fix(vector-sync): address PR review — dict guard, symmetric backstop, metrics

- purge route: 400 (not 500) on a valid-JSON non-object body
- scanner: backstop-purge admin-disabled note/news_item/deck_card points
  (their deletion-tracking lives inside the skipped scan_* fns), mirroring the
  files path; gated on a concrete allow-set so fail-open never deletes
- processor: record_ingest_dropped("admin_disabled") so consent-skipped index
  tasks are observable/alertable
- app.py: list /api/v1/vector-sync/purge in the endpoints log line
- capabilities: drop empty-string doc types; return frozenset throughout
- purge: document the count-before-delete approximation
- tests: non-object body -> 400, ProvisioningRequiredError -> 428, cache TTL
  expiry refetch, and the scanner consent backstop

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Chris Coutinho
2026-06-16 00:54:50 +02:00
co-authored by Claude Opus 4.8
parent ef5b3f3873
commit 477fb02b0a
10 changed files with 258 additions and 43 deletions
+7 -3
View File
@@ -35,6 +35,11 @@ async def purge_doc_types(doc_types: list[str]) -> dict[str, int]:
deletion). Each doc type is purged independently so a failure on one does
not abort the rest; failures re-raise after the loop only if every doc type
failed, otherwise partial progress is returned.
The count is taken just before the delete (two separate Qdrant calls), so
it is approximate — a point indexed in the gap is deleted but not counted.
This is acceptable: indexing of a disabled source is already gated upstream,
so the window is effectively empty in practice.
"""
qdrant_client = await get_qdrant_client()
collection = get_settings().get_collection_name()
@@ -61,10 +66,9 @@ async def purge_doc_types(doc_types: list[str]) -> dict[str, int]:
)
except Exception as exc: # noqa: BLE001 — record and continue
last_error = exc
logger.error(
"Failed to purge indexed points for doc_type=%s: %s",
logger.exception(
"Failed to purge indexed points for doc_type=%s",
doc_type,
exc,
)
if not purged and last_error is not None: