Resource Access Log: Audit Who Accessed What Resource and When

Question

## Problem

You are given a list of access log entries: `{user_id, resource_id, action, timestamp}` where action is `"read"` or `"write"`. Answer two queries:

1. `most_accessed(resource_id)`: return the user_id who accessed the resource the most times total.
2. `unauthorized_writes(allowed_writers: set)`: return a sorted list of user_ids who performed write actions but are NOT in the allowed_writers set.

```python
class AccessLog:
    def __init__(self, entries: list[dict]): ...
    def most_accessed(self, resource_id: str) -> str: ...
    def unauthorized_writes(self, allowed_writers: set) -> list[str]: ...
```

**Example:**
```
entries = [
  {"user_id": "u1", "resource_id": "r1", "action": "read",  "timestamp": 1},
  {"user_id": "u2", "resource_id": "r1", "action": "write", "timestamp": 2},
  {"user_id": "u1", "resource_id": "r1", "action": "write", "timestamp": 3}
]

most_accessed("r1") -> "u1"  (2 accesses vs u2's 1)
unauthorized_writes({"u1"}) -> ["u2"]  (u2 wrote but is not allowed)
```

## Follow-ups

1. How would you build the data structures in `__init__` to make both queries O(1) or O(k) in the result size?
2. What if entries stream in continuously — how do you maintain counts incrementally?
3. How would you generate a time-series report of access counts per hour per resource?
4. How would you design this as a security audit system with alerting on unauthorized writes?

Resource Access Log: Audit Who Accessed What Resource and When

Question Details

Full Details

About This Question

Topics

More Karat Interview Questions