SEZ9 commented on PR #12015:
URL: https://github.com/apache/seatunnel/pull/12015#issuecomment-6431953601

   @luozihen — apologies for the slow reply on the contract question; answering 
it directly now.
   
   1. Agreed on the parser point. Given `PATH_TOKEN_SEPARATOR` is `->` in this 
repo, the row we disagreed on resolves the way you describe: `job.mode` / 
`shade.identifier` stay literal through `parseMap`, `a->b` is the form that 
expands, and `^t_.*$` is the one that actually breaks with `BadPath`. That 
settles PR12015-F1 in favour of the narrow fix rather than the all-literal 
rebuild, so I withdraw my objection to keeping the REST fix in this PR.
   
   2. Quote-only-invalid is the contract I'm comfortable with: quote a key only 
when it cannot be parsed as a path, and leave everything that parses today 
untouched. Please make sure the implementation really is a no-op for every key 
that parses today (dotted keys, already-quoted keys, `a->b`), so no existing 
caller of `toConfig()` sees a change.
   
   3. On tests (PR12015-F7), what I need to see before I can resolve this:
      - the REST-shaped regression (`map -> RestUtil.buildConfig -> 
ConfigShadeUtils.decryptConfig`) plus a direct `toConfig()` case, asserting 
`shade.identifier`, `job.mode`, `a.b`, `a->b` and a regex key together in one 
test so the final behavior is pinned in one place;
      - the existing test's wording aligned with the `->` separator, as you 
proposed.
   
   4. Sequencing: since #12443 (head `509fddab`) carries the config-key 
conversion, please land that first, then sync `dev` and drop `0c43f10dd` from 
this branch so the two PRs don't ship competing conversions. The head here is 
still `d9bb1574`, so my review stands until the rebased head is pushed; once it 
is, I'll re-review the REST point and the test together.
   
   @nzw921rx noted — fine to take this after 3.1.0; nothing here needs to rush 
the release.
   
   Nothing further is needed from you until #12443 merges. Ping me when the 
rebased head is up.
   
   <!-- streview-comment:2054 -->


-- 
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]

Reply via email to