The naming of parse nodes (from the parsers) is somewhat inconsistent and random at the moment. This should be made consistent. Also, names should not try to represent the sub-parse-tree, but make it easy to find specific nodes. Thus, it might be sufficient to just use the type of parser.