An open grid has many equally short paths, yet your search returns exactly one, every time. Today you cash in the determinism you built, confirming the fixed neighbor order and heap tie-break pin down a single stable path.
Confirm the search returns one specific stable path among many equal-length options.
On a wide-open grid the number of shortest paths between two corners is enormous, all the same length. A search that returned “some” shortest path would give a different one on a whim, and no test could pin it. Everything in this project has been building toward the opposite: a single, stable path, the same on every run and in every language.
Two decisions make it so. The neighbor order, fixed as North, East, South, West back in chapter one, means East is always tried before South, so the path hugs the top edge before turning down the last column. The heap tie-break, breaking equal priorities by insertion order, resolves every remaining coin-flip the same way. This lesson writes no new logic; it is the payoff, confirming that determinism is a property you can rely on. That reliability is exactly what lets the capstone assert an exact solved-maze path down to the cell.
// No new production code: this asserts the determinism already built.// With cost-1 cells, dozens of 6-step routes tie in length. The chosen// one is decided by (a) neighbor order N,E,S,W (East is tried before South)// and (b) the heap breaking equal f-values by insertion order.// Run the search a few times; assert the path is byte-for-byte identical.