Revision of Object-Oriented Design Interview: Parking Lot

linkjobai revised this gist 1 day ago. Go to revision

1 file changed, 59 insertions

parking.md(file created)

		@@ -0,0 +1,59 @@
1	+	# Object-Oriented Design Interview: Parking Lot
2	+
3	+	"Design a parking lot" is the most common object-oriented design (OOD) question. The interviewer is checking whether you can turn vague requirements into clean classes and responsibilities, not whether you know a single right answer.
4	+
5	+	## Clarify first
6	+
7	+	- Multiple floors? Spot sizes (motorcycle, compact, large, EV charging)?
8	+	- Entry and exit gates, tickets, payment (hourly rate, flat fee)?
9	+	- Should it show free spots per floor on a display board?
10	+
11	+	## Core classes
12	+
13	+	```python
14	+	from enum import Enum
15	+	from datetime import datetime
16	+
17	+	class Size(Enum):
18	+	SMALL = 1; MEDIUM = 2; LARGE = 3
19	+
20	+	class Vehicle:
21	+	def __init__(self, plate, size): self.plate, self.size = plate, size
22	+
23	+	class Spot:
24	+	def __init__(self, spot_id, size):
25	+	self.id, self.size, self.vehicle = spot_id, size, None
26	+	def fits(self, v): return self.vehicle is None and v.size.value <= self.size.value
27	+
28	+	class Ticket:
29	+	def __init__(self, vehicle, spot):
30	+	self.vehicle, self.spot, self.entered = vehicle, spot, datetime.now()
31	+
32	+	class ParkingLot:
33	+	def __init__(self, floors): self.floors, self.active = floors, {}
34	+	def park(self, v):
35	+	for floor in self.floors:
36	+	for spot in floor:
37	+	if spot.fits(v):
38	+	spot.vehicle = v
39	+	t = Ticket(v, spot); self.active[v.plate] = t
40	+	return t
41	+	return None # lot full
42	+	def leave(self, plate, pricing):
43	+	t = self.active.pop(plate); t.spot.vehicle = None
44	+	return pricing.fee(t.entered, datetime.now())
45	+	```
46	+
47	+	## Design choices to talk about
48	+
49	+	- Strategy pattern for pricing: `HourlyPricing`, `FlatPricing`, weekend rates, all behind one `fee()` method.
50	+	- Spot assignment strategy: nearest to the entrance, or smallest spot that fits, behind an interface.
51	+	- Faster lookup: keep a free-spot set (or min-heap by distance) per size instead of scanning every spot.
52	+	- Concurrency: two gates can grab the same spot; lock the spot or use an atomic compare-and-set.
53	+
54	+	## What interviewers reward
55	+
56	+	Clear responsibilities (a ticket doesn't calculate fees), extensibility (adding EV spots touches one class), and asking about edge cases: lost tickets, full lot, a vehicle bigger than any spot.
57	+
58	+	---
59	+	Disclosure: I'm part of the [Linkjob AI](https://www.linkjob.ai/) team; we build an AI interview assistant.

Newer Older