Round-3 claude-review finding:
- 🟡 Double _ensure_bearer() on gateway.embed_batch(). Round 1 made
OpenAIProvider.embed_batch() delegate to embed_batch_with_usage(); because
GatewayProvider overrode both embed_batch() and embed_batch_with_usage() (each
calling _ensure_bearer), gateway.embed_batch() refreshed the bearer twice
(the second a cache-hit no-op). Remove the now-redundant embed_batch()
override: OpenAI's embed_batch() routes through embed_batch_with_usage(),
which the gateway still overrides, so the bearer refreshes exactly once on
every path. The remaining two overrides (embed + embed_batch_with_usage) cover
all four entrypoints; documented the topology.
- 🟢 Added test_gateway_embed_batch_ensures_bearer_once locking in the single
refresh.
Cohere token-fallback (🟢 nit) is already covered by
test_bedrock_with_usage_estimates_when_token_count_absent.
Deck #67.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Round-2 claude-review findings:
- 🟡 Base-class recursion invariant: documented on embed_with_usage /
embed_batch_with_usage that a provider overriding embed()/embed_batch() to
delegate to the *_with_usage variant MUST also override that variant, or the
two recurse. (No recursion today; the shipped providers pair the overrides.)
- 🟡 Processor metering had no unit test: extracted the two-event recording
into a module-level record_indexing_usage() helper and added
tests/unit/test_processor_metering.py (value mapping, flag/zero-chunk no-ops,
best-effort failure swallowed).
- 🟡 SonarQube hotspots (python:S5332) were 3 http:// URLs in the new test
fixtures (mock hosts, never contacted) blocking the quality gate
(new_security_hotspots_reviewed). Switched them to https:// so no hotspot is
raised.
- 🟢 Zero-chunk guard: record_indexing_usage() no-ops when chunk_count == 0, so
an empty document no longer writes zero-value billing rows.
Deferred (stated on the PR): Mistral x.index-or-0 sort key (pre-existing,
equivalent), CHANGELOG note for the Ollama /api/embed switch (CHANGELOG is
commitizen-generated from commit bodies, which document it), class-var
query_token_count (safe under the per-request instance pattern).
Deck #67.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Round-1 claude-review findings:
- 🔴 Multi-doc_type search billed N embedding calls as 1. nc_semantic_search
loops search() once per doc_type on one BM25HybridSearchAlgorithm instance,
and each call re-embedded the query, so only the last query_token_count was
recorded. Cache the dense embedding per query on the (per-request) instance
so the query is embedded — and metered — exactly once regardless of how many
doc_types are searched. This also removes the redundant per-type embed work
and avoids billing a user N× for one logical query.
- 🟡 Ollama embed() now delegates to embed_with_usage() so single and batch
embeds use the same /api/embed endpoint (was the legacy /api/embeddings),
keeping _detect_dimension and other embed() callers consistent.
- 🟢 round() instead of truncating int() when coercing provider-reported token
counts (forward-compatible if a provider ever returns a float).
Tests: per-instance query-embedding cache (embedded once across 3 doc_types;
re-embeds on a different query).
Deferred (stated on the PR): mistral/openai single-embed dual path (changes
tested error/request semantics on the cloud-critical path — separate refactor),
bedrock boto3 sync-in-async (pre-existing; no new invoke_model calls per doc).
Deck #67.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
embeddings_queries now records the embedding request's token count (the unit
upstream providers bill on) instead of an operation count, and fires on the
indexing path too. Previously only semantic search recorded it (value=1), so a
re-indexing run produced no embeddings_queries events at all — only pages_chunks.
- Provider layer: additive embed_with_usage / embed_batch_with_usage surface the
per-request token count (Mistral/OpenAI usage.total_tokens, Bedrock Titan
inputTextTokenCount, Ollama prompt_eval_count); a char-based estimate is the
fallback (Simple, and any provider/response without a token field). Gateway and
EmbeddingService forward through. The count travels as a return value / a
per-request SearchAlgorithm attribute — never on the singleton — so concurrent
indexing + search can't mis-attribute bills.
- Indexing (vector/processor.py): records embeddings_queries (value=batch tokens)
alongside the existing pages_chunks event.
- Search (server/semantic.py): value is now the query embedding's token count,
relayed from BM25HybridSearchAlgorithm via query_token_count.
The astrolabe_embeddings_queries Stripe meter (sum aggregation) now sums tokens
with no CP/Terraform change. The meter "queries"->tokens naming/unit
clarification (homelab-terraform #254) + CP rollup/portal copy is a follow-up.
Deck #67.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Simplify the OCS status guard to `if status and status != "ok"` — falsy
(missing/None/"") is tolerated more naturally than the explicit tuple.
- Add `test_value_error_from_ocs_failure_returns_none`, covering the
OCS-failure ValueError flowing through `_get_enabled_apps_or_none` to the
scan-all fallback.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Correct the misleading `test_malformed_envelope_returns_empty_set`
docstring: an empty-set return does NOT trigger the
`_get_enabled_apps_or_none` scan-all fallback (which fires only on
exceptions); optional apps are gated off for that cycle, Files unaffected.
- Check OCS `meta.status` in `get_enabled_apps`: a 200 carrying
`status != "ok"` now raises, so a 200-with-failure envelope routes through
the scanner's scan-all fallback instead of silently gating every app off.
Add a test for the failure-status raise.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Extract `_app_enabled` to a module-level helper so the gate predicate is
unit-tested directly instead of via an inline copy that could drift.
- Move `import logging` to module scope in test_scanner_app_gating.py.
- Harden `get_enabled_apps` OCS-envelope parsing (`X or {}` / `or []`) so a
present-but-null `ocs`/`data` coerces to empty instead of raising on
`None.get`; add parametrized malformed-envelope tests.
- Use https:// in the test request URL (SonarCloud S5332 hotspot).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The vector-sync scanner polled every indexed app (Notes, Files, News,
Deck) for every provisioned user on each scan cycle. When a user lacks
an app, its REST API returns 404; these were caught (indexing
continued) but flooded tenant logs with repeated 404s, scaling with
users x disabled-apps x scan-frequency and masking real failures.
Add NextcloudClient.get_enabled_apps(), which reads the per-user
/ocs/v2.php/core/navigation/apps endpoint (respects group
restrictions). Chosen over /cloud/capabilities because the News app
advertises no capability and never appears there.
scan_user_documents now resolves the enabled-app set once per cycle and
skips the Notes/News/Deck scans for apps the user lacks. Files stays
unconditional (core Tags API, not a 404 source). Detection failures
fall back to scanning every app (prior behaviour), so a transient
nav-endpoint blip never silently halts indexing; the per-app 404 guards
remain as the safety net.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Assert the open card stays visible under status="open" in the
deck_get_stack integration test (completes the partition check).
- Move the _append_archived_cards docstring closing quotes to their own line.
deck_get_stack's status="archived" + include_cards=False path is left as-is:
a single get_stack call is the cheapest way to obtain the stack metadata
there — routing it through the archived fast-path would fetch every archived
stack on the board just to strip the cards, which is heavier, not lighter.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- deck_get_stack: fetch active + archived concurrently for status="all", and
for status="archived" source the stack from /stacks/archived in a single
call (skip the active fetch whose open cards are filtered out anyway),
matching deck_get_cards' pattern.
- Type the `client` param of _archived_cards_by_stack as NextcloudClient.
- Extend the stacks/overview integration test to assert status="archived"
(only the archived card) in addition to status="all".
- Document the third_party/astrolabe submodule mount policy in CLAUDE.md:
unmounted by default (CI installs the published app-store version); mount
only for tightly-coupled feature work needing CI integration, then revert.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The active Deck listing endpoints (StackService::findAll /
CardMapper::findAllForStacks and StackService::find / CardMapper::findAll)
filter out archived cards at the SQL level — only the /stacks/archived
endpoint returns them. The client-side status="all"/"archived" filters in
deck_get_cards, deck_get_stacks, deck_get_stack and deck_get_board_overview
therefore operated on a list the server had already stripped of archived
cards, so they could never surface one. deck_get_card (by ID) bypasses the
filter, which is why it appeared to work. Fixes#842.
When status is "all" or "archived", also fetch /stacks/archived
(client.deck.get_archived_stacks) and merge those cards back in per stack —
concurrently with the active fetch where applicable. status="open"/"done"
are unchanged and cost no extra call.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- semantic.py: normalize both None and [] doc_types to null in the
metadata so a future `metadata->'doc_types' IS NULL` query counts the
all-types case consistently.
- test: use a fixed past date in test_occurred_at_roundtrip instead of a
future literal (deterministic, no "why this date" confusion).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- hooks: document why user_id in metadata is safe — it stays tenant-local
(the CP rollup aggregates GROUP BY (day, metric) into usage_daily, which
has no metadata column, so it never reaches Stripe) and is retained to
keep Deck #67's future per-user attribution derivable from the app DB.
- migration: instantiate the SQLite-side column types (sa.Text() etc.) for
visual parity with the instantiated Postgres types.
- tests: assert the WARNING contract in the unserializable-metadata test
too; add an autouse fixture that resets UsageEventStore._shared_instance
so a stray shared() call can't leak across tests.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- remove accidentally-committed .claude/scheduled_tasks.lock (Claude Code
runtime artifact swept in by `git add -A`) and gitignore it; the rest
of .claude/ stays tracked.
- store: cache UsageEventStore.shared() as a process-wide instance so the
hot search path doesn't allocate a fresh wrapper per metered query (the
wrapper is stateless beyond its storage handle).
- hooks: pass enabled=True directly (the outer guard already confirmed
the flag) instead of re-reading settings.usage_metering_enabled.
- migration: document the no-TTL retention design (control-plane rollup
owns the lifecycle; the data plane only appends).
- tests: assert the best-effort error path logs at WARNING (observability
contract).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- store: add optional `enabled` param to record_usage_event so hot-path
callers (nc_semantic_search) pass the already-resolved flag instead of
forcing a second uncached Settings build (ADR-024); falls back to
get_settings() when None so the store stays self-gating for standalone
use.
- hooks: thread enabled= through both call sites; bump the outer
shared()/construction failure log from debug → warning so "metering
enabled but no billing data" is visible at the default INFO level.
- migration: instantiate postgresql.JSONB() to match the sibling
TIMESTAMP(timezone=True) column.
- tests: fix the misleading "asyncpg returns JSONB as a JSON string"
comment; add occurred_at dialect round-trip test and an enabled-param
short-circuit test.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Deck #67 data-plane slice: tenant Pods record billable operations
(embedding queries, pages/chunks embedded) into an app-DB usage_events
table that the control plane later pulls read-only into the billing
ledger and syncs to Stripe Meter Events.
- migration 007: usage_events table (Postgres TIMESTAMPTZ/JSONB/UUID
with portable SQLite fallbacks), indexed (occurred_at, metric) for the
CP rollup's per-day range scan + GROUP BY metric.
- UsageEventStore: best-effort, flag-gated writer reusing the shared
RefreshTokenStorage engine; ON CONFLICT (event_id) DO NOTHING for
idempotent retries; dialect-branched occurred_at bind. All work
(incl. metadata JSON encode) is swallowed so a metering failure never
surfaces to the user op.
- USAGE_METERING_ENABLED flag (default off) wired through Settings +
env map; off-path touches no storage, so OSS self-hosters get an empty
table and zero write overhead.
- two recording hooks: embeddings_queries (per nc_semantic_search, which
nc_semantic_search_answer reuses) and pages_chunks (after dense
embedding succeeds, covering both in-process and procrastinate paths).
- storage.acquire()/.dialect public seams so the sibling store doesn't
reach into the underscored internal.
- tests parametrized over SQLite + Postgres: flag-off no-op, roundtrip,
ON CONFLICT dedup, JSON/NULL metadata, and the best-effort swallow of
both DB errors and unserializable metadata.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address claude-review round 5 on PR #868: add
test_all_blank_pages_returns_empty_list documenting that PageAwareChunker
returns [] when every page is blank — and asserting parity with
DocumentChunker, which already returns [] for whitespace-only non-empty
content. The empty-chunk-list case is therefore pre-existing pipeline
behavior, not new to this PR.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address claude-review round 3 on PR #868: add module-level
`pytestmark = pytest.mark.unit` so TestPageAwareChunker and
TestDocumentChunkerPositions are collected under `-m unit`, matching
test_processor_routing.py.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address claude-review round 2 on PR #868:
- Extract the use_page_aware branching into a pure `should_use_page_aware`
helper and cover the (doc_type, page_boundaries, page_aware_setting) matrix
in tests/unit/test_processor_routing.py (file+boundaries+enabled, empty
list, None, non-file doc types, disabled setting).
- Clarify the PageAwareChunker.chunk_text no-boundaries comment: the processor
pre-filters via should_use_page_aware, so that branch is a direct-call safety
net, not a production indexing path.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address claude-review round 1 on PR #868:
- use_page_aware now gates on `bool(page_boundaries)` instead of
`is not None`, so a PDF that yields an empty boundary list takes the
char-based path explicitly (assign_page_numbers no-ops on []) rather than
the page-aware chunker's no-boundaries fallback. Same result, clearer intent.
- add test_oversized_page_with_leading_whitespace_offsets, exercising the
start+start_index offset path for an oversized page whose sub-chunks have
leading whitespace.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add PageAwareChunker, which splits paginated documents (PDFs) on page
boundaries first and only character-splits pages larger than chunk_size.
No chunk spans a page boundary, so page_number is always exact and stored
excerpts never lead with a neighbouring page's text. When chunk_size is at
least the largest page, this yields exactly one chunk per page: a
predictable vector count (== page count), a flat per-page embedding cost,
and zero cross-page overlap duplication.
Gated by DOCUMENT_CHUNK_PAGE_AWARE (default true). When false, the legacy
char-based DocumentChunker + post-hoc assign_page_numbers path runs
unchanged. Only doc_type="file" with page_boundaries (PDFs) takes the
page-aware path; notes/deck/news are unaffected.
Measured on a 15-page record (query "leadership award louis", target =
top-half of page 15): char-based degraded the target to dense-rank 10 at
cs=2048 (OCR) and mislabeled its page; page-aware restored rank 1 across
every fusion/modality and chunk size, with correct page labels and clean
snippets.
BREAKING CHANGE: PDFs are re-chunked page-aware by default. Existing
deployments will re-index PDF content on the next vector sync (different
chunk counts and page_number labels). Set DOCUMENT_CHUNK_PAGE_AWARE=false
to retain the previous char-based behaviour.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Tagging an existing file/folder emits only OCP\SystemTag\MapperEvent — never a
Node*Event — so tagged PDFs were previously only picked up by the hourly
scanner. Subscribe to the tag event and reconcile membership so adding/removing
the `vector-index` tag (re)indexes in near-real time.
- webhook_presets: add OCP\SystemTag\MapperEvent to the files_sync preset
(NC 32+, where MapperEvent gained getWebhookSerializable(); harmless on older
servers — it just never fires).
- webhook_parser: parse MapperEvent (objectType=files) into a path-less file
"reconcile" task. The payload carries only a fileid + tagIds (no name/path),
so assign and unassign both collapse to a reconcile.
- processor._reconcile_tag_event: resolve the fileid against the user's current
vector-index PDFs (find_files_by_tag). Present -> index with the resolved
path/etag; absent -> flip to delete. Naturally handles "an unrelated tag
changed" and a tagged folder's own fileid (no-op; the scanner still expands
folders to descendants).
- Unit tests for the parser branch and the reconcile.
The matching admin-UI preset change ships separately in the astrolabe app repo.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #863 round 3:
- classify_from_text emits the "scanned" flag (was "no_text_layer") for the
empty-text-layer case -- same name + meaning as classify_pdf, so
astrolabe_document_classifier_flag_total isn't split across two labels for the
same concept (and matches the metric's documented vocab).
- classify_from_text logs at DEBUG when image_coverage length != the expected
min(pages, MAX_SAMPLED_PAGES), so a 1:1-alignment contract break (extractor
reorders/skips pages) surfaces instead of silently misattributing coverage.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #863 round 2:
- classify_pdf now flags a page needs_ocr on the SAME three signals as
classify_from_text (image scan OR low text-quality OR near-empty), not image
coverage alone. Previously a word-merged digital doc with no images routed
"fast" via classify_pdf but "ocr" via the pipeline -- so an operator
reproducing routing offline got a different answer. They now match.
- Add a test that when image_coverage is shorter than the page boundaries (the
MAX_SAMPLED_PAGES cap on large scans), the leading page uses the scan signal
and later pages fall back to text-quality.
Left as-is: overlong_score (>20) partially overlaps merge_score (>12) -- the
double-penalty on very-long tokens is intentional, not a bug (per review).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #863 review:
- MIN_TEXT_QUALITY 0.45 -> 0.5 so the module/diagnostic default matches the
DOCUMENT_OCR_MIN_TEXT_QUALITY setting (registry always passes the setting; this
keeps classify_pdf and the test/default path on the production threshold).
- image_coverage_per_page is bounded to MAX_SAMPLED_PAGES (the image pass is the
costly part, so a 200-page scan isn't fully rasterised on the hot path); pages
beyond the cap fall back to the text-quality signal, and page_fraction still
gates over every page.
- Extracted _page_image_coverage(page) helper, shared by classify_pdf and
image_coverage_per_page (DRY + keeps the tiling-double-count note in one place).
- Scan-detection failure logs at WARNING (not DEBUG) so a systematic failure on
an OCR-enabled tenant is visible at LOG_LEVEL=INFO.
- Add the missing DOCUMENT_OCR_MIN_PAGE_CHARS range-validator test.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The hot-path classifier escalated to OCR purely on character count, so a
scanned/handwritten PDF with a low-quality embedded text layer (>16 chars/page
but garbled) routed `fast` and indexed the junk -- e.g. Student 147.pdf's
"Little Acoms Primary"/"0110912020", which pollutes the vector and demotes the
doc in search (Deck #207).
- classifier: recalibrate `_text_quality` with a long-token-fraction term that
detects word-merging (dropped inter-word spaces) -- the dominant junk-layer
failure the old whitespace/overlong(>20) terms missed. Measured: the Student
147 scan ~0.42 (60% pages junk) vs >=0.94 for clean digital docs.
- classify_from_text now routes on quality + scan: a page is OCR-worthy if
near-empty OR low text-quality OR (when OCR + scan detection are enabled) it's
mostly a raster image. New `image_coverage_per_page` re-opens the PDF for the
scan signal, so that cost is paid only by OCR-opted-in tenants. Thresholds are
passed in from per-tenant settings (keyword-only).
- config: 4 per-tenant settings -- DOCUMENT_OCR_MIN_TEXT_QUALITY (0.5),
DOCUMENT_OCR_PAGE_FRACTION (0.5), DOCUMENT_OCR_MIN_PAGE_CHARS (16),
DOCUMENT_OCR_DETECT_SCANNED (true) -- with range validators.
- metrics: new astrolabe_document_ocr_page_fraction histogram (the value the
page-fraction threshold acts on) alongside document_text_quality, so operators
can tune the OCR escalation per tenant (quality vs cost).
Escalation gate, OCR backends, and off-by-default behavior unchanged (#858).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The high ocr_frac is driven by each segment being shorter than MIN_PAGE_CHARS
(needs_ocr), not by text quality; quality drives bad_text_layer separately.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Follow-up to the tiered document processor (#858), landing the round-4 review
nits the reviewer approved without:
- pypdfium2_fast: free the page handle in an outer finally so a corrupt page
that makes get_textpage() raise can't orphan it.
- test: classify_from_text junk-text-layer path (non-zero chars, low quality,
high ocr_frac) flags bad_text_layer -- the hot-path coverage gap.
- test: build_ocr_backend raises ValueError when the gateway M2M client_id is
set without its token_url/secret.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #858 review round 3:
- document_tier1_engine / document_ocr_provider now validate + normalize via
Settings.__post_init__ _enum_fields (the repo's canonical opt-in-enum pattern;
case-insensitive) instead of dynaconf Validators. A typo now raises ValueError
at load and "Gateway" normalizes to "gateway".
- classify_from_text gates no_text_layer/bad_text_layer on ocr_frac >=
OCR_PAGE_FRACTION, matching classify_pdf -- a "fast"-routed doc with a few junk
pages no longer emits a misleading flag (keeps the shadow vs hot-path
classification metrics consistent).
- build_ocr_backend warns when an EXPLICIT provider is misconfigured
(gateway without EMBEDDING_GATEWAY_URL, mistral without MISTRAL_API_KEY)
instead of silently returning None.
- Pypdfium2FastProcessor.health_check probes the import; documented why
OcrProcessor.health_check is unconditionally True (lazy per-tenant backends).
- Removed the leftover per-boundary / per-chunk debug logging loops.
Tests: enum normalization + rejection for the two new settings.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #858 review round 2:
- OcrProcessor backend resolution is now guarded by an anyio.Lock (lazy-init,
double-checked) so a burst of concurrent first-OCR calls resolves the backend
once instead of each fetching its own gateway M2M token.
- The document_tier1_engine=pymupdf rollback now logs a warning when it falls
back to the fast processor (no 'structured' registered) instead of silently
using the very engine the operator opted out of.
- classify_from_text defaults ocr_frac to 0.0 (not 1.0) for a zero-page PDF, so
the recorded classification metric is "fast" (no OCR evidence) rather than a
misleading "ocr"; the no_text_layer/bad_text_layer flags are gated on having
sampled at least one page.
New tests: zero-page classify routes fast, rollback-fallback warning.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #858 review:
- 🔴 OcrProcessor now resolves its backend once and reuses it. Rebuilding per
call created a fresh GatewayTokenProvider each time -- discarding its M2M-token
cache, so every OCR'd document fetched a new token -- and a new Mistral client.
- 🔴 build_ocr_backend uses explicit ValueError (not assert, which is stripped
under `python -O`) for the gateway M2M triple.
- PIPELINE_TIER in the Qdrant payload now reflects the tier that actually
produced the doc: the registry stamps result.metadata["pipeline_tier"] and the
processor reads it (was hardcoded "fast", wrong for OCR/structured).
- Escalation now requires classification.page_count > 0, so a zero-page
(empty/corrupt) PDF isn't pointlessly sent to OCR; documented that a fast
FAILURE (encrypted/unopenable) is a hard failure and is not OCR-escalated.
- Documented the OCR page_boundaries separator-attribution choice.
- Downgraded the per-document page-boundary / page-assignment INFO logs to debug.
New tests: zero-page no-escalation, pipeline_tier stamping.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
OCR is an enhancement, not a gate. Previously, escalating a scanned doc to the
OCR tier returned the OCR result unconditionally -- so with DOCUMENT_OCR_ENABLED
=true but no backend configured (no gateway URL / no MISTRAL_API_KEY) the OCR
processor returned success=False and the whole document was marked failed and
skipped: strictly worse than leaving OCR off (where it would at least index the
tier-1 text).
Now the registry keeps the tier-1 fast result when the OCR escalation doesn't
succeed (no backend, API down, empty output), logging a warning. A
misconfiguration degrades gracefully instead of dropping scanned docs.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds the OCR escalation target the tiered registry already routes to. Scanned /
no-text-layer PDFs (the tier-0 "ocr" verdict) escalate here when
document_ocr_enabled (default off).
Two interchangeable backends, selected by document_ocr_provider
(auto | gateway | mistral | none):
- gateway: POST to the Astrolabe Cloud model gateway's /v1/ocr -- the same
M2M-authenticated gateway as embeddings, so NO provider keys live in the pod
(the platform default; reuses EMBEDDING_GATEWAY_URL + the M2M creds).
- mistral: call the Mistral OCR API directly from the pod (MISTRAL_API_KEY), for
self-hosters / deployments without the gateway.
"auto" prefers the gateway, then direct Mistral.
Both return per-page markdown joined into text + exact page_boundaries (the
pdf_highlighter contract; bbox re-derived from the PDF bytes as for other tiers).
Validated end-to-end via direct Mistral on the scanned Student 147.pdf:
success, 15 pages, 22k chars, offsets exact, ~4s.
Settings: document_ocr_provider (enum-validated), document_ocr_model
("mistral/mistral-ocr-latest" -- gateway routes on the prefix, the direct mistral
backend strips it). OcrProcessor registered at lowest priority so it is never the
non-tiered default.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replaces single-engine pymupdf4llm extraction with a tiered pipeline (Deck #205,
follows the tier-0 classifier #855). pypdfium2 becomes the default and only
hot-path PDF extractor; pymupdf4llm is deprecated to a rollback toggle.
Why: pymupdf4llm's O(n^2) find_tables drove the OOM (#852) and the form-PDF
parse timeouts (#856), carries AGPL/commercial licensing liability, and -- per
the benchmarks -- recovers near-zero usable tables on the real corpus. pypdfium2
(Apache/BSD) extracts the same text far faster (Student 1a.pdf: 120s timeout ->
0.2s) with no table-detection bomb.
- document_processors/pypdfium2_fast.py: tier-1 "fast" processor emitting text +
exact page_boundaries (the pdf_highlighter contract). pymupdf processor is now
tier "structured" (the rollback engine), registered but not default.
- registry: tiered routing in ProcessorRegistry. tier-1 fast extracts, then
classification is DERIVED from that text (classifier.classify_from_text -- no
PDF re-open), records the classification metrics, and escalates scanned /
no-text-layer docs to the "ocr" tier when document_ocr_enabled (default off;
no provider yet, so fast is terminal). Wires record_document_escalation + the
real "escalated" span attribute (was hardcoded False).
- Removes the separate _shadow_classify pass from vector/processor.py -- it
re-opened every PDF and re-extracted text (~0.5-1.3s/doc of pure duplicated
CPU that lowered throughput); classification now rides the tier-1 extraction.
- Settings: document_tier1_engine ("pypdfium2" default | "pymupdf" rollback,
enum-validated), document_ocr_enabled (default false).
Tests: pypdfium2 extractor, registry tiering (fast routing, rollback, classify
recording, OCR escalation on/off), classify_from_text. Full unit suite green.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Round-1 review follow-ups (PR #857):
- scanner.py: skip the rename-reconcile when the existing metadata point is a
placeholder. reconcile_document_path only touches real chunks, so a not-yet-
indexed file would just incur a 0-point set_payload; the real index writes the
current path anyway.
- test_sharing_state.py: add a dedup-hit case where the file was renamed AND the
user is new to the ACL, asserting both set_payload writes fire (file_path/title
and acl_principals).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The vector-sync pipeline derived an indexed file's display title from the
document's embedded metadata (e.g. a PDF's /Title), falling back to the
filename only when absent. That embedded title frequently disagrees with how
the user named the file in Nextcloud and is confusing in the astrolabe
vector-viz UI (a passive consumer of the `title` payload field).
For files, always derive the title from the Nextcloud filename via a shared
`file_title_from_path` helper. Notes/deck/news keep their metadata titles.
A rename/move in Nextcloud keeps the fileid (doc_id) and content (etag/mtime)
but changes the path, so both the dedup claim and the scanner freshness gate
skip re-embedding and the stored file_path/title go stale. Add
`reconcile_document_path`: a metadata-only set_payload that refreshes
file_path + title on the existing real chunks without re-fetch/re-embed.
Wire it into both skip paths:
- dedup hit (etag unchanged on rename) via claim_existing_index(current_path=...)
- scanner incremental skip (etag changed, mtime stable)
Both reuse already-fetched payloads, so steady-state scans add no extra
round-trip (reconcile is a no-op when the path is unchanged).
Refs: Deck #204
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
SonarQube flagged three float equality checks in the classifier tests
(python:S1244, "do not perform equality checks with floating point values"):
the _text_quality empty case and the ocr_page_fraction 0.0/1.0 assertions now
use pytest.approx.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #855 round 3 (non-blocking test completeness):
- Add a test that a mostly-digital doc with one full-page image carries the
image_heavy flag yet still routes fast (ocr_frac < OCR_PAGE_FRACTION) -- the
flag-vs-routing asymmetry operators read in the metrics, now guarded against
silent regression.
- test_full_page_image_routes_ocr also asserts the scanned flag (no text layer).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #855 round 2:
- 🔴 _shadow_classify swallowed all exceptions at DEBUG, so a systematic
failure (pymupdf bug, memory pressure) is invisible at LOG_LEVEL=INFO and
trips SonarQube S2221/S5754. Log at WARNING instead (still best-effort --
indexing is unaffected).
- classifier: use `with pymupdf.open(...) as doc` instead of manual try/finally.
- tests: release the Pixmap's native memory (del pix) in the image fixtures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #855 review (all non-blocking):
- classifier: _sample_indices now always includes the first AND last page (the
old evenly-spaced sample missed the tail, e.g. last sampled index 95 on a
100-page doc -- a scanned tail could be missed).
- classifier + metrics: document that flags are diagnostic and fire
independently of routing (image_heavy on ANY page vs the ocr route needing a
page FRACTION), so flag{image_heavy} is expected to exceed classified{ocr}.
- classifier: clarify the text-quality whitespace comment (caps at 12%) and note
the image double-count approximation (min() caps coverage).
- tests: add the scanned (no text layer) and bad_text_layer (junk text over an
image) flag paths, and a test pinning first/last-page sampling.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
First step of the tiered document-processor effort (Deck #203): a cheap, local
pre-pass that recommends which extraction tier a PDF should start in, emitting
metrics WITHOUT changing routing yet -- so we gather per-tenant doc-mix data
before turning escalation on.
document_processors/classifier.py: classify_pdf(content) -> DocClassification.
Page-sampled (bounded on large docs), <~1s. Cheap signals only -- text-layer
chars, a text-quality score (catches the "Student 147" failure where a text
layer exists but is mashed/space-less junk), and image coverage. A page that is
mostly a raster image routes to OCR: its content (handwriting, stamps) isn't in
any text layer. Deliberately no get_drawings/graphics-density signal -- it's
slow on the exact pages it'd flag, the hotfix's graphics_limit already makes the
parse safe, and the (future) tier-1 quality gate catches lost tables.
Validated on the sample corpus: born-digital 2-col arxiv and a digital student
record -> fast (tier 1); a scanned+handwritten form -> ocr (tier 3).
Wiring (vector/processor.py): _shadow_classify runs the classifier on PDFs in a
worker thread, best-effort (never blocks/fails indexing), gated by the new
DOCUMENT_CLASSIFY_ENABLED setting. Metrics: astrolabe_document_classified_total
{recommended_tier}, astrolabe_document_classifier_flag_total{flag},
astrolabe_document_text_quality histogram.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The OOM hotfix (#852) set graphics_limit=5000, which caught the 955k-drawing
bomb but let a second pathology through: form/table PDFs (e.g. student records)
have ~1.5k grid-line vector drawings per page -- under 5000, so uncapped. With
those pages uncapped, pymupdf4llm's O(n^2) find_tables grinds ~17s/page, so a
7-page form hits the 120s timeout. All 6 current backfill parse failures in
tenant-blackbox-demo are this exact timeout (zero OOM, zero error).
Measured on a 7-page sample: graphics_limit=2000 -> 119s (timeout), 1000 -> 2.9s
-- with identical extracted text and ZERO recovered tables either way (the
expensive analysis produces nothing useful on these dense forms). Lowering the
default to 1000 makes them index in ~3s; the bomb file (955k >> 1000) stays
capped, and pages with genuine simple tables (<1000 drawings) still get table
detection.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The Astrolabe PHP→Vue settings refactor dropped three stable element ids
(#mcp-enable-background-button, #mcp-revoke-background-button,
#mcp-revoke-background-form) that the multi-user-basic integration suite
drives the background-sync enable/disable/revoke flows through. Their
absence timed out the 5s Playwright locators and failed four tests:
- test_astrolabe_multi_user_background_sync::test_multi_user_astrolabe_background_sync_enablement
- test_astrolabe_multi_user_background_sync::test_revoke_background_sync_access
- test_astrolabe_chunk_context::test_chunk_context_endpoint_uses_app_password
- test_astrolabe_plotly_visualization::test_astrolabe_plotly_visualization_with_basic_auth
(the latter two enable background sync via complete_astrolabe_authorization
before exercising the app-password / indexed-search paths).
Two-part fix:
1. Bump the astrolabe submodule to v0.20.1 (cbcoutinho/astrolabe#116),
which restores the three element ids on the refactored NcButtons.
2. Defense-in-depth in the test helpers: resolve the enable/revoke buttons
by their stable id first, falling back to the button's accessible name
so a future id rename degrades to a slower-but-working lookup instead of
a hard timeout. Avoids a combined `.or_()` locator, which would
strict-mode-violate (the id button also matches by text).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #852 round 2:
- config: DOCUMENT_PDF_GRAPHICS_LIMIT validator is now gte=1 (pymupdf4llm treats
0 as "no cap", which would re-expose the OOM); documented the zero semantics
and that the per-worker mem rlimit needs a pod restart to change.
- processor: annotate `_index_document -> bool | None` and document the contract
so the `if indexed is False` check is explicit/type-checkable.
- tests: add the RLIM_INFINITY-hard branch assertion for _apply_mem_limit
(soft==target, hard stays unbounded).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The document processor crash-looped on one pathological PDF: pymupdf4llm's
table/graphics detection over a page with ~1M vector path items ballooned past
the 2 GiB pod limit. The parse ran in a thread, so nothing could interrupt or
memory-bound it -- a single bad file OOM-killed the whole pod.
Run the parse in an isolated worker subprocess (anyio.to_process, cancellable)
with an RLIMIT_AS memory cap and a wall-clock timeout, so a pathological file
fails THAT document instead of the pod (new document_processors/_isolation.py).
Also pass graphics_limit (default 5000) to to_markdown -- validated to cut the
known trigger page from 112 s to 23 s with bounded memory.
On a permanent parse failure the processor returns success=False (instead of
raising, which would retry 3x); vector/processor.py marks the placeholder
"failed" and skips indexing, and the scanner stops re-queuing failed placeholders
until the file changes -- so a doomed file no longer churns.
New per-tenant (per-pod env) settings: DOCUMENT_PDF_GRAPHICS_LIMIT,
DOCUMENT_PARSE_TIMEOUT_SECONDS, DOCUMENT_PARSE_MEM_LIMIT_MB. New metric
astrolabe_document_parse_failed_total{reason=timeout|oom|error} surfaces hard
failures that previously killed the process before any except ran.
First PR of the tiered document-processor effort (Deck #199); tier 0/1/3
pipeline tracked separately.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Clear the module-singleton ingest references (task_producer,
document_send_stream, document_receive_stream) on lifespan shutdown via a new
_clear_vector_sync_state() helper, mirroring the eviction_task_group cleanup.
Defense-in-depth so a late webhook (or a module-singleton integration test)
can't touch a producer/stream backed by an already-closed resource.
- Add IngestTransport.backend_name ("memory"/"postgres") and use it in both
lifespan log lines, removing the last settings.ingest_queue read from the
background-sync setup — the lifespan no longer inspects the backend at all.
- Cover backend_name in the build_transport adapter-selection tests.
Refs: Deck #196
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
