Processes dictionary, looking for words that are composed of two concatenated smaller words.
Utilises a trie data structure because it is sick for this sort of thing. It's slightly modified such that nodes either contain a null word or a full word, but it still pretty much follows the descriptions here:
http://en.wikipedia.org/wiki/Trie http://www.toptal.com/java/the-trie-a-neglected-data-structure
Coded for Mint Digital job application and got slightly distracted by testing how fast it will go (it's pretty damn fast!).