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 use C4::Charset qw( SanitizeRecord );
  22 use C4::Context;
  23 use DBI;
  24 use C4::Biblio;
  25 use Getopt::Long;
  26 use Pod::Usage;
  27
  28 my ( $help, $verbose, $confirm, $biblionumbers, $reindex, $filename,
  29     $auto_search, $fix_ampersand );
  30 my $result = GetOptions(
  31     'h|help'          => \$help,
  32     'v|verbose'       => \$verbose,
  33     'c|confirm'       => \$confirm,
  34     'biblionumbers:s' => \$biblionumbers,
  35     'reindex'         => \$reindex,
  36     'f|filename:s'    => \$filename,
  37     'auto-search'     => \$auto_search,
  38     'fix-ampersand'   => \$fix_ampersand,
  39 ) || pod2usage(1);
  40
  41 # This script only fix ampersand at the moment.
  42 # It is enabled by default.
  43 $fix_ampersand = 1;
  44
  45 if ($help) {
  46     pod2usage(0);
  47 }
  48
  49 unless ( $filename or $biblionumbers or $auto_search ) {
  50     pod2usage(
  51         -exitval => 1,
  52         -message =>
  53           qq{\n\tAt least one record number source should be provided.\n}
  54     );
  55 }
  56
  57 if (   $filename and $biblionumbers
  58     or $filename and $auto_search
  59     or $biblionumbers and $auto_search )
  60 {
  61     pod2usage(
  62         -exitval => 1,
  63         -message => qq{\n\tOnly one record number source should be provided.\n}
  64     );
  65 }
  66
  67 my @biblionumbers;
  68
  69 # We first detect if we have a file or biblos directly entered by command line
  70 #or if we want to use findAmp() sub
  71 if ($auto_search) {
  72     @biblionumbers = biblios_to_sanitize();
  73 }
  74 elsif ($filename) {
  75     if ( -e $filename ) {
  76         open( my $fh, '<', $filename ) || die("Can't open $filename ($!)");
  77         while (<$fh>) {
  78             chomp;
  79             my $line = $_;
  80             push @biblionumbers, split( " |,", $line );
  81         }
  82         close $fh;
  83     }
  84     else {
  85         pod2usage(
  86             -exitval => 1,
  87             -message =>
  88 qq{\n\tThis filename does not exist. Please verify the path is correct.\n}
  89         );
  90     }
  91 }
  92 else {
  93     @biblionumbers = split m|,|, $biblionumbers if $biblionumbers;
  94 }
  95
  96 # We remove spaces
  97 s/(^\s*|\s*$)//g for @biblionumbers;
  98
  99 # Remove empty lines
 100 @biblionumbers = grep { !/^$/ } @biblionumbers;
 101
 102 say @biblionumbers . " records to process" if $verbose;
 103
 104 my @changes;
 105 for my $biblionumber (@biblionumbers) {
 106     print "processing record $biblionumber..." if $verbose;
 107     unless ( $biblionumber =~ m|^\d+$| ) {
 108         say " skipping. ERROR: Invalid biblionumber." if $verbose;
 109         next;
 110     }
 111     my $record = C4::Biblio::GetMarcBiblio({ biblionumber => $biblionumber });
 112     unless ($record) {
 113         say " skipping. ERROR: Invalid record." if $verbose;
 114         next;
 115     }
 116
 117     my ( $cleaned_record, $has_been_modified ) =
 118       C4::Charset::SanitizeRecord( $record, $biblionumber );
 119     if ($has_been_modified) {
 120         my $frameworkcode = C4::Biblio::GetFrameworkCode($record);
 121
 122         C4::Biblio::ModBiblio( $cleaned_record, $biblionumber, $frameworkcode )
 123             if $confirm;
 124         push @changes, $biblionumber;
 125         say " Done!" if $verbose;
 126     }
 127     else {
 128         say " Nothing to do." if $verbose;
 129     }
 130 }
 131
 132 if ($verbose) {
 133     say "Total: "
 134       . @changes
 135       . " records "
 136       . ( $confirm ? "cleaned!" : "to clean." );
 137 }
 138
 139 if ( $reindex and $confirm and @changes ) {
 140     say "Now, reindexing using -b -v" if $verbose;
 141     my $kohapath = C4::Context->config('intranetdir');
 142     my $cmd      = qq|
 143         $kohapath/misc/migration_tools/rebuild_zebra.pl -b -v -where "biblionumber IN ( |
 144       . join( ',', @changes ) . q| )"
 145     |;
 146     system($cmd);
 147 }
 148
 149 sub biblios_to_sanitize {
 150     my $dbh   = C4::Context->dbh;
 151     my $query = q{
 152         SELECT biblionumber
 153         FROM biblio_metadata
 154         WHERE format = 'marcxml'
 155             AND marcflavour = ?
 156             AND metadata LIKE "%&amp;amp;%"
 157         };
 158     return @{ $dbh->selectcol_arrayref( $query, { Slice => {} }, C4::Context->preference('marcflavour') ) };
 159 }
 160
 161 =head1 NAME
 162
 163 sanitize_records - This script sanitizes a record.
 164
 165 =head1 SYNOPSIS
 166
 167 sanitize_records.pl [-h|--help] [-v|--verbose] [-c|--confirm] [--biblionumbers=BIBLIONUMBER_LIST] [-f|--filename=FILENAME] [--auto-search] [--reindex] [--fix-ampersand]
 168
 169 You can either give some biblionumbers or a file with biblionumbers or ask for an auto-search.
 170
 171 =head1 OPTIONS
 172
 173 =over
 174
 175 =item B<-h|--help>
 176
 177 Print a brief help message
 178
 179 =item B<-v|--verbose>
 180
 181 Verbose mode.
 182
 183 =item B<-c|--confirm>
 184
 185 This flag must be provided in order for the script to actually
 186 sanitize records. If it is not supplied, the script will
 187 only report on the record list to process.
 188
 189 =item B<--biblionumbers=BIBLIONUMBER_LIST>
 190
 191 Give a biblionumber list using this parameter. They must be separated by
 192 commas.
 193
 194 =item B<-f|--filename=FILENAME>
 195
 196 Give a biblionumber list using a filename. One biblionumber by line or separate them with a whitespace character.
 197
 198 =item B<--auto-search>
 199
 200 Automatically search records containing "&amp;" in biblio_metadata.metadata or in the specified fields.
 201
 202 =item B<--fix-ampersand>
 203
 204 Replace '&amp;' by '&' in the records.
 205 Replace '&amp;amp;amp;etc.' with '&amp;' in the records.
 206
 207 =item B<--reindex>
 208
 209 Reindex the modified records.
 210
 211 =back
 212
 213 =head1 AUTHOR
 214
 215 Alex Arnaud <alex.arnaud@biblibre.com>
 216 Christophe Croullebois <christophe.croullebois@biblibre.com>
 217 Jonathan Druart <jonathan.druart@biblibre.com>
 218
 219 =head1 COPYRIGHT
 220
 221 Copyright 2014 BibLibre
 222
 223 =head1 LICENSE
 224
 225 This file is part of Koha.
 226
 227 Koha is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software
 228 Foundation; either version 3 of the License, or (at your option) any later version.
 229
 230 You should have received a copy of the GNU General Public License along
 231 with Koha; if not, write to the Free Software Foundation, Inc.,
 232 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
 233
 234 =cut