|
|
|
|
78
|
$self->{value} = 0; |
13
|
|
|
|
|
|
|
#is this node terminate a prefix of further words? |
14
|
23
|
|
|
|
|
45
|
$self->{daughters} = {}; |
15
|
23
|
|
|
|
|
87
|
bless $self, $class; |
16
|
|
|
|
|
|
|
} |
17
|
|
|
|
|
|
|
|
18
|
|
|
|
|
|
|
#returns all words in the trie |
19
|
|
|
|
|
|
|
sub getAll { |
20
|
0
|
|
|
0
|
0
|
0
|
my $self = shift; |
21
|
|
|
|
|
|
|
#keeps track of the path in the trie up to this point |
22
|
0
|
|
|
|
|
0
|
my $path = ""; |
23
|
|
|
|
|
|
|
#calls a recursive routine to check the path |
24
|
0
|
|
|
|
|
0
|
return $self->_getAllRecurse($path); |
25
|
|
|
|
|
|
|
} |
26
|
|
|
|
|
|
|
|
27
|
|
|
|
|
|
|
#recursive routine to collect all words, called by getAll() |
28
|
|
|
|
|
|
|
#DON'T CALL THIS DIRECTLY; use getAll() instead |
29
|
|
|
|
|
|
|
sub _getAllRecurse { |
30
|
0
|
|
|
0
|
|
0
|
my $self = shift; |
31
|
0
|
|
|
|
|
0
|
my $path = shift; |
32
|
0
|
|
|
|
|
0
|
my $daughters = $self->{daughters}; |
33
|
|
|
|
|
|
|
#the set of words to return |
34
|
0
|
|
|
|
|
0
|
my @result = (); |
35
|
|
|
|
|
|
|
#return the current path if the current node terminates a word |
36
|
0
|
0
|
|
|
|
0
|
if ($self->{value}) { |
37
|
0
|
|
|
|
|
0
|
push @result, $path; |
38
|
|
|
|
|
|
|
} |
39
|
0
|
|
|
|
|
0
|
my @keys = keys %$daughters; |
40
|
|
|
|
|
|
|
#check all daughter nodes recursively adding their results to current ones |
41
|
0
|
|
|
|
|
0
|
foreach my $letter (@keys) { |
42
|
0
|
|
|
|
|
0
|
my $newpath = $path . $letter; |
43
|
0
|
|
|
|
|
0
|
my @letterresult = $daughters->{$letter}->_getAllRecurse($newpath); |
44
|
0
|
|
|
|
|
0
|
push @result, @letterresult; |
45
|
|
|
|
|
|
|
} |
46
|
0
|
|
|
|
|
0
|
return @result; |
47
|
|
|
|
|
|
|
} |
48
|
|
|
|
|
|
|
|
49
|
|
|
|
|
|
|
#adds a word to the tree by recursively checking each letter of the word and |
50
|
|
|
|
|
|
|
#adding nodes as needed. |
51
|
|
|
|
|
|
|
sub add { |
52
|
27
|
|
|
27
|
0
|
35
|
my $self = shift; |
53
|
27
|
|
|
|
|
40
|
my $str = shift; |
54
|
|
|
|
|
|
|
#data can be added or not |
55
|
27
|
|
|
|
|
30
|
my $data = shift; |
56
|
|
|
|
|
|
|
#separates first letter from the rest |
57
|
27
|
|
|
|
|
39
|
my $first = substr $str, 0, 1; |
58
|
27
|
|
|
|
|
32
|
my $rest = substr $str, 1; |
59
|
27
|
|
|
|
|
42
|
my $daughters = $self->{daughters}; |
60
|
|
|
|
|
|
|
#checks if there is a node for the first letter |
61
|
27
|
100
|
|
|
|
61
|
if (not exists $daughters->{$first}) { |
62
|
|
|
|
|
|
|
#adds a node if necessary |
63
|
22
|
|
|
|
|
43
|
$daughters->{$first} = Data::Trie->new; |
64
|
|
|
|
|
|
|
} |
65
|
27
|
|
|
|
|
39
|
my $daughter = $daughters->{$first}; |
66
|
|
|
|
|
|
|
#is the word only one letter long? |
67
|
27
|
100
|
|
|
|
54
|
if (length $rest > 0) { |
68
|
|
|
|
|
|
|
#recurse on the remaining letters |
69
|
23
|
|
|
|
|
64
|
$daughter->add($rest, $data); |
70
|
|
|
|
|
|
|
} else { |
71
|
|
|
|
|
|
|
#set the value to 1 and store the data |
72
|
4
|
|
|
|
|
6
|
$daughter->{value} = 1; |
73
|
4
|
|
|
|
|
8
|
$daughter->{data} = $data; |
74
|
|
|
|
|
|
|
} |
75
|
27
|
|
|
|
|
46
|
return 1; |
76
|
|
|
|
|
|
|
} |
77
|
|
|
|
|
|
|
|
78
|
|
|
|
|
|
|
#removes a word from the trie (does NOT prune nodes) |
79
|
|
|
|
|
|
|
sub remove { |
80
|
8
|
|
|
8
|
0
|
70
|
my $self = shift; |
81
|
8
|
|
|
|
|
10
|
my $str = shift; |
82
|
|
|
|
|
|
|
#splits the word into first letter and rest |
83
|
8
|
|
|
|
|
14
|
my $first = substr $str, 0, 1; |
84
|
8
|
|
|
|
|
13
|
my $rest = substr $str, 1; |
85
|
8
|
|
|
|
|
12
|
my $daughters = $self->{daughters}; |
86
|
8
|
50
|
|
|
|
15
|
if (exists $daughters->{$first}) { |
87
|
8
|
|
|
|
|
12
|
my $daughter = $daughters->{$first}; |
88
|
8
|
100
|
|
|
|
14
|
if (length $rest == 0) { |
89
|
1
|
|
|
|
|
2
|
$daughter->{value} = 0; |
90
|
|
|
|
|
|
|
} else { |
91
|
7
|
|
|
|
|
24
|
$daughter->remove($rest); |
92
|
|
|
|
|
|
|
} |
93
|
|
|
|
|
|
|
} |
94
|
8
|
|
|
|
|
13
|
return $str; |
95
|
|
|
|
|
|
|
} |
96
|
|
|
|
|
|
|
|
97
|
|
|
|
|
|
|
#looks up a word in the trie |
98
|
|
|
|
|
|
|
sub lookup { |
99
|
6
|
|
|
6
|
0
|
10
|
my $self = shift; |
100
|
6
|
|
|
|
|
10
|
my $str = shift; |
101
|
|
|
|
|
|
|
#splits the word into first letter and rest |
102
|
6
|
|
|
|
|
14
|
my $first = substr $str, 0, 1; |
103
|
6
|
|
|
|
|
8
|
my $rest = substr $str, 1; |
104
|
6
|
|
|
|
|
11
|
my $daughters = $self->{daughters}; |
105
|
|
|
|
|
|
|
#checks if the first letter matches a daughter |
106
|
6
|
50
|
|
|
|
18
|
if (not exists $daughters->{$first}) { |
|
|
100
|
|
|
|
|
|
107
|
|
|
|
|
|
|
#if not, lookup fails |
108
|
0
|
|
|
|
|
0
|
return 0; |
109
|
|
|
|
|
|
|
#if it does match, recurse on remaining letters |
110
|
|
|
|
|
|
|
} elsif (length $rest == 0) { |
111
|
1
|
|
|
|
|
7
|
return ($daughters->{$first}->{value}, $daughters->{$first}->{data}); |
112
|
|
|
|
|
|
|
} else { |
113
|
5
|
|
|
|
|
29
|
return $daughters->{$first}->lookup($rest); |
114
|
|
|
|
|
|
|
} |
115
|
|
|
|
|
|
|
} |
116
|
|
|
|
|
|
|
|
117
|
|
|
|
|
|
|
1; |
118
|
|
|
|
|
|
|
|
119
|
|
|
|
|
|
|
=head1 NAME |
120
|
|
|
|
|
|
|
|
121
|
|
|
|
|
|
|
Data::Trie - An implementation of a letter trie |
122
|
|
|
|
|
|
|
|
123
|
|
|
|
|
|
|
=head1 SYNOPSIS |
124
|
|
|
|
|
|
|
|
125
|
|
|
|
|
|
|
use Data::Trie; |
126
|
|
|
|
|
|
|
$t = Data::Trie->new; |
127
|
|
|
|
|
|
|
$t->add('orange', 'kind of fruit'); |
128
|
|
|
|
|
|
|
($result, $data) = $t->lookup->('orange'); |
129
|
|
|
|
|
|
|
$t->remove('orange'); |
130
|
|
|
|
|
|
|
$t->getAll; |
131
|
|
|
|
|
|
|
|
132
|
|
|
|
|
|
|
=head1 DESCRIPTION |
133
|
|
|
|
|
|
|
|
134
|
|
|
|
|
|
|
This module implements a letter trie data structure. This is a linked set of |
135
|
|
|
|
|
|
|
nodes representing a set of words. Starting from the root, each letter of an |
136
|
|
|
|
|
|
|
included word is a daughter node of the trie. Hence, if a word is in the trie, |
137
|
|
|
|
|
|
|
there will be a path from root to leaf for that word. If a word is not in the |
138
|
|
|
|
|
|
|
trie, there will be no such path. |
139
|
|
|
|
|
|
|
|
140
|
|
|
|
|
|
|
This structure allows for a relatively compact representation of a set of words. |
141
|
|
|
|
|
|
|
This particular implementation allows each word to be stored alone or with some |
142
|
|
|
|
|
|
|
associated data item. |
143
|
|
|
|
|
|
|
|
144
|
|
|
|
|
|
|
Note that the C method does I prune nodes and thus a C can |
145
|
|
|
|
|
|
|
only grow in size. |
146
|
|
|
|
|
|
|
|
147
|
|
|
|
|
|
|
=head1 COMPARE |
148
|
|
|
|
|
|
|
|
149
|
|
|
|
|
|
|
This implementation differs from L in that C checks for a |
150
|
|
|
|
|
|
|
match, rather than checking for whether the current string is a prefix. |
151
|
|
|
|
|
|
|
|
152
|
|
|
|
|
|
|
=head1 VERSION |
153
|
|
|
|
|
|
|
|
154
|
|
|
|
|
|
|
0.01 |
155
|
|
|
|
|
|
|
|
156
|
|
|
|
|
|
|
=head1 AUTHOR |
157
|
|
|
|
|
|
|
|
158
|
|
|
|
|
|
|
Michael Hammond, I |
159
|
|
|
|
|
|
|
|
160
|
|
|
|
|
|
|
=cut |
161
|
|
|
|
|
|
|
|