since you know so much regex, why dont you write a regex html parser /s