SEZ9 commented on PR #11983: URL: https://github.com/apache/seatunnel/pull/11983#issuecomment-5460551355
A note on what happened to this PR, since the diff just changed shape: #11984 was stacked on this PR's first commit (`1bf67a91b`), because it needs the `LogLevels` helper introduced here. A squash merge squashes the PR's whole diff against the base branch, so when #11984 landed as 9ec9c4fe8 it carried `1bf67a91b` into `dev` together with its own changes. That is why this PR started showing a conflict: it was trying to add files that `dev` already has. Everything this PR set out to fix for #11981 (a) is on `dev` now: - `Log4j2HttpPostCommandProcessor#setLoggerLevel` rejects an unknown level with `400` and lists the valid names, instead of answering `SUCCESS` - `LogLevels.parse` / `LogLevels.validNames` and every `LogLevelsTest` case from this PR (#11984 added five more on top) - both `incompatible-changes.md` entries The one piece that did not make it is the javadoc correction, which is the blocker raised in the review above. I pushed it here after #11984's branch had already been cut from `1bf67a91b`, and I chose not to rebase #11984 onto it at the time, so the squash could not pick it up. `dev` still reads "which log4j2 silently ignores", and that is precisely the misconception the fix exists to remove: a `null` level clears the logger's explicit level, so the logger falls back to its parent and the root logger falls back to `ERROR`. So I have rebased this PR onto the current `dev` and reduced it to that single correction. The conflict is gone and the diff is now one line in `Log4j2HttpPostCommandProcessor`; the review discussion above still applies to it unchanged. -- 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]
