When parsing descriptions and titles with characters like '&', '<', etc. feed was broken. This regex swaps these characters with their HTML entity equivalent.
