The project said there should be two consective zero in B and C if A has no non-zero elements. But what's the point of that? Since for B it only contains non-zero elements, that simply means all the zero are used for representing null. So one zero is just enough to identify the end of each row. on the other hand, if we allow B to contain some zero elements, there is no way to decide where is the end of each row in B(at least not mentioned in the project).