[FULLSTACK] Add user leaderboard

[Tsandas]

Add a leaderboard for viewing statistics, high scores etc. based on user generated data. The goal is to add a gamification aspect on studying and a sense of community between students.

Requirements

Create minimum the following leaderboards based on individual users:

• Users with the most quizzes completed.
• Users with the best average score on quizes completed.
• Users with the most questions uploaded.
  Create minimum the following leaderboards based on community effort:
• Most "popular" courses by quizes completed.
• Courses with the highest average scores.
• "Hardest" quizes with the lowest average score
  If the user is logged in also show their position in each option e.g. the user is in position 66 of average scores and we only show the top 10 then show them on the bottom along with their position

How to BACKEND

• Create leaderboardService and leaderboardController and relevant response DTOs
• For each leaderboard option create a separate endpoint, service function, repo query
• Each option will return the top 100 users along with minimum their name (username + discordName in parentheses) and position AND the logged in user's position

How to FRONTEND

• Create a separate page along with main page, courses, information etc.
• Show as a table the results of the backend based on each option
• Show all the options as buttons based on category (individual/community)
• Show the results by descending order by default (1->100) and also have the option to show them by ascending order

[dimitriostsakiridis]

Add leaderboard and course/student analytics endpoints

What this adds

Read-only endpoints for course leaderboards and department-wide
quiz analytics.

Endpoint	Description
GET /api/v1/leaderboard/courses/{courseId}	best score per student for a given course
GET /api/v1/leaderboard/courses/popular	Most popular courses (by quizzes completed)
GET /api/v1/leaderboard/courses/avg	Courses ranked by average score
GET /api/v1/leaderboard/helpers/total_questions	Questions created per author (admin/helper authoring stats)

Implementation notes

All endpoints are backed by parameterized native @Query methods
rather than in-memory stream aggregation. This was a deliberate
choice after prototyping both approaches:

Approach	Leaderboard (1 course)	Course avg (all data)
Native query	~80ms	~90ms
Stream aggregation	~200ms	~7s (join fan-out)

The stream-based version required fetching full entity graphs
(@EntityGraph) to avoid N+1 lazy-loading, which caused expensive
join fan-out at real data volume (~2000 bundles). Since every query
here uses bind parameters (no string concatenation), the usual SQL
injection concern with native queries doesn't apply, and query text
lives only in the repository layer — not in a database view visible
to anyone with schema read access.

All displayed names go through User.publicName() to respect the
Discord display-name preference rather than reading display_name
directly.

Testing

• Verified against ~2000 seeded bundles across multiple courses
• Compared native query output against equivalent hand-run SQL to
  confirm correctness
• Load-tested via repeated Postman requests to confirm no N+1
  regressions