Let one file hold both features, since one file already does - #33
Conversation
`policy/hooks.toml` carries the waivers `hooks --identity` reads and the `[[probe]]` fixtures `uphold probe` drives. Both readers use `deny_unknown_fields`, which is what catches `waivee` and a misplaced `reason` -- and which cannot tell a typo from a sibling feature's table unless the sibling is named. `probe`'s reader named `waive`. The waiver reader named nothing, so the first `[[probe]]` written into this repository's own file made `hooks --identity` exit 2 on it. The file was well formed and the command could not look at it. Found while re-running the fleet comparison after four consuming repositories adopted the guard sets, and the way it was found is the point: the report was grepped for findings, the grep counted zero, and zero findings over an error is exactly the reading this tool exists to refuse. The command said so -- exit 2, with the parse error -- and only the count was being read. Both directions have a test now, and so does the case that made `deny_unknown_fields` worth having: a waiver whose `reason` was typed as `resaon` is still refused, because a waiver with no reason is a check switched off with nobody's name on it. The real answer, once the command could read its own file: three divergences across six repositories, all of them `forked` third-party hooks scoped differently on purpose. The five `pinned apart` findings and every `absent` one are gone, which is what the four adoption pull requests were for.
|
Warning Review limit reached
Next review available in: 73 minutes You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (5)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #33 +/- ##
=======================================
Coverage 89.15% 89.15%
=======================================
Files 29 29
Lines 9217 9217
=======================================
Hits 8217 8217
Misses 1000 1000 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
…em noisy `hooks --identity` reports a fork whatever the reason, which is right: the reason is in neither tree, so only a waiver file can hold it. Three remain across this fleet once the guard sets are adopted, and all three are decisions. `ruff-check` and `ruff-format`: this repository runs ruff with `--fix` at commit time, where a workspace whose Python is a handful of scripts runs it check-only over a files scope. Autofixing somebody else's tree on their commit is not a default to share, and where a repository keeps its Python is a fact about that repository. `shellcheck`: a workspace whose shell scripts stand up lab images runs it at four stages and at `--severity=style`; this repository has a handful of scripts and runs the default at commit time. Worth revisiting if these scripts grow, which is why the difference is written down rather than hidden. The fleet report is clean now, and clean means every difference has a sentence somebody wrote rather than every difference having disappeared.
2 participants
policy/hooks.tomlcarries two things: the waivershooks --identityreads, and the[[probe]]fixturesuphold probedrives. Both readers usedeny_unknown_fields-- which is what catcheswaiveeand a misplacedreason, and which cannot tell a typo from a sibling feature's table unless the sibling is named.probe's reader namedwaive. The waiver reader named nothing, so the first[[probe]]written into this repository's own file madehooks --identityexit2on it. The file was well formed and the command could not look at it.How it was found, which is the part worth reading
Re-running the fleet comparison after four consuming repositories adopted the guard sets. The report was grepped for findings, the grep counted zero, and zero findings over an error output is exactly the reading this tool exists to refuse. The command was doing its job -- exit 2, with the parse error on stderr -- and only the count was being read.
The real answer, once the command could read its own file
Three divergences across six repositories, all
forkedthird-party hooks scoped differently on purpose (ruff-check,ruff-format,shellcheck). The fivepinned apartfindings and everyabsentone are gone, which is what the four adoption pull requests were for.Tests
Both directions -- a waiver file carrying probes, a probe file carrying waivers -- and the case that makes
deny_unknown_fieldsworth having: a waiver whosereasonwas typed asresaonis still refused, because a waiver with no reason is a check switched off with nobody's name on it.425 tests pass, 3 of them new.
And the three that remain are recorded
ruff-check,ruff-formatandshellcheckare declared differently across this fleet on purpose, and the reasons are in neither tree -- sopolicy/hooks.tomlnow holds them. This repository runs ruff with--fixat commit time where a workspace whose Python is a handful of scripts runs it check-only over a files scope; autofixing somebody else's tree on their commit is not a default to share. Same shape for shellcheck's severity and stages.The fleet report is clean, and clean here means every difference has a sentence somebody wrote rather than every difference having disappeared.