97bca096e6
Watchdog odtwarzania: origin z >= 300 zywymi zrodlami i zerem udanych odtworzen w 7 dni (hentaimama i xanime padly tak, ze nic nie alarmowalo). latestpornvideo: link podgladu pixhost zamieniany na obrazek, usuniety obrazek i brak miniatury -> plakat odtwarzacza lulu; skrypt dla starych. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
94 lines
3.2 KiB
Python
94 lines
3.2 KiB
Python
"""latestpornvideo: odzyskanie miniatur, ktore nie dzialaja (2026-10-15).
|
|
|
|
Scraper od dzis sam poprawia nowe wpisy (`better_thumb`), tu poprawiamy istniejace:
|
|
- brak miniatury,
|
|
- `pixhost.to/show/...` (strona podgladu zamiast obrazka),
|
|
- obrazek pixhost usuniety (przekierowanie na /image/removed.png),
|
|
- z `--imgds`: miniatury imgds.net (dawne ticdn.net), ktore daja 404.
|
|
|
|
W kazdym z tych przypadkow zapasem jest plakat odtwarzacza lulu z wpisu.
|
|
|
|
python scripts/fix_latestpornvideo_thumbs.py # podglad 20 wierszy
|
|
python scripts/fix_latestpornvideo_thumbs.py --apply --limit 20000 [--imgds]
|
|
|
|
Dlugie, na prodzie przez `docker compose run -d --rm worker python3 /srv/scripts/...`.
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import argparse
|
|
import logging
|
|
import time
|
|
|
|
import httpx
|
|
import sqlalchemy as sa
|
|
|
|
from app.connectors.direct_scrapers.latestpornvideo import better_thumb
|
|
from app.db import session_scope
|
|
|
|
log = logging.getLogger(__name__)
|
|
|
|
|
|
def _imgds_missing(url: str) -> bool:
|
|
try:
|
|
r = httpx.head(url, headers={"User-Agent": "Mozilla/5.0"}, timeout=10.0, follow_redirects=True)
|
|
except Exception:
|
|
return False
|
|
return r.status_code == 404
|
|
|
|
|
|
def run(apply: bool, limit: int, imgds: bool, sleep: float) -> dict[str, int]:
|
|
c = {"checked": 0, "changed": 0, "unchanged": 0, "lost": 0}
|
|
cond = (
|
|
"thumbnail_url IS NULL OR thumbnail_url LIKE '%pixhost%'"
|
|
+ (" OR thumbnail_url LIKE 'https://imgds.net/snaps/%'" if imgds else "")
|
|
)
|
|
with session_scope() as s:
|
|
rows = s.execute(
|
|
sa.text(
|
|
"SELECT id, page_url, thumbnail_url FROM playback_sources "
|
|
"WHERE origin='tube:latestpornvideocom' AND dead_at IS NULL "
|
|
f"AND ({cond}) ORDER BY id LIMIT :lim"
|
|
),
|
|
{"lim": limit},
|
|
).all()
|
|
for pb_id, page_url, thumb in rows:
|
|
c["checked"] += 1
|
|
candidate = thumb
|
|
if thumb and thumb.startswith("https://imgds.net/snaps/") and _imgds_missing(thumb):
|
|
candidate = None
|
|
new = better_thumb(candidate, page_url, timeout=30.0)
|
|
if new == thumb:
|
|
c["unchanged"] += 1
|
|
elif not new:
|
|
c["lost"] += 1
|
|
else:
|
|
c["changed"] += 1
|
|
if c["changed"] <= 5:
|
|
print(f"{page_url}: {thumb} -> {new}")
|
|
if apply:
|
|
with session_scope() as s:
|
|
s.execute(
|
|
sa.text("UPDATE playback_sources SET thumbnail_url=:u WHERE id=:id"),
|
|
{"u": new[:2048], "id": pb_id},
|
|
)
|
|
s.commit()
|
|
if c["checked"] % 200 == 0:
|
|
log.info("lpv thumbs %s", c)
|
|
time.sleep(sleep)
|
|
return c
|
|
|
|
|
|
def main() -> int:
|
|
logging.basicConfig(level=logging.INFO)
|
|
p = argparse.ArgumentParser()
|
|
p.add_argument("--apply", action="store_true")
|
|
p.add_argument("--limit", type=int, default=20)
|
|
p.add_argument("--imgds", action="store_true")
|
|
p.add_argument("--sleep", type=float, default=0.2)
|
|
args = p.parse_args()
|
|
print(run(args.apply, args.limit, args.imgds, args.sleep))
|
|
return 0
|
|
|
|
|
|
if __name__ == "__main__":
|
|
raise SystemExit(main())
|