misc/migration_tools/build_oai_sets.pl

   1 #!/usr/bin/perl
   2
   3 # Copyright 2011 BibLibre
   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 =head1 DESCRIPTION
  21
  22 This script build OAI-PMH sets (to be used by opac/oai.pl) according to sets
  23 and mappings defined in Koha. It reads informations from oai_sets and
  24 oai_sets_mappings, and then fill table oai_sets_biblios with builded infos.
  25
  26 =head1 USAGE
  27
  28     build_oai_sets.pl [-h] [-v] [-r] [-i] [-l LENGTH [-o OFFSET]]
  29         -h          Print help message;
  30         -v          Be verbose
  31         -r          Truncate table oai_sets_biblios before inserting new rows
  32         -i          Embed items informations, mandatory if you defined mappings
  33                     on item fields
  34         -l LENGTH   Process LENGTH biblios
  35         -o OFFSET   If LENGTH is defined, start processing from OFFSET
  36
  37 =cut
  38
  39 use Modern::Perl;
  40 use MARC::Record;
  41 use MARC::File::XML;
  42 use List::MoreUtils qw/uniq/;
  43 use Getopt::Std;
  44
  45 use Koha::Script;
  46 use C4::Context;
  47 use C4::Charset qw/StripNonXmlChars/;
  48 use C4::Biblio;
  49 use C4::OAI::Sets;
  50
  51 my %opts;
  52 $Getopt::Std::STANDARD_HELP_VERSION = 1;
  53 my $go = getopts('vo:l:ihr', \%opts);
  54
  55 if(!$go or $opts{h}){
  56     &print_usage;
  57     exit;
  58 }
  59
  60 my $verbose = $opts{v};
  61 my $offset = $opts{o};
  62 my $length = $opts{l};
  63 my $embed_items = $opts{i};
  64 my $reset = $opts{r};
  65
  66 my $dbh = C4::Context->dbh;
  67
  68 # Get OAI sets mappings
  69 my $mappings = GetOAISetsMappings;
  70
  71 # Get all biblionumbers and marcxml
  72 print "Retrieving biblios... " if $verbose;
  73 my $query = qq{
  74     SELECT biblionumber, metadata
  75     FROM biblio_metadata
  76     WHERE format='marcxml'
  77     AND  `schema` = ?
  78 };
  79 if($length) {
  80     $query .= "LIMIT $length";
  81     if($offset) {
  82         $query .= " OFFSET $offset";
  83     }
  84 }
  85 my $sth = $dbh->prepare($query);
  86 $sth->execute( C4::Context->preference('marcflavour') );
  87 my $results = $sth->fetchall_arrayref({});
  88 print "done.\n" if $verbose;
  89
  90 # Build lists of parents sets
  91 my $sets = GetOAISets;
  92 my $parentsets;
  93 foreach my $set (@$sets) {
  94     my $setSpec = $set->{'spec'};
  95     while($setSpec =~ /^(.+):(.+)$/) {
  96         my $parent = $1;
  97         my $parent_set = GetOAISetBySpec($parent);
  98         if($parent_set) {
  99             push @{ $parentsets->{$set->{'id'}} }, $parent_set->{'id'};
 100             $setSpec = $parent;
 101         } else {
 102             last;
 103         }
 104     }
 105 }
 106
 107 my $num_biblios = scalar @$results;
 108 my $i = 1;
 109 my $sets_biblios = {};
 110 foreach my $res (@$results) {
 111     my $biblionumber = $res->{'biblionumber'};
 112     my $marcxml = $res->{'metadata'};
 113     if($verbose and $i % 1000 == 0) {
 114         my $percent = ($i * 100) / $num_biblios;
 115         $percent = sprintf("%.2f", $percent);
 116         say "Progression: $i/$num_biblios ($percent %)";
 117     }
 118     # The following lines are copied from GetMarcBiblio
 119     # We don't call GetMarcBiblio to avoid a sql query to be executed each time
 120     $marcxml = StripNonXmlChars($marcxml);
 121     MARC::File::XML->default_record_format(C4::Context->preference('marcflavour'));
 122     my $record;
 123     eval {
 124         $record = MARC::Record::new_from_xml($marcxml, "utf8", C4::Context->preference('marcflavour'));
 125     };
 126     if($@) {
 127         warn "(biblio $biblionumber) Error while creating record from marcxml: $@";
 128         next;
 129     }
 130     if($embed_items) {
 131         C4::Biblio::EmbedItemsInMarcBiblio({
 132             marc_record  => $record,
 133             biblionumber => $biblionumber });
 134     }
 135
 136     my @biblio_sets = CalcOAISetsBiblio($record, $mappings);
 137     foreach my $set_id (@biblio_sets) {
 138         push @{ $sets_biblios->{$set_id} }, $biblionumber;
 139         foreach my $parent_set_id ( @{ $parentsets->{$set_id} } ) {
 140             push @{ $sets_biblios->{$parent_set_id} }, $biblionumber;
 141         }
 142     }
 143     $i++;
 144 }
 145 say "Progression: done." if $verbose;
 146
 147 say "Summary:";
 148 foreach my $set_id (keys %$sets_biblios) {
 149     $sets_biblios->{$set_id} = [ uniq @{ $sets_biblios->{$set_id} } ];
 150     my $set = GetOAISet($set_id);
 151     my $setSpec = $set->{'spec'};
 152     say "Set '$setSpec': ". scalar(@{$sets_biblios->{$set_id}}) ." biblios";
 153 }
 154
 155 print "Updating database... ";
 156 if($reset) {
 157     ModOAISetsBiblios( {} );
 158 }
 159 AddOAISetsBiblios($sets_biblios);
 160 print "done.\n";
 161
 162 sub print_usage {
 163     print "build_oai_sets.pl: Build OAI-PMH sets, according to mappings defined in Koha\n";
 164     print "Usage: build_oai_sets.pl [-h] [-v] [-i] [-l LENGTH [-o OFFSET]]\n\n";
 165     print "\t-h\t\tPrint this help and exit\n";
 166     print "\t-v\t\tBe verbose\n";
 167     print "\t-i\t\tEmbed items informations, mandatory if you defined mappings on item fields\n";
 168     print "\t-l LENGTH\tProcess LENGTH biblios\n";
 169     print "\t-o OFFSET\tIf LENGTH is defined, start processing from OFFSET\n\n";
 170 }