Find K Pairs with Smallest Sums

If nums1 and nums2 are sorted, view each index i in nums1 as a sorted row of pair sums with nums2. Seed a min-heap with (i, 0) for i from zero through min(k, nums1.length)-1. Repeatedly pop the smallest pair, append its values, and if j plus one exists, push (i, j+1). Each row only exposes its next unseen candidate, so the heap is a frontier rather than the full Cartesian product. The initial-row limit is safe because any row after k starts with a first pair no smaller than the first pair in an earlier row, so it cannot be needed among k results. Clarify output tie behavior if it matters. Test empty arrays, k zero, k exceeding the pair count, duplicate values, and negative numbers. Heap setup is O(k log k) in the loose bound and output work is O(k log min(k,m)); space is O(min(k,m)).