Mutation Testing: Did my test break my code ? π€
summary
Project shows that 100% code coverage still yields weak tests: a Python demo goes from 100% coverage to a mutation score 64% using the tool called mutePi, requiring an added zero-case to kill a surviving mutant. Java demo runs PIT with 50% coverage and reports a mutation score 70%. Mutation testing generates single-change mutants, produces HTML/XML reports, and scales only with filtering and incremental strategies (Google paper: 2 billion LOC, 150 million tests).
AI-written summary Β· about
chapters (27)
- 0:00Intro
- 1:23Monday
- 1:45History
- 2:58Is 100 code coverage
- 3:28Who is Maha
- 4:12What is Mutation Testing
- 5:21Why Mutation Testing
- 6:02Concept of Mutation Testing
- 6:49Mutant
- 7:41Example
- 10:21Types of Mutation
- 11:24Arithmetic Operator
- 11:56Mutations
- 12:24Mutation Types
- 12:55Dont Panic
- 13:29Mutational Score
- 14:08Demo
- 14:26Python Demo
- 19:17Mutation List
- 19:58MutePi
- 21:03PiTest
- 22:36PiTest IDE
- 27:36Recap
- 29:15Mutation Testing
- 29:47Equivalent Mutations
- 31:18Limitations of Mutation Testing
- 33:55Conclusion
related talks
41mMutation testing in practice
47mMutation Testing: Did my test break my code ? π€
38mMutants to the rescue: How effective are your unit tests?
45mMutation Testing
48mMutation Testing, or What Code Coverage Keeps Silent About
26mMutation Testing
41mMutation Testing with Mutant
07mWho is Testing the Tests?