Schedule stale gallery enrichment

This commit is contained in:
qananasikq
2026-08-18 10:40:26 +03:00
parent 19007dafc8
commit ca1fab1034
15 changed files with 460 additions and 82 deletions
+92 -45
View File
@@ -7,7 +7,7 @@ import signal
import hashlib
import re
import unicodedata
from datetime import datetime, timezone
from datetime import datetime, timedelta, timezone
from threading import Event, Thread
import time
import uuid
@@ -2431,7 +2431,7 @@ def mobilede_sync_detail_task(self, listing_id: str, lane: str = "mobile_de_cars
@shared_task(
name="mobilede.enrich_images_batch",
queue=MOBILEDE_SYNC_QUEUE,
queue=MOBILEDE_IMAGES_QUEUE,
bind=True,
max_retries=1,
default_retry_delay=120,
@@ -2439,53 +2439,100 @@ def mobilede_sync_detail_task(self, listing_id: str, lane: str = "mobile_de_cars
)
def mobilede_enrich_images_batch_task(
self,
limit: int = 50,
lane: str = "mobile_de_cars",
max_existing_images: int = 1,
delay_seconds: float = 1.0,
batch_size: int | None = None,
max_existing_images: int | None = None,
staleness_hours: int | None = None,
delay_seconds: float | None = None,
):
runtime_config = RuntimeConfig.from_file(Settings().runtime_config_file)
enrichment = runtime_config.mobilede.enrichment
if not enrichment.enabled:
return {
"status": "disabled",
"candidates": 0,
"enriched": 0,
"failed": 0,
"skipped": 0,
}
effective_batch_size = enrichment.batch_size if batch_size is None else max(1, int(batch_size))
effective_max_images = (
enrichment.max_existing_images
if max_existing_images is None
else max(0, int(max_existing_images))
)
effective_staleness_hours = (
enrichment.staleness_hours
if staleness_hours is None
else max(0, int(staleness_hours))
)
effective_delay = enrichment.delay_seconds if delay_seconds is None else max(0.0, float(delay_seconds))
redis_client = _get_redis()
owner_token = str(getattr(self.request, "id", None) or uuid.uuid4().hex)
lock_ttl = max(300, int(float(os.getenv("MOBILEDE_ENRICH_LOCK_TTL_SECONDS", "3600"))))
if not _acquire_lock(redis_client, MOBILEDE_ENRICH_IMAGES_LOCK_KEY, owner_token, lock_ttl):
return {
"status": "locked",
"candidates": 0,
"enriched": 0,
"failed": 0,
"skipped": 0,
}
persistence = _get_persistence()
scraper = MobileDeScraper(persistence=persistence)
candidates = persistence.get_active_cars_batch_for_image_enrich(
limit=limit,
max_existing_images=max_existing_images,
)
enriched = 0
failed = 0
skipped = 0
for _car_id, origin_id, origin_url, image_count in candidates:
listing_id = _mobilede_extract_listing_id(origin_url) or str(origin_id).rsplit(":", 1)[-1]
if not listing_id:
skipped += 1
continue
try:
result = scraper.sync_detail(str(listing_id), lane=lane)
enriched += 1
logger.info(
"mobile.de image enrich completed: listing_id=%s origin_id=%s old_images=%s result=%s",
listing_id,
origin_id,
image_count,
result.get("upsert", {}),
)
except Exception as exc:
failed += 1
logger.warning(
"mobile.de image enrich failed: listing_id=%s origin_id=%s error=%s",
listing_id,
origin_id,
exc,
exc_info=True,
)
if delay_seconds:
time.sleep(max(0.0, float(delay_seconds)))
return {
"status": "success",
"candidates": len(candidates),
"enriched": enriched,
"failed": failed,
"skipped": skipped,
}
try:
stale_before = datetime.now(timezone.utc) - timedelta(hours=effective_staleness_hours)
candidates = persistence.get_active_cars_batch_for_image_enrich(
max_existing_images=effective_max_images,
stale_before=stale_before,
)
enriched = 0
failed = 0
skipped = 0
for batch_start in range(0, len(candidates), effective_batch_size):
batch = candidates[batch_start:batch_start + effective_batch_size]
for _car_id, origin_id, origin_url, image_count in batch:
listing_id = _mobilede_extract_listing_id(origin_url) or str(origin_id).rsplit(":", 1)[-1]
if not listing_id:
skipped += 1
continue
try:
result = scraper.sync_detail(str(listing_id), lane=lane)
enriched += 1
logger.info(
"mobile.de image enrich completed: listing_id=%s origin_id=%s old_images=%s result=%s",
listing_id,
origin_id,
image_count,
result.get("upsert", {}),
)
except Exception as exc:
failed += 1
logger.warning(
"mobile.de image enrich failed: listing_id=%s origin_id=%s error=%s",
listing_id,
origin_id,
exc,
exc_info=True,
)
if effective_delay:
time.sleep(effective_delay)
return {
"status": "success",
"candidates": len(candidates),
"enriched": enriched,
"failed": failed,
"skipped": skipped,
}
finally:
_release_lock_if_owner(
redis_client,
MOBILEDE_ENRICH_IMAGES_LOCK_KEY,
owner_token,
)
@shared_task(