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
  49 my $result = GetOptions(
  50     'encoding:s'    => \$encoding,
  51     'file:s'        => \$input_file,
  52     'match|match-bibs:s'  => \$match,
  53     'add-items'     => \$add_items,
  54     'no-replace'    => \$no_replace,
  55     'comment:s'     => \$batch_comment,
  56     'authorities'   => \$authorities,
  57     'h|help'        => \$want_help
  58 );
  59
  60 $record_type = 'auth' if ($authorities);
  61
  62 if ($encoding eq "") {
  63     $encoding = "utf8";
  64 }
  65
  66 if (not $result or $input_file eq "" or $want_help) {
  67     print_usage();
  68     exit 0;
  69 }
  70
  71 unless (-r $input_file) {
  72     die "$0: cannot open input file $input_file: $!\n";
  73 }
  74
  75 my $dbh = C4::Context->dbh;
  76 $dbh->{AutoCommit} = 0;
  77 process_batch($input_file, $record_type, $match, $add_items, $batch_comment);
  78 $dbh->commit();
  79
  80 exit 0;
  81
  82 sub process_batch {
  83     my ($input_file, $record_type, $match, $add_items, $batch_comment) = @_;
  84
  85     open IN, "<$input_file" or die "$0: cannot open input file $input_file: $!\n";
  86     my $marc_records = "";
  87     $/ = "\035";
  88     my $num_input_records = 0;
  89     while (<IN>) {
  90         s/^\s+//;
  91         s/\s+$//;
  92         next unless $_; # skip if record has only whitespace, as might occur
  93                         # if file includes newlines between each MARC record
  94         $marc_records .= $_; # FIXME - this sort of string concatenation
  95                              # is probably rather inefficient
  96         $num_input_records++;
  97     }
  98     close IN;
  99
 100     print "... staging MARC records -- please wait\n";
 101     my ($batch_id, $num_valid_records, $num_items, @import_errors) =
 102         BatchStageMarcRecords($record_type, $encoding, $marc_records, $input_file, $batch_comment, '', $add_items, 0,
 103                               100, \&print_progress_and_commit);
 104     print "... finished staging MARC records\n";
 105
 106     my $num_with_matches = 0;
 107     if ($match) {
 108         my $matcher = C4::Matcher->fetch($match) ;
 109         if (defined $matcher) {
 110             SetImportBatchMatcher($batch_id, $match);
 111         } elsif ($record_type eq 'biblio')  {
 112             $matcher = C4::Matcher->new($record_type);
 113             $matcher->add_simple_matchpoint('isbn', 1000, '020', 'a', -1, 0, '');
 114             $matcher->add_simple_required_check('245', 'a', -1, 0, '',
 115                                             '245', 'a', -1, 0, '');
 116         }
 117         # set default record overlay behavior
 118         SetImportBatchOverlayAction($batch_id, ($no_replace) ? 'ignore' : 'replace');
 119         SetImportBatchNoMatchAction($batch_id, 'create_new');
 120         SetImportBatchItemAction($batch_id, 'always_add');
 121         print "... looking for matches with records already in database\n";
 122         $num_with_matches = BatchFindDuplicates($batch_id, $matcher, 10, 100, \&print_progress_and_commit);
 123         print "... finished looking for matches\n";
 124     }
 125
 126     my $num_invalid_records = scalar(@import_errors);
 127     print <<_SUMMARY_;
 128
 129 MARC record staging report
 130 ------------------------------------
 131 Input file:                 $input_file
 132 Record type:                $record_type
 133 Number of input records:    $num_input_records
 134 Number of valid records:    $num_valid_records
 135 Number of invalid records:  $num_invalid_records
 136 _SUMMARY_
 137     if ($match) {
 138         print "Number of records matched:  $num_with_matches\n";
 139     } else {
 140         print "Incoming records not matched against existing records (--match option not supplied)\n";
 141     }
 142     if ($record_type eq 'biblio') {
 143         if ($add_items) {
 144             print "Number of items parsed:  $num_items\n";
 145         } else {
 146             print "No items parsed (--add-items option not supplied)\n";
 147         }
 148     }
 149
 150     print "\n";
 151     print "Batch number assigned:  $batch_id\n";
 152     print "\n";
 153 }
 154
 155 sub print_progress_and_commit {
 156     my $recs = shift;
 157     $dbh->commit();
 158     print "... processed $recs records\n";
 159 }
 160
 161 sub print_usage {
 162     print <<_USAGE_;
 163 $0: stage MARC file into reservoir.
 164
 165 Use this batch job to load a file of MARC bibliographic
 166 (with optional item information) or authority records into
 167 the Koha reservoir.
 168
 169 After running this program to stage your file, you can use
 170 either the batch job commit_file.pl or the Koha
 171 Tools option "Manage Staged MARC Records" to load the
 172 records into the main Koha database.
 173
 174 Parameters:
 175     --file <file_name>      name of input MARC bib file
 176     --authorities           stage authority records instead of bibs
 177     --encoding <encoding>   encoding of MARC records, default is utf8.
 178                             Other possible options are: MARC-8,
 179                             ISO_5426, ISO_6937, ISO_8859-1, EUC-KR
 180     --match <match_id>      use this option to match records
 181                             in the file with records already in
 182                             the database for future overlay.
 183                             If <match_id> isn't defined, a default
 184                             MARC21 ISBN & title match rule will be applied
 185                             for bib imports.
 186     --add-items             use this option to specify that
 187                             item data is embedded in the MARC
 188                             bibs and should be parsed.
 189     --no-replace            overlay action for record: default is to
 190                             replace extant with the imported record.
 191     --comment <comment>     optional comment to describe
 192                             the record batch; if the comment
 193                             has spaces in it, surround the
 194                             comment with quotation marks.
 195     --help or -h            show this message.
 196 _USAGE_
 197 }