feat(routing): Phase-2 aggregate reclaim — verify_target (slice 3) #21
+404
−8
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
reclaim.py must verify a size-routed target as an AGGREGATE or every routed target
perpetually false-DRIFTs (its newest log is one object of few files). Adds:
- verify_target(manifest, holds, …): reconstruct the whole target from its manifest —
expected set = ⋃ objects' arcnames, archived = Σ n_files, archived_bytes =
Σ source_bytes; require EVERY object tar present (hsi_exists per object); newer-than-run
is PER-OBJECT (a live member drifts iff newer than the run of the object that owns it).
Same verdict vocabulary + INV-E exclusion re-gate as verify() (the manifest carries the
same `exclusion` block, so recheck_exclusions works on it directly). Returns the standard
result dict + all_tars.
- find_manifests(log_dir): load {log_dir}/_vault/*.manifest.json (the aggregate indexes).
- main(): --scan skips routed OBJECT logs (routed_object marker) and verifies each manifest
via verify_target; --config with size_routing resolves its badge → manifest → verify_target;
--match filters manifests too. Legacy/non-routed logs flow through verify() unchanged.
- delete_source: guarded re-confirm of ALL object tars (all_tars) before rm, not just one.
Tests: tests/test_routing_reclaim.py (13) — SAFE / count-drift / newer-drift / bytes-drift /
missing-tar FAILED / no-objects FAILED / GONE / HOLD; INV-E no-longer-covered → DRIFTED and
still-covered subtraction → SAFE; find_manifests; multi-tar delete ABORT. Suite 196 → 209.
Based on main (NOT stacked). See docs/RFC_incremental_v2.md §2.6.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Jarrod S Doucette (jdoucett)
added a commit
that referenced
this pull request
Jul 10, 2026
…holds) Merging slice-3 (#21, verify_target) and typed holds (#22) composed a latent bug: #22 makes load_holds return typed dict entries, but verify_target still did inline string matching (h in source_folder), which raises TypeError on a dict hold — i.e. verify_target would crash on any routed target the moment a reclaim_holds.json is used. Legacy .txt (string) holds were unaffected, so the merged suite stayed green and it went unseen. Route verify_target through the same hold_entry_for + classify_hold helpers verify() uses, so routed targets get the full business/promote-pending/superseded tiering (HOLD / AWAITING- APPROVAL) instead of a flat crash-prone string match. Tests: verify_target under a business dict hold -> HOLD, promote-pending no-receipt -> HOLD, promote-pending valid-receipt -> AWAITING-APPROVAL. Suite 235 -> 238. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Sign in
to join this conversation on GitHub.
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Phase 2, Slice 3 — aggregate reclaim (
verify_target)Required before any routed target can be reclaimed: a size-routed target ships as many
objects, so reclaim's per-log
verify()sees only the newest object's few files and wouldperpetually false-DRIFT every routed target. This makes reclaim reconstruct the whole
target from its manifest. Based directly on
main(not stacked). Suite 196 → 209 green.verify_target(manifest, holds, …)— expected set = ⋃ objects' arcnames, archived =Σ
n_files, archived_bytes = Σsource_bytes; every object tar must be present(
hsi_existsper object); newer-than-run is per-object (a live member drifts iff newerthan the run of the object that owns it). Same verdict vocabulary + INV-E exclusion
re-gate as
verify()(the manifest carries the sameexclusionblock, sorecheck_exclusionsworks on it directly). Returns the standard result dict +all_tars.find_manifests(log_dir)— loads{log_dir}/_vault/*.manifest.json.main()—--scanskips routed object logs (routed_objectmarker) and verifieseach manifest via
verify_target;--configwithsize_routingresolves badge → manifest;--matchfilters manifests too. Legacy / non-routed logs flow throughverify()unchanged.
delete_source— guarded re-confirm of all object tars beforerm, not just one.Tests —
tests/test_routing_reclaim.py(13)SAFE • count-drift (extra file) • newer-drift • bytes-drift • missing-tar → FAILED •
no-objects → FAILED • GONE • HOLD • INV-E no-longer-covered → DRIFTED • INV-E still-covered
subtraction → SAFE •
find_manifests• multi-tar delete ABORT. Existing 196 unchanged.Phase 2 status after this
Slices 1+2 in
main; this is slice 3. Remaining: slice 4 (reship-monitor), slice 5 (fullrestore). With 1–3, a routed target archives (append) and reclaims correctly.
🤖 Generated with Claude Code