Optimal control problem with inequality constraints at discrete times

Question

I am interested in solving the following optimal control problem $$\min_{\mathbf{u}(t)} \ \phi(\mathbf{x}(t_f),t_f)+\int_{t_0}^{t_f}\mathcal{L}(\mathbf{x}(t), \mathbf{u}(t), t)\, \mathrm{d}t$$ subjected to $$\dot{\mathbf{x}}=f(\mathbf{x}(t), \mathbf{u}(t), t), \quad \mathbf{x}(t_0)=\mathbf{x}_0$$ and a path constraint in the form $$g(\mathbf{x}(\tau), \mathbf{u}(\tau), \tau)\le0 \quad \mbox{for some} \quad \tau\in[t_0, t_f].$$ For clarity's sake, this strange path constraint should not be enforced for the whole timespan, but it is sufficient that it is respected in just one point in time $\tau$, unknown a priori.
I thought that one approach could be to exploit a direct optimization method, discretizing the state and controls, and appending $\tau$ to the decision vector.
However, I am wondering: does it exists an alternative (and more elegant) approach to deal with this problem and the "discrete" path constraint?