A*-based Temporal Logic Path Planning with User Preferences on Relaxed Task Satisfaction