This paper presents a new contact-motion planning algorithm which searches not only contacts and motions but also contact transition strategies. Especially, this paper focus on slide-type and detach-type switching. Our idea is to separate the planning process into 3 steps: select contact transition strategy, search contact state, and generate motion. This separation prevents from complicating the motion search process which is generally a non-linear optimization process and is computationally expensive, and enables to manually control the priorities of all transition strategies. The contribution of this paper is to develop a new concept of contact-motion planning algorithm involving contact transition strategies switching, and to examine the algorithm's search capabilities through 2 scenarios: bed seating, and standing up. The bed seating experiment is shown not only in the simulation world but also in the real world with real humanoid robot.