t/Charset.t

   1 #!/usr/bin/perl
   2
   3 # This file is part of Koha.
   4 #
   5 # Koha is free software; you can redistribute it and/or modify it
   6 # under the terms of the GNU General Public License as published by
   7 # the Free Software Foundation; either version 3 of the License, or
   8 # (at your option) any later version.
   9 #
  10 # Koha is distributed in the hope that it will be useful, but
  11 # WITHOUT ANY WARRANTY; without even the implied warranty of
  12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13 # GNU General Public License for more details.
  14 #
  15 # You should have received a copy of the GNU General Public License
  16 # along with Koha; if not, see <http://www.gnu.org/licenses>.
  17
  18 use Modern::Perl;
  19
  20 use Test::More tests => 17;
  21 use Encode qw( is_utf8 );
  22
  23 use MARC::Record;
  24
  25 use t::lib::Mocks;
  26
  27 use utf8;
  28 use open ':std', ':encoding(utf8)';
  29
  30 BEGIN {
  31     use_ok('C4::Charset');
  32 }
  33
  34 my $string;
  35 ok(!defined(NormalizeString($string,undef,1)),'Uninitialized string case 1 normalizes to uninitialized string.');
  36
  37 $string = 'Sample';
  38 ok(defined(NormalizeString($string,undef,0)), 'Initialized string case 1 normalizes to some string.');
  39 ok(defined(NormalizeString($string,undef,1)), 'Initialized string case 2 normalizes to some string.');
  40 ok(defined(NormalizeString($string,1,0)),     'Initialized string case 3 normalizes to some string.');
  41 ok(defined(NormalizeString($string,1,1)),     'Initialized string case 4 normalizes to some string.');
  42
  43 my $octets = "abc";
  44 ok(IsStringUTF8ish($octets), "verify octets are valid UTF-8 (ASCII)");
  45
  46 $octets = "flamb\xc3\xa9";
  47 ok(!Encode::is_utf8($octets), "verify that string does not have Perl UTF-8 flag on");
  48 ok(IsStringUTF8ish($octets), "verify octets are valid UTF-8 (LATIN SMALL LETTER E WITH ACUTE)");
  49 ok(!Encode::is_utf8($octets), "verify that IsStringUTF8ish does not magically turn Perl UTF-8 flag on");
  50
  51 $octets = "a\xc2" . "c";
  52 ok(!IsStringUTF8ish($octets), "verify octets are not valid UTF-8");
  53
  54 ok( !SetUTF8Flag(), 'SetUTF8Flag returns undef if no record passed' );
  55
  56 my $record = MARC::Record->new();
  57 ok( !SetUTF8Flag($record), 'SetUTF8Flag returns undef if the record has no subfields' );
  58 # Add some fields/subfields
  59 $record->append_fields(
  60     MARC::Field->new('100', ' ', ' ', a => 'Julio Cortazar'),
  61     MARC::Field->new('245', ' ', ' ', a => 'Rayuela'),
  62 );
  63 # Verify our data serves its purpose
  64 ok( !Encode::is_utf8($record->subfield('100','a')) &&
  65     !Encode::is_utf8($record->subfield('245','a')),
  66     'Verify that the subfields are NOT set the UTF-8 flag yet' );
  67
  68 SetUTF8Flag($record);
  69
  70 ok( Encode::is_utf8($record->subfield('100','a')) &&
  71     Encode::is_utf8($record->subfield('245','a')),
  72     'SetUTF8Flag sets the UTF-8 flag to all subfields' );
  73
  74 is( nsb_clean("Le Moyen Âge"), "Le Moyen Âge", "nsb_clean removes  and " );
  75
  76 subtest 'SetMarcUnicodeFlag' => sub {
  77     plan tests => 2;
  78     # TODO This should be done in MARC::Record
  79     my $leader                  = '012345678X0             ';
  80     my $expected_marc21_leader  = '012345678a0             '; # position 9 of leader must be 'a'
  81     my $expected_unimarc_leader = '012345678X0             '; # position 9 of leader must not be changed
  82     # Note that position 9 of leader should be blank for UNIMARC, but as it is not related to encoding
  83     # we do not want to change it
  84
  85     t::lib::Mocks::mock_preference( 'marcflavour', 'MARC21' );
  86     my $marc21_record = MARC::Record->new;
  87     $marc21_record->leader($leader);
  88     SetMarcUnicodeFlag( $marc21_record, C4::Context->preference('marcflavour') );
  89     is( $marc21_record->leader, $expected_marc21_leader, 'Leader 9 for MARC21 mush be "a"' );
  90
  91     t::lib::Mocks::mock_preference( 'marcflavour',             'UNIMARC' );
  92     t::lib::Mocks::mock_preference( 'UNIMARCField100Language', 'fre' );
  93     my $unimarc_record = MARC::Record->new;
  94     $unimarc_record->leader($leader);
  95     SetMarcUnicodeFlag( $unimarc_record, C4::Context->preference('marcflavour') );
  96     is( $unimarc_record->leader, $expected_unimarc_leader, 'Leader 9 for UNIMARC must be blank' );
  97 };
  98
  99 1;