misc/migration_tools/import_lexile.pl

   1 #!/usr/bin/perl
   2 #-----------------------------------
   3 # Copyright 2015 ByWater Solutions
   4 #
   5 # This file is part of Koha.
   6 #
   7 # Koha is free software; you can redistribute it and/or modify it under the
   8 # terms of the GNU General Public License as published by the Free Software
   9 # Foundation; either version 3 of the License, or (at your option) any later
  10 # version.
  11 #
  12 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
  13 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
  14 # A PARTICULAR PURPOSE.  See the GNU General Public License for more details.
  15 #
  16 # You should have received a copy of the GNU General Public License along
  17 # with Koha; if not, write to the Free Software Foundation, Inc.,
  18 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  19 #-----------------------------------
  20
  21 =head1 NAME
  22
  23 import_lexile.pl  Import lexile scores for records from csv.
  24
  25 =cut
  26
  27 use utf8;
  28
  29 use Modern::Perl;
  30
  31 use Getopt::Long;
  32 use Text::CSV;
  33
  34 use C4::Context;
  35 use C4::Biblio;
  36 use C4::Koha qw( GetVariationsOfISBN );
  37 use Koha::Database;
  38
  39 binmode STDOUT, ':encoding(UTF-8)';
  40
  41 BEGIN {
  42
  43     # find Koha's Perl modules
  44     # test carefully before changing this
  45     use FindBin;
  46     eval { require "$FindBin::Bin/../kohalib.pl" };
  47 }
  48
  49 my $help;
  50 my $confirm;
  51 my $test;
  52 my $file;
  53 my $verbose;
  54 my $start;
  55 my $end;
  56 my $field_number                  = "521";
  57 my $subfield_target_audience_note = "a";
  58 my $subfield_source               = "b";
  59 my $subfield_source_value         = "Lexile";
  60
  61 GetOptions(
  62     'h|help'                 => \$help,
  63     'c|confirm'              => \$confirm,
  64     't|test'                 => \$test,
  65     'f|file=s'               => \$file,
  66     'v|verbose+'             => \$verbose,
  67     's|start=s'              => \$start,
  68     'e|end=s'                => \$end,
  69     'field=s'                => \$field_number,
  70     'target-audience-note=s' => $subfield_target_audience_note,
  71     'source=s'               => $subfield_source,
  72     'source-value=s'         => $subfield_source_value,
  73 );
  74
  75 my $usage = << 'ENDUSAGE';
  76 import_lexile.pl: Import lexile scores for records from csv.
  77
  78 import_lexile.pl -f /path/to/LexileTitles.txt
  79
  80 This script takes the following parameters :
  81
  82     -h --help               Display this help
  83     -c --confirm            Confirms you want to really run this script ( otherwise print help )
  84     -t --test               Runs the script in test mode ( no changes will be made to your database )
  85     -f --file               CSV file of lexile scores ( acquired from Lexile.com )
  86     -v --verbose            Print data on found matches. Use -v -v for more data, and -v -v -v will give the most data.
  87     --field                 Defines the field number for the Lexile data ( default: 521 )
  88     --target-audience-note  Defines the subfield for the lexile score ( default: a )
  89     --source                Defines the "Source" subfield ( default: b )
  90     --source-value          Defines the value to put stored in the "Source" subfield ( default: "Lexile" )
  91
  92     The CSV file must have the following columns ( with the first line being the column headers ) in tab delimited format:
  93     Title, Author, ISBN, ISBN13, Lexile
  94
  95 ENDUSAGE
  96
  97 if ( $help || !$file || !$confirm ) {
  98     say $usage;
  99     exit(1);
 100 }
 101
 102 my $schema = Koha::Database->new()->schema();
 103
 104 my $csv = Text::CSV->new( { binary => 1, sep_char => "\t" } )
 105   or die "Cannot use CSV: " . Text::CSV->error_diag();
 106
 107 open my $fh, "<:encoding(utf8)", $file or die "test.csv: $!";
 108
 109 my $column_names = $csv->getline($fh);
 110 $csv->column_names(@$column_names);
 111
 112 my $counter = 0;
 113 my $i       = 0;
 114 while ( my $row = $csv->getline_hr($fh) ) {
 115     $i++;
 116
 117     next if ( $start && $i < $start );
 118     last if ( $end   && $i >= $end );
 119
 120     if ( $verbose > 1 ) {
 121         say "Searching for matching record for row $i...";
 122         say "Title: " . $row->{Title};
 123         say "Author: " . $row->{Author};
 124         say "ISBN10: " . $row->{ISBN};
 125         say "ISBN13: " . $row->{ISBN13};
 126         say q{};
 127     }
 128
 129     # Match by ISBN
 130     my @isbns;
 131     for ( 'ISBN', 'ISBN13' ) {
 132         if ( $row->{$_} && $row->{$_} ne "None" ) {
 133             push( @isbns, $row->{$_} );
 134             eval { push( @isbns, GetVariationsOfISBN( $row->{$_} ) ) };
 135         }
 136     }
 137     @isbns = grep( $_, @isbns );
 138     next unless @isbns;
 139
 140     say "Searching for ISBNs: " . join( ' : ', @isbns ) if ( $verbose > 2 );
 141
 142     my @likes = map { { isbn => { like => '%' . $_ . '%' } } } @isbns;
 143
 144     my @biblionumbers =
 145       $schema->resultset('Biblioitem')->search( { -or => \@likes } )
 146       ->get_column('biblionumber')->all();
 147
 148     say "Found matching records! Biblionumbers: " . join( " ,", @biblionumbers )
 149       if ( @biblionumbers && $verbose > 2 );
 150
 151     foreach my $biblionumber (@biblionumbers) {
 152         $counter++;
 153         my $record = GetMarcBiblio($biblionumber);
 154
 155         if ($verbose) {
 156             say "Found matching record! Biblionumber: $biblionumber";
 157
 158             if ( $verbose > 2 ) {
 159                 my $biblio = GetBiblioData($biblionumber);
 160                 say "Title from record: " . $biblio->{title}
 161                   if ( $biblio->{title} );
 162                 say "Author from record: " . $biblio->{author}
 163                   if ( $biblio->{author} );
 164                 say "ISBN from record: " . $biblio->{isbn}
 165                   if ( $biblio->{isbn} );
 166             }
 167             say "Title: " . $row->{Title};
 168             say "Author: " . $row->{Author};
 169             say "ISBN10: " . $row->{ISBN};
 170             say "ISBN13: " . $row->{ISBN13};
 171             say q{};
 172         }
 173
 174         # Check for existing embedded lexile score
 175         my $lexile_score_field;
 176         for my $field ( $record->field($field_number) ) {
 177             if ( defined( $field->subfield($subfield_source) )
 178                 && $field->subfield($subfield_source) eq
 179                 $subfield_source_value )
 180             {
 181                 $lexile_score_field = $field;
 182                 last;    # Each item can only have one lexile score
 183             }
 184         }
 185
 186         if ($lexile_score_field) {
 187             $lexile_score_field->update(
 188                 ind1                           => '8',
 189                 ind2                           => '#',
 190                 $subfield_target_audience_note => $row->{Lexile},
 191                 $subfield_source               => $subfield_source_value,
 192             );
 193         }
 194         else {
 195             my $field = MARC::Field->new(
 196                 $field_number, '8', '#',
 197                 $subfield_target_audience_note => $row->{Lexile},
 198                 $subfield_source               => $subfield_source_value,
 199             );
 200             $record->append_fields($field);
 201         }
 202
 203         ModBiblio( $record, $biblionumber ) unless ( $test );
 204     }
 205
 206 }
 207 say "Update $counter records" if $verbose;