1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
|
# This document contains text in Perl "POD" format.
# Use a POD viewer like perldoc or perlman to render it.
=encoding utf-8
=head1 NAME
Locale::Maketext::Cookbook - recipes for using Locale::Maketext
=head1 INTRODUCTION
This is a work in progress. Not much progress by now :-)
=head1 ONESIDED LEXICONS
I<Adapted from a suggestion by Dan Muey>
It may be common (for example at your main lexicon) that
the hash keys and values coincide. Like that
q{Hello, tell me your name}
=> q{Hello, tell me your name}
It would be nice to just write:
q{Hello, tell me your name} => ''
and have this magically inflated to the first form.
Among the advantages of such representation, that would
lead to
smaller files, less prone to mistyping or mispasting,
and handy to someone translating it which can simply
copy the main lexicon and enter the translation
instead of having to remove the value first.
That can be achieved by overriding C<init>
in your class and working on the main lexicon
with code like that:
package My::I18N;
...
sub init {
my $lh = shift; # a newborn handle
$lh->SUPER::init();
inflate_lexicon(\%My::I18N::en::Lexicon);
return;
}
sub inflate_lexicon {
my $lex = shift;
while (my ($k, $v) = each %$lex) {
$v = $k if !defined $v || $v eq '';
}
}
Here we are assuming C<My::I18N::en> to own the
main lexicon.
There are some downsides here: the size economy
will not stand at runtime after this C<init()>
runs. But it should not be that critical, since
if you don't have space for that, you won't have
space for any other language besides the main one
as well. You could do that too with ties,
expanding the value at lookup time which
should be more time expensive as an option.
=head1 DECIMAL PLACES IN NUMBER FORMATTING
I<After CPAN RT #36136 (https://rt.cpan.org/Ticket/Display.html?id=36136)>
The documentation of L<Locale::Maketext> advises that
the standard bracket method C<numf> is limited and that
you must override that for better results. It even
suggests the use of L<Number::Format>.
One such defect of standard C<numf> is to not be
able to use a certain decimal precision.
For example,
$lh->maketext('pi is [numf,_1]', 355/113);
outputs
pi is 3.14159292035398
Since pi ≈ 355/116 is only accurate
to 6 decimal places, you would want to say:
$lh->maketext('pi is [numf,_1,6]', 355/113);
and get "pi is 3.141592".
One solution for that could use C<Number::Format>
like that:
package Wuu;
use base qw(Locale::Maketext);
use Number::Format;
# can be overridden according to language conventions
sub _numf_params {
return (
-thousands_sep => '.',
-decimal_point => ',',
-decimal_digits => 2,
);
}
# builds a Number::Format
sub _numf_formatter {
my ($lh, $scale) = @_;
my @params = $lh->_numf_params;
if ($scale) { # use explicit scale rather than default
push @params, (-decimal_digits => $scale);
}
return Number::Format->new(@params);
}
sub numf {
my ($lh, $n, $scale) = @_;
# get the (cached) formatter
my $nf = $lh->{__nf}{$scale} ||= $lh->_numf_formatter($scale);
# format the number itself
return $nf->format_number($n);
}
package Wuu::pt;
use base qw(Wuu);
and then
my $lh = Wuu->get_handle('pt');
$lh->maketext('A [numf,_1,3] km de distância', 1550.2222);
would return "A 1.550,222 km de distância".
Notice that the standard utility methods of
C<Locale::Maketext> are irremediably limited
because they could not aim to do everything
that could be expected from them in different languages,
cultures and applications. So extending C<numf>,
C<quant>, and C<sprintf> is natural as soon
as your needs exceed what the standard ones do.
|
