Rotten Oranges / Multi-Source BFS (taxis)

LC 542 / 994 variant — for every cell, distance to the nearest active source (a 'taxi'). Multi-source BFS is the canonical solution. Requirements Grid where some cells are sources (taxis / rotten oranges / gates); compute distance from every other cell to the nearest source. Multi-source BFS: enqueue all sources at distance 0 simultaneously, expand layer by layer. Follow-up reported: handle moving sources / time-varying placement (recompute periodically vs incremental update). Examples Grid [[2,1,1],[1,1,0],[0,1,1]] (rotting-oranges variant): return number of minutes until all reachable 1s become 2 (4). Taxi variant: for each empty cell, distance to nearest taxi. Notes BFS from all sources simultaneously is O(mn); per-source BFS is O(mn × #sources). The moving-source follow-up trades exactness for amortized cost — common answer is periodic rebuild with bounded staleness. Preparation Drill multi-source BFS from memory in under 10 min. Prep a 2-minute pitch on incremental-vs-rebuild trade-off for the moving-source follow-up. Practice the rotting-oranges / walls-and-gates / 01-matrix triplet — all share the multi-source BFS skeleton.