misc/stage_file.pl

   1 #!/usr/bin/perl
   2
   3 # This file is part of Koha.
   4 #
   5 # Copyright (C) 2007 LibLime
   6 # Parts Copyright BSZ 2011
   7 # Parts Copyright C & P Bibliography Services 2012
   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 2 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 this program; if not, write to the Free Software Foundation, Inc.,
  20 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  21
  22 use strict;
  23 use warnings;
  24 BEGIN {
  25     # find Koha's Perl modules
  26     # test carefully before changing this
  27     use FindBin;
  28     eval { require "$FindBin::Bin/kohalib.pl" };
  29 }
  30
  31 use C4::Context;
  32 use C4::ImportBatch;
  33 use C4::Matcher;
  34 use Getopt::Long;
  35
  36 $| = 1;
  37
  38 # command-line parameters
  39 my $record_type = "biblio";
  40 my $encoding = "";
  41 my $authorities = 0;
  42 my $match = 0;
  43 my $add_items = 0;
  44 my $input_file = "";
  45 my $batch_comment = "";
  46 my $want_help = 0;
  47 my $no_replace ;
  48 my $no_create;
  49 my $item_action = 'always_add';
  50
  51 my $result = GetOptions(
  52     'encoding:s'    => \$encoding,
  53     'file:s'        => \$input_file,
  54     'match|match-bibs:s'  => \$match,
  55     'add-items'     => \$add_items,
  56     'item-action:s' => \$item_action,
  57     'no-replace'    => \$no_replace,
  58     'no-create'     => \$no_create,
  59     'comment:s'     => \$batch_comment,
  60     'authorities'   => \$authorities,
  61     'h|help'        => \$want_help
  62 );
  63
  64 $record_type = 'auth' if ($authorities);
  65
  66 if ($encoding eq "") {
  67     $encoding = "utf8";
  68 }
  69
  70 if (not $result or $input_file eq "" or $want_help) {
  71     print_usage();
  72     exit 0;
  73 }
  74
  75 unless (-r $input_file) {
  76     die "$0: cannot open input file $input_file: $!\n";
  77 }
  78
  79 my $dbh = C4::Context->dbh;
  80 $dbh->{AutoCommit} = 0;
  81 process_batch($input_file, $record_type, $match, $add_items, $batch_comment);
  82 $dbh->commit();
  83
  84 exit 0;
  85
  86 sub process_batch {
  87     my ($input_file, $record_type, $match, $add_items, $batch_comment) = @_;
  88
  89     open IN, "<$input_file" or die "$0: cannot open input file $input_file: $!\n";
  90     my $marc_records = "";
  91     $/ = "\035";
  92     my $num_input_records = 0;
  93     while (<IN>) {
  94         s/^\s+//;
  95         s/\s+$//;
  96         next unless $_; # skip if record has only whitespace, as might occur
  97                         # if file includes newlines between each MARC record
  98         $marc_records .= $_; # FIXME - this sort of string concatenation
  99                              # is probably rather inefficient
 100         $num_input_records++;
 101     }
 102     close IN;
 103
 104     print "... staging MARC records -- please wait\n";
 105     #FIXME: We should really allow the use of marc modification frameworks and to_marc plugins here if possible
 106     my ($batch_id, $num_valid_records, $num_items, @import_errors) =
 107         BatchStageMarcRecords($record_type, $encoding, $marc_records, $input_file, undef, undef, $batch_comment, '', $add_items, 0,
 108                               100, \&print_progress_and_commit);
 109     print "... finished staging MARC records\n";
 110
 111     my $num_with_matches = 0;
 112     if ($match) {
 113         my $matcher = C4::Matcher->fetch($match) ;
 114         if (defined $matcher) {
 115             SetImportBatchMatcher($batch_id, $match);
 116         } elsif ($record_type eq 'biblio')  {
 117             $matcher = C4::Matcher->new($record_type);
 118             $matcher->add_simple_matchpoint('isbn', 1000, '020', 'a', -1, 0, '');
 119             $matcher->add_simple_required_check('245', 'a', -1, 0, '',
 120                                             '245', 'a', -1, 0, '');
 121         }
 122         # set default record overlay behavior
 123         SetImportBatchOverlayAction($batch_id, ($no_replace) ? 'ignore' : 'replace');
 124         SetImportBatchNoMatchAction($batch_id, ($no_create) ? 'ignore' : 'create_new');
 125         SetImportBatchItemAction($batch_id, $item_action);
 126         print "... looking for matches with records already in database\n";
 127         $num_with_matches = BatchFindDuplicates($batch_id, $matcher, 10, 100, \&print_progress_and_commit);
 128         print "... finished looking for matches\n";
 129     }
 130
 131     my $num_invalid_records = scalar(@import_errors);
 132     print <<_SUMMARY_;
 133
 134 MARC record staging report
 135 ------------------------------------
 136 Input file:                 $input_file
 137 Record type:                $record_type
 138 Number of input records:    $num_input_records
 139 Number of valid records:    $num_valid_records
 140 Number of invalid records:  $num_invalid_records
 141 _SUMMARY_
 142     if ($match) {
 143         print "Number of records matched:  $num_with_matches\n";
 144     } else {
 145         print "Incoming records not matched against existing records (--match option not supplied)\n";
 146     }
 147     if ($record_type eq 'biblio') {
 148         if ($add_items) {
 149             print "Number of items parsed:  $num_items\n";
 150         } else {
 151             print "No items parsed (--add-items option not supplied)\n";
 152         }
 153     }
 154
 155     print "\n";
 156     print "Batch number assigned:  $batch_id\n";
 157     print "\n";
 158 }
 159
 160 sub print_progress_and_commit {
 161     my $recs = shift;
 162     $dbh->commit();
 163     print "... processed $recs records\n";
 164 }
 165
 166 sub print_usage {
 167     print <<_USAGE_;
 168 $0: stage MARC file into reservoir.
 169
 170 Use this batch job to load a file of MARC bibliographic
 171 (with optional item information) or authority records into
 172 the Koha reservoir.
 173
 174 After running this program to stage your file, you can use
 175 either the batch job commit_file.pl or the Koha
 176 Tools option "Manage Staged MARC Records" to load the
 177 records into the main Koha database.
 178
 179 Parameters:
 180     --file <file_name>      name of input MARC bib file
 181     --authorities           stage authority records instead of bibs
 182     --encoding <encoding>   encoding of MARC records, default is utf8.
 183                             Other possible options are: MARC-8,
 184                             ISO_5426, ISO_6937, ISO_8859-1, EUC-KR
 185     --match <match_id>      use this option to match records
 186                             in the file with records already in
 187                             the database for future overlay.
 188                             If <match_id> isn't defined, a default
 189                             MARC21 ISBN & title match rule will be applied
 190                             for bib imports.
 191     --add-items             use this option to specify that
 192                             item data is embedded in the MARC
 193                             bibs and should be parsed.
 194     --item-action           action to take if --add-items is specifed;
 195                             choices are 'always_add',
 196                             'add_only_for_matches', 'add_only_for_new',
 197                             'ignore', or 'replace'
 198     --no-replace            overlay action for record: default is to
 199                             replace extant with the imported record.
 200     --no-create             nomatch action for record: default is to
 201                             create new record with imported record.
 202     --comment <comment>     optional comment to describe
 203                             the record batch; if the comment
 204                             has spaces in it, surround the
 205                             comment with quotation marks.
 206     --help or -h            show this message.
 207 _USAGE_
 208 }