Make fingerprint integration tools match the same way Nmap does

[dmiller-nmap]

The fingerprint integration tools in nmap-private-dev are used to match and compare user-submitted fingerprints to the reference prints in nmap-os-db. However, there are slight differences in how fingermatch handles duplicate tests and how Nmap would handle it, which cause unnecessary work for the reviewer.

For example, sometimes Nmap will do multiple OS detection attempts. When it does this and the results are different, it will print the output of the different attempts as duplicate probe lines. Duplicate SEQ lines are common. When such a fingerprint is handled by fingermatch, it only looks at the first such test. If that test is an anomaly, then fingermatch will not show a match, even if subsequent lines would exactly match an existing print.

Questions to be answered before attempting a fix:

1. Under what conditions does Nmap do multiple OS detection attempts?
2. How exactly does Nmap perform a match in that case?
3. How can fingermatch best replicate this behavior?
4. Will we lose valuable information about variance by doing this?