open FP, '<', $ARGV[0] or die "Can't open input file: $!\n";
-my $count = 0;
-my $i = 0;
-my ($total) = split /s+/, `wc -l $ARGV[0]`;
-$| = 1;
-
print "Loading and ranking fingerprints\n";
while (<FP>) {
my @fields = split "\t", $_;
my $fp = populate_fingerprint(@fields);
rank_fingerprint($fp);
- $i++;
print "\r", ( int($i / $total) ), "% complete" unless ($i % 1000);
}
-print "$total fingerprints processed\n";
-print "$count records set as leads\n";
print "Writing matchset to disk\n";
dump_records();
score => $fp->{compact} },
recs => [ $id ] };
$recs{$id}{lead} = 1;
- $count++;
} else {
# have seen this fp. push record id onto matchlist
push @{ $fps{$sha1}{recs} }, $id;