it would be most useful to see your code, as well as some samples of your input and desired output.

As a general rule, when you're looking for records (or parts of records) that are in file 2 and not in file 1, the best method is to read file 1 and to store the data of interest of file 1 in a hash; and then to scan file 2 and keep the records not matching the hash. BTW, a hash will remove the duplicates.

Not knowing anything about your input files (e.g., what is the record separator?), I can't suggest any real code. But the basic idea in pseudo-code is:

Open file 1 and read file 1 record by record, and:
for (each record) {
    get column D from the record;
    get the interesting part from column D;
    store the interesting part in a hash: some thing like: $hash{inter
+esting part} = 1;
}
close file 1
open file 2 and read file 2 record by record, and:
for (each record) {
    get column D from the record;
    get the interesting part from column D;
    if (exists $hash{interesting part}) {
        print $hash{interesting part};
    }
}
close file 2;
[download]

I (and other monks here) will be able to provide actual code if you provide the necessary information (especially sample input and desires output).