Depth-first searchEdit

Depth-first search (or DFS) is an aggressive search of a graph where each exploration continues as far as possible, only backtracking when absolutely necessary.

It is commonly implemented using recursion, but can also be implemented in terms of a stack (ie. where freshly expanded nodes are pushed onto a stack).

Exploring a directed acyclic graph using depth-first search gives a topological ordering (ie. every edge has a starting point earlier in the ordering than the ending point of the edge); this is useful for computing dependencies.

See also