line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package String::Normal::Config::BusinessStop; |
2
|
9
|
|
|
9
|
|
30
|
use strict; |
|
9
|
|
|
|
|
12
|
|
|
9
|
|
|
|
|
185
|
|
3
|
9
|
|
|
9
|
|
23
|
use warnings; |
|
9
|
|
|
|
|
10
|
|
|
9
|
|
|
|
|
174
|
|
4
|
|
|
|
|
|
|
|
5
|
9
|
|
|
9
|
|
35
|
use String::Normal::Config; |
|
9
|
|
|
|
|
10
|
|
|
9
|
|
|
|
|
1719
|
|
6
|
|
|
|
|
|
|
|
7
|
|
|
|
|
|
|
sub _data { |
8
|
3
|
|
|
3
|
|
8
|
my %params = @_; |
9
|
|
|
|
|
|
|
|
10
|
3
|
|
|
|
|
3
|
my $fh; |
11
|
3
|
100
|
|
|
|
13
|
if ($params{business_stop}) { |
12
|
1
|
50
|
|
|
|
24
|
open $fh, $params{business_stop} or die "Can't read '$params{business_stop}' $!\n"; |
13
|
|
|
|
|
|
|
} else { |
14
|
2
|
|
|
|
|
7
|
$fh = *DATA; |
15
|
|
|
|
|
|
|
} |
16
|
|
|
|
|
|
|
|
17
|
3
|
|
|
|
|
5
|
my %stop; |
18
|
3
|
|
|
|
|
11
|
for (String::Normal::Config::_slurp( $fh )) { |
19
|
92
|
|
|
|
|
124
|
my ($word,$count) = split ',', $_; |
20
|
92
|
|
100
|
|
|
183
|
$count ||= 1; |
21
|
|
|
|
|
|
|
|
22
|
92
|
100
|
|
|
|
162
|
if (substr( $word, 0, 1 ) eq '^') { |
|
|
100
|
|
|
|
|
|
23
|
4
|
|
|
|
|
3
|
substr( $word, 0, 1 ) = ''; |
24
|
4
|
|
|
|
|
10
|
$stop{first}->{$word} = $count; |
25
|
|
|
|
|
|
|
} elsif (substr( $word, -1, 1 ) eq '$') { |
26
|
22
|
|
|
|
|
17
|
substr( $word, -1, 1 ) = ''; |
27
|
22
|
|
|
|
|
40
|
$stop{last}->{$word} = $count; |
28
|
|
|
|
|
|
|
} else { |
29
|
66
|
|
|
|
|
98
|
$stop{middle}->{$word} = $count; |
30
|
|
|
|
|
|
|
} |
31
|
|
|
|
|
|
|
} |
32
|
|
|
|
|
|
|
|
33
|
3
|
|
|
|
|
17
|
return \%stop; |
34
|
|
|
|
|
|
|
} |
35
|
|
|
|
|
|
|
|
36
|
|
|
|
|
|
|
1; |
37
|
|
|
|
|
|
|
|
38
|
|
|
|
|
|
|
=head1 NAME |
39
|
|
|
|
|
|
|
|
40
|
|
|
|
|
|
|
String::Normal::Config::BusinessStop; |
41
|
|
|
|
|
|
|
|
42
|
|
|
|
|
|
|
=head1 DESCRIPTION |
43
|
|
|
|
|
|
|
|
44
|
|
|
|
|
|
|
This package defines removals to be performed on the business types. |
45
|
|
|
|
|
|
|
|
46
|
|
|
|
|
|
|
=head1 STRUCTURE |
47
|
|
|
|
|
|
|
|
48
|
|
|
|
|
|
|
One entry per line. Each value will be removed. Special characters |
49
|
|
|
|
|
|
|
C<^> and C<$> are used to anchor the first and last tokens respectively: |
50
|
|
|
|
|
|
|
|
51
|
|
|
|
|
|
|
^mr |
52
|
|
|
|
|
|
|
lane$ |
53
|
|
|
|
|
|
|
|
54
|
|
|
|
|
|
|
Only those occurances of C at the beginning of the token list will be |
55
|
|
|
|
|
|
|
removed, and only those occurances of at the end of the token list |
56
|
|
|
|
|
|
|
will be removed. |
57
|
|
|
|
|
|
|
|
58
|
|
|
|
|
|
|
You can prepend C<#> to lines that you wish to skip. For example: |
59
|
|
|
|
|
|
|
|
60
|
|
|
|
|
|
|
the |
61
|
|
|
|
|
|
|
#and |
62
|
|
|
|
|
|
|
|
63
|
|
|
|
|
|
|
Would remove all occurances of C but not C. Special character |
64
|
|
|
|
|
|
|
C<,> is used to override removal if the final number of tokens if less |
65
|
|
|
|
|
|
|
then the digit specified. For example: |
66
|
|
|
|
|
|
|
|
67
|
|
|
|
|
|
|
the,2 |
68
|
|
|
|
|
|
|
|
69
|
|
|
|
|
|
|
Would remove all occurances of C in "The Generic Title" but would |
70
|
|
|
|
|
|
|
not remove any occurances in "The The". |
71
|
|
|
|
|
|
|
|
72
|
|
|
|
|
|
|
You can provide your own data by creating a text file containing your |
73
|
|
|
|
|
|
|
values and provide the path to that file via the constructor: |
74
|
|
|
|
|
|
|
|
75
|
|
|
|
|
|
|
my $normalizer = String::Normal->new( business_stop => '/path/to/values.txt' ); |
76
|
|
|
|
|
|
|
|
77
|
|
|
|
|
|
|
=head1 AUTHOR |
78
|
|
|
|
|
|
|
|
79
|
|
|
|
|
|
|
Jeff Anderson, C<< >> |
80
|
|
|
|
|
|
|
|
81
|
|
|
|
|
|
|
=head1 LICENSE AND COPYRIGHT |
82
|
|
|
|
|
|
|
|
83
|
|
|
|
|
|
|
Copyright 2016 Jeff Anderson. |
84
|
|
|
|
|
|
|
|
85
|
|
|
|
|
|
|
This program is free software; you can redistribute it and/or modify it |
86
|
|
|
|
|
|
|
under the terms of the the Artistic License (2.0). You may obtain a |
87
|
|
|
|
|
|
|
copy of the full license at: |
88
|
|
|
|
|
|
|
|
89
|
|
|
|
|
|
|
L |
90
|
|
|
|
|
|
|
|
91
|
|
|
|
|
|
|
Any use, modification, and distribution of the Standard or Modified |
92
|
|
|
|
|
|
|
Versions is governed by this Artistic License. By using, modifying or |
93
|
|
|
|
|
|
|
distributing the Package, you accept this license. Do not use, modify, |
94
|
|
|
|
|
|
|
or distribute the Package, if you do not accept this license. |
95
|
|
|
|
|
|
|
|
96
|
|
|
|
|
|
|
If your Modified Version has been derived from a Modified Version made |
97
|
|
|
|
|
|
|
by someone other than you, you are nevertheless required to ensure that |
98
|
|
|
|
|
|
|
your Modified Version complies with the requirements of this license. |
99
|
|
|
|
|
|
|
|
100
|
|
|
|
|
|
|
This license does not grant you the right to use any trademark, service |
101
|
|
|
|
|
|
|
mark, tradename, or logo of the Copyright Holder. |
102
|
|
|
|
|
|
|
|
103
|
|
|
|
|
|
|
This license includes the non-exclusive, worldwide, free-of-charge |
104
|
|
|
|
|
|
|
patent license to make, have made, use, offer to sell, sell, import and |
105
|
|
|
|
|
|
|
otherwise transfer the Package with respect to any patent claims |
106
|
|
|
|
|
|
|
licensable by the Copyright Holder that are necessarily infringed by the |
107
|
|
|
|
|
|
|
Package. If you institute patent litigation (including a cross-claim or |
108
|
|
|
|
|
|
|
counterclaim) against any party alleging that the Package constitutes |
109
|
|
|
|
|
|
|
direct or contributory patent infringement, then this Artistic License |
110
|
|
|
|
|
|
|
to you shall terminate on the date that such litigation is filed. |
111
|
|
|
|
|
|
|
|
112
|
|
|
|
|
|
|
Disclaimer of Warranty: THE PACKAGE IS PROVIDED BY THE COPYRIGHT HOLDER |
113
|
|
|
|
|
|
|
AND CONTRIBUTORS "AS IS' AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES. |
114
|
|
|
|
|
|
|
THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR |
115
|
|
|
|
|
|
|
PURPOSE, OR NON-INFRINGEMENT ARE DISCLAIMED TO THE EXTENT PERMITTED BY |
116
|
|
|
|
|
|
|
YOUR LOCAL LAW. UNLESS REQUIRED BY LAW, NO COPYRIGHT HOLDER OR |
117
|
|
|
|
|
|
|
CONTRIBUTOR WILL BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, OR |
118
|
|
|
|
|
|
|
CONSEQUENTIAL DAMAGES ARISING IN ANY WAY OUT OF THE USE OF THE PACKAGE, |
119
|
|
|
|
|
|
|
EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
120
|
|
|
|
|
|
|
|
121
|
|
|
|
|
|
|
=cut |
122
|
|
|
|
|
|
|
|
123
|
|
|
|
|
|
|
__DATA__ |