0 # initial state
[2] # terminal state
(0,0,'!A',ConstantRewardFunction(1))
(0,1,'A',ConstantRewardFunction(1))
(1,1,'!B',ConstantRewardFunction(1))
(1,2,'B',ConstantRewardFunction(1))
(2,2,'!C',ConstantRewardFunction(1))
(2,0,'C',ConstantRewardFunction(1)) 