In this paper, we present a comprehensive, in-depth survey of the literature on reinforcement learning approaches to ridesharing problems. Papers on the topics of rideshare matching, vehicle repositioning, ride-pooling, and dynamic pricing are covered. Popular data sets and open simulation environments are also introduced. Subsequently, we discuss a number of challenges and opportunities for reinforcement learning research on this important domain.