line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package Search::Tokenizer; |
2
|
1
|
|
|
1
|
|
70076
|
use warnings; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
33
|
|
3
|
1
|
|
|
1
|
|
6
|
use strict; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
22
|
|
4
|
1
|
|
|
1
|
|
5
|
use Carp qw(croak); |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
45
|
|
5
|
1
|
|
|
1
|
|
549
|
use Unicode::CaseFold qw(fc); # because CORE::fc only came with Perl 5.16 |
|
1
|
|
|
|
|
1001
|
|
|
1
|
|
|
|
|
116
|
|
6
|
|
|
|
|
|
|
|
7
|
|
|
|
|
|
|
our $VERSION = '1.03'; |
8
|
|
|
|
|
|
|
|
9
|
|
|
|
|
|
|
#====================================================================== |
10
|
|
|
|
|
|
|
# MAIN FUNCTIONALITY |
11
|
|
|
|
|
|
|
#====================================================================== |
12
|
|
|
|
|
|
|
|
13
|
|
|
|
|
|
|
sub new { |
14
|
2
|
|
|
2
|
0
|
641
|
my $class = shift; |
15
|
|
|
|
|
|
|
|
16
|
|
|
|
|
|
|
# defaults |
17
|
1
|
|
|
1
|
|
669
|
my $regex = qr/\p{Word}+/; |
|
1
|
|
|
|
|
15
|
|
|
1
|
|
|
|
|
15
|
|
|
2
|
|
|
|
|
9
|
|
18
|
2
|
|
|
|
|
4
|
my $lower = 1; |
19
|
2
|
|
|
|
|
6
|
my $filter = undef; |
20
|
2
|
|
|
|
|
4
|
my $filter_in_place = undef; |
21
|
2
|
|
|
|
|
3
|
my $stopwords = undef; |
22
|
|
|
|
|
|
|
|
23
|
|
|
|
|
|
|
# parse arguments |
24
|
2
|
50
|
|
|
|
11
|
unshift @_, "regex" if @_ == 1; # positional API |
25
|
2
|
|
|
|
|
13
|
while (my ($arg, $val) = splice(@_, 0, 2)) { |
26
|
7
|
|
100
|
|
|
26
|
$arg .= "=>" . (ref($val) || "SCALAR"); |
27
|
|
|
|
|
|
|
CHECK: |
28
|
7
|
|
|
|
|
13
|
for ($arg) { |
29
|
7
|
100
|
|
|
|
26
|
/^regex=>Regexp$/ and do { $regex = $val; last CHECK}; |
|
2
|
|
|
|
|
7
|
|
|
2
|
|
|
|
|
11
|
|
30
|
5
|
100
|
|
|
|
13
|
/^lower=>SCALAR$/ and do { $lower = !!$val; last CHECK}; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
4
|
|
31
|
4
|
50
|
|
|
|
11
|
/^filter=>CODE$/ and do { $filter = $val; last CHECK}; |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
32
|
4
|
100
|
|
|
|
12
|
/^filter_in_place=>CODE$/ and do { $filter_in_place = $val; last CHECK}; |
|
2
|
|
|
|
|
5
|
|
|
2
|
|
|
|
|
7
|
|
33
|
2
|
50
|
|
|
|
10
|
/^stopwords=>HASH$/ and do { $stopwords = $val; last CHECK}; |
|
2
|
|
|
|
|
4
|
|
|
2
|
|
|
|
|
7
|
|
34
|
0
|
|
|
|
|
0
|
croak "Invalid option or invalid operand: $arg"; |
35
|
|
|
|
|
|
|
} |
36
|
|
|
|
|
|
|
} |
37
|
|
|
|
|
|
|
|
38
|
|
|
|
|
|
|
# check that regex doest not match the empty string |
39
|
2
|
50
|
|
|
|
14
|
not "" =~ $regex |
40
|
|
|
|
|
|
|
or croak "regex $regex matches the empty string: cannot tokenize"; |
41
|
|
|
|
|
|
|
|
42
|
|
|
|
|
|
|
# return tokenizer factory: closure |
43
|
|
|
|
|
|
|
return sub { |
44
|
3
|
|
|
3
|
|
16
|
my ($string, @other_args) = @_; |
45
|
|
|
|
|
|
|
not @other_args |
46
|
3
|
50
|
|
|
|
21
|
or croak "too many args -- just a single string is expected"; |
47
|
|
|
|
|
|
|
|
48
|
3
|
|
|
|
|
5
|
my $term_index = -1; |
49
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
|
# return tokenizer : additional closure on $string and $term_index |
51
|
|
|
|
|
|
|
return sub { |
52
|
|
|
|
|
|
|
|
53
|
|
|
|
|
|
|
# get next occurrence of $regex in $string (thanks to the /g flag) |
54
|
10
|
|
|
|
|
51
|
while ($string =~ /$regex/g) { |
55
|
|
|
|
|
|
|
|
56
|
|
|
|
|
|
|
# index of this term within the input string |
57
|
12
|
|
|
|
|
19
|
$term_index += 1; |
58
|
|
|
|
|
|
|
|
59
|
|
|
|
|
|
|
# boundaries for the match |
60
|
12
|
|
|
|
|
18
|
my $end = pos($string); |
61
|
12
|
|
|
|
|
27
|
my $term = $&; # used to be slow in older perls, but now OK |
62
|
12
|
|
|
|
|
16
|
my $start = $end - length($term); |
63
|
|
|
|
|
|
|
|
64
|
|
|
|
|
|
|
# the old way used to be as follows, but it is ridiculously slow on utf8 strings |
65
|
|
|
|
|
|
|
# .. see https://github.com/Perl/perl5/issues/18786 |
66
|
|
|
|
|
|
|
# |
67
|
|
|
|
|
|
|
# my ($start, $end) = ($-[0], $+[0]); |
68
|
|
|
|
|
|
|
# my $term = substr($string, $start, $end-$start); |
69
|
|
|
|
|
|
|
|
70
|
|
|
|
|
|
|
# apply filtering and stopwords, if any |
71
|
12
|
100
|
|
|
|
38
|
$term = Unicode::CaseFold::fc($term) if $lower; |
72
|
12
|
50
|
|
|
|
72
|
$term = $filter->($term) if $filter; |
73
|
12
|
50
|
|
|
|
169
|
$filter_in_place->($term) if $filter_in_place; |
74
|
12
|
100
|
66
|
|
|
182
|
undef $term if $stopwords and $stopwords->{$term}; |
75
|
|
|
|
|
|
|
|
76
|
|
|
|
|
|
|
# if $term was not cancelled by filters above, return it |
77
|
12
|
100
|
|
|
|
42
|
if ($term) { |
78
|
7
|
50
|
|
|
|
41
|
return wantarray ? ($term, length($term), $start, $end, $term_index) |
79
|
|
|
|
|
|
|
: $term; |
80
|
|
|
|
|
|
|
} |
81
|
|
|
|
|
|
|
} # otherwise, loop again to extract next term |
82
|
|
|
|
|
|
|
|
83
|
|
|
|
|
|
|
# otherwise, that's the end of the input string, return undef or empty list |
84
|
3
|
|
|
|
|
7
|
return; |
85
|
3
|
|
|
|
|
40
|
}; |
86
|
2
|
|
|
|
|
19
|
}; |
87
|
|
|
|
|
|
|
} |
88
|
|
|
|
|
|
|
|
89
|
|
|
|
|
|
|
#====================================================================== |
90
|
|
|
|
|
|
|
# BUILTIN TOKENIZERS |
91
|
|
|
|
|
|
|
#====================================================================== |
92
|
|
|
|
|
|
|
|
93
|
|
|
|
|
|
|
|
94
|
|
|
|
|
|
|
sub word { |
95
|
0
|
|
|
0
|
1
|
0
|
__PACKAGE__->new(regex => qr/\w+/, @_); |
96
|
|
|
|
|
|
|
} |
97
|
|
|
|
|
|
|
|
98
|
|
|
|
|
|
|
sub word_locale { |
99
|
1
|
|
|
1
|
|
23237
|
use locale; |
|
1
|
|
|
|
|
525
|
|
|
1
|
|
|
|
|
6
|
|
100
|
0
|
|
|
0
|
1
|
0
|
__PACKAGE__->new(regex => qr/\w+/, @_); |
101
|
|
|
|
|
|
|
} |
102
|
|
|
|
|
|
|
|
103
|
|
|
|
|
|
|
sub word_unicode { |
104
|
0
|
|
|
0
|
1
|
0
|
__PACKAGE__->new(regex => qr/\p{Word}+/, @_); |
105
|
|
|
|
|
|
|
} |
106
|
|
|
|
|
|
|
|
107
|
|
|
|
|
|
|
sub unaccent { |
108
|
1
|
|
|
1
|
1
|
565
|
require Text::Transliterator::Unaccent; |
109
|
1
|
|
|
|
|
39908
|
my %args = @_; |
110
|
1
|
|
33
|
|
|
7
|
my $want_lower = !exists $args{lower} || $args{lower}; |
111
|
1
|
50
|
|
|
|
5
|
my %unaccenter_args = $want_lower ? () : (upper => 0); |
112
|
1
|
|
|
|
|
9
|
my $unaccenter = Text::Transliterator::Unaccent->new(%unaccenter_args); |
113
|
1
|
|
|
|
|
36028
|
__PACKAGE__->new(regex => qr/\p{Word}+/, |
114
|
|
|
|
|
|
|
filter_in_place => $unaccenter, |
115
|
|
|
|
|
|
|
%args); |
116
|
|
|
|
|
|
|
} |
117
|
|
|
|
|
|
|
|
118
|
|
|
|
|
|
|
|
119
|
|
|
|
|
|
|
#====================================================================== |
120
|
|
|
|
|
|
|
# UTILITY FUNCTION |
121
|
|
|
|
|
|
|
#====================================================================== |
122
|
|
|
|
|
|
|
|
123
|
|
|
|
|
|
|
sub unroll { |
124
|
3
|
|
|
3
|
1
|
15
|
my $iterator = shift; |
125
|
3
|
|
|
|
|
6
|
my $no_details = shift; |
126
|
3
|
|
|
|
|
4
|
my @results; |
127
|
|
|
|
|
|
|
|
128
|
3
|
|
|
|
|
7
|
while (my @r = $iterator->() ) { |
129
|
7
|
100
|
|
|
|
25
|
push @results, $no_details ? $r[0] : \@r; |
130
|
|
|
|
|
|
|
} |
131
|
3
|
|
|
|
|
27
|
return @results; |
132
|
|
|
|
|
|
|
} |
133
|
|
|
|
|
|
|
|
134
|
|
|
|
|
|
|
|
135
|
|
|
|
|
|
|
1; # End of Search::Tokenizer |
136
|
|
|
|
|
|
|
|
137
|
|
|
|
|
|
|
|
138
|
|
|
|
|
|
|
__END__ |