From ac7df30c0ab3bee04d2526765d6078384e5d70c9 Mon Sep 17 00:00:00 2001
From: dbs <dbs@6d9bc8c9-1ec2-4278-b937-99fde70a366f>
Date: Thu, 16 Apr 2009 19:12:35 +0000
Subject: [PATCH] miker says MARC::Batch would be more efficient; okay. Baby
 steps towards pipeline processing.

git-svn-id: svn://svn.open-ils.org/ILS-Contrib/conifer/trunk@341 6d9bc8c9-1ec2-4278-b937-99fde70a366f
---
 tools/migration-scripts/fixURIs.pl | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/tools/migration-scripts/fixURIs.pl b/tools/migration-scripts/fixURIs.pl
index 70c8897513..41d6ce7ecd 100644
--- a/tools/migration-scripts/fixURIs.pl
+++ b/tools/migration-scripts/fixURIs.pl
@@ -1,7 +1,11 @@
 #!/usr/bin/perl -w
 use strict;
+use warnings;
+
+use Error qw/:try/;
 use Getopt::Long;
 use MARC::File::XML( BinaryEncoding => 'utf8', RecordFormat => 'USMARC' );
+use MARC::Batch;
 
 # Clean up URIs from MARCXML records prior to batch ingest
 #   * If we detect a proxy URL:
@@ -23,11 +27,16 @@ if (!$input_file or !$output_file) {
     exit();
 }
 
-my $input = MARC::File::XML->in( $input_file );
+my $input = new MARC::Batch( 'XML', $input_file );
+$input->strict_off();
 my $output = MARC::File::XML->out( $output_file );
 
 my ($touched, $url_cnt, $ind1_cnt, $ind2_cnt, $sub9_cnt) = (0, 0, 0, 0, 0);
-while (my $marc = $input->next()) {
+my $marc;
+while ( try { $marc = $input->next() } otherwise { $marc = -1 } ) {
+    # Skip the record if we couldn't even decode it
+    next if ($marc == -1);
+
     my $edited = 0;
     my @uri_fields = $marc->field('856');
     foreach my $uri (@uri_fields) {
-- 
2.11.0