line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package SNA::Network::Algorithm::Louvain; |
2
|
|
|
|
|
|
|
|
3
|
14
|
|
|
14
|
|
73
|
use strict; |
|
14
|
|
|
|
|
24
|
|
|
14
|
|
|
|
|
436
|
|
4
|
14
|
|
|
14
|
|
114
|
use warnings; |
|
14
|
|
|
|
|
29
|
|
|
14
|
|
|
|
|
456
|
|
5
|
|
|
|
|
|
|
|
6
|
|
|
|
|
|
|
require Exporter; |
7
|
14
|
|
|
14
|
|
84
|
use base qw(Exporter); |
|
14
|
|
|
|
|
35
|
|
|
14
|
|
|
|
|
1259
|
|
8
|
|
|
|
|
|
|
our @EXPORT = qw(identify_communities_with_louvain); |
9
|
|
|
|
|
|
|
|
10
|
14
|
|
|
14
|
|
9628
|
use SNA::Network::Community; |
|
14
|
|
|
|
|
43
|
|
|
14
|
|
|
|
|
139
|
|
11
|
|
|
|
|
|
|
|
12
|
14
|
|
|
14
|
|
428
|
use List::Util qw(sum); |
|
14
|
|
|
|
|
33
|
|
|
14
|
|
|
|
|
825
|
|
13
|
14
|
|
|
14
|
|
81
|
use List::MoreUtils qw(uniq); |
|
14
|
|
|
|
|
26
|
|
|
14
|
|
|
|
|
24914
|
|
14
|
|
|
|
|
|
|
|
15
|
|
|
|
|
|
|
|
16
|
|
|
|
|
|
|
=head1 NAME |
17
|
|
|
|
|
|
|
|
18
|
|
|
|
|
|
|
SNA::Network::Algorithm::Louvain - identifies communities with the Louvain-method developed by Blondel and Guillaume and Lamboitte and Lefebvre |
19
|
|
|
|
|
|
|
|
20
|
|
|
|
|
|
|
|
21
|
|
|
|
|
|
|
=head1 SYNOPSIS |
22
|
|
|
|
|
|
|
|
23
|
|
|
|
|
|
|
use SNA::Network; |
24
|
|
|
|
|
|
|
|
25
|
|
|
|
|
|
|
my $net = SNA::Network->new(); |
26
|
|
|
|
|
|
|
$net->load_from_pajek_net($filename); |
27
|
|
|
|
|
|
|
... |
28
|
|
|
|
|
|
|
my $num_communities = $net->identify_communities_with_louvain; |
29
|
|
|
|
|
|
|
|
30
|
|
|
|
|
|
|
|
31
|
|
|
|
|
|
|
=head1 METHODS |
32
|
|
|
|
|
|
|
|
33
|
|
|
|
|
|
|
The following methods are added to L. |
34
|
|
|
|
|
|
|
|
35
|
|
|
|
|
|
|
=head2 identify_communities_with_louvain |
36
|
|
|
|
|
|
|
|
37
|
|
|
|
|
|
|
Performs the community identification algorithm and returns the number of communities it identified. The corresponding L objects can be accessed with the C method of L. |
38
|
|
|
|
|
|
|
|
39
|
|
|
|
|
|
|
=cut |
40
|
|
|
|
|
|
|
|
41
|
|
|
|
|
|
|
sub identify_communities_with_louvain { |
42
|
3
|
|
|
3
|
1
|
15
|
my ($self, $level) = @_; |
43
|
3
|
|
100
|
|
|
16
|
$level ||= 0; |
44
|
3
|
|
|
|
|
8
|
$self->{louvain_levels} = $level; |
45
|
|
|
|
|
|
|
|
46
|
3
|
|
|
|
|
15
|
$self->{total_weight} = $self->total_weight; |
47
|
|
|
|
|
|
|
|
48
|
|
|
|
|
|
|
# initialize communities |
49
|
3
|
|
|
|
|
179
|
foreach ($self->nodes) { |
50
|
190
|
|
|
|
|
532
|
$_->{community} = $_->index; |
51
|
190
|
|
|
|
|
547
|
$_->{_k_i} = $_->weighted_summed_degree; |
52
|
190
|
100
|
|
|
|
951
|
_propagate_new_community_id($_) if $level > 0; |
53
|
|
|
|
|
|
|
} |
54
|
190
|
100
|
|
|
|
1315
|
my @communities = map { |
|
|
100
|
|
|
|
|
|
55
|
3
|
|
|
|
|
38
|
SNA::Network::Community->new( |
56
|
|
|
|
|
|
|
network => $self, |
57
|
|
|
|
|
|
|
index => $_->index, |
58
|
|
|
|
|
|
|
members_ref => [$_], |
59
|
|
|
|
|
|
|
w_in => $level > 0 ? $_->loop->weight : 0, |
60
|
|
|
|
|
|
|
w_tot => $_->{_k_i} - ($level > 0 ? $_->loop->weight : 0), |
61
|
|
|
|
|
|
|
) |
62
|
|
|
|
|
|
|
} $self->nodes; |
63
|
|
|
|
|
|
|
|
64
|
3
|
|
|
|
|
21
|
$self->{communities_ref} = \@communities; |
65
|
|
|
|
|
|
|
|
66
|
3
|
|
|
|
|
6
|
my $has_changed = 0; |
67
|
3
|
|
|
|
|
4
|
my $has_improved; |
68
|
|
|
|
|
|
|
|
69
|
|
|
|
|
|
|
PHASE_ONE_ITERATIONS: |
70
|
3
|
|
|
|
|
7
|
do { |
71
|
9
|
|
|
|
|
19
|
$has_improved = 0; |
72
|
|
|
|
|
|
|
|
73
|
9
|
|
|
|
|
46
|
foreach my $node_i ($self->nodes) { |
74
|
1096
|
|
|
|
|
1563
|
my $max_gain = 0; |
75
|
1096
|
|
|
|
|
1222
|
my $best_community_id; |
76
|
|
|
|
|
|
|
|
77
|
7738
|
|
|
|
|
22259
|
$node_i->{_k_i_in} = sum( |
78
|
|
|
|
|
|
|
$level > 0 ? - $node_i->loop->weight : 0, |
79
|
|
|
|
|
|
|
map { |
80
|
17528
|
|
|
|
|
77825
|
$_->weight |
81
|
|
|
|
|
|
|
} grep { |
82
|
1096
|
100
|
|
|
|
4632
|
$_->source->community == $_->target->community |
83
|
|
|
|
|
|
|
} $node_i->edges |
84
|
|
|
|
|
|
|
); |
85
|
|
|
|
|
|
|
|
86
|
1096
|
|
|
|
|
5895
|
my $current_community = @communities[ $node_i->community ]; |
87
|
|
|
|
|
|
|
|
88
|
|
|
|
|
|
|
# gain in current community on removal |
89
|
1096
|
|
|
|
|
3462
|
my $current_community_module = _module_value( |
90
|
|
|
|
|
|
|
$current_community->w_in, |
91
|
|
|
|
|
|
|
$current_community->w_tot, |
92
|
|
|
|
|
|
|
$self->{total_weight}, |
93
|
|
|
|
|
|
|
); |
94
|
|
|
|
|
|
|
|
95
|
1096
|
|
|
|
|
5334
|
my $new_current_community_module = _module_value( |
96
|
|
|
|
|
|
|
$current_community->w_in - $node_i->{_k_i_in}, |
97
|
|
|
|
|
|
|
$current_community->w_tot - $node_i->{_k_i} + $node_i->{_k_i_in}, |
98
|
|
|
|
|
|
|
$self->{total_weight}, |
99
|
|
|
|
|
|
|
); |
100
|
|
|
|
|
|
|
|
101
|
1096
|
|
|
|
|
1847
|
my $gain_on_removal = $new_current_community_module - $current_community_module; |
102
|
|
|
|
|
|
|
|
103
|
|
|
|
|
|
|
# pre-calculate k_i_new values for all neighbour communities |
104
|
1096
|
|
|
|
|
4410
|
foreach ($node_i->outgoing_edges) { |
105
|
8764
|
|
|
|
|
48214
|
$node_i->{_k_i_new}->{ $_->target->community } += $_->weight |
106
|
|
|
|
|
|
|
} |
107
|
|
|
|
|
|
|
|
108
|
1096
|
|
|
|
|
4404
|
foreach ($node_i->incoming_edges) { |
109
|
8764
|
|
|
|
|
29485
|
$node_i->{_k_i_new}->{ $_->source->community } += $_->weight |
110
|
|
|
|
|
|
|
} |
111
|
|
|
|
|
|
|
|
112
|
1096
|
100
|
|
|
|
3269
|
if ($level > 0) { |
113
|
16
|
|
|
|
|
45
|
my $loop_weight = $node_i->loop->weight; |
114
|
|
|
|
|
|
|
|
115
|
16
|
|
|
|
|
52
|
foreach ( values %{ $node_i->{_k_i_new} } ) { |
|
16
|
|
|
|
|
54
|
|
116
|
65
|
|
|
|
|
96
|
$_ += $loop_weight; |
117
|
|
|
|
|
|
|
} |
118
|
|
|
|
|
|
|
} |
119
|
|
|
|
|
|
|
|
120
|
17528
|
|
|
|
|
45586
|
my @neighbour_community_ids = uniq grep { |
121
|
17528
|
|
|
|
|
45467
|
$_ != $node_i->community |
122
|
|
|
|
|
|
|
} map { |
123
|
1096
|
|
|
|
|
3249
|
$_->community |
124
|
|
|
|
|
|
|
} $node_i->related_nodes; |
125
|
|
|
|
|
|
|
|
126
|
1096
|
|
|
|
|
6946
|
foreach my $neighbour_community_id (@neighbour_community_ids) { |
127
|
|
|
|
|
|
|
# calculate modularity changes |
128
|
4966
|
|
|
|
|
9752
|
my $gain_on_addition = _modularity_gain($self, $node_i, $neighbour_community_id); |
129
|
4966
|
|
|
|
|
7347
|
my $gain = $gain_on_removal + $gain_on_addition; |
130
|
|
|
|
|
|
|
|
131
|
4966
|
100
|
|
|
|
12910
|
if ( $gain > $max_gain ) { |
132
|
500
|
|
|
|
|
570
|
$max_gain = $gain; |
133
|
500
|
|
|
|
|
1364
|
$best_community_id = $neighbour_community_id; |
134
|
|
|
|
|
|
|
} |
135
|
|
|
|
|
|
|
} |
136
|
|
|
|
|
|
|
|
137
|
1096
|
100
|
|
|
|
2670
|
if ( $max_gain > 0.0000001 ) { |
138
|
|
|
|
|
|
|
# merge node to best community |
139
|
|
|
|
|
|
|
|
140
|
275
|
|
|
|
|
697
|
_switch_community($self, $node_i, $best_community_id); |
141
|
275
|
|
|
|
|
922
|
$has_improved = 1; |
142
|
275
|
|
|
|
|
783
|
$has_changed = 1; |
143
|
|
|
|
|
|
|
} |
144
|
|
|
|
|
|
|
|
145
|
1096
|
|
|
|
|
1742
|
undef $node_i->{_k_i_in}; |
146
|
1096
|
|
|
|
|
4925
|
undef $node_i->{_k_i_new}; |
147
|
|
|
|
|
|
|
} |
148
|
|
|
|
|
|
|
} while ($has_improved); |
149
|
|
|
|
|
|
|
|
150
|
3
|
100
|
|
|
|
13
|
return @communities unless $has_changed; |
151
|
2
|
|
|
|
|
6
|
$self->{louvain_levels} += 1; |
152
|
|
|
|
|
|
|
|
153
|
|
|
|
|
|
|
# consolidate community structure; |
154
|
2
|
|
|
|
|
8
|
_consolidate_community_structure($self); |
155
|
|
|
|
|
|
|
|
156
|
|
|
|
|
|
|
|
157
|
2
|
|
|
|
|
8
|
PHASE_TWO: |
158
|
|
|
|
|
|
|
|
159
|
|
|
|
|
|
|
# merge communities to new nodes in a new network |
160
|
|
|
|
|
|
|
my $next_level_network = _create_next_level_network($self); |
161
|
|
|
|
|
|
|
|
162
|
|
|
|
|
|
|
# apply the algorithm recursively to the new network |
163
|
2
|
|
|
|
|
34
|
my @new_community_structure = $next_level_network->identify_communities_with_louvain($level + 1); |
164
|
|
|
|
|
|
|
|
165
|
2
|
|
|
|
|
6
|
$self->{louvain_levels} = $next_level_network->{louvain_levels}; |
166
|
2
|
100
|
|
|
|
10
|
return @new_community_structure if $level > 0; |
167
|
|
|
|
|
|
|
|
168
|
1
|
|
|
|
|
4
|
foreach my $community (@new_community_structure) { |
169
|
720
|
|
|
|
|
1657
|
$community->{members_ref} = [ grep { |
170
|
4
|
|
|
|
|
29
|
$_->community == $community->index |
171
|
|
|
|
|
|
|
} $self->nodes ]; |
172
|
|
|
|
|
|
|
}; |
173
|
|
|
|
|
|
|
|
174
|
1
|
|
|
|
|
3
|
$self->{communities_ref} = \@new_community_structure; |
175
|
1
|
|
|
|
|
7
|
return int $self->communities; |
176
|
|
|
|
|
|
|
} |
177
|
|
|
|
|
|
|
|
178
|
|
|
|
|
|
|
|
179
|
|
|
|
|
|
|
sub _modularity_gain { |
180
|
4966
|
|
|
4966
|
|
7994
|
my ($net, $node_i, $new_community_id) = @_; |
181
|
4966
|
|
|
|
|
9081
|
my $new_community = $net->{communities_ref}->[$new_community_id]; |
182
|
|
|
|
|
|
|
|
183
|
4966
|
|
|
|
|
9030
|
my $k_i_new = $node_i->{_k_i_new}->{ $new_community_id }; |
184
|
|
|
|
|
|
|
|
185
|
4966
|
|
|
|
|
16597
|
my $neighbour_community_module = _module_value( |
186
|
|
|
|
|
|
|
$new_community->w_in, |
187
|
|
|
|
|
|
|
$new_community->w_tot, |
188
|
|
|
|
|
|
|
$net->{total_weight}, |
189
|
|
|
|
|
|
|
); |
190
|
|
|
|
|
|
|
|
191
|
4966
|
|
|
|
|
20251
|
my $new_neighbour_community_module = _module_value( |
192
|
|
|
|
|
|
|
$new_community->w_in + $k_i_new, |
193
|
|
|
|
|
|
|
$new_community->w_tot + $node_i->{_k_i} - $k_i_new, |
194
|
|
|
|
|
|
|
$net->{total_weight}, |
195
|
|
|
|
|
|
|
); |
196
|
|
|
|
|
|
|
|
197
|
4966
|
|
|
|
|
12394
|
return $new_neighbour_community_module - $neighbour_community_module; |
198
|
|
|
|
|
|
|
} |
199
|
|
|
|
|
|
|
|
200
|
|
|
|
|
|
|
|
201
|
|
|
|
|
|
|
sub _module_value { |
202
|
12124
|
|
|
12124
|
|
19828
|
my ($w_in, $w_tot, $w_net) = @_; |
203
|
12124
|
|
|
|
|
39499
|
return $w_in / $w_net - ( ( $w_in + $w_tot ) / ( 2 * $w_net ) ) ** 2; |
204
|
|
|
|
|
|
|
} |
205
|
|
|
|
|
|
|
|
206
|
|
|
|
|
|
|
|
207
|
|
|
|
|
|
|
sub _switch_community { |
208
|
275
|
|
|
275
|
|
399
|
my ($net, $node, $new_community_id) = @_; |
209
|
275
|
|
|
|
|
710
|
my $current_community = $net->{communities_ref}->[ $node->community ]; |
210
|
275
|
|
|
|
|
415
|
my $new_community = $net->{communities_ref}->[$new_community_id]; |
211
|
|
|
|
|
|
|
|
212
|
275
|
|
|
|
|
1523
|
@{ $current_community->members_ref } = grep { |
|
686
|
|
|
|
|
1616
|
|
213
|
275
|
|
|
|
|
1077
|
$_ != $node |
214
|
|
|
|
|
|
|
} $current_community->members; |
215
|
|
|
|
|
|
|
|
216
|
275
|
|
|
|
|
442
|
push @{ $new_community->{members_ref} }, $node; |
|
275
|
|
|
|
|
793
|
|
217
|
|
|
|
|
|
|
|
218
|
275
|
|
|
|
|
608
|
$node->{community} = $new_community_id; |
219
|
|
|
|
|
|
|
|
220
|
275
|
|
|
|
|
459
|
$current_community->{w_in} -= $node->{_k_i_in}; |
221
|
275
|
|
|
|
|
560
|
$current_community->{w_tot} -= $node->{_k_i} - $node->{_k_i_in}; |
222
|
|
|
|
|
|
|
|
223
|
275
|
|
|
|
|
631
|
$new_community->{w_in} += $node->{_k_i_new}->{ $new_community_id }; |
224
|
275
|
|
|
|
|
687
|
$new_community->{w_tot} += $node->{_k_i} - $node->{_k_i_new}->{ $new_community_id }; |
225
|
|
|
|
|
|
|
} |
226
|
|
|
|
|
|
|
|
227
|
|
|
|
|
|
|
|
228
|
|
|
|
|
|
|
sub _consolidate_community_structure { |
229
|
2
|
|
|
2
|
|
4
|
my ($self) = @_; |
230
|
|
|
|
|
|
|
|
231
|
2
|
|
|
|
|
16
|
@{$self->{communities_ref}} = grep { |
|
186
|
|
|
|
|
425
|
|
232
|
2
|
|
|
|
|
14
|
int $_->members > 0 |
233
|
|
|
|
|
|
|
} $self->communities; |
234
|
|
|
|
|
|
|
|
235
|
2
|
|
|
|
|
106
|
for my $index (0 .. int($self->communities) - 1) { |
236
|
10
|
|
|
|
|
18
|
$self->{communities_ref}->[$index]->{index} = $index; |
237
|
10
|
|
|
|
|
34
|
foreach my $member ($self->{communities_ref}->[$index]->members) { |
238
|
186
|
|
|
|
|
349
|
$member->{community} = $index; |
239
|
|
|
|
|
|
|
} |
240
|
|
|
|
|
|
|
} |
241
|
|
|
|
|
|
|
} |
242
|
|
|
|
|
|
|
|
243
|
|
|
|
|
|
|
|
244
|
|
|
|
|
|
|
sub _create_next_level_network { |
245
|
2
|
|
|
2
|
|
5
|
my ($net) = @_; |
246
|
|
|
|
|
|
|
|
247
|
2
|
|
|
|
|
15
|
my $next_level_network = SNA::Network->new; |
248
|
2
|
|
|
|
|
9
|
foreach my $community ($net->communities) { |
249
|
10
|
|
|
|
|
39
|
my $new_node = $next_level_network->create_node( |
250
|
|
|
|
|
|
|
community => $community->index, |
251
|
|
|
|
|
|
|
subcommunity => $community, |
252
|
|
|
|
|
|
|
); |
253
|
10
|
|
|
|
|
54
|
$next_level_network->create_edge( |
254
|
|
|
|
|
|
|
source_index => $new_node->index, |
255
|
|
|
|
|
|
|
target_index => $new_node->index, |
256
|
|
|
|
|
|
|
weight => $community->w_in, |
257
|
|
|
|
|
|
|
); |
258
|
|
|
|
|
|
|
} |
259
|
|
|
|
|
|
|
|
260
|
2
|
|
|
|
|
8
|
my $nc = int $net->communities; |
261
|
2
|
|
|
|
|
9
|
my @edge_weights = map { {} } 1 .. $nc; |
|
10
|
|
|
|
|
17
|
|
262
|
2
|
|
|
|
|
10
|
foreach my $edge ($net->edges) { |
263
|
1478
|
|
|
|
|
4538
|
$edge_weights[ $edge->source->community ]->{ $edge->target->community } += $edge->weight; |
264
|
|
|
|
|
|
|
} |
265
|
|
|
|
|
|
|
|
266
|
2
|
|
|
|
|
72
|
foreach my $meta_node ($next_level_network->nodes) { |
267
|
|
|
|
|
|
|
PEERS: |
268
|
10
|
|
|
|
|
26
|
foreach my $peer_node ($next_level_network->nodes) { |
269
|
52
|
100
|
|
|
|
117
|
next PEERS if $meta_node == $peer_node; |
270
|
42
|
|
|
|
|
114
|
my $weight = $edge_weights[ $meta_node->index ]->{ $peer_node->index }; |
271
|
|
|
|
|
|
|
|
272
|
42
|
100
|
|
|
|
84
|
if ($weight) { |
273
|
36
|
|
|
|
|
117
|
$next_level_network->create_edge( |
274
|
|
|
|
|
|
|
source_index => $meta_node->index, |
275
|
|
|
|
|
|
|
target_index => $peer_node->index, |
276
|
|
|
|
|
|
|
weight => $weight, |
277
|
|
|
|
|
|
|
); |
278
|
|
|
|
|
|
|
} |
279
|
|
|
|
|
|
|
} |
280
|
|
|
|
|
|
|
} |
281
|
|
|
|
|
|
|
|
282
|
2
|
|
|
|
|
14
|
return $next_level_network; |
283
|
|
|
|
|
|
|
} |
284
|
|
|
|
|
|
|
|
285
|
|
|
|
|
|
|
|
286
|
|
|
|
|
|
|
sub _propagate_new_community_id { |
287
|
376
|
|
|
376
|
|
387
|
my ($node) = @_; |
288
|
376
|
100
|
|
|
|
980
|
if ($node->{subcommunity}) { |
289
|
16
|
|
|
|
|
51
|
foreach my $sub_member ($node->{subcommunity}->members) { |
290
|
366
|
|
|
|
|
750
|
$sub_member->{community} = $node->community; |
291
|
366
|
|
|
|
|
559
|
_propagate_new_community_id($sub_member); |
292
|
|
|
|
|
|
|
} |
293
|
|
|
|
|
|
|
} |
294
|
|
|
|
|
|
|
} |
295
|
|
|
|
|
|
|
|
296
|
|
|
|
|
|
|
|
297
|
|
|
|
|
|
|
=head1 AUTHOR |
298
|
|
|
|
|
|
|
|
299
|
|
|
|
|
|
|
Darko Obradovic, C<< >> |
300
|
|
|
|
|
|
|
|
301
|
|
|
|
|
|
|
=head1 BUGS |
302
|
|
|
|
|
|
|
|
303
|
|
|
|
|
|
|
Please report any bugs or feature requests to C, or through |
304
|
|
|
|
|
|
|
the web interface at L. I will be notified, and then you'll |
305
|
|
|
|
|
|
|
automatically be notified of progress on your bug as I make changes. |
306
|
|
|
|
|
|
|
|
307
|
|
|
|
|
|
|
|
308
|
|
|
|
|
|
|
|
309
|
|
|
|
|
|
|
|
310
|
|
|
|
|
|
|
=head1 SUPPORT |
311
|
|
|
|
|
|
|
|
312
|
|
|
|
|
|
|
You can find documentation for this module with the perldoc command. |
313
|
|
|
|
|
|
|
|
314
|
|
|
|
|
|
|
perldoc SNA::Network |
315
|
|
|
|
|
|
|
|
316
|
|
|
|
|
|
|
|
317
|
|
|
|
|
|
|
You can also look for information at: |
318
|
|
|
|
|
|
|
|
319
|
|
|
|
|
|
|
=over 4 |
320
|
|
|
|
|
|
|
|
321
|
|
|
|
|
|
|
=item * RT: CPAN's request tracker |
322
|
|
|
|
|
|
|
|
323
|
|
|
|
|
|
|
L |
324
|
|
|
|
|
|
|
|
325
|
|
|
|
|
|
|
=item * AnnoCPAN: Annotated CPAN documentation |
326
|
|
|
|
|
|
|
|
327
|
|
|
|
|
|
|
L |
328
|
|
|
|
|
|
|
|
329
|
|
|
|
|
|
|
=item * CPAN Ratings |
330
|
|
|
|
|
|
|
|
331
|
|
|
|
|
|
|
L |
332
|
|
|
|
|
|
|
|
333
|
|
|
|
|
|
|
=item * Search CPAN |
334
|
|
|
|
|
|
|
|
335
|
|
|
|
|
|
|
L |
336
|
|
|
|
|
|
|
|
337
|
|
|
|
|
|
|
=back |
338
|
|
|
|
|
|
|
|
339
|
|
|
|
|
|
|
|
340
|
|
|
|
|
|
|
=head1 ACKNOWLEDGEMENTS |
341
|
|
|
|
|
|
|
|
342
|
|
|
|
|
|
|
|
343
|
|
|
|
|
|
|
=head1 COPYRIGHT & LICENSE |
344
|
|
|
|
|
|
|
|
345
|
|
|
|
|
|
|
Copyright 2012 Darko Obradovic, all rights reserved. |
346
|
|
|
|
|
|
|
|
347
|
|
|
|
|
|
|
This program is free software; you can redistribute it and/or modify it |
348
|
|
|
|
|
|
|
under the same terms as Perl itself. |
349
|
|
|
|
|
|
|
|
350
|
|
|
|
|
|
|
|
351
|
|
|
|
|
|
|
=cut |
352
|
|
|
|
|
|
|
|
353
|
|
|
|
|
|
|
1; |
354
|
|
|
|
|
|
|
|