Test Severity Filtering with Contract-Distinguishing Cases

Design unit tests for severity-threshold log filtering. Explain what each test would detect and how you would validate test suggestions from an AI assistant. ### Constraints & Assumptions Use the practice contract: levels are ordered low to high; include the threshold and higher levels; drop empty records and missing/unknown levels; keep an empty message when its known level qualifies; preserve order and duplicates. Do not derive expected results by calling the implementation being tested. ### Clarifying Questions What is the intended behavior for missing levels? Is matching based on structured severity or message text? Are threshold boundaries inclusive? Which test distinguishes the competing interpretations? ### What a Strong Answer Covers Boundary and malformed-record cases, adversarial messages, ordering/duplicate checks, independent expected values, and a reason for each assertion. ### Follow-up Questions How can a test suite pass while encoding the wrong requirement? What properties should hold when the threshold is raised? How would you separate filtering tests from concurrent logging tests?