Approach: Keep filling the receive buffer until we have found enough data to extract the first HTTP header and body