//Iterate all nodes by breadth first search,start with the entry nodes and resolve the dependencies
Bfs,
//Iterate all nodes by deep first search,but run the dependencies that are not ready first
Dfs,
///Start from the entry node,if multiple entry nodes exist,run first of them ,and then ,from this entry node,transit to nodes that met the condition,if
///there is no node that met the condition ,stay in the state ,if there are multiple nodes that met the condition,run first of them
/// If the run node depends on other nodes,run the dependencies first.