misc/maintenance/sanitize_records.pl

   1 #!/usr/bin/perl
   2
   3 # This file is part of Koha.
   4 #
   5 # Copyright 2014 BibLibre
   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
  22 use Koha::Script;
  23 use C4::Charset qw( SanitizeRecord );
  24 use C4::Context;
  25 use DBI;
  26 use C4::Biblio;
  27 use Getopt::Long;
  28 use Pod::Usage;
  29
  30 my ( $help, $verbose, $confirm, $biblionumbers, $reindex, $filename,
  31     $auto_search, $fix_ampersand );
  32 my $result = GetOptions(
  33     'h|help'          => \$help,
  34     'v|verbose'       => \$verbose,
  35     'c|confirm'       => \$confirm,
  36     'biblionumbers:s' => \$biblionumbers,
  37     'reindex'         => \$reindex,
  38     'f|filename:s'    => \$filename,
  39     'auto-search'     => \$auto_search,
  40     'fix-ampersand'   => \$fix_ampersand,
  41 ) || pod2usage(1);
  42
  43 # This script only fix ampersand at the moment.
  44 # It is enabled by default.
  45 $fix_ampersand = 1;
  46
  47 if ($help) {
  48     pod2usage(0);
  49 }
  50
  51 unless ( $filename or $biblionumbers or $auto_search ) {
  52     pod2usage(
  53         -exitval => 1,
  54         -message =>
  55           qq{\n\tAt least one record number source should be provided.\n}
  56     );
  57 }
  58
  59 if (   $filename and $biblionumbers
  60     or $filename and $auto_search
  61     or $biblionumbers and $auto_search )
  62 {
  63     pod2usage(
  64         -exitval => 1,
  65         -message => qq{\n\tOnly one record number source should be provided.\n}
  66     );
  67 }
  68
  69 my @biblionumbers;
  70
  71 # We first detect if we have a file or biblos directly entered by command line
  72 #or if we want to use findAmp() sub
  73 if ($auto_search) {
  74     @biblionumbers = biblios_to_sanitize();
  75 }
  76 elsif ($filename) {
  77     if ( -e $filename ) {
  78         open( my $fh, '<', $filename ) || die("Can't open $filename ($!)");
  79         while (<$fh>) {
  80             chomp;
  81             my $line = $_;
  82             push @biblionumbers, split( " |,", $line );
  83         }
  84         close $fh;
  85     }
  86     else {
  87         pod2usage(
  88             -exitval => 1,
  89             -message =>
  90 qq{\n\tThis filename does not exist. Please verify the path is correct.\n}
  91         );
  92     }
  93 }
  94 else {
  95     @biblionumbers = split m|,|, $biblionumbers if $biblionumbers;
  96 }
  97
  98 # We remove spaces
  99 s/(^\s*|\s*$)//g for @biblionumbers;
 100
 101 # Remove empty lines
 102 @biblionumbers = grep { !/^$/ } @biblionumbers;
 103
 104 say @biblionumbers . " records to process" if $verbose;
 105
 106 my @changes;
 107 for my $biblionumber (@biblionumbers) {
 108     print "processing record $biblionumber..." if $verbose;
 109     unless ( $biblionumber =~ m|^\d+$| ) {
 110         say " skipping. ERROR: Invalid biblionumber." if $verbose;
 111         next;
 112     }
 113     my $record = C4::Biblio::GetMarcBiblio({ biblionumber => $biblionumber });
 114     unless ($record) {
 115         say " skipping. ERROR: Invalid record." if $verbose;
 116         next;
 117     }
 118
 119     my ( $cleaned_record, $has_been_modified ) =
 120       C4::Charset::SanitizeRecord( $record, $biblionumber );
 121     if ($has_been_modified) {
 122         my $frameworkcode = C4::Biblio::GetFrameworkCode($record);
 123
 124         C4::Biblio::ModBiblio( $cleaned_record, $biblionumber, $frameworkcode )
 125             if $confirm;
 126         push @changes, $biblionumber;
 127         say " Done!" if $verbose;
 128     }
 129     else {
 130         say " Nothing to do." if $verbose;
 131     }
 132 }
 133
 134 if ($verbose) {
 135     say "Total: "
 136       . @changes
 137       . " records "
 138       . ( $confirm ? "cleaned!" : "to clean." );
 139 }
 140
 141 if ( $reindex and $confirm and @changes ) {
 142     say "Now, reindexing using -b -v" if $verbose;
 143     my $kohapath = C4::Context->config('intranetdir');
 144     my $cmd      = qq|
 145         $kohapath/misc/migration_tools/rebuild_zebra.pl -b -v -where "biblionumber IN ( |
 146       . join( ',', @changes ) . q| )"
 147     |;
 148     system($cmd);
 149 }
 150
 151 sub biblios_to_sanitize {
 152     my $dbh   = C4::Context->dbh;
 153     my $query = q{
 154         SELECT biblionumber
 155         FROM biblio_metadata
 156         WHERE format = 'marcxml'
 157             AND `schema` = ?
 158             AND metadata LIKE "%&amp;amp;%"
 159         };
 160     return @{ $dbh->selectcol_arrayref( $query, { Slice => {} }, C4::Context->preference('marcflavour') ) };
 161 }
 162
 163 =head1 NAME
 164
 165 sanitize_records - This script sanitizes a record.
 166
 167 =head1 SYNOPSIS
 168
 169 sanitize_records.pl [-h|--help] [-v|--verbose] [-c|--confirm] [--biblionumbers=BIBLIONUMBER_LIST] [-f|--filename=FILENAME] [--auto-search] [--reindex] [--fix-ampersand]
 170
 171 You can either give some biblionumbers or a file with biblionumbers or ask for an auto-search.
 172
 173 =head1 OPTIONS
 174
 175 =over
 176
 177 =item B<-h|--help>
 178
 179 Print a brief help message
 180
 181 =item B<-v|--verbose>
 182
 183 Verbose mode.
 184
 185 =item B<-c|--confirm>
 186
 187 This flag must be provided in order for the script to actually
 188 sanitize records. If it is not supplied, the script will
 189 only report on the record list to process.
 190
 191 =item B<--biblionumbers=BIBLIONUMBER_LIST>
 192
 193 Give a biblionumber list using this parameter. They must be separated by
 194 commas.
 195
 196 =item B<-f|--filename=FILENAME>
 197
 198 Give a biblionumber list using a filename. One biblionumber by line or separate them with a whitespace character.
 199
 200 =item B<--auto-search>
 201
 202 Automatically search records containing "&amp;" in biblio_metadata.metadata or in the specified fields.
 203
 204 =item B<--fix-ampersand>
 205
 206 Replace '&amp;' by '&' in the records.
 207 Replace '&amp;amp;amp;etc.' with '&amp;' in the records.
 208
 209 =item B<--reindex>
 210
 211 Reindex the modified records.
 212
 213 =back
 214
 215 =head1 AUTHOR
 216
 217 Alex Arnaud <alex.arnaud@biblibre.com>
 218 Christophe Croullebois <christophe.croullebois@biblibre.com>
 219 Jonathan Druart <jonathan.druart@biblibre.com>
 220
 221 =head1 COPYRIGHT
 222
 223 Copyright 2014 BibLibre
 224
 225 =head1 LICENSE
 226
 227 This file is part of Koha.
 228
 229 # Koha is free software; you can redistribute it and/or modify it
 230 # under the terms of the GNU General Public License as published by
 231 # the Free Software Foundation; either version 3 of the License, or
 232 # (at your option) any later version.
 233 #
 234 # Koha is distributed in the hope that it will be useful, but
 235 # WITHOUT ANY WARRANTY; without even the implied warranty of
 236 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
 237 # GNU General Public License for more details.
 238 #
 239 # You should have received a copy of the GNU General Public License
 240 # along with Koha; if not, see <http://www.gnu.org/licenses>.
 241
 242 =cut