Learning List-Level Domain-Invariant Representations for Ranking