--- trunk/tools/mods2unimarc.pl	2004/05/14 17:22:39	326
+++ trunk/tools/mods2unimarc.pl	2004/05/15 18:54:41	327
@@ -24,9 +24,15 @@
 
 Feel free to hack this script and convert it to your own needs.
 
-=head1 WARNING
+=head1 CAVEAT
 
-This script is in state of flux.
+This script will parse imput XML twice: once with C<XML::Twig> and
+then each entry with C<XML::Simple> to produce in-memory structure.
+That's because I wanted to keep node selection logical (and perl-like).
+
+If you don't like it, you can rewrite this script to use XPATH. I tried
+and failed (it seems that MODS is too complicated for my limited knowledge
+of XPATH).
 
 =cut
 
@@ -40,6 +46,7 @@
 
 my $xml_file = "/data/tehnika/fer/all.xml";
 $xml_file = "/data/tehnika/fer/modsFER_1.xml";
+$xml_file = "/data/tehnika/fer/mods-small.xml";
 my $marc_file = "fer.marc";
 
 $|=1;
@@ -47,33 +54,35 @@
 
 my $marc = MARC->new;
 
+my $ENCODING = 'ISO-8859-2';
+
 my $twig=XML::Twig->new(
-	twig_roots => { 'mods' => \&item },
-	output_encoding => 'iso-8859-2',
+	twig_roots => { 'mods' => \&mods },
+	output_encoding => 'UTF8',
 );
 
-my $utf2iso = Text::Iconv->new("UTF-8", "ISO-8859-2");
+my $utf2iso = Text::Iconv->new("UTF8", $ENCODING);
 
+print "$xml_file: ";
 $twig->parsefile($xml_file);
 $twig->purge;
+print "$nr\nSaving MARC file...\n";
 
 $marc->output({file=>"> $marc_file",'format'=>"usmarc"});
 
-sub item {
+sub mods {
 	my( $t, $elt)= @_;
 
 	my $xml=$elt->xml_string;
-	my $ref = XMLin("<xml>".$xml."</xml>", 
+	my $ref = XMLin('<xml>'.$xml.'</xml>', 
 		ForceArray => [
 			'name',
 			'classification',
 			'topic',
-			'udc',
 		],
 		KeyAttr => {
 			'namePart' => 'type',
 			'identifier' => 'type',
-			'classification' => 'authority',
 			'namePart' => 'type',
 			'role' => 'type',
 		},
@@ -88,49 +97,95 @@
 	my $m_cache;
 
 	sub marc_add {
-		my $m_cache = \shift;
+		my $m_cache = \shift || die "need m_cache";
 		my $fld = shift || die "need field!";
-		my $sf = shift || '';
+		my $sf = shift;
+		my $data = shift || return;
 
-		return if (! @_);
+#print "add: $fld",($sf ? "^".$sf : ''),": $data\n";
 
-		my @a;
-		foreach (@_) {
-			next if (! $_);
-			push @a,$sf if ($sf);
-#			push @a,$utf2iso->convert($_) || $_;
-			push @a,$_;
+		if ($sf) {
+			push @{$$m_cache->{tmp}->{$fld}}, $sf;
 		}
+		push @{$$m_cache->{tmp}->{$fld}}, $utf2iso->convert($data);
+	}
 
-		return if (! @a);
+	sub marc_rep {
+		my $m_cache = \shift || die "need m_cache";
+		foreach my $fld (@_) {
+#print "marc_rep: $fld\n";	
+			push @{$$m_cache->{array}->{$fld}}, [ @{$$m_cache->{tmp}->{$fld}} ] if ($$m_cache->{tmp}->{$fld});
+			delete $$m_cache->{tmp}->{$fld};
+		}
+	}
 
-#		print "storing $fld: ",join("|",@a),"\n";
+	sub marc_single {
+		my $m_cache = \shift || die "need m_cache";
+		foreach my $fld (@_) {
+#print "marc_single: $fld\n";	
 
-		push @{$$m_cache->{$fld}}, @a;
+			die "$fld already defined! not single?" if ($$m_cache->{single}->{$fld});
 
+			$$m_cache->{single}->{$fld} = \@{$$m_cache->{tmp}->{$fld}} if ($$m_cache->{tmp}->{$fld});
+			delete $$m_cache->{tmp}->{$fld};
+		}
+	}
+
+	sub marc_add_rep {
+		my $m_cache = \shift || die "need m_cache";
+		my $fld = shift || die "need field!";
+		my $sf = shift;
+		my $data = shift || return;
+
+		marc_add($$m_cache,$fld,$sf,$data);
+		marc_rep($$m_cache,$fld);
+	}
+
+	sub marc_add_single {
+		my $m_cache = \shift || die "need m_cache";
+		my $fld = shift || die "need field!";
+		my $sf = shift;
+		my $data = shift || return;
+
+		marc_add($$m_cache,$fld,$sf,$data);
+		marc_single($$m_cache,$fld);
 	}
 
 	my $journal = 0;
+	# Journals start with c- in our MODS
 	$journal = 1 if ($ref->{recordInfo}->{recordIdentifier} =~ m/^c-/);
-	
-	marc_add($m_cache,'610','a',@{$ref->{subject}->{topic}});
 
-	my $fld = '700';
+	foreach my $t (@{$ref->{subject}->{topic}}) {
+		marc_add($m_cache,'610','a', $t);
+		marc_rep($m_cache,'610');
+	}
+
+	my $fld_700 = '700';
+	my $fld_710 = '710';
 
 	foreach my $name (@{$ref->{name}}) {
 		my $role = $name->{role}->{roleTerm}->{content};
 		next if (! $role);
 		if ($role eq "author") {
-			marc_add($m_cache,$fld,'a',$name->{namePart}->{family});
-			marc_add($m_cache,$fld,'b',$name->{namePart}->{given});
-			marc_add($m_cache,$fld,'4',$role);
+			marc_add($m_cache,$fld_700,'a',$name->{namePart}->{family});
+			marc_add($m_cache,$fld_700,'b',$name->{namePart}->{given});
+			marc_add($m_cache,$fld_700,'4',$role);
+
+			marc_rep($m_cache,$fld_700);
 
 			# first author goes in 700, others in 701
-			$fld = '701';
+			$fld_700 = '701';
 		} elsif ($role eq "editor" or $role eq "illustrator") {
 			marc_add($m_cache,'702','a',$name->{namePart}->{family});
 			marc_add($m_cache,'702','b',$name->{namePart}->{given});
 			marc_add($m_cache,'702','4',$role);
+			marc_rep($m_cache,'702');
+		} elsif ($role eq "corporate") {
+			marc_add_single($m_cache,"$fld_710\t0 ",'a',$name->{namePart});
+			$fld_710 = '711';
+		} elsif ($role eq "conference") {
+			marc_add_single($m_cache,"$fld_710\t1 ",'a',$name->{namePart});
+			$fld_710 = '711';
 		} else {
 			die "FATAL: don't know how to map role '$role'" if ($role);
 		}
@@ -141,23 +196,23 @@
 	if ($note) {
 		foreach my $n (split(/\s*;\s+/, $note)) {
 			if ($n =~ s/bibliogr:\s+//i) {
-				marc_add($m_cache,'320','a',"Bibliografija: $n");
+				marc_add_rep($m_cache,'320','a',"Bibliografija: $n");
 			} elsif ($n =~ s/ilustr:\s+//i) {
 				marc_add($m_cache,'215','c', $n);
 			} else {
-				marc_add($m_cache,'320','a',$n);
+				marc_add_rep($m_cache,'320','a',$n);
 			}
 		}
 	}
-			
+
 
 	my $type = $ref->{identifier}->{type};
 
 	if ($type) {
 		if ($type eq "isbn") {
-			marc_add($m_cache,'010','a',$ref->{identifier}->{content});
+			marc_add_rep($m_cache,'010','a',$ref->{identifier}->{content});
 		} elsif ($type eq "issn") {
-			marc_add($m_cache,'011','a',$ref->{identifier}->{content});
+			marc_add_rep($m_cache,'011','a',$ref->{identifier}->{content});
 		} else {
 			die "unknown identifier type $type";
 		}
@@ -181,27 +236,34 @@
 		marc_add($m_cache,'215','a', $data) if ($data);
 		marc_add($m_cache,'215','d', $tmp->{visina});
 	}
+	marc_rep($m_cache,'215');
 
-	marc_add($m_cache,'001',undef,$ref->{recordInfo}->{recordIdentifier});
+	marc_add_single($m_cache,'001',undef,$ref->{recordInfo}->{recordIdentifier});
 
 	marc_add($m_cache,'200','a',$ref->{titleInfo}->{title});
 	marc_add($m_cache,'200','e',$ref->{titleInfo}->{subTitle});
+	marc_single($m_cache,'200');
 
-	marc_add($m_cache,'675','a',$ref->{classification}->{udc});
+	foreach my $c (@{$ref->{classification}}) {
+		if ($c->{'authority'} eq "udc") {
+			marc_add_rep($m_cache,'675','a', $c->{'content'});
+		}
+	}
 
 	my $related = $ref->{relatedItem}->{type};
 	if ($related) {
 		if ($related eq "series") {
 			marc_add($m_cache,'225','a',$ref->{relatedItem}->{titleInfo}->{title});
 			marc_add($m_cache,'999','a',$ref->{relatedItem}->{titleInfo}->{partNumber});
+			marc_rep($m_cache,'225','999');
 		} elsif ($related eq "preceding") {
-			marc_add($m_cache,'430','a',$ref->{relatedItem}->{titleInfo}->{title});
+			marc_add_rep($m_cache,'430','a',$ref->{relatedItem}->{titleInfo}->{title});
 		} else {
 			die "can't parse related item type $related" if ($related);
 		}
 	}
 
-	marc_add($m_cache,'205','a',$ref->{originInfo}->{edition});
+	marc_add_single($m_cache,'205','a',$ref->{originInfo}->{edition});
 
 	my $publisher = $ref->{originInfo}->{publisher};
 	if ($publisher =~ m,^(.+?)\s*/\s*(.+)$,) {
@@ -211,25 +273,55 @@
 		marc_add($m_cache,'210','c', $publisher);
 	}
 
-	marc_add($m_cache,'326','a',$ref->{originInfo}->{frequency}) if ($journal);
-
 	marc_add($m_cache,'210','a',$ref->{originInfo}->{place});
-
 	marc_add($m_cache,'210','d',$ref->{originInfo}->{dateIssued});
 
+	marc_single($m_cache,'210');
+
+	marc_add_single($m_cache,'326','a',$ref->{originInfo}->{frequency}) if ($journal);
+
 	$nr++;
 	print "$nr " if ($nr % 100 == 0);
 
 	# dump record
-	my $m=$marc->createrecord();
-	foreach my $fld (keys %{$m_cache}) {
-#		print "$fld: ",join(" * ",@{$m_cache->{$fld}}),"\n";
+	my $m=$marc->createrecord({leader=>"00000nam  2200000 a 4500"});
+
+	foreach my $fld (keys %{$m_cache->{array}}) {
+		foreach my $arr (@{$m_cache->{array}->{$fld}}) {
+#print "array = ",Dumper($arr);
+			my ($i1,$i2);
+			# do we have indicators?
+			if ($fld =~ m/^(.+)\t(.)(.)$/) {
+				$fld = $1;
+				($i1,$i2) = ($2,$3);
+			}
+			$marc->addfield({record=>$m,
+				field=>$fld,
+				i1=>$i1,
+				i2=>$i2,
+				value=>$arr
+			});
+		}
+	}
+
+	foreach my $fld (keys %{$m_cache->{single}}) {
+#print "single = ",Dumper($m_cache->{single}->{$fld});
+		my ($i1,$i2);
+		# do we have indicators?
+		if ($fld =~ m/^(.+)\t(.)(.)$/) {
+			$fld = $1;
+			($i1,$i2) = ($2,$3);
+		}
 		$marc->addfield({record=>$m,
 			field=>$fld,
-			value=>\@{$m_cache->{$fld}}
+			i1=>$i1,
+			i2=>$i2,
+			value=>$m_cache->{single}->{$fld}
 		});
 	}
 
+	$m_cache = {};
+
 	$t->purge;           # frees the memory
 }