Files
FabledCurator/backend/app/services/download_service.py
T
bvandeusen 19aece1fc4
CI / lint (push) Successful in 3s
CI / backend-lint-and-test (push) Successful in 16s
CI / frontend-build (push) Successful in 27s
CI / intimp (push) Successful in 3m42s
CI / intapi (push) Successful in 7m38s
CI / intcore (push) Successful in 8m20s
feat(download): tick/backfill modes + partial-success classifier (plan #544)
Routine subscription polls walked the entire post history every tick
even when nothing had changed, because gallery-dl's default `skip: True`
continues iterating archived posts. A creator with ~550 archived posts
(Knuxy patreon) saturates the 870s wall-clock cap before completing,
even with zero downloads needed. Plus, a tier-limited run that
downloaded hundreds of files but ran out the clock should be a
warning, not an error.

Two coupled changes, both operator-flagged 2026-06-01:

* **Tick mode (default, cron polls).** New `TICK_SKIP_VALUE = "exit:20"`
  asks gallery-dl to exit after 20 contiguous archived items. Fresh
  subscriptions + new-content cases still walk normally; established
  subscription with zero new content exits in ~30s of HEAD requests
  instead of pegging the timeout. 20 (not 5) gives headroom against
  paywall warnings interleaving with archived items.
* **Backfill mode (explicit, operator-triggered).** Sticky for N runs
  via new `Source.backfill_runs_remaining` (alembic 0031). While > 0,
  downloads use `skip: True` + 1800s timeout. Auto-decrements per run
  with early-reset to 0 when a clean run finds zero files (queue
  drained). N defaults to 3 — multiple runs give the system enough
  budget to finish a deep walk across timeout boundaries. New
  `POST /api/sources/{id}/backfill` arms the source; "Deep scan"
  button on each SourceRow (chip shows remaining count) wires it.

Plus partial-success classifier: non-zero gallery-dl exit + ≥1 file
downloaded + no source-level error fires `ErrorType.PARTIAL`, which
download_service maps to `status=\"ok\"`. The run did real work; the
next tick continues via gallery-dl's archive. No more red events for
"timed out mid-walk after downloading 300 files."

Retires `SourceConfig.skip_existing` — skip value is now derived from
the source state and passed as a separate `skip_value` parameter
through download() / _build_config_for_source(). `GD_DEFAULTS` drops
the now-dead key (was inert data after this refactor).

Tests cover:
* tick + backfill skip-value emission in _build_config_for_source
* PARTIAL classifier branch + TIER_LIMITED-wins-over-PARTIAL ordering
* SourceService.set_backfill_runs validation + persistence
* /api/sources/{id}/backfill 200/400/404 paths
* download_service auto-decrement / auto-reset / tick-mode-no-touch
* PARTIAL → status=ok in the orchestrator (no consecutive_failures bump)
2026-06-01 18:23:28 -04:00

380 lines
15 KiB
Python

"""FC-3c download orchestrator.
Three-phase pipeline:
Phase 1 — DB setup (brief session): load source, mark event running,
fetch credentials + settings.
Phase 2 — Execute (no DB connection): call GalleryDLService; on
Patreon campaign-ID failure, resolve vanity and retry once.
Phase 3 — Import + persist (fresh session): attach_in_place each
written file; clean up duplicates; persist rich metadata.
"""
from __future__ import annotations
import asyncio
import logging
import re
from datetime import UTC, datetime
from pathlib import Path
from typing import Any
from sqlalchemy import select
from sqlalchemy.ext.asyncio import AsyncSession
from sqlalchemy.orm import Session as SyncSession
from sqlalchemy.orm import joinedload
from ..models import Artist, DownloadEvent, Source
from .credential_service import CredentialService
from .gallery_dl import (
BACKFILL_SKIP_VALUE,
BACKFILL_TIMEOUT_SECONDS,
TICK_SKIP_VALUE,
ErrorType,
GalleryDLService,
SourceConfig,
)
from .importer import Importer
from .patreon_resolver import resolve_campaign_id
from .scheduler_service import set_platform_cooldown
log = logging.getLogger(__name__)
_PATREON_VANITY_RE = re.compile(
r"^https?://(?:www\.)?patreon\.com/(?:c/)?(?!id:)([^/?#]+)",
re.IGNORECASE,
)
_CAMPAIGN_ID_FAILURE_PATTERN = "failed to extract campaign id"
def _extract_patreon_vanity(url: str) -> str | None:
m = _PATREON_VANITY_RE.match(url)
return m.group(1) if m else None
def _looks_like_campaign_id_failure(stdout: str, stderr: str) -> bool:
return _CAMPAIGN_ID_FAILURE_PATTERN in f"{stdout}\n{stderr}".lower()
def _effective_url(platform: str, source_url: str, overrides: dict) -> str:
if platform == "patreon" and overrides.get("patreon_campaign_id"):
return f"https://www.patreon.com/id:{overrides['patreon_campaign_id']}"
return source_url
class DownloadService:
"""Async orchestrator. The Celery task runs `asyncio.run(svc.download_source(N))`.
The Importer is sync (existing FC convention); calls to attach_in_place
happen via run_in_executor.
"""
def __init__(
self,
async_session: AsyncSession,
sync_session: SyncSession,
gdl: GalleryDLService,
importer: Importer,
cred_service: CredentialService,
):
self.async_session = async_session
self.sync_session = sync_session
self.gdl = gdl
self.importer = importer
self.cred_service = cred_service
async def download_source(self, source_id: int) -> int:
"""Returns DownloadEvent.id. Idempotent: in-flight events are returned as-is."""
setup = await self._phase1_setup(source_id)
if setup["status"] in ("skipped", "in_flight"):
return setup["event_id"]
ctx = setup
source_config = SourceConfig.from_dict(ctx["config_overrides"] or {})
# alembic 0031 / plan #544: derive skip_value + timeout from the
# source's backfill_runs_remaining counter. When > 0, walk the full
# post history (skip: True + 1800s); when 0, exit gallery-dl after
# 20 contiguous archived items (skip: "exit:20" + the default
# 870s). Operator sets backfill via POST /api/sources/{id}/backfill.
backfill_remaining = ctx.get("backfill_runs_remaining", 0) or 0
if backfill_remaining > 0:
skip_value: bool | str = BACKFILL_SKIP_VALUE
source_config.timeout = BACKFILL_TIMEOUT_SECONDS
else:
skip_value = TICK_SKIP_VALUE
effective_url = _effective_url(
ctx["platform"], ctx["url"], ctx["config_overrides"] or {}
)
dl_result = await self.gdl.download(
url=effective_url,
artist_slug=ctx["artist_slug"],
platform=ctx["platform"],
source_config=source_config,
cookies_path=ctx["cookies_path"],
auth_token=ctx["auth_token"],
skip_value=skip_value,
)
resolved_campaign_id: str | None = None
if (
ctx["platform"] == "patreon"
and not dl_result.success
and "patreon_campaign_id" not in (ctx["config_overrides"] or {})
and _looks_like_campaign_id_failure(dl_result.stdout, dl_result.stderr)
):
vanity = _extract_patreon_vanity(ctx["url"])
if vanity:
log.info(
"Attempting campaign-ID resolution for %s (%s)",
ctx["artist_slug"], vanity,
)
resolved_campaign_id = await resolve_campaign_id(
vanity, ctx["cookies_path"]
)
if resolved_campaign_id:
dl_result = await self.gdl.download(
url=f"https://www.patreon.com/id:{resolved_campaign_id}",
artist_slug=ctx["artist_slug"],
platform=ctx["platform"],
source_config=source_config,
cookies_path=ctx["cookies_path"],
auth_token=ctx["auth_token"],
skip_value=skip_value,
)
return await self._phase3_persist(
ctx["event_id"], ctx, dl_result, resolved_campaign_id,
)
async def _phase1_setup(self, source_id: int) -> dict[str, Any]:
source = (await self.async_session.execute(
select(Source).options(joinedload(Source.artist)).where(Source.id == source_id)
)).scalar_one_or_none()
if source is None:
raise LookupError(f"source {source_id} not found")
if not source.enabled:
ev = DownloadEvent(source_id=source_id, status="pending")
self.async_session.add(ev)
await self.async_session.commit()
await self.async_session.refresh(ev)
await self._finalize_event_and_source(
event_id=ev.id, source_id=source_id, status="skipped",
error_message="source disabled",
)
return {"status": "skipped", "event_id": ev.id}
existing = (await self.async_session.execute(
select(DownloadEvent).where(
DownloadEvent.source_id == source_id,
DownloadEvent.status.in_(["pending", "running"]),
).order_by(DownloadEvent.id.desc()).limit(1)
)).scalar_one_or_none()
if existing and existing.status == "running":
return {"status": "in_flight", "event_id": existing.id}
if existing and existing.status == "pending":
existing.status = "running"
await self.async_session.commit()
event_id = existing.id
else:
ev = DownloadEvent(source_id=source_id, status="running")
self.async_session.add(ev)
await self.async_session.commit()
await self.async_session.refresh(ev)
event_id = ev.id
artist = source.artist
if source.platform in ("discord", "pixiv"):
cookies_path = None
auth_token = await self.cred_service.get_token(source.platform)
else:
cookies_path_obj = await self.cred_service.get_cookies_path(source.platform)
cookies_path = str(cookies_path_obj) if cookies_path_obj else None
auth_token = None
return {
"status": "ok",
"event_id": event_id,
"source_id": source_id,
"url": source.url,
"platform": source.platform,
"artist_slug": artist.slug if artist else "_unknown",
"artist_id": artist.id if artist else None,
"config_overrides": dict(source.config_overrides or {}),
"cookies_path": cookies_path,
"auth_token": auth_token,
"backfill_runs_remaining": source.backfill_runs_remaining or 0,
}
async def _phase3_persist(
self,
event_id: int,
ctx: dict,
dl_result,
resolved_campaign_id: str | None,
) -> int:
# Cache Patreon campaign id on the source.
if resolved_campaign_id:
src = self.sync_session.get(Source, ctx["source_id"])
if src is not None:
src.config_overrides = {
**(src.config_overrides or {}),
"patreon_campaign_id": resolved_campaign_id,
}
self.sync_session.commit()
artist = None
source_row = None
if ctx.get("artist_id"):
artist = self.sync_session.get(Artist, ctx["artist_id"])
source_row = self.sync_session.get(Source, ctx["source_id"])
import_summary = {"attached": 0, "skipped": 0, "errors": 0}
bytes_downloaded = 0
loop = asyncio.get_running_loop()
for path_str in dl_result.written_paths or []:
if path_str in (dl_result.quarantined_paths or []):
continue
path = Path(path_str)
# Sync stdlib filesystem ops are intentional here: this orchestrator
# runs under `asyncio.run` inside a Celery task — no other awaitables
# compete for the loop. ASYNC240 suppressed below.
if not path.exists(): # noqa: ASYNC240
continue
def _attach(p=path):
return self.importer.attach_in_place(
p, artist=artist, source=source_row,
)
result = await loop.run_in_executor(None, _attach)
if result.status in ("imported", "superseded"):
import_summary["attached"] += 1
try:
bytes_downloaded += path.stat().st_size # noqa: ASYNC240
except OSError:
pass
elif result.status == "skipped" and result.skip_reason and result.skip_reason.value in (
"duplicate_hash", "duplicate_phash",
):
import_summary["skipped"] += 1
try:
path.unlink(missing_ok=True) # noqa: ASYNC240
except OSError:
pass
else:
import_summary["errors"] += 1
ev = (await self.async_session.execute(
select(DownloadEvent).where(DownloadEvent.id == event_id)
)).scalar_one()
run_stats = self.gdl._compute_run_stats(
dl_result.return_code, dl_result.stdout, dl_result.stderr
)
run_stats["quarantined_count"] = dl_result.files_quarantined
stderr_summary = self.gdl._extract_errors_warnings(dl_result.stderr)
# Plan #544: PARTIAL means the run downloaded ≥1 file but the
# subprocess didn't finish in budget (typically wall-clock timeout
# mid-walk). Real work happened; the next tick continues via
# gallery-dl's archive. NOT a failure for status purposes.
if dl_result.success and import_summary["errors"] == 0:
status = "ok"
elif dl_result.error_type == ErrorType.PARTIAL and import_summary["errors"] == 0:
status = "ok"
else:
status = "error"
ev.status = status
ev.finished_at = datetime.now(UTC)
ev.files_count = import_summary["attached"]
ev.bytes_downloaded = bytes_downloaded
ev.error = dl_result.error_message if status == "error" else None
ev.metadata_ = {
"run_stats": run_stats,
"error_type": dl_result.error_type.value if dl_result.error_type else None,
"stdout": self.gdl._truncate_log(dl_result.stdout) or None,
"stderr": self.gdl._truncate_log(dl_result.stderr) or None,
"stderr_errors_warnings": stderr_summary or None,
"duration_seconds": dl_result.duration_seconds,
"quarantined_paths": dl_result.quarantined_paths or None,
"import_summary": import_summary,
}
await self._update_source_health(
source_id=ctx["source_id"], status=status, error_message=ev.error,
error_type=dl_result.error_type.value if dl_result.error_type else None,
)
# Plan #544: backfill lifecycle — auto-complete when a clean
# backfill run drained the queue (gallery-dl exited 0 + zero files
# downloaded means there was nothing to fetch); otherwise decrement
# the counter. Next tick falls back to tick mode once it hits 0.
backfill_remaining = ctx.get("backfill_runs_remaining", 0) or 0
if backfill_remaining > 0:
src = (await self.async_session.execute(
select(Source).where(Source.id == ctx["source_id"])
)).scalar_one()
if dl_result.return_code == 0 and dl_result.files_downloaded == 0:
src.backfill_runs_remaining = 0
else:
src.backfill_runs_remaining = max(0, backfill_remaining - 1)
await self.async_session.commit()
return event_id
async def _update_source_health(
self, *, source_id: int, status: str, error_message: str | None,
error_type: str | None = None,
) -> None:
"""FC-3d: update Source.{consecutive_failures, last_error, last_checked_at}.
ok -> failures = 0, error = None, checked_at = now
error -> failures += 1, error = error_message, checked_at = now
skipped -> failures unchanged, error = None, checked_at = now
When error_type == 'rate_limited', also stamps a platform-wide
cooldown via scheduler_service.set_platform_cooldown so the next
scan tick skips every source on this platform until the cooldown
expires. Preventive half of the burst-prevention pair —
consecutive_failures still backs the offending source off across
ticks.
"""
source = (await self.async_session.execute(
select(Source).where(Source.id == source_id)
)).scalar_one()
now = datetime.now(UTC)
if status == "ok":
source.consecutive_failures = 0
source.last_error = None
elif status == "error":
source.consecutive_failures = (source.consecutive_failures or 0) + 1
source.last_error = error_message
if error_type == "rate_limited":
await set_platform_cooldown(self.async_session, source.platform)
elif status == "skipped":
source.last_error = None
source.last_checked_at = now
async def _finalize_event_and_source(
self, *, event_id: int, source_id: int, status: str,
error_message: str | None,
) -> None:
"""Finalize the event AND the source-health columns in one
transactional step. Used by the skipped early-out path and by
unit tests that exercise the source-health writes directly.
"""
ev = (await self.async_session.execute(
select(DownloadEvent).where(DownloadEvent.id == event_id)
)).scalar_one()
ev.status = status
ev.finished_at = datetime.now(UTC)
ev.error = error_message
await self._update_source_health(
source_id=source_id, status=status, error_message=error_message,
)
await self.async_session.commit()