ByteDance QA Engineer Interview Questions

ByteDance aims to enhance video viewing experience by recommending similar videos based on user preferences. Given a directed graph where nodes represent videos and edges represent user preferences, apply DFS to find all reachable videos from a starting video node.
You need to implement the following function:
- List<Integer> recommendVideos(int[][] graph, int startVideo) - returns a list of all videos reachable from startVideo.
Example 1:
Input: graph = [[1, 2], [3], [3], []], startVideo = 0
Output: [1, 2, 3]
Explanation: From video 0, you can reach videos 1, 2, and 3.
Example 2:
Input: graph = [[1], [2], []], startVideo = 1
Output: [2]
Constraints:
- 0 <= graph.length <= 100
- 0 <= graph[i].length <= 100

In ByteDance’s video application, we need a system to filter video playlists based on user history. The goal is to maximize the number of unique videos in the playlist while adhering to a maximum length constraint.
You are given an int[] videoIds representing the videos a user has viewed and an int maxLength which is the maximum length that the playlist can be.
The function signature is:
- List<Integer> filterPlaylist(int[] videoIds, int maxLength) - returns a list of video IDs in the filtered playlist avoiding repetition.
Example 1:
Input: videoIds = [1, 2, 3, 1, 2, 3], maxLength = 4
Output: [1, 2, 3]
Explanation: The unique videos that fit within the length 4 are 1, 2, and 3.
Example 2:
Input: videoIds = [5, 6, 5, 7, 8, 6], maxLength = 5
Output: [5, 6, 7, 8]
Explanation: The unique videos that fit within the length 5 are 5, 6, 7, and 8.
Constraints:
- 0 <= videoIds.length <= 10000
- 1 <= maxLength <= 1000

In the context of ByteDance's backend performance testing, design a class LoadTestOrchestrator that can schedule and execute load tests across multiple services. The orchestrator must manage test execution and gather metrics.
Class Signature:
- class LoadTestOrchestrator:
- def __init__(self): - Initializes the orchestrator.
- def schedule_load_test(self, service_name: str, duration: int): - Schedules a load test for a given service.
- def execute_tests(self) -> None: - Executes all scheduled tests concurrently.
- def get_test_results(self) -> Dict[str, Any]: - Returns a dictionary of test results for each service.
Example 1:
- Input: orchestrator = LoadTestOrchestrator(), orchestrator.schedule_load_test('serviceA', 60)
- Output: After calling execute_tests() the results can be accessed via get_test_results().
Constraints:
- 1 <= service_name.length <= 100
- 0 <= duration <= 3600

In testing environments common to ByteDance, it's essential to identify flaky tests to improve the reliability of the automation pipelines. Design a class FlakyTestDetector to help identify tests that fail intermittently and should be reviewed or fixed.
Class Signature:
- class FlakyTestDetector:
- def __init__(self): - Initializes the detector.
- def report_test_result(self, test_id: str, success: bool): - Reports the result of a test execution.
- def get_flaky_tests(self) -> List[str]: - Returns a list of flaky test ids.
Example 1:
- Input: detector = FlakyTestDetector(), detector.report_test_result('test1', True), detector.report_test_result('test1', False)
- Output: detector.get_flaky_tests()
- Returns: ['test1']
Constraints:
- 1 <= test_id <= 100000