Design TTL Cache

Design a cache where every entry has a time-to-live (TTL): it is readable for a fixed window after insertion, then disappears. Implement the TTLCache class:

• TTLCache() Initializes an empty cache.
• void put(int currentTime, String key, int value, int ttl) Stores key -> value with an expiry time of currentTime + ttl. If key already exists, the entry (and its TTL) is refreshed with the new value and new expiry.
• int get(int currentTime, String key) Returns the value for key if it is present and not yet expired at currentTime. Otherwise returns -1.
• int size(int currentTime) Returns the number of entries that are still alive at currentTime. Expiry semantics (this is the part the interviewer will press on):
• An entry is alive at time t iff t < expiry.
• At t == expiry the entry is already expired (it lived for exactly ttl units, from putTime up to but not including putTime + ttl).
• You may assume currentTime values passed in across calls are non-decreasing. Design constraints:
• put and get should be O(1) average.
• get must not return expired values — if the key is present but expired, return -1 (and you may remove it then, or defer deletion to size).
• size must return an accurate count at currentTime, so expired entries must actually be removed before it answers.

At t=3, a (expiry=4) and b (expiry=12) are both alive → size=2. put at t=5 refreshes "a" to value=30, new expiry=10. get at t=6 sees the new value. At t=7, a (10>7) and b (12>7) still alive → size=2.

"a" (expiry=3) is expired by t=4; size(4) sweeps it away and returns 1 (only "b" left). get(4,"a") confirms it is gone. By t=10, "b" (expiry=6) is also expired → size=0.