pub fn simple_tokenize(text: &str) -> Vec<String>
Simple tokenizer for demo purposes - splits on whitespace and punctuation