my $sth = $dbh->prepare("INSERT INTO HLDdata (Data) VALUES (?)");
my $count = $dbh->prepare( "select count(*) from HLDdata where Data =
+?");
foreach $txt (@TXT) {
print "$txt\n";
open(IN,"$txt") || warn("cant open $txt");
while(<IN>){
$line = $_;
$count->execute($line);
my ($data) = $count->fetchrow();
print "$data \n";
if ($data != 0){
print "record exist not adding\n";
Logit("record exist not adding");
next;
}
my ($Data) = $line;
$sth->execute($Data);
$line =~s/\|/\t/g;
# print "$line\n";
print OUT $line;
}
}
Yes, you're probably correct that the checking is taking the most of the time. This is because you're missing an index on HLDdata (data) -- sqlite has to scan the whole table every time you check for a row's existence. You can create the missing index with CREATE UNIQUE INDEX HLDdata_uniq ON HLDdata (Data);
There's a trade-off: you'll now use double the disk space. If $Data tends to be big (above a hundred bytes or so), you are better off storing a hash (sha1, md5, or similar) of $Data in the table in a second column and checking for its existence instead. (SQLite does not appear to have support for functional indices or a hashing function, so you'll have to do it on the Perl side.) |