PreviousNext
AI-assisted engineering/Verifying AI output

Fix the surviving mutant report for a coverage push

Review the AI's codeEasy12 min

Your task

An AI agent took the PricingEngine module to 91 percent test coverage, all green. Coverage shows which lines ran, not whether any test would notice a bug in them. So your team runs mutation testing: a tool plants small bugs called mutants, and a test that fails on a mutant kills it. The mutants nobody kills are the bugs the suite would ship.

An assistant wrote surviving_mutants(mutant_ids, kill_lines). It is in the editor. It has a bug. Find it and fix it.

Input and output:

  • mutant_ids: the ids of the planted mutants.
  • kill_lines: the tool's report, one line per test, written test_name>id,id,.... The list after > can be empty.
  • Return the ids of the surviving mutants, in the order they first appear in mutant_ids.

Rules the tests check:

  • In a kill line, the killed ids are the items after the first >, split on ,, each with whitespace trimmed from both ends. Empty items are ignored. A line with no > kills nothing.
  • Trim whitespace from both ends of each entry in mutant_ids. Skip an entry that is empty, and skip a repeat of an id already seen.
  • A mutant is killed when some killed id equals its id exactly, case included.
  • Return every id that is not killed.

Examples

surviving_mutants(mutant_ids: list[str], kill_lines: list[str]) -> list[str]
Example 1
Inputsurviving_mutants(["m_lt_to_lte", "m_round_half_down", "m_discount_skip"], ["test_price_basic>m_lt_to_lte", "test_zero_quantity>", "test_snapshot_output>"])
Output["m_round_half_down", "m_discount_skip"]
Example 2
Inputsurviving_mutants(["m_lt_to_lte", "m_round_half_down"], ["test_boundary_price>m_lt_to_lte,m_round_half_down", "test_smoke>"])
Output[]
Example 3
Inputsurviving_mutants(["m_currency_floor"], ["test_calls_engine>m_ghost_mutant", "test_no_assertions>"])
Output["m_currency_floor"]