functionally removing case b from the matchsets

[migration-tools.git] / fingerprinter

diff --git a/fingerprinter b/fingerprinter
index 71376b9..b87c59a 100755 (executable)
--- a/fingerprinter
+++ b/fingerprinter
@@ -26,7 +26,7 @@ for my $file (@ARGV) {
     my $batch = undef; my $record = undef;
 
     #$batch = MARC::Batch->new('XML', $records); # The other part of the hack
-    $batch = MARC::Batch->new('XML', $file);
+    $batch = MARC::Batch->new($conf->{marctype}, $file);
     $batch->strict_off();
     $batch->warnings_off();
 
@@ -80,6 +80,15 @@ sub populate_marc {
     $marc{date1} = substr($my_008,7,4) if ($my_008);
     $marc{date2} = substr($my_008,11,4) if ($my_008); # UNUSED
 
+    unless ($marc{date1} and $marc{date1} =~ /\d{4}/) {
+        my $my_260 = $record->field('260');
+        my $date1 = $my_260->subfield('c') if $my_260;
+        if (defined $date1 and $date1 =~ /\d{4}/) {
+            $marc{date1} = $date1;
+            print XF ">> using 260c as date1 on $id\n";
+        }
+    }
+
     # item_form
     if ( $marc{record_type} =~ /[gkroef]/ ) { # MAP, VIS
         $marc{item_form} = substr($my_008,29,1) if ($my_008);
@@ -204,6 +213,13 @@ sub dump_fingerprints {
             }
         }
 
+
+=pod
+
+Looks like case b isn't a terribly good idea most of the time -- it
+will generate many spurious matches in series books in particular. So
+for now, simply turning it off.
+
         if ($marc->{edition}) { # case b : edition
             print OF join("\t", $marc->{id}, "case b",
                               $marc->{item_form}, $marc->{date1},
@@ -211,6 +227,8 @@ sub dump_fingerprints {
                               $marc->{title}, $marc->{edition}), "\n";
         }
 
+=cut
+
         if ($marc->{issn}) { # case c : issn
             print OF join("\t", $marc->{id}, "case c",
                               $marc->{item_form}, $marc->{date1},