line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package Search::Tokenizer; |
2
|
1
|
|
|
1
|
|
78408
|
use warnings; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
39
|
|
3
|
1
|
|
|
1
|
|
6
|
use strict; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
19
|
|
4
|
1
|
|
|
1
|
|
520
|
use Unicode::CaseFold (); |
|
1
|
|
|
|
|
1080
|
|
|
1
|
|
|
|
|
477
|
|
5
|
|
|
|
|
|
|
|
6
|
|
|
|
|
|
|
our $VERSION = '1.02'; |
7
|
|
|
|
|
|
|
|
8
|
|
|
|
|
|
|
sub new { |
9
|
2
|
|
|
2
|
0
|
682
|
my $class = shift; |
10
|
|
|
|
|
|
|
|
11
|
|
|
|
|
|
|
# defaults |
12
|
2
|
|
|
|
|
10
|
my $regex = qr/\w+/; |
13
|
2
|
|
|
|
|
6
|
my $lower = 1; |
14
|
2
|
|
|
|
|
6
|
my $filter = undef; |
15
|
2
|
|
|
|
|
4
|
my $filter_in_place = undef; |
16
|
2
|
|
|
|
|
5
|
my $stopwords = undef; |
17
|
|
|
|
|
|
|
|
18
|
|
|
|
|
|
|
# parse arguments |
19
|
2
|
50
|
|
|
|
13
|
unshift @_, "regex" if @_ == 1; # positional API |
20
|
2
|
|
|
|
|
10
|
while (my $arg = shift) { |
21
|
7
|
|
|
|
|
11
|
my $val = shift; |
22
|
7
|
|
100
|
|
|
30
|
$arg .= "=>" . (ref($val) || "NOREF"); |
23
|
7
|
|
|
|
|
15
|
for ($arg) { |
24
|
7
|
100
|
|
|
|
23
|
/^regex=>Regexp$/ and do { $regex = $val; last}; |
|
2
|
|
|
|
|
10
|
|
|
2
|
|
|
|
|
10
|
|
25
|
5
|
100
|
|
|
|
14
|
/^lower=>NOREF$/ and do { $lower = !!$val; last}; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
3
|
|
26
|
4
|
50
|
|
|
|
13
|
/^filter=>CODE$/ and do { $filter = $val; last}; |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
27
|
4
|
100
|
|
|
|
14
|
/^filter_in_place=>CODE$/ and do { $filter_in_place = $val; last}; |
|
2
|
|
|
|
|
6
|
|
|
2
|
|
|
|
|
6
|
|
28
|
2
|
50
|
|
|
|
10
|
/^stopwords=>HASH$/ and do { $stopwords = $val; last}; |
|
2
|
|
|
|
|
3
|
|
|
2
|
|
|
|
|
6
|
|
29
|
0
|
|
|
|
|
0
|
die "Invalid option or invalid operand: $arg"; |
30
|
|
|
|
|
|
|
} |
31
|
|
|
|
|
|
|
} |
32
|
|
|
|
|
|
|
|
33
|
|
|
|
|
|
|
# check that regex doest not match the empty string |
34
|
2
|
50
|
|
|
|
14
|
not "" =~ $regex |
35
|
|
|
|
|
|
|
or die "regex $regex matches the empty string: cannot tokenize"; |
36
|
|
|
|
|
|
|
|
37
|
|
|
|
|
|
|
# return tokenizer factory: closure |
38
|
|
|
|
|
|
|
return sub { |
39
|
5
|
|
|
5
|
|
2383
|
my $string = shift; |
40
|
5
|
|
|
|
|
10
|
my $term_index = -1; |
41
|
|
|
|
|
|
|
|
42
|
|
|
|
|
|
|
# return tokenizer : additional closure on $string and $term_index |
43
|
|
|
|
|
|
|
return sub { |
44
|
|
|
|
|
|
|
|
45
|
|
|
|
|
|
|
# get next occurrence of $regex in $string (thanks to the /g flag) |
46
|
3026
|
|
|
|
|
21341
|
while ($string =~ /$regex/g) { |
47
|
|
|
|
|
|
|
|
48
|
|
|
|
|
|
|
# index of this term within the input string |
49
|
3040
|
|
|
|
|
4629
|
$term_index += 1; |
50
|
|
|
|
|
|
|
|
51
|
|
|
|
|
|
|
# boundaries for the match |
52
|
3040
|
|
|
|
|
4460
|
my $end = pos($string); |
53
|
3040
|
|
|
|
|
5081
|
my $term = $&; # used to be slow in older perls, but now OK |
54
|
3040
|
|
|
|
|
4680
|
my $start = $end - length($term); |
55
|
|
|
|
|
|
|
|
56
|
|
|
|
|
|
|
# the old way used to be as follows, but it is ridiculously slow on utf8 strings |
57
|
|
|
|
|
|
|
# .. see https://github.com/Perl/perl5/issues/18786 |
58
|
|
|
|
|
|
|
# |
59
|
|
|
|
|
|
|
# my ($start, $end) = ($-[0], $+[0]); |
60
|
|
|
|
|
|
|
# my $term = substr($string, $start, $end-$start); |
61
|
|
|
|
|
|
|
|
62
|
|
|
|
|
|
|
# apply filtering and stopwords, if any |
63
|
3040
|
100
|
|
|
|
7293
|
$term = Unicode::CaseFold::fc($term) if $lower; |
64
|
3040
|
50
|
|
|
|
19762
|
$term = $filter->($term) if $filter; |
65
|
3040
|
50
|
|
|
|
54267
|
$filter_in_place->($term) if $filter_in_place; |
66
|
3040
|
100
|
66
|
|
|
20991
|
undef $term if $stopwords and $stopwords->{$term}; |
67
|
|
|
|
|
|
|
|
68
|
|
|
|
|
|
|
# if $term was not cancelled by filters above, return it |
69
|
3040
|
100
|
|
|
|
5191
|
if ($term) { |
70
|
3021
|
50
|
|
|
|
11756
|
return wantarray ? ($term, length($term), $start, $end, $term_index) |
71
|
|
|
|
|
|
|
: $term; |
72
|
|
|
|
|
|
|
} |
73
|
|
|
|
|
|
|
} # otherwise, loop again to extract next term |
74
|
|
|
|
|
|
|
|
75
|
|
|
|
|
|
|
# otherwise, no more term in input string, return undef or empty list |
76
|
5
|
|
|
|
|
12
|
return; |
77
|
5
|
|
|
|
|
52
|
}; |
78
|
2
|
|
|
|
|
20
|
}; |
79
|
|
|
|
|
|
|
} |
80
|
|
|
|
|
|
|
|
81
|
|
|
|
|
|
|
sub word { |
82
|
0
|
|
|
0
|
1
|
0
|
__PACKAGE__->new(regex => qr/\w+/, @_); |
83
|
|
|
|
|
|
|
} |
84
|
|
|
|
|
|
|
|
85
|
|
|
|
|
|
|
sub word_locale { |
86
|
1
|
|
|
1
|
|
562
|
use locale; |
|
1
|
|
|
|
|
622
|
|
|
1
|
|
|
|
|
5
|
|
87
|
0
|
|
|
0
|
1
|
0
|
__PACKAGE__->new(regex => qr/\w+/, @_); |
88
|
|
|
|
|
|
|
} |
89
|
|
|
|
|
|
|
|
90
|
|
|
|
|
|
|
sub word_unicode { |
91
|
1
|
|
|
1
|
1
|
783
|
__PACKAGE__->new(regex => qr/\p{Word}+/, @_); |
|
1
|
|
|
0
|
|
15
|
|
|
1
|
|
|
|
|
16
|
|
|
0
|
|
|
|
|
0
|
|
92
|
|
|
|
|
|
|
} |
93
|
|
|
|
|
|
|
|
94
|
|
|
|
|
|
|
sub unaccent { |
95
|
1
|
|
|
1
|
1
|
1831
|
require Text::Transliterator::Unaccent; |
96
|
1
|
|
|
|
|
41170
|
my %args = @_; |
97
|
1
|
|
33
|
|
|
8
|
my $want_lower = !exists $args{lower} || $args{lower}; |
98
|
1
|
50
|
|
|
|
5
|
my %unaccenter_args = $want_lower ? () : (upper => 0); |
99
|
1
|
|
|
|
|
10
|
my $unaccenter = Text::Transliterator::Unaccent->new(%unaccenter_args); |
100
|
1
|
|
|
|
|
40022
|
__PACKAGE__->new(regex => qr/\p{Word}+/, |
101
|
|
|
|
|
|
|
filter_in_place => $unaccenter, |
102
|
|
|
|
|
|
|
%args); |
103
|
|
|
|
|
|
|
} |
104
|
|
|
|
|
|
|
|
105
|
|
|
|
|
|
|
|
106
|
|
|
|
|
|
|
1; # End of Search::Tokenizer |
107
|
|
|
|
|
|
|
|
108
|
|
|
|
|
|
|
|
109
|
|
|
|
|
|
|
__END__ |