Defer publication of `_qdrant_client` until after the in-lock backfill +
payload-index migration awaits complete. The fast-path check at the top of
`get_qdrant_client` reads the singleton without holding the init lock, so
publishing the constructed-but-unmigrated client let concurrent fast-path
callers fire filtered searches before `_ensure_payload_indexes` ran —
producing HTTP 400 ("Index required but not found") on Qdrant Cloud strict
mode. Local `provisional` is now used for every await inside the lock; the
global is assigned exactly once, last.
Replace the five hand-rolled `scroll(..., limit=10000)` calls in
`vector/scanner.py` (notes / files / news / deck-cards deletion tracking,
plus the timestamp scroll) with a single paginated `_scroll_all_points`
helper. The previous single-page cap silently dropped deletion-tracking
points beyond the first 10 k for any user past that threshold. Pagination
follows Qdrant's documented contract (loop until `next_page_offset is
None`) with a fixed per-page `_DELETION_TRACKING_PAGE_SIZE = 1024`.
Extract `_create_one_payload_index` from `_ensure_payload_indexes` to drop
its cognitive complexity below the SonarQube limit (17 → ≤ 15) without
losing the per-field error-containment rationale; every comment is
preserved verbatim on the helper.
Drop the stale `SearchResult.id` `int | str` comment and the redundant
`str(d)` coercion in `_verify_news_items` — the contract has been
str-only since the producer-side stringification landed earlier in this
PR.
Fix eight `doc_id=<int>` test calls in `test_chunk_context_offset_gate.py`
that violated the `doc_id: str` signature of `get_chunk_with_context`,
plus align `_make_result` in `test_verification.py` to coerce `id=str(...)`
matching the production contract — and update 30+ assertions from int
sets (`{1, 2, 3}`) to str sets (`{"1", "2", "3"}`) so the tests now model
the post-PR `SearchResult.id: str` reality end-to-end. Previously these
were masked by the `str(d)` coercion now removed from production.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- pdf_highlighter.compute_chunk_bboxes_batch: drop unused chunk_text
destructure (SonarQube finding), and replace positional
page_boundaries[page_num - 1] with a key-based next() match so
reordered or non-1-indexed boundaries can't silently shift the bbox.
Convert touched f-string log to lazy %s formatting.
- vector/processor: rename the trace_operation span from
"vector_sync.generate_highlights" to "vector_sync.compute_chunk_bboxes"
to match what the function actually does.
- Add test_compute_chunk_bboxes_handles_unordered_page_boundaries —
reverses the boundaries list and asserts identical results to the
in-order case, guarding the boundary-lookup regression class.
- Pin pre-push-review skill to sonnet model.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Add a fixed-UUID sentinel point written after a successful doc_id
backfill so subsequent restarts retrieve it and short-circuit the
O(N) scroll. Sentinel has no user_id/doc_id/doc_type payload so
production search filters never see it.
- Pre-fetch payload_schema in _ensure_keyword_payload_indexes and
silently skip fields that are already indexed; the "Created KEYWORD
payload index" INFO log fires only on actual creation.
- Narrow stale `int | str` doc_id annotations to `str` across
search/verification.py (BatchVerifier return type, per-verifier
accessible sets, by_type / accessible_by_type / inaccessible
collections); drop the now-redundant `type(d).__name__` prefix in
the dropped-docs log.
- Align the backfill log message with the PR description's
"Running doc_id backfill" promise; add a caller cross-reference to
the wait=True comment.
- Fix _get_file_path_from_qdrant docstring (file_id is str, not numeric).
- Convert legacy `id=1` to `id="1"` in test_search_result.py to match
the SearchResult.id: str annotation.
Three new unit tests cover sentinel-found, sentinel-written, and
skip-existing-index branches; existing backfill tests pass dimension
and explicit retrieve.return_value=[] for the no-sentinel path.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- chunk_bboxes is now dict[int, list[tuple[...]]] holding the bbox list
directly, not {"bbox": ..., "page": ...}. The page from text-search was
stored but never read; page_number from offset-based assignment is
authoritative for the Qdrant payload.
- Add two unit tests for the documented omission contract: chunks whose
offsets fall outside every page boundary, and chunks whose text cannot
be located on the rendered page, are silently dropped from the result.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- _ensure_keyword_payload_indexes: distinguish 400 (schema conflict, warning)
from other status codes (5xx/network, error) so a transient outage doesn't
silently leave the collection unindexed.
- build_search_result_from_point: use .get("doc_id") + return None on missing
instead of KeyError-crashing the search; reverse metadata merge order so
payload-derived chunk_index/total_chunks win over caller-supplied extras.
- docs/configuration.md: restore the OpenAI/Mistral/Bedrock/Simple provider
sections + reference-table rows that were dropped in the rebase. Reword
the "Startup migrations" bullet to describe what the code actually does
(no sampling — full scroll, zero writes when clean). Add operator note
about the SemanticSearchResult.id TypeError path.
- tests: pytest.approx for float equality (Sonar python:S1244); coverage
for non-400 → ERROR, payload={doc_id: None}, and missing doc_id key.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Drop chunk_bbox_page from Qdrant payload — viz endpoints never read it
(page_number is the canonical PDF page field).
- Bump upsert BATCH_SIZE 10 → 100 now that payloads no longer carry PNGs.
- compute_chunk_bboxes_batch: move doc.close() into finally, replace
unused stored_page_num with _.
- purge_page_images.py: switch to anyio.run() per project convention,
and wrap AsyncQdrantClient in try/finally so the aiohttp session is
always closed (the class doesn't implement async-context-manager).
- Decorate new bbox unit tests with @pytest.mark.unit so they run under
the fast-feedback selector.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Per-chunk PDF page renders (~150–700 KB base64 PNG each) were the dominant
disk consumer in production, repeatedly tripping `No space left on device:
WAL buffer size exceeds available disk space` on welcomed-malamute Qdrant.
Replace the inline highlighted_page_image / highlighted_page_number /
highlight_count fields with a small `chunk_bbox` field:
list[(x0, y0, x1, y1)] of normalized [0, 1] floats, ~32 bytes per chunk.
Astrolabe (the only known consumer) renders the highlight client-side as
a percentage-positioned overlay on top of the existing /api/v1/pdf-preview
render-on-demand path (cbcoutinho/astrolabe#76).
- pdf_highlighter: new compute_chunk_bboxes_batch() that reuses the
existing _find_chunk_bbox text-search path, skipping all pixmap/PIL/PNG
work.
- processor: store chunk_bbox + chunk_bbox_page in the Qdrant payload,
drop highlighted_page_image + friends, drop the base64 import.
- visualization /api/v1/chunk-context and auth/viz_routes: read
chunk_bbox instead of highlighted_page_image.
- vector/__init__: stop eagerly re-exporting `processor`/`scanner` —
fixes a pre-existing circular import (search.algorithms ->
vector.placeholder -> vector/__init__ -> processor -> scanner ->
server.semantic -> search.bm25_hybrid -> search.algorithms partial).
Test suite that was broken on master (test_bm25_hybrid.py et al.) now
collects and passes.
- scripts/purge_page_images.py: ad-hoc, idempotent migration that
delete_payload's the legacy keys from existing points. No reindex
required; legacy chunks render the page with no overlay.
Pairs with cbcoutinho/astrolabe#76. Frontend handles missing chunk_bbox
gracefully, so this can land in either order.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Addresses reviewer feedback on PR #773:
- Backfill set_payload now uses wait=True to avoid a race where
_ensure_keyword_payload_indexes builds the KEYWORD index before
fire-and-forget writes have committed, leaving int payloads
invisible to filters.
- Batch points sharing the same int doc_id into a single set_payload
call (one document → many chunks → one round-trip instead of N).
- Drop _has_int_doc_id_sample short-circuit. The sample's false-negative
window (clean first 256 results, ints further in) is gone; full scroll
is the dominant cost on first run anyway.
- Simplify _ensure_keyword_payload_indexes: the "already exists" 400
branch was dead code (Qdrant returns 200 on identical re-create); any
400 now logs a warning and continues.
- search/context.py: comment the broadened file-type guard. Add explicit
not doc_id.isdigit() checks at the top of note/news_item/deck_card
branches in _fetch_document_text so malformed payloads surface as
warnings instead of being swallowed by the broad except.
Also extracts build_search_result_from_point into search/algorithms.py
to deduplicate the 71-line payload-extraction loop shared by
SemanticSearchAlgorithm and BM25HybridSearchAlgorithm. This fixes
SonarQube's quality-gate failure (4.0% new-code duplication, max 3%).
Test coverage:
- 7 new unit tests for build_search_result_from_point covering missing
payload, note/file/deck_card metadata, int doc_id coercion, and
metadata_extras merging.
- Replace _has_int_doc_id_sample tests with clean-collection no-op and
per-batch grouping tests.
- Update set_payload assertions from wait=False to wait=True.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Six review items raised; four required code changes (#3, #4, #5, #6) and
two were resolved without code changes (#1 audit-only, #2 informational).
* search/verification.py — clarify the granularity asymmetry between the
whole-batch fail-open (structural API failure) and the per-item fail-open
(single bad stored doc_id). Future readers no longer need to derive why
the two paths have different blast radii from the code alone.
* models/semantic.py — `dropped_document_count` description now explicitly
notes that subtracting it from `verified_chunk_count` is not a meaningful
operation, since the two fields count different units (documents vs
chunks). Surfaces the unit mismatch where MCP clients actually see it.
* server/semantic.py — clarify the per-doc_type over-fetch comment so the
N×2 pre-merge Qdrant cost (vs the cross-app branch's 1×2) is explicit
rather than implied by "same 2× over-fetch budget".
* tests/unit/search/test_verification.py — add four new 429 unit tests
(notes/news/files/deck) mirroring the existing 5xx-keeps pattern. Locks
in that `_is_definitive_404_or_403` returns False for 429 so a future
refactor cannot accidentally treat rate-limit responses as permanent
revocations.
Audit confirmation for review item #1: all four `WebDAVClient.get_file_info`
call sites already handle the new `HTTPStatusError`-on-404 contract
(verification.py:156, tests/integration/test_rag.py:139,
tests/unit/client/test_webdav.py:153/190). No silent breakage internal to
this repo.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Address findings surfaced by `pre-push-review` after the round 8 sweep:
- Add deck verifier symmetry tests (404, transient 5xx, unexpected
exception, non-numeric metadata) so deck has the same shape as the
notes/news/files verifiers. Also add unexpected-exception tests for
the news and file verifiers, which had `except Exception` branches
no test was reaching. Keeps the registry-style verifier coverage
uniform.
- Modernize sibling field types in `VectorSyncState`, `AppContext`,
and `OAuthAppContext` from `Optional[X]` to `X | None`, matching the
`eviction_task_group: TaskGroup | None` field added in the round 8
diff (resolves the inconsistency flagged by A6). The lone remaining
`Optional` import is dropped.
- Reverse cross-reference direction in the verifier docstrings: the
later-defined `_verify_deck_cards` and `_verify_news_items` now
point at `_verify_notes` as the canonical hoisted-cast pattern,
rather than `_verify_notes` forward-referring to verifiers defined
below it.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Rename `verified_count` → `verified_chunk_count` to make the count
granularity explicit at the field name (chunks vs unique docs).
- News verifier now fails open *per-item* on non-numeric stored doc_ids
(matches notes/files/deck shape); a single bad id no longer rescues
definitively-missing siblings from eviction.
- Update note-verifier integration test to use string doc_ids end-to-end
to match production storage (scanner.py:241 stringifies note ids).
- Add regression test for the closed-task-group race guard in
`verify_search_results` so the RuntimeError swallow is locked in.
- Convert remaining f-string logger calls in `server/semantic.py` to
lazy %-style formatting (per repo convention).
- Document `evict_on_missing` as a developer/test flag (no env var) and
flag the `get_file_info` 404→raise contract change in its docstring.
- Add a TODO(ADR-019) breadcrumb for the hardcoded 2× over-fetch so
future tuning has a clear hook.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Round 7 raised 5 issues; this round addresses all of them and fixes
the underlying causes (not just the comments) where applicable so
they don't get re-flagged in future passes.
Critical:
- verified_count description in SemanticSearchResponse said "unique
documents" but the value is len(verified_results), a chunk count.
Description rewritten to accurately document chunk-level granularity
AND explicitly call out the asymmetry with dropped_count (which
counts unique (doc_id, doc_type) pairs).
- _verify_files false-eviction risk: the round-6 doc-only fix was
re-flagged. Address at the source — widen WebDAVClient.get_file_info
to raise HTTPStatusError on 404 (matching the rest of the client
convention) and reserve None for the genuinely ambiguous
malformed-PROPFIND case. _verify_files now keeps the result on None
(cannot tell whether the file exists) and evicts only on a
definitive HTTPStatusError 404. Tests updated; new test added for
the malformed-XML keep-result path.
Non-critical:
- News verifier semaphore lifetime now explicitly documented: one
slot held for one deduplicated fetch per search is the correct
backpressure behaviour.
- Cross-reference comments in _verify_notes / _verify_deck_cards no
longer claim "Mirrors X" pointing at functions defined later in
the file; now use direction-neutral "parallel implementation in".
- accessible_by_type is mutated by concurrent run_verifier tasks; a
comment explains why this is race-free under anyio's cooperative
multitasking (distinct keys per task, no await between read and
write) so a future reader doesn't add a redundant lock.
- Knock-on: tests/integration/test_rag.py wraps get_file_info in a
try/except for the new contract.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Closes out the remaining nits flagged in the round-6 review.
Critical:
- _verify_files contract comment now enumerates all None-return cases
(404 + malformed PROPFIND XML) and documents the false-eviction
trade-off; self-healing via re-indexing recovers
- int(r.id) cast at the SemanticSearchResult boundary now raises a
TypeError with explicit doc_type/value context instead of bubbling
up as an opaque "Search failed: ..." McpError
Design observations:
- nc_semantic_search_answer docstring documents the per-note
round-trip cost from the post-verification race guard
- News verification latency hint added to configuration.md
- SemanticSearchResponse exposes verified_count + dropped_count so
short result pages on high-ghost-density indexes are
distinguishable from genuine scarcity. verify_search_results now
returns (kept, dropped_count); production caller and tests updated
Minor:
- Comment clarifies the .get() fallback in verify_search_results is
defensive only (run_verifier always populates the entry)
- Eviction task-group guard narrowed from except Exception to
except RuntimeError (the only documented failure mode of
TaskGroup.start_soon on a closed group)
- Indexer logs a warning when a deck_card task is missing
board_id/stack_id, surfacing data-quality issues at index time
rather than at verification time
- New unit test covers the news verifier's non-numeric-id fail-open
path (one bad doc_id keeps the entire batch)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Tightens verifier consistency, closes test gaps, hardens the fire-and-forget
eviction snapshot, and routes the new concurrency knob through Settings.
- Pre-flight ``int()`` guard in ``_verify_notes`` mirrors ``_verify_deck_cards``,
so a non-numeric note id produces a type-specific log line instead of
falling through to the generic "unexpected error" branch.
- Adds explicit 403 tests for the file and news verifiers (symmetry with the
existing notes/deck 403 tests) plus a ``non_numeric_id_keeps`` test.
- ``AppContext`` and ``OAuthAppContext`` no longer snapshot
``_vector_sync_state.eviction_task_group`` at lifespan-yield time. Both
expose it as a ``@property`` that reads the singleton dynamically, removing
the order-sensitive race where a future startup-ordering change could
silently degrade fire-and-forget eviction to inline forever.
- Adds ``verification_concurrency`` (env var ``VERIFICATION_CONCURRENCY``,
default 20) to ``Settings`` with a dynaconf validator; ``verify_search_results``
resolves the cap lazily from settings when the caller doesn't override it.
- Enriches the news verifier TODO to call out that ``batch_size=-1`` is
intentional — a numeric ceiling would silently break correctness because
any item beyond the cap would be missing from ``present_ids`` and dropped.
- Updates ``Optional[TaskGroup]`` to ``TaskGroup | None`` per project style.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Implements fire-and-forget eviction (ADR-019 §"Lazy eviction"): the
search response no longer waits on Qdrant deletes, instead spawning
evict() on a long-lived lifespan-owned task group. Falls back to inline
eviction in modes without vector sync and in unit tests.
Also: harden _verify_news_items against non-numeric ids (fail open
instead of crashing the verifier); document the get_file_info None-on-404
contract; add INDEXED_DOC_TYPES single source of truth in vector/scanner.py
referenced by the CI-guard test; write a Verify-on-Read Latency Budget
section in docs/configuration.md covering the unbounded news.get_items
fetch. Closes the two remaining ADR-019 implementation checklist items.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Cap all_results to limit*2 after sort in the per-doc_types branch of
nc_semantic_search to bound over-verification (was unbounded N-types).
- Switch BatchVerifier from (client, doc_ids, user_id) to (client, results,
semaphore). Verifiers now read file paths and deck board/stack ids from
SearchResult.metadata instead of doing fresh Qdrant scrolls — eliminates
one duplicate round-trip per file/deck-card verification.
- Bound per-id verification concurrency with a shared anyio.Semaphore
(default 20, matching server/semantic.py context-expansion convention).
Prevents httpx pool exhaustion / rate limiting on large search pages.
- Propagate stack_id from Qdrant payload to SearchResult.metadata in both
bm25_hybrid.py and semantic.py (board_id was already propagated).
- Drop now-unused _resolve_file_path / _resolve_deck_metadata helpers.
- Drop redundant int(d) in requested predicate from _verify_news_items.
- Rewrite eviction comment to be honest about inline (not background)
execution and the resulting latency coupling.
- ADR-019 status: Proposed -> Accepted.
- Add news property to NextcloudClientProtocol.
- Widen SearchResult.id and SemanticSearchResult.id to int | str to match
BatchVerifier signature and document support for future string-id types.
- Flip openWorldHint to True on nc_semantic_search_answer (it calls into
Nextcloud via nc_semantic_search).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The vector index lags Nextcloud (5-min webhook cron + scanner interval),
producing ghost records for deleted/unshared documents until the next
reconciliation. Verify each unique document against Nextcloud at query
time, drop inaccessible results, and lazily evict the corresponding
Qdrant points.
Per-doc_type batch verifiers: notes/files/deck cards run concurrently
per id; news items use a single fetch + intersect to avoid the per-item
fetch-all amplification. Transient errors fail open (keep result, log
warning) — only definitive 4xx drops. Multiple chunks of the same doc
collapse to one verification call.
Wired into nc_semantic_search before the limit trim and before context
expansion. nc_semantic_search_answer's per-note re-fetch retained as a
sub-second race guard since verification now happens upstream.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Added support for two fusion algorithms (RRF and DBSF) to combine dense
semantic and sparse BM25 search results, with comprehensive documentation
and unit tests.
Changes:
- Added fusion parameter to nc_semantic_search and nc_semantic_search_answer tools
- Updated ADR-014 with detailed comparison of RRF vs DBSF fusion algorithms
- Added unit tests for fusion algorithm initialization and validation
- Updated search_method in responses to include fusion type (e.g., "bm25_hybrid_rrf")
Fusion Algorithms:
- RRF (Reciprocal Rank Fusion): Default, rank-based, general-purpose
- DBSF (Distribution-Based Score Fusion): Score normalization using statistics
RRF is recommended for most use cases due to its robustness and established
track record. DBSF may provide better results when retrieval systems have
very different score distributions.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Fix false-positive validation error where DBSF (Distribution-Based Score
Fusion) correctly produces scores > 1.0 but SearchResult validation
incorrectly rejected them.
**Root Cause**: SearchResult.__post_init__() enforced scores in [0.0, 1.0]
range, but DBSF sums normalized scores from multiple retrieval systems
(dense semantic + sparse BM25), resulting in scores like 1.55 when both
systems strongly agree a document is relevant.
**Changes**:
- Relaxed validation to allow any score ≥ 0.0 (algorithms.py:147-157)
- Updated SearchResult and SemanticSearchResult documentation to explain
score ranges for RRF ([0.0, 1.0]) vs DBSF (unbounded)
- Added comprehensive test coverage for both fusion methods
- Added DBSF fusion option to vector visualization UI
- Updated viz routes and vizApp() to support fusion parameter selection
**Testing**: All 157 unit tests pass, type checking passes, ruff passes
Fixes error: "Configuration error: Score must be between 0.0 and 1.0, got 1.1528953"
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
