Repository navigation
fix: complete the firefox recipe removal and gate partial removals - #131
hivecommons-hive[bot] wants to merge 1 commit into
Conversation
Commit 0312fbe deleted packages/firefox/ and stopped there. docs/contributing.md requires a recipe removal to reach four places; three were skipped, leaving 191 recipe directories against 192 source locks and 192 .packit.yaml blocks. Three unit tests fail on main as a result, so prepare is red there and on every PR branched from it. Delete the orphaned firefox lock, regenerate .packit.yaml, and move the four hardcoded package counts to the new set size. The counts only report that the set changed size. Add tests/test_recipe_set_agreement.py, which compares packages/, config/upstream-sources.json and .packit.yaml against each other and names the place a removal did not reach, and point docs/contributing.md at it. Assisted-by: Claude Opus 5 via GitHub Copilot Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Signed-off-by: sec-check[bot] <sec-check[bot]@users.noreply.github.com>
|
Important Held for human review by the hive's ACMM level gate. This PR was opened by the "scanner" agent while Hive policy required a human checkpoint for that agent. Non-outreach agents are held at ACMM L3–L5; the Hive will automatically remove the |
|
Closing: it conflicts across Generated by Claude Code |
Closes #130
Fix
0312fbedeletedpackages/firefox/and stopped there.docs/contributing.md§ Removing a package requires a recipe removal to reach four places; three were skipped, somaincurrently holds 191 recipe directories against 192 source locks and 192.packit.yamlblocks. Three unit tests fail as a result, which is whyprepareis red onmainand on every PR branched from it.Reproduced before the change:
Files and functions this PR claims
config/upstream-sources.json— delete the orphanedfirefox155.0lock (192 → 191 entries)..packit.yaml— regenerated withpython3 tools/render_packit_config.py --write, which drops thefirefoxblock (192 → 191specfile_path:blocks).tests/test_package_inventory.pyPackageInventoryTests.test_inventory_reports_every_recipe_fully_source_locked— count 192 → 191.tests/test_render_packit_config.pyRenderPackitConfigTests.test_rendered_config_matches_repository_file— count 192 → 191.tests/test_packit_srpm.pyPackitSrpmTests.test_workflow_stages_verified_sources_for_every_configured_package— count 192 → 191. This one was passing only because.packit.yamlwas also stale; regenerating without touching it would have turned one red test into two.tests/test_source_inventory.pyTargetSelectionTests.test_hummingbird_supplied_packages_are_skipped_by_default— count 191 → 190.tests/test_recipe_set_agreement.py— new,RecipeSetAgreementTests.docs/contributing.md§ Removing a package — point at the new test, and say to regenerate.packit.yamlrather than hand-edit the block.Why a new test rather than only bumping the numbers
tests/test_package_inventory.pyalready carried# 192: gcc was dropped from the rebuild set (...#69). The same partial-removal breakage happened at thegccremoval and was fixed by hand-bumping the magic number, which left nothing behind to catch the next one. The procedure is documented; nothing enforced it.RecipeSetAgreementTestscomparespackages/,config/upstream-sources.jsonand.packit.yamlagainst each other and fails by name:test_every_recipe_has_a_source_lock— a recipe with no lock is not eligible to build.test_no_source_lock_outlives_its_recipe— the exact state this PR is cleaning up.test_packit_config_covers_exactly_the_recipes— tells you to runrender_packit_config.py --write.Verified it actually catches the failure mode, by deleting
packages/fish/and leaving the other two places alone:The count assertions are kept — they still gate a deliberate change in set size — but they now sit behind a check that says where.
Verification
Scope
firefoxis correctly out of the rebuild set — #113 ships it as a Flatpak. This PR does not revisit that decision; it only completes the three steps the deletion did not reach.It does not touch
precedence/detect/verify, which are red onmainfor an unrelated reason (manifest for quay.io/fedora/fedora@sha256:a58a82ed... not found, exit 125). #116 and #122 already cover that and are awaiting review.Hold-gated: human review required.
— hive: agent=scanner backend=copilot model=claude-opus-5 copilot=1.0.78