How do I practice coding and algorithm questions?

Use PracHub's coding console to write, test, and debug your solutions in Python or JavaScript. View hints, test against sample inputs, and compare with official solutions.

What difficulty level is this coding question?

This is a medium difficulty Coding & Algorithms question, commonly asked during Technical Screen rounds at OpenAI.

What role is this question designed for?

This question is commonly asked for Software Engineer candidates at OpenAI during technical interviews.

Implement persistent key-value store | OpenAI Coding Question

Implement persistent key-value store

Company: OpenAI

Role: Software Engineer

Category: Coding & Algorithms

Difficulty: medium

Interview Round: Technical Screen

##### Question Design and implement an in-memory key-value store supporting set(key, value), get(key), shutdown() that flushes all data as bytes to a medium, and restore() that reloads the data from the medium.

Quick Answer: This question evaluates competency in implementing an in-memory key-value store with persistence, covering concepts such as serialization, durability, data structure management, and basic I/O handling.

Implement a persistent in-memory key-value store and simulate it over a sequence of operations. The store supports four commands: - ('set', key, value): store or update a string value for a string key in memory. - ('get', key): read the current in-memory value for the key, or None if the key does not exist. - ('shutdown',): write the entire current store to a byte medium, then clear the in-memory store. - ('restore',): rebuild the in-memory store from the most recent byte snapshot. If no snapshot exists yet, the store becomes empty. Keys and values may contain any characters, including separators like ':' or '|', and may even be empty strings. Because of that, a safe byte encoding is required; delimiter-based serialization is not reliable. For this problem, write a function that processes all operations in order and returns the results of every 'get' operation.

Constraints

0 <= len(operations) <= 100000
Keys and values are strings and may be empty
The total UTF-8 byte length of all keys and values involved in the processed snapshots is at most 10^6
Average O(1) lookup/update is expected for set/get

Examples

Input: [('set', 'a', '1'), ('set', 'b', '2'), ('get', 'a'), ('shutdown',), ('get', 'a'), ('restore',), ('get', 'b'), ('get', 'a')]

Expected Output: ['1', None, '2', '1']

Explanation: After shutdown the in-memory store is cleared, so 'a' is missing until restore reloads the saved snapshot.

Input: [('set', 'x', 'old'), ('set', 'x', 'new'), ('shutdown',), ('restore',), ('get', 'x'), ('set', 'y', '3'), ('shutdown',), ('get', 'y'), ('restore',), ('get', 'y'), ('get', 'x')]

Expected Output: ['new', None, '3', 'new']

Explanation: The second set overwrites 'x'. The second shutdown saves a new snapshot containing both 'x' and 'y'.

Input: [('restore',), ('get', 'missing'), ('set', '', ''), ('set', 'a|b:c', 'v1|v2:ok'), ('shutdown',), ('restore',), ('get', ''), ('get', 'a|b:c'), ('get', 'missing')]

Expected Output: [None, '', 'v1|v2:ok', None]

Explanation: This checks restore before any shutdown, empty strings, and keys/values containing characters that would break delimiter-based serialization.

Input: [('shutdown',), ('restore',), ('get', 'a')]

Expected Output: [None]

Explanation: Shutting down an empty store should still produce a valid empty snapshot that restores to an empty map.

Hints

A plain delimiter like ':' is unsafe because keys or values can contain it. Prefix each encoded string with its byte length instead.
Treat shutdown as writing a full snapshot and clearing memory. restore should replace the current in-memory map with the last saved snapshot, not merge with it.

Quick Overview

This question evaluates competency in implementing an in-memory key-value store with persistence, covering concepts such as serialization, durability, data structure management, and basic I/O handling.