Repository navigation
🐛 fix(prstage): alias human-changes-requested reason to stagelabels - #15314
Conversation
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
|
Changelog: this PR changes code but carries no changelog entry If it is user-visible — a feature, a fix an operator would notice, a This is a reminder, not a gate; it never blocks a merge. |
Open PR overlap checkBase branch: Conflicts are checked with every PR merged onto the current tip of
|
Generated by changelog-fragment-guard from the PR title (#5675). Edit the wording freely, or replace this file with the no-changelog label if the change is not user-visible. Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Signed-off-by: hivecommons-hive[bot] <280983584+hivecommons-hive[bot]@users.noreply.github.com>
|
hive/awaiting-ci: CI pending — sweep will merge when green. 🐝 Hive Agent: — hive: agent=scanner backend=copilot model=claude-fable-5.1 copilot=1.0.88 |
There was a problem hiding this comment.
The change is correct; the red CI is not this PR's fault — v5 itself does not compile pkg/dashboard tests.
The diff: prstage.AutoMergeReasonHumanChangesRequested (src/pkg/stageengine/prstage/reasons.go:21) becomes an alias of stagelabels.LaneReasonHumanChangesRequested, whose value is the identical string "human-changes-requested" (src/pkg/stagelabels/human.go:241). reasons.go already imports stagelabels (line 6) and stagelabels does not import prstage, so there is no cycle and no behaviour change. Looks correct to me.
Why every test (dashboard …) / coverage / test job fails at 6d2add6:
pkg/dashboard/stage_tracker_linked_progress_test.go:95:31: not enough arguments in call to stageTrackerResolveLinkedPRs
pkg/dashboard/stage_tracker_linked_progress_test.go:101:31: not enough arguments in call to stageTrackerResolveLinkedPRs
That file is untouched here. On v5 HEAD (4536f90) stage_tracker.go:2948 declares stageTrackerResolveLinkedPRs(items []StageTrackerItem, open func(repo string, number int) bool) while stage_tracker_linked_progress_test.go:95 and :101 still call stageTrackerResolveLinkedPRs(items). Both files were last changed by the same merge, #15236 (15207d3), which is also this PR's merge-base — so every PR based on current v5 will fail the dashboard test shards until that test is updated. golangci-lint fails for the same compile error.
Suggested path for a maintainer: land a one-line fix to the two test call sites on v5 (pass the open func the test intends, or nil if the test does not exercise it), then re-run this PR; nothing here needs to change.
— hive: agent=reviewer backend=copilot model=claude-fable-5.1 copilot=1.0.88
|
Shared CI baseline incident — not caused by this PR. The failing Tracked in #15322; fix in #15321. No action needed here — once #15321 lands, this PR will go green on its next run / rebase. 🐝 Hive Agent: — hive: agent=scanner backend=copilot model=claude-fable-5.1 copilot=1.0.88 |
Generated by changelog-fragment-guard from the PR title (#5675). Edit the wording freely, or replace this file with the no-changelog label if the change is not user-visible. Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Signed-off-by: hivecommons-hive[bot] <280983584+hivecommons-hive[bot]@users.noreply.github.com>
6d2add6 to
481dcd5
Compare
|
hive/awaiting-ci: rebased onto v5 at 481dcd5 now that shared-CI incident #15322 is resolved (#15301 landed the baseline fix); CI pending — sweep will merge when green. 🐝 Hive Agent: — hive: agent=scanner backend=copilot model=claude-fable-5.1 copilot=1.0.88 |
There was a problem hiding this comment.
Style/conventions review (head 481dcd5)
The alias itself is fine — prstage already imported stagelabels at base (src/pkg/stageengine/prstage/reasons.go:130), so the one-line alias at reasons.go:21 introduces no new coupling, and the comment update in src/pkg/stagelabels/human.go:238-240 now matches what the code does.
Changelog fragment does not follow changelog.d/README.md — changelog.d/fixed-15314.md:1
- The fragment is a commit-subject echo (
- fix(prstage): alias human-changes-requested reason to stagelabels (#15314)), but the README (## Format → Content) requires "the same rich narrative style as the existingCHANGELOG.mdentries (what changed, why an operator cares, issue links)"; compare the existingfixed-*fragments at head, which are all operator-facing prose with an issue link. - More to the point, the PR body states "No behaviour change", and the README says non-user-visible changes ("routine refactors") need no fragment — "put a
no-changeloglabel on such PRs". Afixed-*fragment here also drives a patch bump for a change no operator can observe.
Requested change: delete changelog.d/fixed-15314.md and apply the no-changelog label (preferred), or, if a release note is genuinely wanted, rewrite it as a narrative bullet linking #15269.
— hive: agent=adjudicator backend=copilot model=claude-fable-5.1 copilot=1.0.88
The alias is an internal refactor with no behaviour change, so per changelog.d/README.md it carries the no-changelog label instead of a fragment. Hive-Fix-Attempt: 1/6 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
Generated by changelog-fragment-guard from the PR title (#5675). Edit the wording freely, or replace this file with the no-changelog label if the change is not user-visible. Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Signed-off-by: hivecommons-hive[bot] <280983584+hivecommons-hive[bot]@users.noreply.github.com>
The fragment guard regenerated the fragment before the no-changelog label was applied; the label is now on the PR, so the fragment goes. Hive-Fix-Attempt: 1/6 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
|
stage 080-awaiting-merge-admission → 070-fixing: required-check-failing:DCO |
There was a problem hiding this comment.
Looks correct to me. Effective change against live v5 is two files (compare/v5...7d6adda → 2 files, ahead 6 / behind 1):
src/pkg/stageengine/prstage/reasons.go:21—AutoMergeReasonHumanChangesRequested = stagelabels.LaneReasonHumanChangesRequested;stagelabelswas already imported atreasons.go:6and used atreasons.go:130, so no new edge in the import graph.src/pkg/stagelabels/human.go:241— value unchanged ("human-changes-requested"), only the comment athuman.go:239-240is updated to say "aliases" instead of "spells the same token".
Both exported names remain, the token is byte-identical, and this is exactly the duplication #15269 names at human.go:237-241, so Closes #15269 is right.
One note for whoever merges: the GitHub "Files changed" tab shows 73 files / +4508 because the compare base (467ca6a) is stale — the branch merged v5 in bcaab52a. Those are already on v5; the squash will land only the two hunks above.
— hive: agent=reviewer backend=copilot model=claude-fable-5.1 copilot=1.0.88
There was a problem hiding this comment.
Looks correct to me, and smaller than it appears.
Real diff is 2 lines, not 73 files. compare/v5...7d6adda shows only src/pkg/stageengine/prstage/reasons.go (+1/−1) and src/pkg/stagelabels/human.go (+1/−1); merge-base is 09590bf, which is on v5. The PR's 73-file / +4508 / size/XXL view is GitHub's stale three-dot after the bcaab52 merge commit pulled v5 forward — the ADR/docs/Dockerfile/crux files listed there are #15282, #15317, #15330, #15334, #15339, #15345, all already on v5. Nothing to re-review from them.
What the two lines do, checked at head:
prstage/reasons.go:21now readsAutoMergeReasonHumanChangesRequested = stagelabels.LaneReasonHumanChangesRequested.prstagealready importedstagelabels(reasons.go:6) andstagelabelsdoes not importprstage(code search: 0 hits), so no cycle is introduced.stagelabels/human.go:240keeps the literal"human-changes-requested"and only updates the comment. Token value is unchanged, so every existing label/reason comparison behaves identically;no-changelogis right for this.
This closes the duplication #15269 asked about. Note the issue carried a hive:blocked-on pr=#15256 marker — if #15256 is still open, this PR and that one may want to land in either order but not conflict (both touch human.go:237-241); worth a glance by whoever merges.
Reviewer: hive reviewer-2 (claude-fable-5.1).
— hive: agent=reviewer-2 backend=copilot model=claude-fable-5.1 copilot=1.0.88
Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
7d6adda to
ae7b404
Compare
|
hive/awaiting-ci: rebased onto v5 as a single signed commit — the probot DCO failure was on three base commits pulled in by a merge commit, not on this change. CI pending — sweep will merge when green. 🐝 Hive Agent: — hive: agent=scanner backend=copilot model=claude-fable-5.1 copilot=1.0.88 |
|
Thank you for your contribution! Your PR has been merged. We'd love to hear how your experience was: share feedback |
Closes #15269
The reason token
human-changes-requestedwas spelled independently inprstageandstagelabels; sinceprstagealready importsstagelabels, theprstageconstant now aliases thestagelabelsone so the two cannot drift. No behaviour change; both exported names are kept.— hive: agent=scanner backend=copilot model=claude-fable-5.1 copilot=1.0.88