Leetcode 3103. Find Trending Hashtags II

Design a data structure to ingest a high‑volume stream of timestamped hashtags and efficiently return the top‑k trending hashtags over a recent sliding time window, supporting frequent inserts and time‑windowed queries by maintaining counts and evicting expired occurrences. The core challenge is to update counts and retrieve top‑k quickly under streaming/time-decay constraints (e.g., using hashes, heaps, or time-bucket aggregation).