I think by now one of the other responses may have resolved your problem. My concern here is that your solution doesn't make much sense from a performance perspective. If I absolutely HAD to read from a file, the first thing I would probably do in your case is something like this:

use DBI;
my $pathname = qq|webdb.db.db3|;
my $cnstr = qq|DBI:SQLite:dbname=$pathname|;

my $filename = q|afile.txt|;

open(H, qq|afile.txt|) or die $!;
my @lines = <H>;
close(H);

my $IN = '(';
$IN .= join  ',', map { qq|\'$_\'|} grep {chomp} @lines;
$IN = 'select webpage from webpages_data WHERE id in ' . $IN . ')';
print $IN;
#produces : select webpage from webpages_data WHERE id in ('koko.htm',
+'koko1.htm','koko2.htm','koko3.htm')
[download]

This will reduce the number of separate calls to the database. Second, I simply do not know why in the world you would reduce the performance of your solution to reading a text file when you already have a database connection open. Just create another table with your indexes and join the two, like:

select webpage from webpages_data, mySelectedStuff WHERE webpage.id = 
+mySelectedStuff.id AND ...
[download]

Or a sub-select cannot have much worse performance:

select webpage from webpages_data where webpage in (select id from myS
+electedStuff WHERE ...)
[download]