How to do an incremental knn search?

Incremental knn search save the priority queue in each round generated by knn search if best-first strategy is adopted, and use priority queue saved in the previous round to search knn in the current round.
I want to do an incremental knn search. Does anyone can teach me, thank you very much!

No files, tests, or entry points are identified. Start by locating the repository's existing KNN-search API and its tests, then determine whether incremental reuse is supported; done requires a maintainer-confirmed implementation or documented usage path.

Written by the indexing model from the issue text.