What it does
- Builds a prefix tree from a dictionary file, one node per letter.
- Marks the end of a valid word on the node itself rather than storing whole strings.
- Scans a document, strips punctuation from each token, and counts and prints the words that are not in the tree.