pjfanning commented on PR #4006: URL: https://github.com/apache/fory/pull/4006#issuecomment-5471271871
### AI review loop: rounds 5 and 6 Following the earlier summary, the loop ran two further rounds and has converged. **Round 5 (`a86a870b4`)** — both reviewers independently found the same gap: guarding only `Class.forName` did not make case-class recognition free of linkage failures, because `getField` and `getMethods` resolve member descriptors. A companion or case class declaring a member whose type is absent at runtime could throw `NoClassDefFoundError` out of `isCaseClass`, a predicate consulted for every `Product` reaching the module, including types it does not own. Fixed in `13a62665c`: recognition declines such a type, the committing path still reports it, and ambiguity stays loud with the comment saying so. **Round 6 (`13a62665c`, current head)** — Fory-guided reviewer: no blocking findings. Independent reviewer: no actionable findings, "I would approve this". The independent reviewer traced whether the widened `catch` could swallow a genuine problem and confirmed it cannot; the Fory-guided reviewer compiled and inspected the class files to verify the `getLiteralName` reasoning, which also corrected one of its own earlier nits. Three optional round-6 nits were reviewed and deliberately not actioned, so that the clean result stands on the head that is actually being merged. They are listed in the gist index. **Artifacts updated** — all twelve reviewer reports, verbatim, with a per-round outcome table: https://gist.github.com/pjfanning/6b9fcd687e22450399df72dabc368e99 **Verification added since the earlier comment:** `fory-json` maven suite, 964 passed, 0 failures. Full results, and the gaps that remain (the GraalVM path is verified by inspection only, and `sbt +test`/`spotless`/`checkstyle` did not run locally), are in the PR description. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
