At the early commercialization stage of hydrogen fuel cell vehicles (HFCVs), there is a strong causal relationship between HFCV parc and hydrogen refueling stations (HRSs) planning, also known as indirect network effects. This is a crucial factor that hasn't been considered and cannot be ignored in the hydrogen supply infrastructure planning (HSIP) problem. To model the indirect network effects, we quantify the effect of HRS planning on HFCV parc in detail by proposing the calculation method of drivers' fuel buffer, defining hydrogen refueling convenience, and incorporating coupled constraints between the hydrogen network and transportation network. Further, we characterize the effect of HFCVs on HRS planning by constructing a decision-dependent distributionally robust optimization problem considering the HFCV parc uncertainty. We innovatively quantify the upper bound of the ambiguity set of Wasserstein associated with both the endogenous and exogenous factors. We propose effective linearization methods to deal with the nonlinear constraints in the indirect network effects model and transform the two-stage optimization problem into a solvable MILP problem. The simulation results show that the proposed method can obtain more economical investment decisions and achieve higher energy utilization efficiency. More importantly, in this planning framework, investors can set model parameters more flexibly according to the situation of the HFCV market, thereby promoting the popularization of HFCVs.