line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package KSx::Highlight::Summarizer; |
2
|
|
|
|
|
|
|
|
3
|
|
|
|
|
|
|
$VERSION = '0.06'; |
4
|
|
|
|
|
|
|
|
5
|
|
|
|
|
|
|
@ISA = KinoSearch::Highlight::Highlighter; |
6
|
1
|
|
|
1
|
|
1500
|
use KinoSearch::Highlight::Highlighter; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
57
|
|
7
|
|
|
|
|
|
|
|
8
|
1
|
|
|
1
|
|
6
|
use strict; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
44
|
|
9
|
|
|
|
|
|
|
|
10
|
1
|
|
|
1
|
|
24
|
use List::Util qw 'min'; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
169
|
|
11
|
1
|
|
|
1
|
|
1073
|
use Number::Range; |
|
1
|
|
|
|
|
26897
|
|
|
1
|
|
|
|
|
76
|
|
12
|
|
|
|
|
|
|
|
13
|
1
|
|
|
1
|
|
1509
|
use Hash::Util::FieldHash::Compat 'fieldhashes'; |
|
1
|
|
|
|
|
25608
|
|
|
1
|
|
|
|
|
10
|
|
14
|
|
|
|
|
|
|
fieldhashes \my( %ellipsis, %summ_len, %page_h, %encoder ); |
15
|
|
|
|
|
|
|
|
16
|
|
|
|
|
|
|
sub _range_endpoints { |
17
|
0
|
|
|
0
|
|
0
|
my $range = shift; |
18
|
0
|
|
|
|
|
0
|
my @range = $range->range; |
19
|
0
|
|
|
|
|
0
|
my $previous = shift @range; |
20
|
0
|
|
|
|
|
0
|
my $subrange = [($previous) x 2]; |
21
|
0
|
|
|
|
|
0
|
my @arrays; |
22
|
0
|
|
|
|
|
0
|
foreach my $current (@range) { |
23
|
0
|
0
|
|
|
|
0
|
if ($current == ($previous + 1)) { |
24
|
0
|
|
|
|
|
0
|
$subrange->[1] = $current; |
25
|
|
|
|
|
|
|
} |
26
|
|
|
|
|
|
|
else { |
27
|
0
|
|
|
|
|
0
|
push @arrays, $subrange; |
28
|
0
|
|
|
|
|
0
|
$subrange = [($current) x 2]; |
29
|
|
|
|
|
|
|
} |
30
|
0
|
|
|
|
|
0
|
$previous = $current; |
31
|
|
|
|
|
|
|
} |
32
|
0
|
|
|
|
|
0
|
return @arrays, $subrange; # Make sure the last subrange isn’t left out! |
33
|
|
|
|
|
|
|
} |
34
|
|
|
|
|
|
|
|
35
|
|
|
|
|
|
|
sub new { |
36
|
1
|
|
|
1
|
1
|
31123
|
my($pack, %args) = @_; |
37
|
1
|
50
|
|
|
|
10
|
my $ellipsis = exists $args{ellipsis} ? delete $args{ellipsis} |
38
|
|
|
|
|
|
|
: ' ... '; |
39
|
1
|
50
|
|
|
|
4
|
my $summ_len = exists $args{summary_length} |
40
|
|
|
|
|
|
|
? delete $args{summary_length} : 0; |
41
|
1
|
|
|
|
|
4
|
my $page_h = delete $args{page_handler}; |
42
|
1
|
|
|
|
|
3
|
my $encoder = delete $args{encoder}; |
43
|
|
|
|
|
|
|
|
44
|
|
|
|
|
|
|
# accept args that the superclass only allows one to set through |
45
|
|
|
|
|
|
|
# accessor methods: |
46
|
1
|
|
|
|
|
3
|
my $pre_tag = delete $args{pre_tag}; |
47
|
1
|
|
|
|
|
2
|
my $post_tag = delete $args{post_tag}; |
48
|
|
|
|
|
|
|
|
49
|
1
|
|
|
|
|
201
|
my $self = SUPER::new $pack %args; |
50
|
|
|
|
|
|
|
|
51
|
0
|
|
|
|
|
|
$ellipsis{$self} = $ellipsis; |
52
|
0
|
|
|
|
|
|
$summ_len{$self} = $summ_len; |
53
|
0
|
|
|
|
|
|
$page_h{$self} = $page_h; |
54
|
0
|
|
|
|
|
|
$encoder{$self} = $encoder; |
55
|
|
|
|
|
|
|
|
56
|
0
|
0
|
|
|
|
|
defined $pre_tag and $self->set_pre_tag($pre_tag); |
57
|
0
|
0
|
|
|
|
|
defined $post_tag and $self->set_post_tag($post_tag); |
58
|
|
|
|
|
|
|
|
59
|
0
|
|
|
|
|
|
return $self; |
60
|
|
|
|
|
|
|
} |
61
|
|
|
|
|
|
|
|
62
|
|
|
|
|
|
|
sub create_excerpt { |
63
|
0
|
|
|
0
|
1
|
|
my ($self, $hitdoc) = @_; |
64
|
|
|
|
|
|
|
|
65
|
0
|
|
|
|
|
|
my $field = $self->get_field; |
66
|
0
|
|
|
|
|
|
my $x_len = $self->get_excerpt_length; |
67
|
0
|
|
|
|
|
|
my $limit = int($x_len /3 ); |
68
|
|
|
|
|
|
|
|
69
|
|
|
|
|
|
|
# retrieve the text from the chosen field |
70
|
0
|
|
|
|
|
|
my $text = $hitdoc->{$field}; |
71
|
0
|
0
|
|
|
|
|
return unless defined $text; |
72
|
0
|
|
|
|
|
|
my $text_length = length $text; |
73
|
0
|
0
|
|
|
|
|
return '' unless $text_length; |
74
|
|
|
|
|
|
|
|
75
|
|
|
|
|
|
|
# get offsets and weights of words that match |
76
|
0
|
|
|
|
|
|
my $searcher = $self->get_searchable; |
77
|
0
|
|
|
|
|
|
my $posits = $self->get_compiler->highlight_spans( |
78
|
|
|
|
|
|
|
searchable => $searcher, |
79
|
|
|
|
|
|
|
field => $field, |
80
|
|
|
|
|
|
|
doc_vec => $searcher->fetch_doc_vec( |
81
|
|
|
|
|
|
|
$hitdoc->get_doc_id |
82
|
|
|
|
|
|
|
), |
83
|
|
|
|
|
|
|
); |
84
|
0
|
|
|
|
|
|
my @locs = map [$_->get_offset,$_->get_weight], @{ |
85
|
0
|
|
|
|
|
|
KinoSearch::Highlight::HeatMap->new( |
86
|
|
|
|
|
|
|
spans => $posits, |
87
|
|
|
|
|
|
|
window => $limit*2 |
88
|
|
|
|
|
|
|
)->get_spans |
89
|
|
|
|
|
|
|
}; |
90
|
0
|
|
|
|
|
|
@locs = map $$_[0], sort { $$b[1] <=> $$a[1] } @locs; |
|
0
|
|
|
|
|
|
|
91
|
|
|
|
|
|
|
|
92
|
0
|
0
|
|
|
|
|
@locs or @locs = 0; |
93
|
|
|
|
|
|
|
|
94
|
|
|
|
|
|
|
#warn "@locs" if $summ_len{$self}; |
95
|
|
|
|
|
|
|
# determine the rough boundaries of the excerpts |
96
|
0
|
|
|
|
|
|
my $range = new Number::Range; |
97
|
0
|
|
|
|
|
|
my $summ_len = $summ_len{$self}; |
98
|
0
|
|
|
|
|
|
for(@locs) { |
99
|
1
|
|
|
1
|
|
1316
|
no warnings; # suppress Number::Range’s nasty warnings |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
1701
|
|
100
|
0
|
|
|
|
|
|
my $start = $_-$limit; |
101
|
0
|
0
|
|
|
|
|
$start = 0 if $start < 0; |
102
|
0
|
|
|
|
|
|
$range->addrange($start . '..' . min($start+$x_len, $text_length)); |
103
|
0
|
0
|
0
|
|
|
|
last if !$summ_len || $range->size >= $summ_len; |
104
|
|
|
|
|
|
|
} |
105
|
0
|
|
|
|
|
|
my @excerpt_bounds = _range_endpoints($range); |
106
|
|
|
|
|
|
|
#use DDS; warn Dump \@excerpt_bounds if $summ_len; |
107
|
|
|
|
|
|
|
|
108
|
|
|
|
|
|
|
# close small gaps between ranges |
109
|
0
|
|
|
|
|
|
for(my $c = 1; $c < @excerpt_bounds;++$c) { |
110
|
0
|
0
|
|
|
|
|
$excerpt_bounds[$c][0] - $excerpt_bounds[$c-1][1] <= 10 and |
111
|
|
|
|
|
|
|
$excerpt_bounds[$c-1][1] = $excerpt_bounds[$c][1], |
112
|
|
|
|
|
|
|
splice(@excerpt_bounds, $c, 1), |
113
|
|
|
|
|
|
|
--$c; |
114
|
|
|
|
|
|
|
} |
115
|
|
|
|
|
|
|
|
116
|
|
|
|
|
|
|
# extract the offsets from the highlight spans |
117
|
0
|
|
|
|
|
|
my(@starts, @ends); |
118
|
0
|
|
|
|
|
|
for(@$posits) { |
119
|
0
|
|
|
|
|
|
push(@starts, my $start = $_->get_offset); |
120
|
0
|
|
|
|
|
|
push(@ends, $start + $_->get_length); |
121
|
|
|
|
|
|
|
} |
122
|
|
|
|
|
|
|
|
123
|
|
|
|
|
|
|
# make the summary |
124
|
0
|
|
|
|
|
|
my $summary = ''; |
125
|
0
|
|
|
|
|
|
my $ellipsis = $ellipsis{$self}; |
126
|
0
|
|
|
|
|
|
my $token_re = qr/\b\w+(?:'\w+)?\b/; |
127
|
0
|
|
|
|
|
|
my $prev_ellipsis; # whether the previous excerpt ended with an ellip. |
128
|
0
|
|
|
|
|
|
my $prev_page = 0; # last page number of previous excerpt |
129
|
0
|
|
|
|
|
|
my $page_h = $page_h{$self}; |
130
|
0
|
|
|
|
|
|
for(@excerpt_bounds) { |
131
|
|
|
|
|
|
|
# make the excerpt |
132
|
0
|
|
|
|
|
|
my ($start,$end) = @$_; |
133
|
|
|
|
|
|
|
|
134
|
|
|
|
|
|
|
# determine the page number that $start falls within |
135
|
0
|
|
|
|
|
|
my $page_no; |
136
|
0
|
0
|
|
|
|
|
$page_h and $page_no = |
137
|
|
|
|
|
|
|
substr($text, 0,$start) =~ y/\014// + 1; |
138
|
|
|
|
|
|
|
|
139
|
0
|
|
|
|
|
|
my $x; # short for x-cerpt |
140
|
|
|
|
|
|
|
my $need_ellipsis; |
141
|
|
|
|
|
|
|
|
142
|
|
|
|
|
|
|
#warn "<<".substr($text,$start,$limit).">>"; |
143
|
|
|
|
|
|
|
# look for a page break within $limit chars from $start (except we |
144
|
|
|
|
|
|
|
# shouldn’t do it if $start is 0 because there’s a good chance |
145
|
|
|
|
|
|
|
# we’ll go past the very word for whose sake this excerpt exists) |
146
|
|
|
|
|
|
|
# ~~~ What about a case in which a page break plus maybe a few |
147
|
|
|
|
|
|
|
# spaces occur just *before* $start. That shouldn’t get an |
148
|
|
|
|
|
|
|
# ellipsis (as in the elsif block below), should it? |
149
|
0
|
0
|
0
|
|
|
|
if($page_h && $start && |
|
|
0
|
0
|
|
|
|
|
150
|
|
|
|
|
|
|
substr($text,$start,$limit) =~ /^(.*)\014/s) { |
151
|
0
|
|
|
|
|
|
$start += length($1) + 1; |
152
|
0
|
|
|
|
|
|
$page_no += 1 + $1 =~ y/\014//; |
153
|
0
|
|
|
|
|
|
$x = substr $text, $start; |
154
|
|
|
|
|
|
|
} |
155
|
|
|
|
|
|
|
elsif( $start ) { # if this is not the beginning of the doc |
156
|
0
|
|
|
|
|
|
my $sb = $self->find_sentences( |
157
|
|
|
|
|
|
|
text => $text, offset => $start, length => $limit |
158
|
|
|
|
|
|
|
); |
159
|
0
|
0
|
|
|
|
|
if(@$sb) { |
160
|
0
|
|
|
|
|
|
$start = $$sb[0]; |
161
|
|
|
|
|
|
|
} |
162
|
0
|
|
|
|
|
|
else { ++ $need_ellipsis } |
163
|
0
|
|
|
|
|
|
$x = substr $text, $start; |
164
|
0
|
0
|
|
|
|
|
if($need_ellipsis) { |
165
|
|
|
|
|
|
|
# skip past possible partial tokens, but don’t insert an |
166
|
|
|
|
|
|
|
# ellipsis yet, because it might need to come after a |
167
|
|
|
|
|
|
|
# page marker |
168
|
0
|
0
|
|
|
|
|
if ($x =~ s/ |
169
|
|
|
|
|
|
|
\A |
170
|
|
|
|
|
|
|
( |
171
|
|
|
|
|
|
|
.{1,$limit}? # don't go outside the window |
172
|
|
|
|
|
|
|
) |
173
|
|
|
|
|
|
|
(?=$token_re) # just b4 the start of a full token |
174
|
|
|
|
|
|
|
//xsm |
175
|
|
|
|
|
|
|
) |
176
|
|
|
|
|
|
|
{ |
177
|
0
|
|
|
|
|
|
$start += length($1); |
178
|
|
|
|
|
|
|
} |
179
|
|
|
|
|
|
|
} |
180
|
|
|
|
|
|
|
} |
181
|
0
|
|
|
|
|
|
else { $x = substr $text, $start } |
182
|
|
|
|
|
|
|
|
183
|
|
|
|
|
|
|
# trim unwanted text from the end of the excerpt |
184
|
0
|
|
|
|
|
|
$x = substr $x, 0, $end-$start+1; # +1 ’cos we need that extra |
185
|
|
|
|
|
|
|
# char later |
186
|
0
|
|
|
|
|
|
my $end_with_ellipsis = 0; |
187
|
|
|
|
|
|
|
|
188
|
|
|
|
|
|
|
# if we’ve trimmed the end of the text |
189
|
0
|
0
|
|
|
|
|
if ( $end < $text_length) {{ # doubled so ‘last’ will work |
190
|
|
|
|
|
|
|
# check to see whether there are page breaks after the high- |
191
|
|
|
|
|
|
|
# lighted word, and stop at the first one if so |
192
|
0
|
0
|
0
|
|
|
|
if ($page_h and substr($x, $limit*-2) =~ s/(\014[^\014]*)//) { |
|
0
|
|
|
|
|
|
|
193
|
0
|
|
|
|
|
|
$end -= length $1; last; |
|
0
|
|
|
|
|
|
|
194
|
|
|
|
|
|
|
} |
195
|
|
|
|
|
|
|
|
196
|
|
|
|
|
|
|
# remove possible partial tokens from the end of the excerpt |
197
|
0
|
|
|
|
|
|
my $extra_char = chop $x; # the char we left dangling earlier |
198
|
|
|
|
|
|
|
# if the extra char wasn't part of a token, then we’re not |
199
|
|
|
|
|
|
|
# splitting one |
200
|
0
|
0
|
|
|
|
|
if ( $extra_char =~ $token_re ) { |
201
|
0
|
|
|
|
|
|
$x =~ s/$token_re$//; # if this fails, that's fine |
202
|
|
|
|
|
|
|
} |
203
|
|
|
|
|
|
|
|
204
|
|
|
|
|
|
|
# if the excerpt doesn't end with a full stop, end with |
205
|
|
|
|
|
|
|
# an ellipsis |
206
|
0
|
0
|
|
|
|
|
if ( $x !~ /\.\s*\Z/xsm ) { |
207
|
0
|
|
|
|
|
|
$x =~ s/\W+\Z//xsm; |
208
|
0
|
|
|
|
|
|
$x .= $ellipsis; |
209
|
0
|
|
|
|
|
|
++$end_with_ellipsis; |
210
|
|
|
|
|
|
|
} |
211
|
|
|
|
|
|
|
}} |
212
|
|
|
|
|
|
|
#warn $x if $page_h; |
213
|
|
|
|
|
|
|
|
214
|
|
|
|
|
|
|
# get the offsets that are within range for the excerpt, and make |
215
|
|
|
|
|
|
|
# them relative to $start |
216
|
0
|
|
|
|
|
|
my @relative_starts = map $_-$start, @starts; |
217
|
0
|
|
|
|
|
|
my @relative_ends = map $_-$start, @ends; |
218
|
0
|
|
|
|
|
|
my $this_x_len = $end - $start; |
219
|
0
|
|
0
|
|
|
|
while ( @relative_starts and $relative_starts[0] < 0 ) { |
220
|
0
|
|
|
|
|
|
shift @relative_starts; |
221
|
0
|
|
|
|
|
|
shift @relative_ends; |
222
|
|
|
|
|
|
|
} |
223
|
0
|
|
0
|
|
|
|
while ( @relative_ends and $relative_ends[-1] > $this_x_len ) { |
224
|
0
|
|
|
|
|
|
pop @relative_starts; |
225
|
0
|
|
|
|
|
|
pop @relative_ends; |
226
|
|
|
|
|
|
|
} |
227
|
|
|
|
|
|
|
|
228
|
|
|
|
|
|
|
# insert highlight tags and page break markers |
229
|
|
|
|
|
|
|
# sstart and send stand for span start and end |
230
|
0
|
|
|
|
|
|
my ( $sstart, $send, $last_sstart, $last_send ) = |
231
|
|
|
|
|
|
|
( undef, undef, 0, 0 ); |
232
|
0
|
0
|
|
|
|
|
if($page_h) { # Some of this code *is* repeated redundantly, but it |
233
|
|
|
|
|
|
|
# should theoretically run faster since the |
234
|
|
|
|
|
|
|
# if($page_h) check doesn’t have to be made every |
235
|
|
|
|
|
|
|
# time through the loop. |
236
|
0
|
0
|
0
|
|
|
|
$prev_page != $page_no |
|
|
|
0
|
|
|
|
|
237
|
|
|
|
|
|
|
? ( |
238
|
|
|
|
|
|
|
$summary .= &$page_h($hitdoc, $page_no), |
239
|
|
|
|
|
|
|
$need_ellipsis && ($summary .= $ellipsis) |
240
|
|
|
|
|
|
|
) : $need_ellipsis && !$prev_ellipsis && |
241
|
|
|
|
|
|
|
($summary .= $ellipsis) |
242
|
|
|
|
|
|
|
; |
243
|
0
|
|
|
|
|
|
while (@relative_starts) { |
244
|
0
|
|
|
|
|
|
$send = shift @relative_ends; |
245
|
0
|
|
|
|
|
|
$sstart = shift @relative_starts; |
246
|
0
|
0
|
0
|
|
|
|
$summary .= _encode_with_pb( $self, |
247
|
|
|
|
|
|
|
substr( $x, $last_send, $sstart - $last_send ), |
248
|
|
|
|
|
|
|
$page_h, \$page_no, $hitdoc |
249
|
|
|
|
|
|
|
) unless !$last_send && !$sstart; |
250
|
0
|
|
|
|
|
|
$summary .= $self->highlight( |
251
|
|
|
|
|
|
|
_encode_with_pb( $self, |
252
|
|
|
|
|
|
|
substr( $x, $sstart, $send - $sstart ), |
253
|
|
|
|
|
|
|
$page_h, \$page_no, $hitdoc |
254
|
|
|
|
|
|
|
) |
255
|
|
|
|
|
|
|
); |
256
|
0
|
|
|
|
|
|
$last_send = $send; |
257
|
|
|
|
|
|
|
} |
258
|
0
|
0
|
|
|
|
|
$summary .= _encode_with_pb( $self, |
259
|
|
|
|
|
|
|
substr( $x, $last_send ), |
260
|
|
|
|
|
|
|
$page_h, \$page_no, $hitdoc |
261
|
|
|
|
|
|
|
) unless $last_send == length $x; |
262
|
0
|
|
|
|
|
|
$prev_page = $page_no; |
263
|
|
|
|
|
|
|
} |
264
|
|
|
|
|
|
|
else { |
265
|
0
|
0
|
0
|
|
|
|
$need_ellipsis and !$prev_ellipsis and $summary .= $ellipsis; |
266
|
0
|
|
|
|
|
|
while (@relative_starts) { |
267
|
0
|
|
|
|
|
|
$send = shift @relative_ends; |
268
|
0
|
|
|
|
|
|
$sstart = shift @relative_starts; |
269
|
0
|
0
|
0
|
|
|
|
$summary .= $self->encode( |
270
|
|
|
|
|
|
|
substr( $x, $last_send, $sstart - $last_send ) ) |
271
|
|
|
|
|
|
|
unless !$last_send && !$sstart; |
272
|
0
|
|
|
|
|
|
$summary .= $self->highlight( |
273
|
|
|
|
|
|
|
$self->encode( |
274
|
|
|
|
|
|
|
substr( $x, $sstart, $send - $sstart ) |
275
|
|
|
|
|
|
|
) |
276
|
|
|
|
|
|
|
); |
277
|
0
|
|
|
|
|
|
$last_send = $send; |
278
|
|
|
|
|
|
|
} |
279
|
0
|
0
|
|
|
|
|
$summary .= $self->encode( substr( $x, $last_send ) ) |
280
|
|
|
|
|
|
|
unless $last_send == length $x; |
281
|
|
|
|
|
|
|
} |
282
|
|
|
|
|
|
|
|
283
|
0
|
|
|
|
|
|
$prev_ellipsis = $end_with_ellipsis; |
284
|
|
|
|
|
|
|
|
285
|
|
|
|
|
|
|
} |
286
|
|
|
|
|
|
|
|
287
|
0
|
|
|
|
|
|
return $summary; |
288
|
|
|
|
|
|
|
} |
289
|
|
|
|
|
|
|
|
290
|
|
|
|
|
|
|
# This is not called as a method above, because it’s a private routine that |
291
|
|
|
|
|
|
|
# should not be overridden (it is not guaranteed to exist in future ver- |
292
|
|
|
|
|
|
|
# sions), and it’s faster to call it as a function. |
293
|
|
|
|
|
|
|
sub _encode_with_pb { # w/page breaks |
294
|
0
|
|
|
0
|
|
|
my ($self, $text, $page_h, $page_no_ref, $hitdoc) = @_; |
295
|
0
|
|
|
|
|
|
my @to_encode = split /\014/, $text, -1; # -1 to allow trailing |
296
|
0
|
|
|
|
|
|
my $ret = ''; # null fields |
297
|
0
|
0
|
|
|
|
|
$ret .= $self->encode(shift @to_encode) if length $to_encode[0]; |
298
|
0
|
|
|
|
|
|
for(@to_encode) { |
299
|
0
|
|
|
|
|
|
$ret .= &$page_h($hitdoc, ++$$page_no_ref); |
300
|
0
|
0
|
|
|
|
|
$ret .= $self->encode($_) if length; |
301
|
|
|
|
|
|
|
} |
302
|
0
|
|
|
|
|
|
$ret; |
303
|
|
|
|
|
|
|
} |
304
|
|
|
|
|
|
|
|
305
|
|
|
|
|
|
|
sub encode { |
306
|
0
|
|
|
0
|
1
|
|
my @__ = @_; # workaround for perl5.8.8 bug |
307
|
|
|
|
|
|
|
&{ |
308
|
0
|
0
|
|
|
|
|
$encoder{$__[0]} or return shift(@__)->SUPER::encode(@__) |
|
0
|
|
|
|
|
|
|
309
|
|
|
|
|
|
|
}($__[1]) |
310
|
|
|
|
|
|
|
} |
311
|
|
|
|
|
|
|
|
312
|
|
|
|
|
|
|
1; |
313
|
|
|
|
|
|
|
|
314
|
|
|
|
|
|
|
__END__ |