Courier Delivery Cost Tracker

Courier Delivery Cost Tracker

You are building a backend service for an on-demand logistics platform to track courier activity and delivery payouts. Design a class CourierCostTracker that records delivery shifts, maintains courier balances, and analyzes peak delivery concurrency.

Each delivery takes place over a half-open time interval [start_time, end_time) (meaning the courier is active from start_time up until, but not including, end_time) and incurs a set delivery payout.

Implement the CourierCostTracker class:

• CourierCostTracker(): Initializes the tracker.
• record_delivery(courier_id: str, start_time: int, end_time: int, fee: int) -> None: Records a completed delivery for courier courier_id active during [start_time, end_time) with an associated payout of fee.
• get_courier_total(courier_id: str) -> int: Returns the total cumulative payout earned by courier_id. If courier_id has no recorded deliveries, return 0.
• get_peak_concurrent(window_start: int, window_end: int) -> int: Returns the maximum number of deliveries active at any single point in time t within the query window [window_start, window_end). Deliveries that overlap with the window should be counted only for the portion of time that falls strictly within [window_start, window_end).

Constraints

• 0 <= start_time < end_time <= 10^9
• 0 <= window_start < window_end <= 10^9
• 1 <= fee <= 10^5
• courier_id consists of lowercase English letters and digits, length between 1 and 30.
• At most 2,000 calls to record_delivery and get_peak_concurrent combined.
• At most 10,000 calls to get_courier_total.

Example 1

tracker = CourierCostTracker()
tracker.record_delivery("c1", 10, 30, 25)
tracker.record_delivery("c2", 20, 50, 40)
tracker.record_delivery("c1", 40, 60, 30)

# c1 earned 25 + 30 = 55
assert tracker.get_courier_total("c1") == 55

# c3 has no records
assert tracker.get_courier_total("c3") == 0

# In window [15, 45):
# - Delivery 1 [10, 30) is active in window during [15, 30)
# - Delivery 2 [20, 50) is active in window during [20, 45)
# - Delivery 3 [40, 60) is active in window during [40, 45)
# Peak concurrency occurs between time 20 and 30 (c1, c2), and between 40 and 45 (c2, c1), max is 2.
assert tracker.get_peak_concurrent(15, 45) == 2

# In window [30, 40), only Delivery 2 [20, 50) is active. Peak is 1.
assert tracker.get_peak_concurrent(30, 40) == 1

# In window [70, 100), no deliveries overlap. Peak is 0.
assert tracker.get_peak_concurrent(70, 100) == 0

Example 2

tracker = CourierCostTracker()
tracker.record_delivery("c1", 10, 20, 15)
tracker.record_delivery("c2", 20, 30, 15)

# Because intervals are half-open [10, 20) and [20, 30), they do not overlap at t = 20.
assert tracker.get_peak_concurrent(0, 50) == 1