gistfile1.txt

So what you would do is find any letters within the string that show up at least twice. From there you look at the next letter beside it and save that in some sort of entry. That is the substring to beat. Then you run through the rest of the letters to see if anything can beat that record. 

You do that so on and so forth until you reach the end of your substrings and then return the longest one the program found.