From 469b84c98616215fc6500f0a49797a47bf0a567b Mon Sep 17 00:00:00 2001
From: Nick Clemens <nick@bywatersolutions.com>
Date: Thu, 31 Oct 2024 19:23:08 +0000
Subject: [PATCH] Bug 37478: Rename strict to skip_bad_records and add POD

Option name changed, POD updated.
The reason not to do this by default is that it is an extra record conversion
to check the validity, so may slow down the import.

Signed-off-by: Martin Renvoize <martin.renvoize@ptfs-europe.com>
---
 misc/migration_tools/bulkmarcimport.pl | 80 ++++++++++++++------------
 1 file changed, 43 insertions(+), 37 deletions(-)

diff --git a/misc/migration_tools/bulkmarcimport.pl b/misc/migration_tools/bulkmarcimport.pl
index 691e9c9e363..4406252fc7d 100755
--- a/misc/migration_tools/bulkmarcimport.pl
+++ b/misc/migration_tools/bulkmarcimport.pl
@@ -77,45 +77,45 @@ my $localcust;
 my $marc_mod_template    = '';
 my $marc_mod_template_id = -1;
 my $skip_indexing        = 0;
-my $strict_mode;
+my $skip_bad_records;
 $| = 1;
 
 GetOptions(
-    'commit:f'          => \$commit,
-    'file:s'            => \$input_marc_file,
-    'n:f'               => \$number,
-    'o|offset:f'        => \$offset,
-    'h|help'            => \$version,
-    'd|delete'          => \$delete,
-    't|test'            => \$test_parameter,
-    's'                 => \$skip_marc8_conversion,
-    'c:s'               => \$char_encoding,
-    'v|verbose:+'       => \$verbose,
-    'fk'                => \$fk_off,
-    'm:s'               => \$format,
-    'l:s'               => \$logfile,
-    'append'            => \$append,
-    'k|keepids:s'       => \$keepids,
-    'b|biblios'         => \$biblios,
-    'a|authorities'     => \$authorities,
-    'authtypes:s'       => \$authtypes,
-    'filter=s@'         => \$filters,
-    'insert'            => \$insert,
-    'update'            => \$update,
-    'all'               => \$all,
-    'match=s@'          => \$match,
-    'i|isbn'            => \$isbn_check,
-    'x:s'               => \$sourcetag,
-    'y:s'               => \$sourcesubfield,
-    'idmap:s'           => \$idmapfl,
-    'cleanisbn!'        => \$cleanisbn,
-    'yaml:s'            => \$yamlfile,
-    'dedupbarcode'      => \$dedup_barcode,
-    'framework=s'       => \$framework,
-    'custom:s'          => \$localcust,
-    'marcmodtemplate:s' => \$marc_mod_template,
-    'si|skip_indexing'  => \$skip_indexing,
-    'st|strict'         => \$strict_mode,
+    'commit:f'            => \$commit,
+    'file:s'              => \$input_marc_file,
+    'n:f'                 => \$number,
+    'o|offset:f'          => \$offset,
+    'h|help'              => \$version,
+    'd|delete'            => \$delete,
+    't|test'              => \$test_parameter,
+    's'                   => \$skip_marc8_conversion,
+    'c:s'                 => \$char_encoding,
+    'v|verbose:+'         => \$verbose,
+    'fk'                  => \$fk_off,
+    'm:s'                 => \$format,
+    'l:s'                 => \$logfile,
+    'append'              => \$append,
+    'k|keepids:s'         => \$keepids,
+    'b|biblios'           => \$biblios,
+    'a|authorities'       => \$authorities,
+    'authtypes:s'         => \$authtypes,
+    'filter=s@'           => \$filters,
+    'insert'              => \$insert,
+    'update'              => \$update,
+    'all'                 => \$all,
+    'match=s@'            => \$match,
+    'i|isbn'              => \$isbn_check,
+    'x:s'                 => \$sourcetag,
+    'y:s'                 => \$sourcesubfield,
+    'idmap:s'             => \$idmapfl,
+    'cleanisbn!'          => \$cleanisbn,
+    'yaml:s'              => \$yamlfile,
+    'dedupbarcode'        => \$dedup_barcode,
+    'framework=s'         => \$framework,
+    'custom:s'            => \$localcust,
+    'marcmodtemplate:s'   => \$marc_mod_template,
+    'si|skip_indexing'    => \$skip_indexing,
+    'sk|skip_bad_records' => \$skip_bad_records,
 );
 
 $biblios ||= !$authorities;
@@ -343,7 +343,7 @@ RECORD: while () {
     }
     if ($record) {
 
-        if ($strict_mode) {
+        if ($skip_bad_records) {
             my $xml = $record->as_xml_record();
             eval { MARC::Record::new_from_xml( $xml, 'UTF-8', "MARC21" ); };
             if ($@) {
@@ -1044,6 +1044,12 @@ If set, do not index the imported records with Zebra or Elasticsearch.
 Use this when you plan to do a complete reindex of your data after running
 bulkmarciport. This can increase performance and avoid unnecessary load.
 
+=item B<-sk, --skip_bad_records>
+
+If set, check the validity of records before adding. If they are invalid we will
+print the outout of MARC::Lint->check_record and skip them during the import. Without
+this option bad records may kill the job.
+
 =back
 
 =cut
-- 
2.47.0