c55596dd6f
- _search_base: hak _thumbnail_map (mapa url -> miniatura z calej strony) przed szukaniem <img> w oknie tekstu; xhamster mial 54% zrodel bez miniatury - mypornerleak search: karty maja teraz src zamiast data-src; browse (REST) bez featured media -> og:image ze strony posta - scripts/backfill_thumbs.py: uzupelnia miniatury ze strony sceny, od najnowszych Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
32 lines
1.2 KiB
Python
32 lines
1.2 KiB
Python
"""XHamster.com, direct HTML scrape search results.
|
|
|
|
Search: `https://xhamster.com/search/<q>?page=<n>`
|
|
Scene URL: `https://xhamster.com/videos/<slug>-<id>`
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import re
|
|
|
|
# Karty sa renderowane z JSON-a w HTML (`window.initials`): kazda ma `pageURL` i
|
|
# `thumbURL` (526x298 webp). Szukanie `<img>` w oknie tekstu gubilo polowe scen, bo
|
|
# ten sam link wystepuje dwa razy (miniatura i tytul), a img jest tylko przy pierwszym.
|
|
_JSON_CARD_RE = re.compile(
|
|
r'"pageURL":"(https:\\/\\/xhamster\.com\\/videos\\/[^"]+)","thumbURL":"([^"]+)"'
|
|
)
|
|
|
|
from app.connectors.direct_scrapers._search_base import BaseSearchScraper
|
|
|
|
|
|
class XHamsterScraper(BaseSearchScraper):
|
|
sitetag = "xhamstercom"
|
|
_search_url_template = "https://xhamster.com/search/{query}?page={page}"
|
|
_scene_url_re = re.compile(
|
|
r'href="(?P<url>https://xhamster\.com/videos/(?P<slug>[a-z0-9_\-]+))"',
|
|
re.IGNORECASE,
|
|
)
|
|
|
|
def _thumbnail_map(self, listing_html: str) -> dict[str, str]:
|
|
out: dict[str, str] = {}
|
|
for page_url, thumb in _JSON_CARD_RE.findall(listing_html):
|
|
out[page_url.replace("\\/", "/")] = thumb.replace("\\/", "/")
|
|
return out
|