AI code review challenge tests if edits preserve program behavior.
A programming challenge asks reviewers to analyze three proposed edits to a function. The function scores integer inputs from -64 to 64, plus a None value. Each edit must be judged as preserving behavior, changing it, or being unsupported by the review system. The correct answers, confirmed via testing, are that one edit preserves behavior, one changes it, and one is unsupported. The exercise demonstrates principles behind automated code review and mutation analysis tools.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.


Discussion (0)
Log in to join the discussion and vote.
Log in