There may be a learning curve if you are unfamiliar with tree structures, but html::tree contains a wealth of methods for loading html into a suitable memory structure and extracting the bits you want.