--- trunk/Fast.pm	2005/10/08 16:33:09	5
+++ trunk/Fast.pm	2007/06/21 10:24:12	17
@@ -1,5 +1,5 @@
-
 package MARC::Fast;
+
 use strict;
 use Carp;
 use Data::Dumper;
@@ -7,7 +7,7 @@
 BEGIN {
 	use Exporter ();
 	use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS);
-	$VERSION     = 0.02;
+	$VERSION     = 0.07;
 	@ISA         = qw (Exporter);
 	#Give a hoot don't pollute, do not export more than needed by default
 	@EXPORT      = qw ();
@@ -42,6 +42,11 @@
 	quiet => 0,
 	debug => 0,
 	assert => 0,
+	hash_filter => sub {
+		my ($t, $record_number) = @_;
+		$t =~ s/foo/bar/;
+		return $t;
+	},
   );
 
 =cut
@@ -59,6 +64,7 @@
 	print STDERR "# opening ",$self->{marcdb},"\n" if ($self->{debug});
 
 	open($self->{fh}, $self->{marcdb}) || croak "can't open ",$self->{marcdb},": $!";
+	binmode($self->{fh});
 
 	$self->{count} = 0;
 
@@ -69,7 +75,12 @@
 		push @{$self->{fh_offset}}, tell($self->{fh});
 
 		my $leader;
-		read($self->{fh}, $leader, 24);
+		my $len = read($self->{fh}, $leader, 24);
+
+		if ($len < 24) {
+			carp "short read of leader, aborting\n";
+			last;
+		}
 
 		# Byte        Name
 		# ----        ----
@@ -103,7 +114,12 @@
 		push @{$self->{leaders}}, $leader;
 
 		# skip to next record
-		seek($self->{fh},substr($leader,0,5)-24,1);
+		my $o = substr($leader,0,5);
+		if ($o > 24) {
+			seek($self->{fh},$o-24,1) if ($o);
+		} else {
+			last;
+		}
 
 	}
 
@@ -196,11 +212,7 @@
 		my $f = substr($fields,$addr,$len);
 		print STDERR "tag/len/addr $tag [$len] $addr: '$f'\n" if ($self->{debug});
 
-		if ($row->{$tag}) {
-			$row->{$tag} .= $f;
-		} else {
-			$row->{$tag} = $f;
-		}
+		push @{	$row->{$tag} }, $f;
 
 		my $del = substr($fields,$addr+$len-1,1);
 
@@ -221,16 +233,112 @@
 	return $row;
 }
 
-1;
-__END__
 
-=head1 BUGS
+=head2 to_hash
+
+Read record with specified MFN and convert it to hash
+
+  my $hash = $marc->to_hash($mfn);
+
+It has ability to convert characters (using C<hash_filter>) from MARC
+database before creating structures enabling character re-mapping or quick
+fix-up of data.
+
+This function returns hash which is like this:
+
+  '200' => [
+             {
+               'i1' => '1',
+               'i2' => ' '
+               'a' => 'Goa',
+               'f' => 'Valdo D\'Arienzo',
+               'e' => 'tipografie e tipografi nel XVI secolo',
+             }
+           ],
+
+This method will also create additional field C<000> with MFN.
+
+=cut
+
+sub to_hash {
+	my $self = shift;
 
+	my $mfn = shift || confess "need mfn!";
 
+	# init record to include MFN as field 000
+	my $rec = { '000' => [ $mfn ] };
 
-=head1 SUPPORT
+	my $row = $self->fetch($mfn) || return;
 
+	foreach my $rec_nr (keys %{$row}) {
+		foreach my $l (@{$row->{$rec_nr}}) {
+
+			# remove end marker
+			$l =~ s/\x1E$//;
+
+			# filter output
+			$l = $self->{'hash_filter'}->($l, $rec_nr) if ($self->{'hash_filter'});
+
+			my $val;
+
+			# has identifiers?
+			($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\x1F/\x1F/);
+
+			# has subfields?
+			if ($l =~ m/\x1F/) {
+				foreach my $t (split(/\x1F/,$l)) {
+					next if (! $t);
+					my $f = substr($t,0,1);
+					# repeatable subfileds. When we hit first one,
+					# store CURRENT (up to that) in first repetition
+					# of this record. Then, new record with same
+					# identifiers will be created.
+					if ($val->{$f}) {
+						push @{$rec->{$rec_nr}}, $val;
+						$val = {
+							i1 => $val->{i1},
+							i2 => $val->{i2},
+						};
+					}
+					$val->{substr($t,0,1)} = substr($t,1);
+				}
+			} else {
+				$val = $l;
+			}
 
+			push @{$rec->{$rec_nr}}, $val;
+		}
+	}
+
+	return $rec;
+}
+
+=head2 to_ascii
+
+  print $marc->to_ascii( 42 );
+
+=cut
+
+sub to_ascii {
+	my $self = shift;
+
+	my $mfn = shift || confess "need mfn";
+	my $row = $self->fetch($mfn) || return;
+
+	my $out;
+
+	foreach my $f (sort keys %{$row}) {
+		my $dump = join('', @{ $row->{$f} });
+		$dump =~ s/\x1e$//;
+		$dump =~ s/\x1f/\$/g;
+		$out .= "$f\t$dump\n";
+	}
+
+	return $out;
+}
+
+1;
+__END__
 
 =head1 AUTHOR
 
@@ -250,6 +358,6 @@
 
 =head1 SEE ALSO
 
-perl(1).
+L<Biblio::Isis>, perl(1).
 
 =cut