An artist first downloaded by gallery-dl gets Post rows keyed by the per-
attachment id; a later native walk keys the SAME real post by the post id. They
never dedup (uq_post_source_external_id is on external_post_id) → duplicate post
rows (cheunart: 943→1109). The real post id is recoverable in-DB from
raw_metadata['post_id'] (both eras store the sidecar there).
reconcile_duplicate_posts (cleanup_service): group posts by (source_id, canonical
post_id = raw_metadata.post_id else external_post_id); for each group >1, keep the
row already keyed by the post id (the format the CURRENT native downloader
produces, so future walks dedup and this can't recur), re-point
ImageRecord.primary_post_id / ImageProvenance / PostAttachment / ExternalLink onto
it conflict-safe (drop the loser's row where the keeper already has the equivalent,
per each table's uniqueness), backfill the keeper's empty date/title/body/raw_meta
from a loser, set external_post_id=post_id + derive post_url, delete losers.
IMAGES ARE NOT TOUCHED (content-addressed/deduped already; operator-confirmed).
Preview/apply share find_duplicate_post_groups (rule 93). API
/api/admin/posts/reconcile-duplicates (dry_run→{groups,posts_to_merge,sample};
apply→{groups,merged,sample}; optional source_id). UI: a second section on
PostMaintenanceCard (preview groups+sample → confirm merge). Tests: merge +
metadata backfill + image move, no-op when unique, provenance-collision dedup.
Design: milestone #73. Forensics: note #917. Out of scope (flagged): cheunart vs
Cheunart case-variant artist dirs/rows.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Image posts wrap the post date in an <a> permalink
(<div class="post-date"><a href="/posts/ID">DATE</a></div>); text-only posts
don't. Our hand-written <div class="post-date">([^<]+)</div> regex matched ONLY
the unwrapped case, so every image post got a null published_at and sorted to the
top of the feed looking broken (cheunart 2026-06-17). Port gallery-dl's
_data_from_post method: text up to the first </, then after the last > — handles
both. Verified against the live raw feed (all 6 dates now parse).
Robust logging (operator request): _parse_posts now logs per-page parse stats
(posts / dated / with-body) and a WARNING canary when posts parse but NONE get a
date or body while the raw markers are present — i.e. our extraction diverged
from the live markup. Makes this failure class diagnosable from the worker log
alone, no authed re-fetch needed.
Test: a permalink-wrapped date parses to ISO.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Some SubscribeStar posts deliver content only through document/audio attachments,
which live OUTSIDE data-gallery. Port gallery-dl's _media_from_post for them:
- docs: scope uploads-docs..post-edit_form, split on doc_preview blocks, take the
href URL + doc_preview-title + data-upload-id (kind=attachment).
- audio: scope uploads-audios..post-edit_form, split on audio_preview-data
blocks, take the src URL + audio_preview-title + data-upload-id (kind=audio).
The existing downloader handles them unchanged (plain streaming GET; the file
validator only inspects image/video extensions via is_validatable, so PDFs/zips/
audio pass straight through, no quarantine).
Test covers doc + audio extraction (the cheunart sample has none, so this pins
gallery-dl's documented markup shape).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Body rendered as a bogus '264 / 265' on every post: our balanced-</div> body
regex either returned empty or over-captured into sibling upload divs and the
'View next posts (N / M)' pagination counter. Replace it with gallery-dl's exact
_data_from_post rule — content between the post_content-text wrapper and the
youtube-uploads div (literal markers), then strip the trix editor's
<html><body>…</body></html> document wrapper to its inner. Verified against the
live cheunart sample: clean per-post bodies, empty for genuinely text-less posts.
Also port gallery-dl's _media_from_post preview guard: skip gallery items whose
URL is under /previews (locked/blurred teasers) — the SubscribeStar analog of
the Patreon gated-preview bug (#874); this is why a locked post yields no media.
Tests: body must not bleed into the pagination counter; trix html-document
wrapper stripped; /previews items skipped. Fixture now includes the youtube-
uploads close marker present in real markup.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
After the delimiter fix, the live cheunart fetch 302'd to /cheunart/verify_
subscriber even with valid .adult cookies (confirmed present: _personalization_id
+ _subscribestar_session on .subscribestar.adult, logged in). Walking gallery-dl's
ENTIRE flow — including the base Extractor._init_session I'd not read — the
divergence is the HTTP request profile, not the cookies or parser.
gallery-dl's default (cookies-only) mode sends, on EVERY request including the
first creator-page GET: a Firefox UA, Accept: */*, Accept-Language, and a same-
site Referer (root/), with NO X-Requested-With anywhere (the load-more endpoint
is a plain GET parsed as JSON). Our Chrome UA + missing Referer + XHR toggling
looked unlike a browser → SubscribeStar gated the adult-creator page.
Make our SubscribeStar session identical: Firefox UA + Accept */* + Accept-
Language via make_session extra_headers; stamp Referer=<base>/ per walk; drop the
per-request XHR headers (both feed and load-more now use the shared profile).
Test updated to assert the gallery-dl-parity profile instead of the old
navigation-vs-XHR split.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The native client split the feed on `<div class="post is-shown`, but `is-shown`
is added by SubscribeStar's infinite-scroll JS when a post scrolls into view —
present in a browser-SAVED page (what the Step-0 characterization used) but
ABSENT from the raw server HTML we and gallery-dl actually fetch. So the live
feed (cheunart) parsed to zero posts and raised a false SubscribeStarDriftError.
Align with gallery-dl's proven `_pagination`: split on the generic
`<div class="post ` (trailing space rules out the hyphenated post-content/
post-date/post-body siblings). Also mirror gallery-dl's redirect-based gating
detection (/verify_subscriber, /age_confirmation_warning => auth, not drift).
Regression tests: raw server markup without is-shown now parses; an age-wall
redirect raises SubscribeStarAuthError.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The XHR fix worked (we now get a real 93KB HTML page, not JSON) but cheunart
still drifts — we're being served a full HTML page that isn't the feed. Add
_describe_page(): the drift error now reports the page <title> + which known
interstitial it resembles (cloudflare/bot-challenge, age-gate, login, captcha),
so the next run names the actual cause instead of "markup changed". Strong
suspicion: a Cloudflare challenge (python-requests has no JS; cf_clearance is
UA-locked and our hardcoded UA likely differs from the cookie-capturing browser).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
First live run (cheunart) tripped the drift guard: "no posts and no recognizable
feed container". The browser-saved page was normal (6 posts + posts_container-list),
so the parser was fine — our live HTTP fetch got a different response. Cause: the
client set X-Requested-With: XMLHttpRequest (+ a JSON Accept) session-wide, so the
initial creator-page GET was sent as an XHR. SubscribeStar (Rails) content-
negotiates an XHR full-page request to a non-HTML body → no container → drift.
Fix: the session now uses browser-like navigation headers (Accept: html, NO
X-Requested-With); the XHR header + JSON Accept are applied PER-REQUEST only on
the "load more" endpoint (which is a genuine XHR). Drift message now reports the
response length + a JSON hint so a recurrence is self-explaining. Regression test
pins the header split (navigation initial GET, XHR load-more).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The _NATIVE_INGESTERS dict captured PatreonIngester/SubscribeStarIngester at
import, so test_download_service's monkeypatch.setattr(db_mod, "PatreonIngester",
_FakeIngester) no longer affected dispatch → the fake's run() never ran →
KeyError 'campaign_id' on empty run_kwargs (integration run 1215). Replace the
dict with a _native_ingester_cls() call-time lookup that reads the module globals,
so monkeypatching the class names works again.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
SubscribeStar now downloads + verifies through the native core ingester instead
of gallery-dl — the go-live switch for milestone #71.
- download_backends: subscribestar added to NATIVE_INGESTER_PLATFORMS; a
_NATIVE_INGESTERS registry + _resolve_native_campaign_id make _run_native_ingester
/ preview_source / verify_source_credential platform-aware. SubscribeStar's
campaign_id IS the creator URL (no resolver); Patreon still resolves the vanity.
preview now catches the shared NativeIngestError (covers both platforms).
- platform_lock: subscribestar serialized (one paced walk at a time).
- gallery_dl: subscribestar entry removed from PLATFORM_DEFAULTS (rule 22 — no
fallback once native works).
- frontend SourceActions: isPatreon → isNative (patreon|subscribestar) so the
recover/recapture actions show for subscribestar; download_service's
cursor/mode/post_first + the preview endpoint already key on
uses_native_ingester, so backfill/recovery/recapture/preview light up for free.
- tests: download_backends (subscribestar native), platform_lock (serialized),
and three gallery-dl-sample tests repointed to hentaifoundry (api_credentials
verify, gallery_dl_service skip-value, api_sources arm-no-preflight).
post_is_gated stays best-effort (can't cause junk downloads); not gating this.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
DRY pass commit 3 — the observability half. ingest_core accumulated ALL
human-readable progress in log_lines → DownloadResult.stdout, persisted to the
DownloadEvent ONLY at phase 3; the real logger was used almost nowhere. So a
worker SIGKILL/OOM/hard-time-limit mid-walk left NO trace (the "task died,
no trace" mode from the recovery-sweep work).
Route run milestones through the container log too, each carrying source_id (L3
context):
- run START (platform/mode/source/campaign/resume_cursor)
- per-PAGE breadcrumb (posts/downloaded/skipped/errors/quarantined/gated/cursor)
at each page boundary — pages are minutes apart on big backfills, so this shows
how far a since-died walk got
- final SUMMARY (same string as the stdout summary)
- operator STOP
(L2 — quarantines log.warning'd — already landed in commit 1's base
_validate_path; failures log.warning via the base _failure_result; the #862 body
canary already log.error's.) log_lines/stdout content is unchanged (summary just
captured in a var), so existing assertions hold.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
DRY pass commit 2. The two adapters re-implemented the same auth→drift→429→404
→http→network mapping in _failure_result; only the exception classes + drift
phrasing differed (divergence-bug risk: a new error_type handled in one and not
the other).
- native_ingest_common gains NativeIngestError / NativeAuthError / NativeDriftError
(status_code + retry_after on the base). Patreon{API,Auth,Drift}Error and
SubscribeStar{API,Auth,Drift}Error now subclass them via multiple inheritance,
keeping their isinstance-distinct platform names.
- Ingester._failure_result (base) does the whole mapping via the shared
NativeAuthError/NativeDriftError taxonomy + status_code; a new platform gets it
free. New drift_label kwarg supplies the per-platform API_DRIFT phrasing
("Patreon API" / "SubscribeStar markup"), preserving the existing message
(test asserts "Patreon API changed").
- Both adapters drop their near-identical _failure_result overrides and their now
-unused DownloadResult/ErrorType/*Auth/*Drift imports.
Verified at every consumer (rule 93/§8b): test_patreon_ingester (auth/drift/429/
404/network) and test_subscribestar_native (_failure_result mapping) both exercise
the base method now. Remaining: ingest_core L1/L3 logging (3/3).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
DRY pass commit 1 (process #594). Consolidate the helpers + download plumbing
the Patreon and SubscribeStar adapters had duplicated (SubscribeStar was
importing patreon privates — wrong owner). New backend/app/services/
native_ingest_common.py is the neutral home for:
- make_session (was _load_session ×2), retry_after_seconds + 429 constants,
sanitize_segment, basename_from_url, post_dir_name, MediaOutcome /
PostRecordOutcome.
- BaseNativeDownloader: the shared streaming GET (transient-retry + Range-resume)
and validation/quarantine. Patreon + SubscribeStar downloaders now subclass it;
each keeps only what differs (Patreon's Mux/yt-dlp video branch + detail-fetch
enrichment; SubscribeStar nothing extra). Behavior preserved exactly; the
divergence-bug risk (a fix to one _fetch_to_file not reaching the other) is gone.
- Folds in #899 L2: a quarantine now log.warning's path+reason (was counted only).
post_dir_name merges both date handlers (accepts trailing-Z and pre-parsed ISO).
Tests repointed to the single source at every consumer (rule 93 / §8b parity):
patreon_client/downloader, subscribestar_native. Exception-trio consolidation +
base _failure_result (2/3) and the remaining ingest_core logging (3/3) follow.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Phase-1 steps 2-4 of moving SubscribeStar off gallery-dl onto the native core
ingester. SubscribeStar has no JSON:API, so the client scrapes HTML; the
platform-agnostic core (ingest_core) is unchanged.
- subscribestar_client.py: HTML-scrape read path. iter_posts pages via the
creator page → infinite_scroll-next_page href → JSON {html} fragments
(campaign_id = creator URL; no resolver). extract_media reads the per-post
data-gallery JSON manifest (id/original_filename/type/url). post_record_key,
post_meta, and post_is_gated (best-effort locked-teaser marker, pending a live
locked sample). Loud auth/drift taxonomy (SubscribeStar{API,Auth,Drift}Error).
Parser validated against the real Step-0 fixtures.
- subscribestar_downloader.py: mirrors PatreonDownloader minus the Mux/yt-dlp
branch (SubscribeStar serves files directly via /post_uploads). gallery-dl
on-disk layout so existing downloads dedup on disk at cutover. Post-first:
_post.json owns the body/links; per-media sidecar carries image identity only.
- subscribestar_ingester.py: thin adapter wiring client/downloader/the
SubscribeStar ledgers into the core; ledger_key = filehash else
post_id:media_id; SubscribeStar failure mapping. verify_subscribestar_credential.
- tests: client parsing/pagination/media/gating/record-key/dates, downloader
layout/sidecar/post-record/skip-seen, ingester ledger_key + failure mapping.
Not yet wired into dispatch (Step 5) — these modules are inert until then.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Phase 1, step 1 of moving SubscribeStar off gallery-dl onto the native core
ingester (milestone: SubscribeStar native). Mirror of the Patreon ledger:
SubscribeStarSeenMedia (skip already-ingested media on routine walks; recovery
bypasses) and SubscribeStarFailedMedia (dead-letter so persistently-failing
media stops re-burning backfill chunks). Per operator decision, dedicated
per-platform tables (not a generalized shared ledger).
filehash is String(128): a CDN content hash when the URL carries one, else a
synthesized <post_id>:<filename> key. UNIQUE (source_id, filehash) upsert key.
Registered in models/__init__; migration 0054 creates both tables (down 0053).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
test_downloaded_archive_gets_provenance_and_tagging's fake_fetch still had the
old `*, timeout` signature; the task now calls fetch_external() without it, so
the stub raised TypeError in the integration lane (run 1191). Switch it to
**kwargs like the other two.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The single _FETCH_TIMEOUT=3000s meant different things per host: a TOTAL
wall-clock for mega (subprocess), but only a per-read socket timeout for HTTP
hosts (requests' timeout is the idle gap between bytes, never a total). So a
stalled HTTP connection tied up a download-worker slot AND the per-host
serialize lock for ~50 min before failing (operator-flagged 2026-06-17).
Split into two limits in external_fetch:
- read timeout (_READ_TIMEOUT=60s, with _CONNECT_TIMEOUT=30s) → requests gets
(connect, read); a stalled socket now fails in ~60s.
- total budget (_TOTAL_TIMEOUT=30min) → enforced as a wall-clock deadline
across chunks in _stream_to_file (HTTP has no total-download timeout), and
passed as the subprocess total for mega.
fetch_external() signature: timeout= → read_timeout=/total_timeout=. gdrive
(gdown) self-manages; the celery hard limit is the outer backstop.
Also lowered the per-host lock TTL 3600→2400 so a worker that dies holding it
can't wedge a host's links much past one fetch's budget.
Each external link is already one Celery task (sweep enqueues one
fetch_external_link.delay per link), so these budgets are per-link.
Tests: total-budget-exceeded cleans the .part; HTTP gets (connect, read);
mega gets the total. Worker fakes updated to **kwargs.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
celery_signals._queue_for is a hand-maintained mirror of task_routes that
stamps TaskRun.queue in the prerun signal. It was missing the
backend.app.tasks.external. prefix, so external fetches recorded
queue='default' even though celery routes external.* → download and runs
them on the download worker. The dashboard's per-queue filters and the
per-queue recovery-sweep threshold therefore missed them — the same
'queue column lies default' gap the 2026-06-02 audit fixed for
backup/admin/library_audit.
Map external.* → download in _queue_for. Composes with the fetch_external_link
task-name sweep override (#883), which wins by precedence regardless of the
recorded queue. Pinned test asserts the mirror agrees with the actual route.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
External file-host fetches run to a 60-min hard limit (time_limit=3600,
per-fetch _FETCH_TIMEOUT=3000s), far longer than the recovery sweep's 5-min
default. recover_stalled_task_runs was phantom-flagging healthy in-flight
fetches as "RecoverySweep: no completion signal received within 5 min"
before the task's own timeout/error handling could surface the real error
(operator-flagged: target 414 swept at 6.6min).
The sweep already has per-queue/per-task overrides for long tasks, but
fetch_external_link was never added and its TaskRun records queue='default'
(no queue override) despite external.* routing to download. Add a task-name
override of 65 min (time_limit 60 + 5 buffer); task-name precedence makes it
robust regardless of the recorded queue. No new internal timeout needed —
the existing _FETCH_TIMEOUT + soft_time_limit + except-block log.exception
already capture the real failure once the sweep stops preempting.
Pinned tests: external-fetch override survives a 10-min row / flags a 70-min
row on queue='default'; invariant guard asserts override >= hard time_limit.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
test_artist_directory_service._seed_image built sha256 from
abs(hash(suffix)) % 10000 — PYTHONHASHSEED-randomized hash() over only 10k
buckets, so two suffixes in one test could birthday-collide and violate
uq_image_record_sha256. Flaky per process seed: passed on dev (run 1179),
failed on main (run 1182) with identical code. Use
hashlib.sha256(suffix).hexdigest() for a stable, collision-free digest.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
recompute_centroid + recompute_centroids were the only tasks still using
the process-wide singleton extensions.get_session() under asyncio.run().
The async engine's asyncpg pool is bound to the loop it was created on;
each Celery task runs a fresh asyncio.run() loop, so after the first
invocation the cached engine handed loop-A connections to loop B and raised
"Future attached to a different loop" — every recompute after the first in
a worker process failed (~35ms, fails on first DB await).
Convert both to the established per-task async_session_factory() pattern
(NullPool engine created + disposed inside the task's own loop), matching
scan/download/admin tasks. No get_session usages remain in tasks/.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Long-running maintenance tasks must survive navigating away or reloading
the page. VideoDedupCard + GatedPurgeCard held the in-flight Celery task id
only in component refs and polled task-result inline, so leaving the page
mid-run lost the id and the result was never shown — even though the task
finished on the worker.
New shared composable useMaintenanceTask: persists {taskId, mode, startedAt}
to localStorage on dispatch, re-attaches on mount, and re-shows the result
when the task finishes (the celery result backend retains the summary well
under result_expires). Stale-guard skips resume past 3h. Both cards refactored
onto it; card-specific computeds + confirm dialog kept.
Also fixed the QueueStatusBar lane: both cards watched queue="maintenance"
but tasks.admin.* routes to maintenance_long, so the bar never reflected
their own task — now queue="maintenance_long".
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A one-shot Maintenance action to remove the blurred locked-preview images
the ingester downloaded from tier-gated Patreon posts before #874.
current_user_can_view was never persisted, so the cleanup re-walks each
enabled Patreon source (read-only) to re-derive which posts are gated now
and the blurred filehashes Patreon serves for them, then matches by
CONTENT HASH against stored source_filehash. Because the hash is
content-addressed, a real file downloaded when access existed has a
different hash and can never match — regained-then-lost-access content is
provably spared (operator's hard requirement). NULL source_filehash =>
unverifiable, kept + reported.
On apply: delete matched ImageRecords + files (provenance cascades),
clear seen/dead-letter ledger rows for those hashes so the real media
re-ingests if access returns, and delete gated posts left bare. Shares
one match predicate between preview and apply (rule 93).
- cleanup_service: collect_gated_previews + purge_gated_previews
- tasks.admin: purge_gated_previews_task (async re-walk bridge, timeboxed)
- api.admin: POST /maintenance/purge-gated-previews
- GatedPurgeCard.vue in Settings > Maintenance (preview -> confirm -> apply)
- tests: collect predicate, hash-match delete/spare/unverifiable, ledger
clear, bare-post removal, no-op
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The prior commit's Edit orphaned the recapture test's relink/stdout
assertions into the new preview test (F821 res_recap/downloader2/m1) and
the gated-skip test's written_paths check matched 'gated' in the tmp dir
name. Restore the recapture assertions to their function and assert on
the media basename instead.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Patreon serves only blurred locked-preview thumbnails for posts the
authenticated account can't fully view; the native ingester was
downloading those as real media. current_user_can_view was already in
_FIELDS_POST but never read.
Add PatreonClient.post_is_gated (gate ONLY on explicit
current_user_can_view=False; missing/None → viewable, never over-filter)
and skip gated posts at the top of the ingest_core run() and preview()
loops — no media download AND no post-record stub (operator: 'no stub
for gated content'). Skipped before the post-record block so gated posts
never inflate the #862 body canary; surfaced as 'N gated-skipped' in the
run summary. Same gate in preview() for preview/apply parity (rule 93).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
GPU enablement (#872) cancelled — not worth the Pascal-specific build for a
modest CPU→GPU win on an old P4. Remove the dead GPU code (device.py, the CUDA
provider branch in tagger, the .to('cuda') path in embedder) so nothing carries
it forward.
Instead, bound CPU inference threads by default so the ml-worker is a predictable
core consumer on a SHARED node — the intended scaling model is multiple worker
replicas (each --concurrency=1, each its own cgroup limit), not one big
container. ONNX Runtime and torch otherwise size their thread pools to ALL host
cores, so each replica would grab every core and oversubscribe / starve the
co-located DB+web. Cap both to _INTRA_OP_THREADS=4 (matches the prior per-worker
cpus:4 unit): run N replicas where N×4 stays within the cores allotted to ML.
- tagger: ort.SessionOptions().intra_op_num_threads = 4 (CPUExecutionProvider).
- embedder: torch.set_num_threads(4).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Step 1 of GPU enablement (code only — CPU-safe, CI-green; the CUDA image is a
separate step pending the host driver version).
- New services/ml/device.py: FC_ML_DEVICE (auto|cuda|cpu) intent + VRAM knobs
(FC_ML_ONNX_GPU_MEM_GB, FC_ML_TORCH_MEM_FRACTION). Per-worker-host bootstrap →
env, not a DB setting (the GPU host runs CUDA, others CPU).
- tagger: use CUDAExecutionProvider (with gpu_mem_limit) when requested AND the
provider is actually present (onnxruntime-gpu), else CPUExecutionProvider. Logs
the active providers.
- embedder: move model + inputs to cuda when requested AND torch.cuda is
available; cap torch's VRAM share; .detach().cpu() before numpy. fp32 kept so
GPU embeddings stay in the same space as existing CPU ones.
Both AND the env intent with the framework's real availability, so on CPU
(CI / CPU onnxruntime / no GPU) they fall back cleanly — behavior unchanged.
The 8GB P4 is shared by both frameworks, hence the conservative default caps.
Tests: device env parsing. (tagger/embedder GPU paths are operator-verified on
the GPU host — models aren't in CI.)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Video tag noise root cause: frames were a FIXED count (6) max-pooled — a tag
firing on one frame survived at peak confidence, and a fixed count under-samples
long multi-scene videos so real scene-local tags looked like noise.
Redesign (operator-steered):
- Sample at a fixed CADENCE — one frame every `video_frame_interval_seconds`
(default 4) across the 5–95% window — so a tag's frame-presence reflects real
screen time independent of video length. Capped at `video_max_frames` (default
64): a long video stretches the spacing instead of exploding into hundreds of
inferences, bounding per-video cost on the single ml-worker (per-frame ffmpeg
timeout also cut 60s→30s).
- Aggregate with `_aggregate_video_predictions`: keep a tag only if it appears in
>= `video_min_tag_frames` sampled frames (≈ that many × interval seconds on
screen — duration-independent noise rejection), with confidence = MEAN over the
frames it appears in (not max). Clamps the threshold to the sample count so a
1–2-frame short video still tags.
- All three knobs are DB-backed ml_settings (migration 0053), patchable via
/api/ml/settings + sliders in the ML settings card — replaces the
VIDEO_ML_FRAMES env var (product-not-project).
Tests: aggregation drops one-frame noise + means corroborated tags + clamps on
short videos; settings round-trip + min>max validation. Replaced the
_maxpool_predictions unit test.
NOTE: this is the QUALITY half of #747. The perf half — the ml-worker runs
CPU-only — is GPU enablement, tracked separately in #872.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Phase 2 of #871: clean up the duplicate videos already in the library (the #859
"same video from multiple sources" clutter). Import-time dedup (Phase 1) only
prevents NEW dups; this is the operator-triggered cleanup of existing ones.
cleanup_service.dedup_videos(dry_run):
- backfill_video_durations: re-probe NULL-duration videos (pre-#871 rows) so the
existing library participates; idempotent (only NULL rows), writes a negative
sentinel for un-probeable files so they're neither re-probed forever nor matched.
- find_video_dup_groups: cluster same-artist videos by duration (±tol) + aspect,
anchored per cluster to bound the span (no chain drift); keeper = highest pixel
area then bytes. Reuses the importer's _VIDEO_DUP_* tolerances.
- apply: re-point each loser's post links to the keeper (so no post loses the
video) THEN delete the redundant records + files via delete_images (cascade).
dry_run shares the same discovery predicate and returns the projection only
(rule 93). Tags on a loser are NOT merged (noted; videos rarely hand-curated).
- dedup_videos_task (maintenance queue; summary → task_run.metadata).
- POST /maintenance/dedup-videos {dry_run} + GET /maintenance/task-result/<id> so
the card shows the dry-run projection before the destructive apply.
- VideoDedupCard: Preview → shows groups/redundant/reclaimable, then Apply behind
a confirm dialog. Mounted in the Maintenance panel.
Tests: dedup collapses + re-links the loser's post to the keeper + removes the
file; dry-run deletes nothing; distinct durations aren't grouped; task registered.
(Migration 0052 for duration_seconds already shipped with Phase 1.)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Videos deduped on sha256 only (pHash is images-only), so a different encode/remux
of the same clip imported as a distinct record — the "same video from multiple
sources" clutter surfaced by #859.
Tier-1 metadata fingerprint: identity = container duration (±1.0s) + matching
aspect ratio, scoped to the same artist; quality axis = pixel dimensions (mirrors
image pHash: larger_exists→skip+link, smaller_exists→supersede). Codec/bitrate
are deliberately NOT part of identity (the point is matching across re-encodes).
Tight tolerances because a wrong video merge is destructive.
- image_record.duration_seconds (Float, nullable; migration 0052). NULL for images.
- safe_probe.probe_video also reads format=duration (one extra ffprobe field on the
call that already runs); ProbeResult.duration.
- _find_similar_video(duration,w,h,artist) shared by both import pipelines.
- _import_media (filesystem/archive path): captures duration, video near-dup
branch, persists duration.
- attach_in_place (download path — handles #859's videos, previously didn't probe
video at all): best-effort probe for dims+duration (LENIENT — never newly rejects
a downloaded video on probe failure), video near-dup branch, persists duration.
- _supersede carries duration onto the kept row.
Reuses SkipReason.duplicate_phash so the existing download/external dup-cleanup
(path-safe unlink, #859) applies unchanged. Tests: skip-smaller, supersede-larger
(+ duration adopted), and distinct-durations-not-merged (false-merge guard).
Follow-up (Phase 2, #871): a backfill to re-probe NULL-duration existing videos so
the current library participates in dedup; retroactive merge of existing dups is a
separate destructive maintenance action.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The dedup branches of _import_media linked the existing image to the new post via
_apply_sidecar(artist=None), relying on the SIDECAR to carry the artist. But an
archive member's artist comes from its path, and under post-first the per-media
sidecar is minimal (no artist) — so a re-packed / cross-posted archive image
deduped and was left UNLINKED from the new post, i.e. the post showed "no images".
Resolve the path-anchored artist (derive_top_level_artist) up-front in
_import_media and pass it to both enrich-on-duplicate branches (sha256 + phash
larger_exists) and the new-record path. Drop the now-dead _attach_artist helper
(its logic is inlined at the single new-record call site).
Surfaced by the new test_archive_all_deduped_is_benign_not_flagged (was asserting
2==4: the second post got no provenance links).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Root cause (operator-confirmed via event metadata + lsar): a "High Resolution
files" pack often wraps a per-chapter .rar/.zip INSIDE one outer archive (incase).
_import_archive only extracted one level — a nested-archive member failed
is_supported and was skipped, so the real pages were silently dropped and the post
showed "archive but no images". The disk scan found this pattern recurring across
the attachment store.
- Recurse into nested archives via _collect_archive_members: a member that is
itself an archive is bomb-probed and extracted too, depth-capped at
_ARCHIVE_MAX_DEPTH=3. Nested members attribute to the OUTER archive's sidecar so
they link to the right Post. Each level is wrapped so one bad nested archive
can't abort the import. The shared path means external (mega/gdrive) archives
recurse too.
- Replace the catch-all "held no supported members" string with a per-outcome
tally (media/deduped/unsupported/failed/nested/nested_rejected). The all-deduped
case is now recognised as BENIGN — images already in the library, re-linked to
this post via enrich-on-duplicate — and returns attached WITHOUT error, so it no
longer false-flags in event metadata.unextracted_archives. Genuine failures
carry the precise breakdown.
Tests: nested zip-in-cbz imports both inner images + links them to the outer post;
all-deduped archive returns attached with error=None and links images to both posts.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Milestone #67 step 3. Spell out, at the IngestCore.run seam resolution, that
post_record_key + write_post_record are the post-first contract a platform
implements when migrating onto the native core ingester — the post-record owns
the body/links, the per-media sidecar carries image identity only. The import
side is already self-enforcing via uses_native_ingester → importer.post_first.
Durable directive recorded as FC project rule #120.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
bvandeusen