--- trunk/lib/WebPAC/Normalize.pm	2006/01/08 20:27:11	364
+++ trunk/lib/WebPAC/Normalize.pm	2006/01/08 21:16:27	371
@@ -2,6 +2,8 @@
 
 use warnings;
 use strict;
+use blib;
+use WebPAC::Common;
 use base 'WebPAC::Common';
 use Data::Dumper;
 
@@ -306,7 +308,7 @@
 sub parse {
 	my $self = shift;
 
-	my ($rec, $format_utf8, $i) = @_;
+	my ($rec, $format_utf8, $i, $rec_size) = @_;
 
 	return if (! $format_utf8);
 
@@ -355,7 +357,7 @@
 		}
 
 		my $found = 0;
-		my $tmp = $self->get_data(\$rec,$3,$4,$r,\$found);
+		my $tmp = $self->get_data(\$rec,$3,$4,$r,\$found,$rec_size);
 
 		if ($found) {
 			$found_any->{$fld_type} += $found;
@@ -429,8 +431,11 @@
 	my $i = 0;
 	my @arr;
 
-	while (my $v = $self->parse($rec,$format_utf8,$i++)) {
+	my $rec_size = { '_' => '_' };
+
+	while (my $v = $self->parse($rec,$format_utf8,$i++,\$rec_size)) {
 		push @arr, $v;
+		warn "parse rec_size = ", Dumper($rec_size);
 	}
 
 	$log->debug("format '$format_utf8' returned ",--$i," elements: ", sub { join(" | ",@arr) }) if (@arr);
@@ -460,6 +465,11 @@
 This method will automatically decode UTF-8 string to local code page
 if needed.
 
+There is optional parametar C<$record_size> which can be used to get sizes of
+all C<field^subfield> combinations in this format.
+
+ my $text = $webpac->fill_in($rec,'got: v900^a v900^x',0,\$rec_size); 
+
 =cut
 
 sub fill_in {
@@ -467,10 +477,13 @@
 
 	my $log = $self->_get_logger();
 
-	my $rec = shift || $log->logconfess("need data record");
-	my $format = shift || $log->logconfess("need format to parse");
+	my ($rec,$format,$i,$rec_size) = @_;
+
+	$log->logconfess("need data record") unless ($rec);
+	$log->logconfess("need format to parse") unless($format);
+
 	# iteration (for repeatable fields)
-	my $i = shift || 0;
+	$i ||= 0;
 
 	$log->logdie("infitite loop in format $format") if ($i > ($self->{'max_mfn'} || 9999));
 
@@ -494,12 +507,12 @@
 
 	# do actual replacement of placeholders
 	# repeatable fields
-	if ($format =~ s/v(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,$i,\$found)/ges) {
+	if ($format =~ s/v(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,$i,\$found,$rec_size)/ges) {
 		$just_single = 0;
 	}
 
 	# non-repeatable fields
-	if ($format =~ s/s(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,0,\$found)/ges) {
+	if ($format =~ s/s(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,0,\$found,$rec_size)/ges) {
 		return if ($i > 0 && $just_single);
 	}
 
@@ -556,8 +569,11 @@
 	my $i = 0;
 	my @arr;
 
-	while (my $v = $self->fill_in($rec,$format_utf8,$i++)) {
+	my $rec_size;
+
+	while (my $v = $self->fill_in($rec,$format_utf8,$i,\$rec_size)) {
 		push @arr, $v;
+		warn "rec_size = ", Dumper($rec_size);
 	}
 
 	$log->debug("format '$format_utf8' returned ",--$i," elements: ", sub { join(" | ",@arr) }) if (@arr);
@@ -570,51 +586,76 @@
 
 Returns value from record.
 
- my $text = $self->get_data(\$rec,$f,$sf,$i,\$found);
+ my $text = $self->get_data(\$rec,$f,$sf,$i,\$found,\$rec_size);
+
+Required arguments are:
+
+=over 8
+
+=item C<$rec>
+
+record reference
+
+=item C<$f>
+
+field
+
+=item C<$sf>
+
+optional subfield
 
-Arguments are:
-record reference C<$rec>,
-field C<$f>,
-optional subfiled C<$sf>,
-index for repeatable values C<$i>.
+=item C<$i>
 
-Optinal variable C<$found> will be incremeted if there
-is field.
+index offset for repeatable values ( 0 ... $rec_size->{'400^a'} )
 
-Returns value or empty string.
+=item C<$found>
+
+optional variable that will be incremeted if preset
+
+=item C<$rec_size>
+
+hash to hold maximum occurances of C<field^subfield> combinations
+(which can be accessed using keys in same format)
+
+=back
+
+Returns value or empty string, updates C<$found> and C<rec_size>
+if present.
 
 =cut
 
 sub get_data {
 	my $self = shift;
 
-	my ($rec,$f,$sf,$i,$found) = @_;
+	my ($rec,$f,$sf,$i,$found,$cache) = @_;
+
+	return '' unless ($$rec->{$f} && ref($$rec->{$f}) eq 'ARRAY');
 
-	if ($$rec->{$f}) {
-		return '' if (! $$rec->{$f}->[$i]);
+	if (defined($$cache)) {
+		$$cache->{ $f . ( $sf ? '^' . $sf : '' ) } ||= scalar @{ $$rec->{$f} };
+	}
+
+	return '' unless ($$rec->{$f}->[$i]);
+
+	{
 		no strict 'refs';
-		if ($sf && $$rec->{$f}->[$i]->{$sf}) {
-			$$found++ if (defined($$found));
+		if (defined($sf)) {
+			$$found++ if (defined($$found) && $$rec->{$f}->[$i]->{$sf});
 			return $$rec->{$f}->[$i]->{$sf};
-		} elsif (! $sf && $$rec->{$f}->[$i]) {
+		} else {
 			$$found++ if (defined($$found));
-			# it still might have subfield, just
-			# not specified, so we'll dump all
+			# it still might have subfields, just
+			# not specified, so we'll dump some debug info
 			if ($$rec->{$f}->[$i] =~ /HASH/o) {
 				my $out;
 				foreach my $k (keys %{$$rec->{$f}->[$i]}) {
-					my $v = $$rec->{$f}->[$i]->{$k};
-					$out .= "$v " if ($v);
+					$out .= '$' . $k .':' . $$rec->{$f}->[$i]->{$k}." ";
 				}
 				return $out;
 			} else {
 				return $$rec->{$f}->[$i];
 			}
-		} else {
-			return '';
 		}
-	} else {
-		return '';
 	}
 }