fix retrieval of correct column headers and data from non-holdings tags

[migration-tools.git] / Equinox-Migration / lib / Equinox / Migration / MapDrivenMARCXMLProc.pm

diff --git a/Equinox-Migration/lib/Equinox/Migration/MapDrivenMARCXMLProc.pm b/Equinox-Migration/lib/Equinox/Migration/MapDrivenMARCXMLProc.pm
index 194a593..77a96dc 100644 (file)
--- a/Equinox-Migration/lib/Equinox/Migration/MapDrivenMARCXMLProc.pm
+++ b/Equinox-Migration/lib/Equinox/Migration/MapDrivenMARCXMLProc.pm
@@ -4,7 +4,6 @@ use warnings;
 use strict;
 
 use XML::Twig;
-use DBM::Deep;
 use Equinox::Migration::SubfieldMapper 1.004;
 
 
@@ -14,15 +13,16 @@ Equinox::Migration::MapDrivenMARCXMLProc
 
 =head1 VERSION
 
-Version 1.003
+Version 1.005
 
 =cut
 
-our $VERSION = '1.003';
+our $VERSION = '1.005';
 
 my $dstore;
 my $sfmap;
-my @mods = qw( multi bib required );
+my @modlist = qw( multi ignoremulti bib required first concatenate );
+my %allmods = ();
 my $multis = {};
 my $reccount;
 my $verbose = 0;
@@ -60,16 +60,10 @@ sub new {
     # initialize map and taglist
     die "Argument 'mapfile' must be specified\n" unless ($args{mapfile});
     $sfmap = Equinox::Migration::SubfieldMapper->new( file => $args{mapfile},
-                                                      mods => \@mods );
+                                                      mods => \@modlist );
 
     # initialize datastore
-    die "Datastore file 'EMMXSSTORAGE.dbmd' already exists. Exiting.\n"
-      if (-e "EMMXSSTORAGE.dbmd");
-    $dstore = DBM::Deep->new( file => "EMMXSSTORAGE.dbmd",
-                              max_buckets => 64,
-                              #data_sector_size => 256,
-                              autoflush => 0,
-                            );
+    $dstore = {};
     $reccount = 0;            # next record ptr
     $dstore->{tags} = $sfmap->tags; # list of all tags
     $self->{data} = $dstore;
@@ -86,8 +80,6 @@ sub new {
     return $self;
 }
 
-sub DESTROY { unlink "EMMXSSTORAGE.dbmd" }
-
 =head2 parse_record
 
 Extracts data from the next record, per the mapping file.
@@ -102,12 +94,28 @@ sub parse_record {
     for my $f (@fields)
       { process_field($f, $crec) }
 
+    # fill in blank values if needed
+    for my $mappedtag ( @{ $sfmap->tags }) {
+        unless (exists $crec->{tmap}{$mappedtag}) {
+            push @{ $crec->{tags} }, {};
+            for my $mappedsub ( @{ $sfmap->subfields($mappedtag) } ) {
+                my $fieldname = $sfmap->field($mappedtag, $mappedsub);
+                my $mods = $sfmap->mods($fieldname);
+                next if $mods->{multi};
+                $crec->{tags}[-1]{uni}{$mappedsub} = '';
+                $crec->{tags}[-1]{multi} = undef;
+                $crec->{tags}[-1]{tag} = $mappedtag;
+            }
+            push @{ $crec->{tmap}{$mappedtag} }, $#{ $crec->{tags} };
+        }
+    }
+
     # cleanup memory and increment pointer
     $record->purge;
     $reccount++;
 
     # check for required fields
-    check_required();
+    check_required($crec);
     push @{ $dstore->{recs} }, $crec;
 
     print STDERR "$reccount\n"
@@ -145,10 +153,6 @@ sub process_field {
             $crec->{tags}[-1]{uni}{$mappedsub} = ''
               unless defined $crec->{tags}[-1]{uni}{$mappedsub};
         }
-        for my $mappedtag ( @{ $sfmap->tags }) {
-            $crec->{tmap}{$mappedtag} = undef
-              unless defined $crec->{tmap}{$mappedtag};
-        }
     }
 }
 
@@ -159,9 +163,12 @@ sub process_subs {
     # handle unmapped tag/subs
     return unless ($sfmap->has($tag, $code));
 
-    # fetch our datafield struct and fiel
+    # fetch our datafield struct and field and mods
     my $dataf = $crec->{tags}[-1];
     my $field = $sfmap->field($tag, $code);
+    my $sep = $sfmap->sep($field);
+    $allmods{$field} = $sfmap->mods($field) unless $allmods{$field};
+    my $mods = $allmods{$field};
 
     # test filters
     for my $filter ( @{$sfmap->filters($field)} ) {
@@ -169,17 +176,34 @@ sub process_subs {
     }
 
     # handle multi modifier
-    if (my $mods = $sfmap->mods($field)) {
-        if ($mods->{multi}) {
-            push @{$dataf->{multi}{$code}}, $sub->text;
+    if ($mods->{multi}) {
+        $multis->{$tag}{$code} = 1;
+        if ($mods->{concatenate}) {
+            if (exists($dataf->{multi}{$code})) {
+                $dataf->{multi}{$code}[0] .= $sep . $sub->text;
+            } else {
+                push @{$dataf->{multi}{$code}}, $sub->text;
+            }
             $multis->{$tag}{$code} = 1;
-            return;
+        } else {
+            push @{$dataf->{multi}{$code}}, $sub->text;
         }
+        return;
+    }
+
+
+    if ($mods->{concatenate}) {
+        if (exists($dataf->{uni}{$code})) {
+            $dataf->{uni}{$code} .= $sep . $sub->text;
+        } else {
+            $dataf->{uni}{$code} = $sub->text;
+        }
+        return;
     }
 
     # if this were a multi field, it would be handled already. make sure its a singleton
     die "Multiple occurances of a non-multi field: $tag$code at rec ",
-      ($reccount + 1),"\n" if (defined $dataf->{uni}{$code});
+      ($reccount + 1),"\n" if (defined $dataf->{uni}{$code} and !$mods->{ignoremulti});
 
     # everything seems okay
     $dataf->{uni}{$code} = $sub->text;
@@ -187,8 +211,8 @@ sub process_subs {
 
 
 sub check_required {
+    my ($crec) = @_;
     my $mods = $sfmap->mods;
-    my $crec = $dstore->{crec};
 
     for my $tag_id (keys %{$mods->{required}}) {
         for my $code (@{$mods->{required}{$tag_id}}) {
@@ -224,6 +248,20 @@ Returns mapped fieldname when passed a tag, and code
 
 sub name { my ($self, $t, $c) = @_; return $sfmap->field($t, $c) }
 
+=head2 first_only
+
+Returns whether mapped fieldname is to be applied only to first
+item in a bib
+
+=cut
+
+sub first_only {
+    my ($self, $t, $c) = @_;
+    my $field = $sfmap->field($t, $c);
+    my $mods = $sfmap->mods($field);
+    return exists($mods->{first});
+}
+
 =head2 get_multis
 
 Returns hashref of C<{tag}{code}> for all mapped multi fields