From: Dobrica Pavlinusic <dpavlin@rot13.org>
Date: Sun, 18 Dec 2005 23:12:26 +0000 (+0000)
Subject: added to_hash, small fix to test, better output in dump_fastmarc.pl [0.02]
X-Git-Url: http://git.rot13.org/?p=MARC-Fast;a=commitdiff_plain;h=8300e4b53e83d8a355dba26d74624d521f67232f

added to_hash, small fix to test, better output in dump_fastmarc.pl [0.02]


git-svn-id: svn+ssh://llin/home/dpavlin/private/svn/MARC-Fast/trunk@6 49f9634a-d7ec-0310-8e6b-ec35c6cc8804
---

diff --git a/Fast.pm b/Fast.pm
index 9a981be..6116b14 100644
--- a/Fast.pm
+++ b/Fast.pm
@@ -59,6 +59,7 @@ sub new {
 	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 +70,12 @@ sub new {
 		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 +109,12 @@ sub new {
 		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 +207,7 @@ sub fetch {
 		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);
 
@@ -221,6 +228,75 @@ sub fetch {
 	return $row;
 }
 
+
+=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 ] };
+
+	my $row = $self->fetch($mfn) || return;
+
+	foreach my $k (keys %{$row}) {
+		foreach my $l (@{$row->{$k}}) {
+
+			# remove end marker
+			$l =~ s/\x1E$//;
+
+			# filter output
+			$l = $self->{'hash_filter'}->($l) 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);
+					$val->{substr($t,0,1)} = substr($t,1);
+				}
+			} else {
+				$val = $l;
+			}
+
+			push @{$rec->{$k}}, $val;
+		}
+	}
+
+	return $rec;
+}
+
+
 1;
 __END__
 
diff --git a/META.yml b/META.yml
index 9a928f9..468eebc 100644
--- a/META.yml
+++ b/META.yml
@@ -1,7 +1,7 @@
 # http://module-build.sourceforge.net/META-spec.html
 #XXXXXXX This is a prototype!!!  It will change in the future!!! XXXXX#
 name:         MARC-Fast
-version:      0.01
+version:      0.02
 version_from: Fast.pm
 installdirs:  site
 requires:
diff --git a/scripts/dump_fastmarc.pl b/scripts/dump_fastmarc.pl
index a37171c..538f8ea 100755
--- a/scripts/dump_fastmarc.pl
+++ b/scripts/dump_fastmarc.pl
@@ -17,22 +17,24 @@ my $marc = new MARC::Fast(
 	debug => $opt{'d'},
 );
 
-print STDERR "$file has ",$marc->count," records...\n";
 
 my $min = 1;
 my $max = $marc->count;
 
 if (my $mfn = $opt{'n'}) {
 	$min = $max = $mfn;
+	print STDERR "Dumping $mfn only\n";
+} else {
+	print STDERR "$file has $max records...\n";
 }
 
-print STDERR "Dumping $min - $max\n" if ($opt{'d'});
-
 for my $mfn ($min .. $max) {
 	my $rec = $marc->fetch($mfn) || next;
+	print Dumper($rec);
 	print "REC $mfn\n";
 	foreach my $f (sort keys %{$rec}) {
-		print "$f\t",$rec->{$f},"\n";
+		print "$f\t", join('', $rec->{$f}) ,"\n";
 	}
 	print "\n";
+	print Dumper($marc->to_hash($mfn));
 }
diff --git a/t/001_marc.t b/t/001_marc.t
index 60254b5..691a3a2 100755
--- a/t/001_marc.t
+++ b/t/001_marc.t
@@ -21,7 +21,7 @@ throws_ok { $marc = MARC::Fast->new(%param); } qr/foo.bar/, "marcdb exist";
 $param{marcdb} = 'data/unimarc.iso';
 
 SKIP: {
-	skip "no ",$param{marcdb}," test file ", 18 unless (-e $param{marcdb});
+	skip "no $param{marcdb} test file ", 17 unless (-e $param{marcdb});
 
 	ok($marc = MARC::Fast->new(%param), "new");