Lists are parsed in linear instead of exponential time now. Contents of block tags, such as <quote>, is parsed directly, without storing it in a string and parsing with parseFromString. Fixed a bug: headers did not terminate lists.