Delete the dead hosted-vlt redirect-ledger helpers (#782)

[Mikola Lysenko (mikolalysenko)]

LLM Description written by Claude Code:claude-opus-5-5

Refs #782 (slice 1; the issue stays open for the items whose files open PRs change).

Summary

Deletes the hosted-vlt helpers that merged and healed from the pre-v5 redirect ledger. #277 removed their last production callers (rebase_vlt_edits and the ledger heal), and v5 never writes that ledger. Also gates the test-only jvm::apply::read_project_file behind #[cfg(test)].

Why

• Issue #782, register row E58 (register), living document part 3 (hosted, "vlt hosted" row and the E58 bullet).
• These pub functions read as if hosted vlt still heals from the redirect ledger. That misleads anyone working on hosted rollback (E33/E45).
• Leverage: B 0, U 0, D ≈1 (194 dead production lines), R L. After Make the vendored-to-hosted takeover atomic #1039 merged, this was the only backlog item whose files no open arch-refactor/* or agent/fix-* PR changes.

What changed

• patch/redirect/vlt.rs: deleted edit_dep_id, lock_node_ids, claims_key, carried_pin_original, carried_pin_ids, carried_pin_lines, same_slots, slot_value and recorded (the whole former "revert" section), along with their tests (claims_stop_at_the_version_boundary, a_superseding_edit_keeps_the_pristine_slots_of_a_carried_pin, the carried-brotli tail of the 1.3 brotli test and the claims_key loop) and the vlt_edit test helper.
• patch/redirect/vlt_heal.rs: deleted ledger_targets and its three tests. LedgerTarget's docs now describe lock_targets, the only constructor left, and say that record is always None. Removing that field means editing commands/scan/hosted/vlt.rs, which open PRs change, so it is left for a follow-up.
• New test lock_targets_name_every_owned_instance_of_the_purls. It covers what the deleted ledger_targets tests used to: peer contexts, slot [0] flags, the purl filter and the absent record. lock_targets had no unit test of its own.
• vendor/jvm/apply.rs: read_project_file is now #[cfg(test)] pub(crate). All of its callers are tests in jvm/{mod,gradle,scala_cli}.rs.

Deleted

git diff --stat origin/main: 3 files, +49 / −344.

• Production: +13 / −194.
• Tests: +36 / −150.

Behavior

None. Every deleted item had no production caller, which the build and clippy confirm. Wrappers (npm/, pypi/, gem/) are unaffected.

Test evidence

• cargo clippy --workspace --all-features -- -D warnings: clean.
• cargo test -p socket-patch-core --all-features --lib: 5772 passed, 4 failed. The failures are the known root-sandbox ones that also fail on main (copy_tree::relax_loop_must_not_traverse_symlinked_root, vlt_heal::an_unremovable_hidden_lock_keeps_every_store_entry, pypi_poetry::wire_write_failure_maps_error_and_leaves_lock_untouched, pypi_requirements::wire_failure_rolls_back_already_written_files). They chmod files, so they pass in CI.
• cargo test -p socket-patch-cli --all-features --test mode_migration_vlt --test in_process_rollback_hosted --test e2e_vlt --test e2e_redirect_vlt_build: 26 + 28 + 27 + 27 passed, 0 failed (the ignored real-vlt e2e tests run in CI).
• rustfmt was run on the three touched files only; all three were rustfmt-clean on main.
• CI on 948939c: 98 of 100 checks green. The two red ones are coverage-docker (sbt), which failed with a transient Maven Central 404 while building its Docker image, before any test ran, and the ci-ok aggregate it fed. Those jobs have been re-run once.

Risk

Low: only deletions and a cfg(test) gate.

Remaining in #782

VendorEntry::committed_artifact_intact (vendor/state.rs), go_sum_edit::remove_lines (#1103), UpstreamClient::seed_rubygems_sha256 and cargo_tag::copy_manifest_tag, plus the folded #746, #800 and #801 items. Each is in a file that an open PR changes.

🤖 Generated with Claude Code

---

Note

Low Risk
Deletion-only and a test-only visibility gate; no production callers of removed APIs.

Overview
Removes dead hosted-vlt redirect-ledger code left over after v5 stopped writing that ledger and #277 dropped its merge/heal callers. Hosted vlt rewrite/heal paths now rely on the lock itself (lock_targets), not ledger edits.

In patch/redirect/vlt.rs, deletes ledger/revert helpers (edit_dep_id, lock_node_ids, claims_key, carried_pin_*, recorded, slot comparison helpers) and their unit tests; trims unused vlt_lock_text imports. The hosted rewriter behavior is unchanged.

In patch/redirect/vlt_heal.rs, removes ledger_targets and three ledger-focused tests; documents that LedgerTarget is only built via lock_targets with record always None. Adds lock_targets_name_every_owned_instance_of_the_purls to cover peer contexts, flags, and purl filtering.

In vendor/jvm/apply.rs, read_project_file is #[cfg(test)] pub(crate) because only JVM planner tests call it.

^{Reviewed by Cursor Bugbot for commit 948939c. Configure here.}

---

Generated by Claude Code

[Mikola Lysenko (mikolalysenko)]

BugBot review

---

Generated by Claude Code

[cursor]

Stale Bugbot comment from a previous run.

[Mikola Lysenko (mikolalysenko)]

[agent] coverage-docker (sbt) failed while building tests/docker/Dockerfile.sbt: curl got a 404 from Maven Central for mill-dist-0.12.17-mill.sh before any test ran. This PR doesn't touch the Dockerfile or anything sbt/Mill, and the same job on main (823810a) passed at 15:18Z with the same pinned URL, so it is a transient registry miss. No fix to port. I'll re-run the failed job once when the workflow run completes.

---

Generated by Claude Code

[Mikola Lysenko (mikolalysenko)]

bugbot run

---

Generated by Claude Code

[cursor]

✅ Bugbot reviewed your changes and found no new issues!

Comment @cursor review or bugbot run to trigger another review on this PR

^{Reviewed by Cursor Bugbot for commit 948939c. Configure here.}

[Mikola Lysenko (mikolalysenko)]

Burn-down agent: labeled Ready for review at 948939c.

• CI: all non-skipped checks green on this head (ci-ok success).
• Bugbot: re-reviewed 948939c on request, no findings; no unresolved review threads.
• Mergeable: yes (merges cleanly with current main).

---

Generated by Claude Code