fix(projection): a re-keyed record is new, not a renamed file
All checks were successful
CI Smoke / host-smoke (push) Successful in 0s
CI Smoke / container-smoke (push) Successful in 1s
Build and Publish Multi-Context Image / build-and-push (push) Successful in 27s

The ad-hoc requalification changed 30 identifiers while deliberately keeping
the filenames, per canon. The matcher's path fallback therefore matched every
old row to its new record and updated it in place — so the files said
CUST-WP-ADHOC-2026-08-25 while the hub still said adhoc-2026-08-25, with the
row's UUID still encoding the old identifier. Exactly the file/hub identity
divergence ADR-007 exists to prevent.

Path matching cannot distinguish a re-key from a rename when the path does not
change. Identity can: a derived row is UUIDv5 over the record id, so a changed
id is a different record — the old row retires and the new one is created.
Rows predating derived identity carry v4 UUIDs, where the identifier is a label
rather than the identity; those keep path matching, so rename detection still
works where it is the right answer.

The unmatched old row keeps its own slug as key, so it lands in `stale` and
becomes a retirement candidate rather than disappearing from the outcome.

720 pass.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 2583210@bnt-lap001
Assistant-Session: f2bff2d5-e9b2-4338-92ca-10282a927006
This commit is contained in:
tegwick 2026-08-28 00:38:50 +02:00
parent 4f68c5609f
commit b2efadf254
2 changed files with 71 additions and 0 deletions

View file

@ -32,6 +32,18 @@ _HEADING_RE = re.compile(r"^(#{1,4})\s+(.+?)$", re.MULTILINE)
DEFAULT_FORGE_BASE = "https://forgejo.coulomb.social/coulomb"
def _identity_is_derived(row: Any) -> bool:
"""Whether this row's UUID was derived from its identifier (ADR-007).
Derived rows are UUIDv5 over the work-record namespace. Rows predating
derived identity carry a random v4 UUID, and for those the identifier is a
label rather than the identity so the path fallback still applies, which
is what keeps rename detection working for legacy records.
"""
rid = getattr(row, "id", None)
return getattr(rid, "version", None) == 5
def derived_record_uuid(record_id: str) -> str:
return str(uuid.uuid5(_WORK_RECORD_NAMESPACE, f"helixforge\n{record_id}"))
@ -629,6 +641,24 @@ async def reset_repository_projection(
for row in rows:
key = (row.slug or "").strip().lower()
if key not in want:
if _identity_is_derived(row):
# ADR-007: a derived record's identity *is* a function of its
# identifier, so a changed identifier means a different record —
# the old one retires and the new one is created. Matching it to
# the file by path instead would update the row in place while
# its UUID still encodes the old identifier, leaving the file
# and the hub disagreeing about what the record is called.
#
# This is not hypothetical: the ad-hoc requalification
# (CUST-WP-0066) deliberately keeps the filename, so for those
# records a re-key *never* changes the path. Path matching
# cannot tell a re-key from a rename there, and silently chose
# rename for all 30 of them.
#
# The row keeps its own slug as key, so it lands in `stale` and
# becomes a retirement candidate rather than vanishing.
matched[key] = row
continue
bp = row.backing_relative_path
if bp and _path_key(bp) in want_paths:
key = want_paths[_path_key(bp)]

View file

@ -684,3 +684,44 @@ class TestForgeCredential:
monkeypatch.setattr(fp.subprocess, "run", fake_run)
assert fp._run_git("status") == "ok"
assert "GIT_CONFIG_COUNT" not in seen["env"]
class TestRekeyIsNotRename:
"""A changed identifier is a new record, not a moved file.
The ad-hoc requalification (CUST-WP-0066) keeps the filename by design, so
for those records a re-key never changes the path. Matching on path there
cannot distinguish the two cases, and choosing rename updates a row whose
UUID still encodes the old identifier leaving the file and the hub
disagreeing about what the record is called.
"""
class _Row:
def __init__(self, id, slug, path):
self.id = id
self.slug = slug
self.backing_relative_path = path
self.projection_retired_at = None
self.status = "finished"
def test_a_derived_row_is_not_matched_by_path(self):
"""uuid5 means identity is a function of the identifier."""
import uuid as _u
old = self._Row(
_u.UUID(fp.derived_record_uuid("ADHOC-2026-08-25")),
"adhoc-2026-08-25",
"workplans/ADHOC-2026-08-25.md",
)
assert fp._identity_is_derived(old)
def test_a_legacy_row_keeps_path_matching(self):
"""v4 rows predate derived identity; there the identifier is a label."""
import uuid as _u
legacy = self._Row(_u.uuid4(), "repo-wp-0001", "workplans/REPO-WP-0001.md")
assert not fp._identity_is_derived(legacy)
def test_the_two_uuids_actually_differ(self):
"""Guards the premise: if they were equal the distinction is moot."""
assert fp.derived_record_uuid("ADHOC-2026-08-25") != fp.derived_record_uuid(
"CUST-WP-ADHOC-2026-08-25"
)