Implement sliding-window rate limiter with dual thresholds

Implement getRequestStatus(urls: List[str]) -> List[str]. The i-th entry in urls represents a single incoming request at timestamp t = i seconds (t starts at 0); the value is the URL being requested. For each request, return '200' if it can be served, else '429' if serving it would violate either limit for that same URL: at most 2 successful requests in any sliding 5-second window [t-4, t], and at most 5 successful requests in any sliding 30-second window [t-29, t]. Important: rejected requests must NOT be recorded as successful and therefore must NOT count toward future windows; successful requests do count. Process the stream online in arrival order. Constraints: 0 ≤ n ≤ 2e5; URLs are arbitrary strings; target O(1) amortized time per event and O(U) memory, where U is the number of distinct URLs seen in the past 30 seconds only. Use only built-in data structures; avoid third-party rate-limiting libraries. Provide: 1) A correct, production-friendly implementation that uses per-URL deques (or equivalent) to maintain only the timestamps still inside the respective windows; 2) Proof of correctness with respect to the two sliding-window constraints and the reject-does-not-count semantics; 3) Tight time and space complexity analysis; 4) Unit tests that cover edge cases such as empty input, alternating URLs, bursts exactly at boundaries (t=4/5 and t=29/30), rapid sequences that flip between '200' and '429', and very long runs where old timestamps must be evicted efficiently.