This involves a manual left corner transform, so I would probably call this a variant of a left corner parser (LC parser)?