t/db_dependent/Sitemapper.t

   1 #!/usr/bin/perl
   2
   3 # Copyright 2015 Tamil s.a.r.l.
   4 #
   5 # This file is part of Koha.
   6 #
   7 # Koha is free software; you can redistribute it and/or modify it
   8 # under the terms of the GNU General Public License as published by
   9 # the Free Software Foundation; either version 3 of the License, or
  10 # (at your option) any later version.
  11 #
  12 # Koha is distributed in the hope that it will be useful, but
  13 # WITHOUT ANY WARRANTY; without even the implied warranty of
  14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  15 # GNU General Public License for more details.
  16 #
  17 # You should have received a copy of the GNU General Public License
  18 # along with Koha; if not, see <http://www.gnu.org/licenses>.
  19
  20 use Modern::Perl;
  21 use File::Basename;
  22 use File::Path;
  23 use DateTime;
  24 use Koha::DateUtils;
  25 use Test::MockModule;
  26 use Test::More tests => 16;
  27 use Carp qw/croak carp/;
  28
  29
  30 BEGIN {
  31     use_ok('Koha::Sitemapper');
  32     use_ok('Koha::Sitemapper::Writer');
  33 }
  34
  35 my $now_value       = dt_from_string();
  36 my $mocked_datetime = Test::MockModule->new('DateTime');
  37 $mocked_datetime->mock( 'now', sub { return $now_value->clone; } );
  38
  39 sub slurp {
  40     my $file = shift;
  41     open my $fh, '<', $file or croak;
  42     local $/ = undef;
  43     my $cont = <$fh>;
  44     close $fh;
  45     return $cont;
  46 }
  47
  48 use Test::DBIx::Class;
  49
  50 sub fixtures {
  51     my ($data) = @_;
  52     fixtures_ok [
  53         Biblio => [ [qw/ biblionumber datecreated timestamp  /], @{$data}, ],
  54     ], 'add fixtures';
  55     return;
  56 }
  57
  58 # Make the code in the module use our mocked Koha::Schema/Koha::Database
  59 my $db = Test::MockModule->new('Koha::Database');
  60 $db->mock(
  61
  62     # Schema() gives us the DB connection set up by Test::DBIx::Class
  63     _new_schema => sub { return Schema(); }
  64 );
  65
  66 my $dir = C4::Context::temporary_directory;
  67
  68 my $data = [
  69     [qw/ 1         2013-11-15 2013-11-15/],
  70     [qw/ 2         2015-08-31 2015-08-31/],
  71 ];
  72 fixtures($data);
  73
  74 # Create a sitemap for a catalog containg 2 biblios, with option 'long url'
  75 my $sitemapper = Koha::Sitemapper->new(
  76     verbose => 0,
  77     url     => 'http://www.mylibrary.org',
  78     dir     => $dir,
  79     short   => 0,
  80 );
  81 $sitemapper->run();
  82
  83 my $file = "$dir/sitemapindex.xml";
  84 ok( -e "$dir/sitemapindex.xml", 'File sitemapindex.xml created' );
  85 my $file_content     = slurp($file);
  86 my $now              = DateTime->now->ymd;
  87 my $expected_content = <<"EOS";
  88 <?xml version="1.0" encoding="UTF-8"?>
  89
  90 <sitemapindex xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
  91   <sitemap>
  92     <loc>http://www.mylibrary.org/sitemap0001.xml</loc>
  93     <lastmod>$now</lastmod>
  94   </sitemap>
  95 </sitemapindex>
  96 EOS
  97 chop $expected_content;
  98 is( $file_content, $expected_content, 'Its content is valid' );
  99
 100 $file = "$dir/sitemap0001.xml";
 101 ok( -e $file, 'File sitemap0001.xml created' );
 102 $file_content     = slurp($file);
 103 $expected_content = <<"EOS";
 104 <?xml version="1.0" encoding="UTF-8"?>
 105
 106 <urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.sitemaps.org/schemas/sitemap/0.9 http://www.sitemaps.org/schemas/sitemap/0.9/sitemap.xsd">
 107   <url>
 108     <loc>http://www.mylibrary.org/cgi-bin/koha/opac-detail.pl?biblionumber=1</loc>
 109     <lastmod>2013-11-15</lastmod>
 110   </url>
 111   <url>
 112     <loc>http://www.mylibrary.org/cgi-bin/koha/opac-detail.pl?biblionumber=2</loc>
 113     <lastmod>2015-08-31</lastmod>
 114   </url>
 115 </urlset>
 116 EOS
 117 is( $file_content, $expected_content, 'Its content is valid' );
 118
 119 # Create a sitemap for a catalog containg 2 biblios, with option 'short url'.
 120 # Test that 2 files are created.
 121 $sitemapper = Koha::Sitemapper->new(
 122     verbose => 0,
 123     url     => 'http://www.mylibrary.org',
 124     dir     => $dir,
 125     short   => 1,
 126 );
 127 $sitemapper->run();
 128
 129 $file = "$dir/sitemap0001.xml";
 130 ok( -e $file, 'File sitemap0001.xml with short URLs created' );
 131 $file_content     = slurp($file);
 132 $expected_content = <<"EOS";
 133 <?xml version="1.0" encoding="UTF-8"?>
 134
 135 <urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.sitemaps.org/schemas/sitemap/0.9 http://www.sitemaps.org/schemas/sitemap/0.9/sitemap.xsd">
 136   <url>
 137     <loc>http://www.mylibrary.org/bib/1</loc>
 138     <lastmod>2013-11-15</lastmod>
 139   </url>
 140   <url>
 141     <loc>http://www.mylibrary.org/bib/2</loc>
 142     <lastmod>2015-08-31</lastmod>
 143   </url>
 144 </urlset>
 145 EOS
 146 is( $file_content, $expected_content, 'Its content is valid' );
 147
 148 # Create a sitemap for a catalog containing 75000 biblios, with option 'short
 149 # url'. Test that 3 files are created: index file + 2 urls file with
 150 # respectively 50000 et 25000 urls.
 151 $data = [];
 152 for my $count ( 3 .. 75_000 ) {
 153     push @{$data}, [ $count, '2015-08-31', '2015-08-31' ];
 154 }
 155 fixtures($data);
 156 $sitemapper = Koha::Sitemapper->new(
 157     verbose => 0,
 158     url     => 'http://www.mylibrary.org',
 159     dir     => $dir,
 160     short   => 1,
 161 );
 162 $sitemapper->run();
 163
 164 $file = "$dir/sitemapindex.xml";
 165 ok( -e "$dir/sitemapindex.xml",
 166     'File sitemapindex.xml for 75000 bibs created' );
 167 $file_content     = slurp($file);
 168 $expected_content = <<"EOS";
 169 <?xml version="1.0" encoding="UTF-8"?>
 170
 171 <sitemapindex xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
 172   <sitemap>
 173     <loc>http://www.mylibrary.org/sitemap0001.xml</loc>
 174     <lastmod>$now</lastmod>
 175   </sitemap>
 176   <sitemap>
 177     <loc>http://www.mylibrary.org/sitemap0002.xml</loc>
 178     <lastmod>$now</lastmod>
 179   </sitemap>
 180 </sitemapindex>
 181 EOS
 182 chop $expected_content;
 183 is( $file_content, $expected_content, 'Its content is valid' );
 184
 185 $file = "$dir/sitemap0001.xml";
 186 ok( -e $file, 'File sitemap0001.xml created' );
 187
 188 open my $fh, '<', $file or croak;
 189 my $count = 0;
 190 while (<$fh>) {
 191     if ( $_ =~ /<loc>/xsm ) { $count++; }
 192 }
 193 close $fh;
 194 is( $count, 50_000, 'It contains 50000 URLs' );
 195
 196 $file = "$dir/sitemap0002.xml";
 197 ok( -e $file, 'File sitemap0002.xml created' );
 198
 199 open $fh, '<', $file or croak;
 200 $count = 0;
 201 while (<$fh>) {
 202     if ( $_ =~ /<loc>/xsm ) { $count++; }
 203 }
 204 close $fh;
 205 is( $count, 25_000, 'It contains 25000 URLs' );
 206
 207 # Cleanup
 208 for my $file (qw/sitemapindex.xml sitemap0001.xml sitemap0002.xml/) {
 209     unlink "$dir/$file";
 210 }