line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package KinoSearch1::Highlight::Highlighter; |
2
|
18
|
|
|
18
|
|
103
|
use strict; |
|
18
|
|
|
|
|
44
|
|
|
18
|
|
|
|
|
667
|
|
3
|
18
|
|
|
18
|
|
99
|
use warnings; |
|
18
|
|
|
|
|
45
|
|
|
18
|
|
|
|
|
453
|
|
4
|
18
|
|
|
18
|
|
127
|
use KinoSearch1::Util::ToolSet; |
|
18
|
|
|
|
|
36
|
|
|
18
|
|
|
|
|
2828
|
|
5
|
18
|
|
|
18
|
|
104
|
use base qw( KinoSearch1::Util::Class ); |
|
18
|
|
|
|
|
68
|
|
|
18
|
|
|
|
|
1578
|
|
6
|
18
|
|
|
18
|
|
11460
|
use locale; |
|
18
|
|
|
|
|
2752
|
|
|
18
|
|
|
|
|
126
|
|
7
|
|
|
|
|
|
|
|
8
|
|
|
|
|
|
|
BEGIN { |
9
|
18
|
|
|
18
|
|
1841
|
__PACKAGE__->init_instance_vars( |
10
|
|
|
|
|
|
|
# constructor params / members |
11
|
|
|
|
|
|
|
excerpt_field => undef, |
12
|
|
|
|
|
|
|
analyzer => undef, |
13
|
|
|
|
|
|
|
formatter => undef, |
14
|
|
|
|
|
|
|
encoder => undef, |
15
|
|
|
|
|
|
|
terms => undef, |
16
|
|
|
|
|
|
|
excerpt_length => 200, |
17
|
|
|
|
|
|
|
pre_tag => undef, # back compat |
18
|
|
|
|
|
|
|
post_tag => undef, # back compat |
19
|
|
|
|
|
|
|
token_re => qr/\b\w+(?:'\w+)?\b/, |
20
|
|
|
|
|
|
|
|
21
|
|
|
|
|
|
|
# members |
22
|
|
|
|
|
|
|
limit => undef, |
23
|
|
|
|
|
|
|
); |
24
|
18
|
|
|
|
|
181
|
__PACKAGE__->ready_get_set(qw( terms )); |
25
|
|
|
|
|
|
|
} |
26
|
|
|
|
|
|
|
|
27
|
18
|
|
|
18
|
|
10961
|
use KinoSearch1::Highlight::SimpleHTMLFormatter; |
|
18
|
|
|
|
|
51
|
|
|
18
|
|
|
|
|
463
|
|
28
|
18
|
|
|
18
|
|
10697
|
use KinoSearch1::Highlight::SimpleHTMLEncoder; |
|
18
|
|
|
|
|
52
|
|
|
18
|
|
|
|
|
52856
|
|
29
|
|
|
|
|
|
|
|
30
|
|
|
|
|
|
|
sub init_instance { |
31
|
1
|
|
|
1
|
1
|
199
|
my $self = shift; |
32
|
1
|
50
|
|
|
|
177
|
croak("Missing required arg 'excerpt_field'") |
33
|
|
|
|
|
|
|
unless defined $self->{excerpt_field}; |
34
|
1
|
|
|
|
|
3
|
$self->{terms} = []; |
35
|
|
|
|
|
|
|
|
36
|
|
|
|
|
|
|
# assume HTML |
37
|
1
|
50
|
|
|
|
174
|
if ( !defined $self->{encoder} ) { |
38
|
1
|
|
|
|
|
11
|
$self->{encoder} = KinoSearch1::Highlight::SimpleHTMLEncoder->new; |
39
|
|
|
|
|
|
|
} |
40
|
1
|
50
|
|
|
|
7
|
if ( !defined $self->{formatter} ) { |
41
|
1
|
|
|
|
|
3
|
my ( $pre_tag, $post_tag ) = @{$self}{qw( pre_tag post_tag )}; |
|
1
|
|
|
|
|
3
|
|
42
|
1
|
50
|
|
|
|
5
|
$pre_tag = '' unless defined $pre_tag; |
43
|
1
|
50
|
|
|
|
4
|
$post_tag = '' unless defined $post_tag; |
44
|
1
|
|
|
|
|
17
|
$self->{formatter} = KinoSearch1::Highlight::SimpleHTMLFormatter->new( |
45
|
|
|
|
|
|
|
pre_tag => $pre_tag, |
46
|
|
|
|
|
|
|
post_tag => $post_tag, |
47
|
|
|
|
|
|
|
); |
48
|
|
|
|
|
|
|
} |
49
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
|
# scoring window is 1.66 * excerpt_length, with the loc in the middle |
51
|
1
|
|
|
|
|
5
|
$self->{limit} = int( $self->{excerpt_length} / 3 ); |
52
|
|
|
|
|
|
|
} |
53
|
|
|
|
|
|
|
|
54
|
|
|
|
|
|
|
sub generate_excerpt { |
55
|
4
|
|
|
4
|
0
|
8
|
my ( $self, $doc ) = @_; |
56
|
4
|
|
|
|
|
11
|
my $excerpt_length = $self->{excerpt_length}; |
57
|
4
|
|
|
|
|
8
|
my $limit = $self->{limit}; |
58
|
4
|
|
|
|
|
8
|
my $token_re = $self->{token_re}; |
59
|
|
|
|
|
|
|
|
60
|
|
|
|
|
|
|
# retrieve the text from the chosen field |
61
|
4
|
|
|
|
|
17
|
my $field = $doc->get_field( $self->{excerpt_field} ); |
62
|
4
|
|
|
|
|
13
|
my $text = $field->get_value; |
63
|
4
|
|
|
|
|
18
|
my $text_length = bytes::length $text; |
64
|
4
|
50
|
|
|
|
21
|
return '' unless $text_length; |
65
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
# determine the rough boundaries of the excerpt |
67
|
4
|
|
|
|
|
15
|
my $posits = $self->_starts_and_ends($field); |
68
|
4
|
|
|
|
|
14
|
my $best_location = $self->_calc_best_location($posits); |
69
|
4
|
|
|
|
|
10
|
my $top = $best_location - $limit; |
70
|
|
|
|
|
|
|
|
71
|
|
|
|
|
|
|
# expand the excerpt if the best location is near the end |
72
|
4
|
100
|
|
|
|
10
|
$top |
73
|
|
|
|
|
|
|
= $text_length - $excerpt_length < $top |
74
|
|
|
|
|
|
|
? $text_length - $excerpt_length |
75
|
|
|
|
|
|
|
: $top; |
76
|
|
|
|
|
|
|
|
77
|
|
|
|
|
|
|
# if the best starting point is the very beginning, cool... |
78
|
4
|
100
|
|
|
|
10
|
if ( $top <= 0 ) { |
79
|
2
|
|
|
|
|
4
|
$top = 0; |
80
|
|
|
|
|
|
|
} |
81
|
|
|
|
|
|
|
# ... otherwise ... |
82
|
|
|
|
|
|
|
else { |
83
|
|
|
|
|
|
|
# lop off $top bytes |
84
|
2
|
|
|
|
|
9
|
$text = bytes::substr( $text, $top ); |
85
|
|
|
|
|
|
|
|
86
|
|
|
|
|
|
|
# try to start the excerpt at a sentence boundary |
87
|
2
|
50
|
|
|
|
74
|
if ($text =~ s/ |
88
|
|
|
|
|
|
|
\A |
89
|
|
|
|
|
|
|
( |
90
|
|
|
|
|
|
|
\C{0,$limit}? |
91
|
|
|
|
|
|
|
\.\s+ |
92
|
|
|
|
|
|
|
) |
93
|
|
|
|
|
|
|
//xsm |
94
|
|
|
|
|
|
|
) |
95
|
|
|
|
|
|
|
{ |
96
|
0
|
|
|
|
|
0
|
$top += bytes::length($1); |
97
|
|
|
|
|
|
|
} |
98
|
|
|
|
|
|
|
# no sentence boundary, so we'll need an ellipsis |
99
|
|
|
|
|
|
|
else { |
100
|
|
|
|
|
|
|
# skip past possible partial tokens, prepend an ellipsis |
101
|
2
|
50
|
|
|
|
54
|
if ($text =~ s/ |
102
|
|
|
|
|
|
|
\A |
103
|
|
|
|
|
|
|
( |
104
|
|
|
|
|
|
|
\C{0,$limit}? # don't go outside the window |
105
|
|
|
|
|
|
|
$token_re # match possible partial token |
106
|
|
|
|
|
|
|
.*? # ... and any junk following that token |
107
|
|
|
|
|
|
|
) |
108
|
|
|
|
|
|
|
(?=$token_re) # just before the start of a full token... |
109
|
|
|
|
|
|
|
/... /xsm # ... insert an ellipsis |
110
|
|
|
|
|
|
|
) |
111
|
|
|
|
|
|
|
{ |
112
|
2
|
|
|
|
|
7
|
$top += bytes::length($1); |
113
|
2
|
|
|
|
|
11
|
$top -= 4 # three dots and a space |
114
|
|
|
|
|
|
|
} |
115
|
|
|
|
|
|
|
} |
116
|
|
|
|
|
|
|
} |
117
|
|
|
|
|
|
|
|
118
|
|
|
|
|
|
|
# remove possible partial tokens from the end of the excerpt |
119
|
4
|
|
|
|
|
16
|
$text = bytes::substr( $text, 0, $excerpt_length + 1 ); |
120
|
4
|
100
|
|
|
|
28
|
if ( bytes::length($text) > $excerpt_length ) { |
121
|
2
|
|
|
|
|
10
|
my $extra_char = chop $text; |
122
|
|
|
|
|
|
|
# if the extra char wasn't part of a token, we aren't splitting one |
123
|
2
|
50
|
|
|
|
16
|
if ( $extra_char =~ $token_re ) { |
124
|
2
|
|
|
|
|
93
|
$text =~ s/$token_re$//; # if this is unsuccessful, that's fine |
125
|
|
|
|
|
|
|
} |
126
|
|
|
|
|
|
|
} |
127
|
|
|
|
|
|
|
|
128
|
|
|
|
|
|
|
# if the excerpt doesn't end with a full stop, end with an an ellipsis |
129
|
4
|
100
|
|
|
|
40
|
if ( $text !~ /\.\s*\Z/xsm ) { |
130
|
3
|
|
|
|
|
40
|
$text =~ s/\W+\Z//xsm; |
131
|
3
|
|
|
|
|
10
|
while ( bytes::length($text) + 4 > $excerpt_length ) { |
132
|
4
|
|
|
|
|
18
|
my $extra_char = chop $text; |
133
|
4
|
50
|
|
|
|
29
|
if ( $extra_char =~ $token_re ) { |
134
|
4
|
|
|
|
|
151
|
$text =~ s/\W+$token_re\Z//xsm; # if unsuccessful, that's fine |
135
|
|
|
|
|
|
|
} |
136
|
4
|
|
|
|
|
73
|
$text =~ s/\W+\Z//xsm; |
137
|
|
|
|
|
|
|
} |
138
|
3
|
|
|
|
|
17
|
$text .= ' ...'; |
139
|
|
|
|
|
|
|
} |
140
|
|
|
|
|
|
|
|
141
|
|
|
|
|
|
|
# remap locations now that we know the starting and ending bytes |
142
|
4
|
|
|
|
|
11
|
$text_length = bytes::length($text); |
143
|
4
|
|
|
|
|
17
|
my @relative_starts = map { $_->[0] - $top } @$posits; |
|
5
|
|
|
|
|
13
|
|
144
|
4
|
|
|
|
|
8
|
my @relative_ends = map { $_->[1] - $top } @$posits; |
|
5
|
|
|
|
|
10
|
|
145
|
|
|
|
|
|
|
|
146
|
|
|
|
|
|
|
# get rid of pairs with at least one member outside the text |
147
|
4
|
|
66
|
|
|
25
|
while ( @relative_starts and $relative_starts[0] < 0 ) { |
148
|
0
|
|
|
|
|
0
|
shift @relative_starts; |
149
|
0
|
|
|
|
|
0
|
shift @relative_ends; |
150
|
|
|
|
|
|
|
} |
151
|
4
|
|
66
|
|
|
22
|
while ( @relative_ends and $relative_ends[-1] > $text_length ) { |
152
|
0
|
|
|
|
|
0
|
pop @relative_starts; |
153
|
0
|
|
|
|
|
0
|
pop @relative_ends; |
154
|
|
|
|
|
|
|
} |
155
|
|
|
|
|
|
|
|
156
|
|
|
|
|
|
|
# insert highlight tags |
157
|
4
|
|
|
|
|
9
|
my $formatter = $self->{formatter}; |
158
|
4
|
|
|
|
|
10
|
my $encoder = $self->{encoder}; |
159
|
4
|
|
|
|
|
6
|
my $output_text = ''; |
160
|
4
|
|
|
|
|
8
|
my ( $start, $end, $last_start, $last_end ) = ( undef, undef, 0, 0 ); |
161
|
4
|
|
|
|
|
11
|
while (@relative_starts) { |
162
|
5
|
|
|
|
|
7
|
$end = shift @relative_ends; |
163
|
5
|
|
|
|
|
6
|
$start = shift @relative_starts; |
164
|
5
|
|
|
|
|
14
|
$output_text .= $encoder->encode( |
165
|
|
|
|
|
|
|
bytes::substr( $text, $last_end, $start - $last_end ) ); |
166
|
5
|
|
|
|
|
18
|
$output_text |
167
|
|
|
|
|
|
|
.= $formatter->highlight( |
168
|
|
|
|
|
|
|
$encoder->encode( bytes::substr( $text, $start, $end - $start ) ) |
169
|
|
|
|
|
|
|
); |
170
|
5
|
|
|
|
|
13
|
$last_end = $end; |
171
|
|
|
|
|
|
|
} |
172
|
4
|
|
|
|
|
12
|
$output_text .= $encoder->encode( bytes::substr( $text, $last_end ) ); |
173
|
|
|
|
|
|
|
|
174
|
4
|
|
|
|
|
25
|
return $output_text; |
175
|
|
|
|
|
|
|
} |
176
|
|
|
|
|
|
|
|
177
|
|
|
|
|
|
|
=for comment |
178
|
|
|
|
|
|
|
Find all points in the text where a relevant term begins and ends. For terms |
179
|
|
|
|
|
|
|
that are part of a phrase, only include points that are part of the phrase. |
180
|
|
|
|
|
|
|
|
181
|
|
|
|
|
|
|
=cut |
182
|
|
|
|
|
|
|
|
183
|
|
|
|
|
|
|
sub _starts_and_ends { |
184
|
4
|
|
|
4
|
|
5
|
my ( $self, $field ) = @_; |
185
|
4
|
|
|
|
|
6
|
my @posits; |
186
|
|
|
|
|
|
|
my %done; |
187
|
|
|
|
|
|
|
|
188
|
4
|
|
|
|
|
6
|
TERM: for my $term ( @{ $self->{terms} } ) { |
|
4
|
|
|
|
|
12
|
|
189
|
16
|
100
|
|
|
|
39
|
if ( a_isa_b( $term, 'KinoSearch1::Index::Term' ) ) { |
190
|
10
|
|
|
|
|
33
|
my $term_text = $term->get_text; |
191
|
|
|
|
|
|
|
|
192
|
10
|
100
|
|
|
|
33
|
next TERM if $done{$term_text}; |
193
|
5
|
|
|
|
|
10
|
$done{$term_text} = 1; |
194
|
|
|
|
|
|
|
|
195
|
|
|
|
|
|
|
# add all starts and ends |
196
|
5
|
|
|
|
|
17
|
my $term_vector = $field->term_vector($term_text); |
197
|
5
|
100
|
|
|
|
15
|
next TERM unless defined $term_vector; |
198
|
4
|
|
|
|
|
12
|
my $starts = $term_vector->get_start_offsets; |
199
|
4
|
|
|
|
|
10
|
my $ends = $term_vector->get_end_offsets; |
200
|
4
|
|
|
|
|
10
|
while (@$starts) { |
201
|
4
|
|
|
|
|
22
|
push @posits, [ shift @$starts, shift @$ends, 1 ]; |
202
|
|
|
|
|
|
|
} |
203
|
|
|
|
|
|
|
} |
204
|
|
|
|
|
|
|
# intersect positions for phrase terms |
205
|
|
|
|
|
|
|
else { |
206
|
|
|
|
|
|
|
# if not a Term, it's an array of Terms representing a phrase |
207
|
6
|
|
|
|
|
10
|
my @term_texts = map { $_->get_text } @$term; |
|
18
|
|
|
|
|
40
|
|
208
|
|
|
|
|
|
|
|
209
|
6
|
|
|
|
|
17
|
my $phrase_text = join( ' ', @term_texts ); |
210
|
6
|
100
|
|
|
|
19
|
next TERM if $done{$phrase_text}; |
211
|
3
|
|
|
|
|
7
|
$done{$phrase_text} = 1; |
212
|
|
|
|
|
|
|
|
213
|
3
|
|
|
|
|
42
|
my $posit_vec = KinoSearch1::Util::BitVector->new; |
214
|
3
|
|
|
|
|
6
|
my @term_vectors = map { $field->term_vector($_) } @term_texts; |
|
9
|
|
|
|
|
21
|
|
215
|
|
|
|
|
|
|
|
216
|
|
|
|
|
|
|
# make sure all terms are present |
217
|
3
|
100
|
|
|
|
16
|
next TERM unless scalar @term_vectors == scalar @term_texts; |
218
|
|
|
|
|
|
|
|
219
|
2
|
|
|
|
|
4
|
my $i = 0; |
220
|
2
|
|
|
|
|
4
|
for my $tv (@term_vectors) { |
221
|
|
|
|
|
|
|
# one term missing, ergo no phrase |
222
|
6
|
50
|
|
|
|
14
|
next TERM unless defined $tv; |
223
|
6
|
100
|
|
|
|
13
|
if ( $i == 0 ) { |
224
|
2
|
|
|
|
|
5
|
$posit_vec->set( @{ $tv->get_positions } ); |
|
2
|
|
|
|
|
7
|
|
225
|
|
|
|
|
|
|
} |
226
|
|
|
|
|
|
|
else { |
227
|
|
|
|
|
|
|
# filter positions using logical "and" |
228
|
4
|
|
|
|
|
27
|
my $other_posit_vec = KinoSearch1::Util::BitVector->new; |
229
|
4
|
|
|
|
|
15
|
$other_posit_vec->set( |
230
|
4
|
|
|
|
|
9
|
grep { $_ >= 0 } |
231
|
4
|
|
|
|
|
10
|
map { $_ - $i } @{ $tv->get_positions } |
|
4
|
|
|
|
|
11
|
|
232
|
|
|
|
|
|
|
); |
233
|
4
|
|
|
|
|
33
|
$posit_vec->logical_and($other_posit_vec); |
234
|
|
|
|
|
|
|
} |
235
|
6
|
|
|
|
|
11
|
$i++; |
236
|
|
|
|
|
|
|
} |
237
|
|
|
|
|
|
|
|
238
|
|
|
|
|
|
|
# add only those starts/ends that belong to a valid position |
239
|
2
|
|
|
|
|
8
|
my $tv_start_positions = $term_vectors[0]->get_positions; |
240
|
2
|
|
|
|
|
9
|
my $tv_starts = $term_vectors[0]->get_start_offsets; |
241
|
2
|
|
|
|
|
6
|
my $tv_end_positions = $term_vectors[-1]->get_positions; |
242
|
2
|
|
|
|
|
6
|
my $tv_ends = $term_vectors[-1]->get_end_offsets; |
243
|
2
|
|
|
|
|
5
|
$i = 0; |
244
|
2
|
|
|
|
|
3
|
my $j = 0; |
245
|
2
|
|
|
|
|
3
|
my $last_token_index = $#term_vectors; |
246
|
2
|
|
|
|
|
4
|
for my $valid_position ( @{ $posit_vec->to_arrayref } ) { |
|
2
|
|
|
|
|
11
|
|
247
|
|
|
|
|
|
|
|
248
|
2
|
|
|
|
|
9
|
while ( $i <= $#$tv_start_positions ) { |
249
|
2
|
50
|
|
|
|
7
|
last if ( $tv_start_positions->[$i] >= $valid_position ); |
250
|
0
|
|
|
|
|
0
|
$i++; |
251
|
|
|
|
|
|
|
} |
252
|
2
|
|
|
|
|
3
|
$valid_position += $last_token_index; |
253
|
2
|
|
|
|
|
5
|
while ( $j <= $#$tv_end_positions ) { |
254
|
2
|
50
|
|
|
|
6
|
last if ( $tv_end_positions->[$j] >= $valid_position ); |
255
|
0
|
|
|
|
|
0
|
$j++; |
256
|
|
|
|
|
|
|
} |
257
|
2
|
|
|
|
|
7
|
push @posits, |
258
|
|
|
|
|
|
|
[ $tv_starts->[$i], $tv_ends->[$j], scalar @$term ]; |
259
|
2
|
|
|
|
|
4
|
$i++; |
260
|
2
|
|
|
|
|
29
|
$j++; |
261
|
|
|
|
|
|
|
} |
262
|
|
|
|
|
|
|
} |
263
|
|
|
|
|
|
|
} |
264
|
|
|
|
|
|
|
|
265
|
|
|
|
|
|
|
# sort, uniquify and return |
266
|
4
|
50
|
|
|
|
13
|
@posits = sort { $a->[0] <=> $b->[0] || $b->[1] <=> $a->[1] } @posits; |
|
3
|
|
|
|
|
10
|
|
267
|
4
|
|
|
|
|
6
|
my @unique; |
268
|
4
|
|
|
|
|
7
|
my $last = ~0; |
269
|
4
|
|
|
|
|
8
|
for (@posits) { |
270
|
6
|
100
|
|
|
|
14
|
push @unique, $_ if $_->[0] != $last; |
271
|
6
|
|
|
|
|
10
|
$last = $_->[0]; |
272
|
|
|
|
|
|
|
} |
273
|
4
|
|
|
|
|
12
|
return \@unique; |
274
|
|
|
|
|
|
|
} |
275
|
|
|
|
|
|
|
|
276
|
|
|
|
|
|
|
=for comment |
277
|
|
|
|
|
|
|
Select the byte address representing the greatest keyword density. Because |
278
|
|
|
|
|
|
|
the algorithm counts bytes rather than characters, it will degrade if the |
279
|
|
|
|
|
|
|
number of bytes per character is larger than 1. |
280
|
|
|
|
|
|
|
|
281
|
|
|
|
|
|
|
=cut |
282
|
|
|
|
|
|
|
|
283
|
|
|
|
|
|
|
sub _calc_best_location { |
284
|
4
|
|
|
4
|
|
6
|
my ( $self, $posits ) = @_; |
285
|
4
|
|
|
|
|
6
|
my $window = $self->{limit} * 2; |
286
|
|
|
|
|
|
|
|
287
|
|
|
|
|
|
|
# if there aren't any keywords, take the excerpt from the top of the text |
288
|
4
|
100
|
|
|
|
10
|
return 0 unless @$posits; |
289
|
|
|
|
|
|
|
|
290
|
3
|
|
|
|
|
3
|
my %locations = map { ( $_->[0] => 0 ) } @$posits; |
|
5
|
|
|
|
|
19
|
|
291
|
|
|
|
|
|
|
|
292
|
|
|
|
|
|
|
# if another keyword is in close proximity, add to the loc's score |
293
|
3
|
|
|
|
|
9
|
for my $loc_index ( 0 .. $#$posits ) { |
294
|
|
|
|
|
|
|
# only score positions that are in range |
295
|
5
|
|
|
|
|
8
|
my $location = $posits->[$loc_index][0]; |
296
|
5
|
|
|
|
|
7
|
my $other_loc_index = $loc_index - 1; |
297
|
5
|
|
|
|
|
16
|
while ( $other_loc_index > 0 ) { |
298
|
0
|
|
|
|
|
0
|
my $diff = $location - $posits->[$other_loc_index][0]; |
299
|
0
|
0
|
|
|
|
0
|
last if $diff > $window; |
300
|
0
|
|
|
|
|
0
|
my $num_tokens_at_pos = $posits->[$other_loc_index][2]; |
301
|
0
|
|
|
|
|
0
|
$locations{$location} |
302
|
|
|
|
|
|
|
+= ( 1 / ( 1 + log($diff) ) ) * $num_tokens_at_pos; |
303
|
0
|
|
|
|
|
0
|
--$other_loc_index; |
304
|
|
|
|
|
|
|
} |
305
|
5
|
|
|
|
|
7
|
$other_loc_index = $loc_index + 1; |
306
|
5
|
|
|
|
|
20
|
while ( $other_loc_index <= $#$posits ) { |
307
|
2
|
|
|
|
|
4
|
my $diff = $posits->[$other_loc_index] - $location; |
308
|
2
|
50
|
|
|
|
9
|
last if $diff > $window; |
309
|
0
|
|
|
|
|
0
|
my $num_tokens_at_pos = $posits->[$other_loc_index][2]; |
310
|
0
|
|
|
|
|
0
|
$locations{$location} |
311
|
|
|
|
|
|
|
+= ( 1 / ( 1 + log($diff) ) ) * $num_tokens_at_pos; |
312
|
0
|
|
|
|
|
0
|
++$other_loc_index; |
313
|
|
|
|
|
|
|
} |
314
|
|
|
|
|
|
|
} |
315
|
|
|
|
|
|
|
|
316
|
|
|
|
|
|
|
# return the highest scoring position |
317
|
3
|
|
|
|
|
13
|
return ( sort { $locations{$b} <=> $locations{$a} } keys %locations )[0]; |
|
2
|
|
|
|
|
10
|
|
318
|
|
|
|
|
|
|
} |
319
|
|
|
|
|
|
|
|
320
|
|
|
|
|
|
|
1; |
321
|
|
|
|
|
|
|
|
322
|
|
|
|
|
|
|
__END__ |