misc/search_tools/rebuild_elastic_search.pl

   1 #!/usr/bin/perl
   2
   3 # This inserts records from a Koha database into elastic search
   4
   5 # Copyright 2014 Catalyst IT
   6 #
   7 # This file is part of Koha.
   8 #
   9 # Koha is free software; you can redistribute it and/or modify it under the
  10 # terms of the GNU General Public License as published by the Free Software
  11 # Foundation; either version 3 of the License, or (at your option) any later
  12 # version.
  13 #
  14 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
  15 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
  16 # A PARTICULAR PURPOSE.  See the GNU General Public License for more details.
  17 #
  18 # You should have received a copy of the GNU General Public License along
  19 # with Koha; if not, write to the Free Software Foundation, Inc.,
  20 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  21
  22 =head1 NAME
  23
  24 rebuild_elastic_search.pl - inserts records from a Koha database into Elasticsearch
  25
  26 =head1 SYNOPSIS
  27
  28 B<rebuild_elastic_search.pl>
  29 [B<-c|--commit>=C<count>]
  30 [B<-v|--verbose>]
  31 [B<-h|--help>]
  32 [B<--man>]
  33
  34 =head1 DESCRIPTION
  35
  36 Inserts records from a Koha database into Elasticsearch.
  37
  38 =head1 OPTIONS
  39
  40 =over
  41
  42 =item B<-c|--commit>=C<count>
  43
  44 Specify how many records will be batched up before they're added to Elasticsearch.
  45 Higher should be faster, but will cause more RAM usage. Default is 5000.
  46
  47 =item B<-d|--delete>
  48
  49 Delete the index and recreate it before indexing.
  50
  51 =item B<-a|--authorities>
  52
  53 Index the authorities only. Combining this with B<-b> is the same as
  54 specifying neither and so both get indexed.
  55
  56 =item B<-b|--biblios>
  57
  58 Index the biblios only. Combining this with B<-a> is the same as
  59 specifying neither and so both get indexed.
  60
  61 =item B<-bn|--bnumber>
  62
  63 Only index the supplied biblionumber, mostly for testing purposes. May be
  64 repeated. This also applies to authorities via authid, so if you're using it,
  65 you probably only want to do one or the other at a time.
  66
  67 =item B<-v|--verbose>
  68
  69 By default, this program only emits warnings and errors. This makes it talk
  70 more. Add more to make it even more wordy, in particular when debugging.
  71
  72 =item B<-h|--help>
  73
  74 Help!
  75
  76 =item B<--man>
  77
  78 Full documentation.
  79
  80 =back
  81
  82 =cut
  83
  84 use autodie;
  85 use Getopt::Long;
  86 use C4::Context;
  87 use Koha::MetadataRecord::Authority;
  88 use Koha::BiblioUtils;
  89 use Koha::SearchEngine::Elasticsearch::Indexer;
  90 use MARC::Field;
  91 use MARC::Record;
  92 use Modern::Perl;
  93 use Pod::Usage;
  94
  95 use Data::Dumper; # TODO remove
  96
  97 my $verbose = 0;
  98 my $commit = 5000;
  99 my ($delete, $help, $man);
 100 my ($index_biblios, $index_authorities);
 101 my (@biblionumbers);
 102
 103 GetOptions(
 104     'c|commit=i'       => \$commit,
 105     'd|delete'         => \$delete,
 106     'a|authorities' => \$index_authorities,
 107     'b|biblios' => \$index_biblios,
 108     'bn|bnumber=i' => \@biblionumbers,
 109     'v|verbose+'       => \$verbose,
 110     'h|help'           => \$help,
 111     'man'              => \$man,
 112 );
 113
 114 # Default is to do both
 115 unless ($index_authorities || $index_biblios) {
 116     $index_authorities = $index_biblios = 1;
 117 }
 118
 119 pod2usage(1) if $help;
 120 pod2usage( -exitstatus => 0, -verbose => 2 ) if $man;
 121
 122 sanity_check();
 123
 124 my $next;
 125 if ($index_biblios) {
 126     _log(1, "Indexing biblios\n");
 127     if (@biblionumbers) {
 128         $next = sub {
 129             my $r = shift @biblionumbers;
 130             return () unless defined $r;
 131             return ($r, Koha::BiblioUtils->get_from_biblionumber($r, item_data => 1 ));
 132         };
 133     } else {
 134         my $records = Koha::BiblioUtils->get_all_biblios_iterator();
 135         $next = sub {
 136             $records->next();
 137         }
 138     }
 139     do_reindex($next, $Koha::SearchEngine::Elasticsearch::BIBLIOS_INDEX);
 140 }
 141 if ($index_authorities) {
 142     _log(1, "Indexing authorities\n");
 143     if (@biblionumbers) {
 144         $next = sub {
 145             my $r = shift @biblionumbers;
 146             return () unless defined $r;
 147             my $a = Koha::MetadataRecord::Authority->get_from_authid($r);
 148             return ($r, $a->record);
 149         };
 150     } else {
 151         my $records = Koha::MetadataRecord::Authority->get_all_authorities_iterator();
 152         $next = sub {
 153             $records->next();
 154         }
 155     }
 156     do_reindex($next, $Koha::SearchEngine::Elasticsearch::AUTHORITIES_INDEX);
 157 }
 158
 159 sub do_reindex {
 160     my ( $next, $index_name ) = @_;
 161
 162     my $indexer = Koha::SearchEngine::Elasticsearch::Indexer->new( { index => $index_name } );
 163     if ($delete) {
 164
 165         # We know it's safe to not recreate the indexer because update_index
 166         # hasn't been called yet.
 167         $indexer->drop_index();
 168     }
 169
 170     my $count        = 0;
 171     my $commit_count = $commit;
 172     my ( @id_buffer, @commit_buffer );
 173     while ( my $record = $next->() ) {
 174         my $id     = $record->id;
 175         my $record = $record->record;
 176         _log( 1, "$id\n" );
 177         $count++;
 178
 179         push @id_buffer,     $id;
 180         push @commit_buffer, $record;
 181         if ( !( --$commit_count ) ) {
 182             _log( 2, "Committing...\n" );
 183             $indexer->update_index( \@id_buffer, \@commit_buffer );
 184             $commit_count  = $commit;
 185             @id_buffer     = ();
 186             @commit_buffer = ();
 187         }
 188     }
 189
 190     # There are probably uncommitted records
 191     $indexer->update_index( \@id_buffer, \@commit_buffer );
 192     _log( 1, "$count records indexed.\n" );
 193 }
 194
 195 # Checks some basic stuff to ensure that it's sane before we start.
 196 sub sanity_check {
 197     # Do we have an elasticsearch block defined?
 198     my $conf = C4::Context->config('elasticsearch');
 199     die "No 'elasticsearch' block is defined in koha-conf.xml.\n" if ( !$conf );
 200 }
 201
 202 # Output progress information.
 203 #
 204 #   _log($level, $msg);
 205 #
 206 # Will output $msg if the verbosity setting is set to $level or more. Will
 207 # not include a trailing newline.
 208 sub _log {
 209     my ($level, $msg) = @_;
 210
 211     print $msg if ($verbose >= $level);
 212 }