[Koha-patches] [PATCH] Bug 6199 - allow bulkmarkimport.pl to remove duplicate barcodes

[Robin Sheat]

This adds the --dedupbarcode option that allows bulkmarkimport to erase
a barcode but keep the item of any items it finds with duplicate
barcodes.

This patch applies against master only.
---
 C4/Items.pm                            |    5 ++-
 misc/migration_tools/bulkmarcimport.pl |   56 ++++++++++++++++++++++++++++++-
 2 files changed, 58 insertions(+), 3 deletions(-)

diff --git a/C4/Items.pm b/C4/Items.pm
index 01509c2..872fa13 100644
--- a/C4/Items.pm
+++ b/C4/Items.pm
@@ -303,7 +303,7 @@ Sequence number of original item tag in the MARC record.
 Item barcode, provide to assist in the construction of
 useful error messages.
 
-=item error_condition
+=item error_code
 
 Code representing the error condition.  Can be 'duplicate_barcode',
 'invalid_homebranch', or 'invalid_holdingbranch'.
@@ -323,6 +323,9 @@ sub AddItemBatchFromMarc {
     my @errors = ();
     my $dbh = C4::Context->dbh;
 
+    # We modify the record, so lets work on a clone so we don't change the
+    # original.
+    $record = $record->clone();
     # loop through the item tags and start creating items
     my @bad_item_fields = ();
     my ($itemtag, $itemsubfield) = &GetMarcFromKohaField("items.itemnumber",'');
diff --git a/misc/migration_tools/bulkmarcimport.pl b/misc/migration_tools/bulkmarcimport.pl
index ddfa5e0..8393f7e 100755
--- a/misc/migration_tools/bulkmarcimport.pl
+++ b/misc/migration_tools/bulkmarcimport.pl
@@ -33,8 +33,8 @@ use Pod::Usage;
 binmode(STDOUT, ":utf8");
 my ( $input_marc_file, $number, $offset) = ('',0,0);
 my ($version, $delete, $test_parameter, $skip_marc8_conversion, $char_encoding, $verbose, $commit, $fk_off,$format,$biblios,$authorities,$keepids,$match, $isbn_check, $logfile);
-my ($sourcetag,$sourcesubfield,$idmapfl);
 my $cleanisbn = 1;
+my ($sourcetag,$sourcesubfield,$idmapfl, $dedup_barcode);
 
 $|=1;
 
@@ -61,6 +61,7 @@ GetOptions(
     'y:s' => \$sourcesubfield,
     'idmap:s' => \$idmapfl,
     'cleanisbn!'     => \$cleanisbn,
+    'dedupbarcode' => \$dedup_barcode,
 );
 $biblios=!$authorities||$biblios;
 
@@ -317,6 +318,7 @@ RECORD: while (  ) {
  			else{
 				printlog({id=>$id||$originalid||$biblionumber, op=>"insert",status=>"ok"}) if ($logfile);
 			}
+			my $clone_record = $record->clone();
             eval { ( $itemnumbers_ref, $errors_ref ) = AddItemBatchFromMarc( $record, $biblionumber, $biblioitemnumber, '' ); };
             if ( $@ ) {
                 warn "ERROR: Adding items to bib $biblionumber failed: $@\n";
@@ -329,7 +331,49 @@ RECORD: while (  ) {
  			else{
 				printlog({id=>$id||$originalid||$biblionumber, op=>"insert",status=>"ok"}) if ($logfile);
 			}
-            if ($#{ $errors_ref } > -1) { 
+            if ($dedup_barcode && grep { exists $_->{error_code} && $_->{error_code} eq 'duplicate_barcode' } @$errors_ref) {
+                # Find the record called 'barcode'
+                my ($tag, $sub) = C4::Biblio::GetMarcFromKohaField('items.barcode', '');
+                # Now remove any items that didn't have a duplicate_barcode error,
+                # erase the barcodes on items that did, and re-add those items.
+                my %dupes;
+                foreach my $i (0 .. $#{$errors_ref}) {
+                    my $ref = $errors_ref->[$i];
+                    if ($ref && $ref->{error_code} eq 'duplicate_barcode') {
+                        $dupes{$ref->{item_sequence}} = 1;
+                        # Delete the error message because we're going to
+                        # retry this one.
+                        delete $errors_ref->[$i];
+                    }
+                }
+                my $seq = 0;
+                foreach my $field ($record->field($tag)) {
+                    $seq++;
+                    if ($dupes{$seq}) {
+                        # Here we remove the barcode
+                        $field->delete_subfield(code => $sub);
+                    } else {
+                        # otherwise we delete the field because we don't want
+                        # two of them
+                        $record->delete_fields($field);
+                    }
+                }
+                # Now re-add the record as before, adding errors to the prev list
+                my $more_errors;
+                eval { ( $itemnumbers_ref, $more_errors ) = AddItemBatchFromMarc( $record, $biblionumber, $biblioitemnumber, '' ); };
+                if ( $@ ) {
+                    warn "ERROR: Adding items to bib $biblionumber failed: $@\n";
+                    printlog({id=>$id||$originalid||$biblionumber, op=>"insertitem",status=>"ERROR"}) if ($logfile);
+                    # if we failed because of an exception, assume that
+                    # the MARC columns in biblioitems were not set.
+                    ModBiblioMarc( $record, $biblionumber, '' );
+                    next RECORD;
+                } else {
+                    printlog({id=>$id||$originalid||$biblionumber, op=>"insert",status=>"ok"}) if ($logfile);
+                }
+                push @$errors_ref, @{ $more_errors };
+            }
+            if ($#{ $errors_ref } > -1) {
                 report_item_errors($biblionumber, $errors_ref);
             }
         }
@@ -400,6 +444,7 @@ sub report_item_errors {
     my $errors_ref = shift;
 
     foreach my $error (@{ $errors_ref }) {
+        next if !$error;
         my $msg = "Item not added (bib $biblionumber, item tag #$error->{'item_sequence'}, barcode $error->{'item_barcode'}): ";
         my $error_code = $error->{'error_code'};
         $error_code =~ s/_/ /g;
@@ -536,6 +581,13 @@ Store ids in 009 (usefull for authorities, where 001 contains the authid for
 Koha, that can contain a very valuable info for authorities coming from LOC or
 BNF. useless for biblios probably)
 
+=item B<-dedupbarcodes>
+
+If set, whenever a duplicate barcode is detected, it is removed and the attempt
+to add the record is retried, thereby giving the record a blank barcode. This
+is useful when something has set barcodes to be a biblio ID, or similar
+(usually other software.)
+
 =back
 
 =cut
-- 
1.7.1