8c0edbdf7b
Two bug-report fixes (2026-06-07): - sxyprn returns HTTP 200 "Post Not Found" for deleted posts (soft-404), so the extractor returned None → resolve treated it as transient and never marked the source dead, leaving a dead link offered forever. Now raise HosterDead on the marker so resolve marks it dead. - Scene playback sources were ordered alphabetically by origin, so a WebView- fallback hoster (fpoxxx, IP-bound + ad-heavy) ranked above a working native source (freshporno) on the same scene. Add is_vps_blocked_fallback() and sort native-resolve origins ahead of WebView-fallback ones. Verified on prod: sxyprn dead URL → HosterDead; scene sources reorder freshpornoorg before fpoxxx. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
215 lines
11 KiB
Python
215 lines
11 KiB
Python
"""Stream URL extractors per-tube.
|
|
|
|
Public API:
|
|
- `try_extract(sitetag, page_url) -> list[StreamSource] | None`
|
|
- `StreamSource` (dataclass)
|
|
- `HosterDead` (exception)
|
|
- `extract_stream_from_hoster(iframe_url, *, referer)` — generic packer-based hoster extract
|
|
- `fetch_tube_html(url)` — Chrome TLS fingerprint fetch (curl_cffi)
|
|
- `browser_get(url)` — low-level
|
|
|
|
Architektura: każdy tube ma osobny moduł `app.extractors.tubes.<tube>` który eksportuje
|
|
`extract(page_url) -> list[StreamSource] | None`. Registry niżej mapuje sitetag →
|
|
modułowy extractor. `try_extract()` to thin wrapper z exception handlingiem.
|
|
|
|
Po removalu porn-app dependency, ten moduł jest jedynym mechanizmem rozwiązywania
|
|
streamów — playback.py nie wpada już do porn-app /stream API.
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import logging
|
|
from collections.abc import Callable
|
|
|
|
from app.extractors._fetch import browser_get, fetch_tube_html
|
|
from app.extractors._models import HosterDead, StreamSource, TubePageError
|
|
from app.extractors.hoster import extract_stream_from_hoster, unpack_packer
|
|
from app.extractors.tubes import (
|
|
_embed_iframe,
|
|
_vps_blocked_fallback,
|
|
_ytdlp,
|
|
eporner,
|
|
freshporno,
|
|
fullmovies,
|
|
hdporngg,
|
|
hqporner,
|
|
latestpornvideo,
|
|
paradisehill,
|
|
porn00,
|
|
pornditt,
|
|
pornhat,
|
|
porntrex,
|
|
sxyprn,
|
|
yespornvip,
|
|
)
|
|
|
|
log = logging.getLogger(__name__)
|
|
|
|
|
|
# Sitetag → extractor function. Sitetag pasuje do format'u z origin: `pornapp:<sitetag>`
|
|
# (lub po Fazie 2 migracji: `tube:<sitetag>`).
|
|
#
|
|
# Mainstream tubes (pornhub/xvideos/xnxx/xhamster/redtube/youporn/porntrex) używają
|
|
# yt-dlp jako extractor — battle-tested, aktualizowane przez upstream przy zmianach
|
|
# HTML. Aggregator tubes (xmoviesforyou/watchporn/siska/...) używają generic
|
|
# embed-iframe extractor (page → /e/<id> iframe → P.A.C.K.E.R. unpack). Custom kod
|
|
# tylko tam gdzie tube ma niestandardowy schemat (eporner XHR, sxyprn URL transform).
|
|
_REGISTRY: dict[str, Callable[[str], list[StreamSource] | None]] = {
|
|
# hqporner — dedicated extractor zwraca multi-quality `<source>` mp4 URLs
|
|
# (bigcdn.cc / hqwo.cc / flyflv) z `force_proxy=True`. CDN URLs IP-bound do
|
|
# VPS, więc playback.py routuje przez proxy — mobile dostaje quality picker
|
|
# + natywny ExoPlayer, bez WebView.
|
|
# Bug-report e8ddd8d4: WebView fallback (`_vps_blocked_fallback`) ładował
|
|
# hqporner.com scene page w WebView, ale ta strona ma ad-iframes (adtng,
|
|
# goaserv, mavrtracktor) + pop-under-triggery → user klikał i widział
|
|
# reklamę zamiast video. INJECTED_JS w PlayerScreen.tsx nie chwytał
|
|
# popupów dośc szybko. Powrót do natywnego = `<source>` mp4 picker omija
|
|
# tę ścieżkę całkowicie.
|
|
"hqpornercom": hqporner.extract,
|
|
"epornercom": eporner.extract,
|
|
"sxyprncom": sxyprn.extract,
|
|
# Mainstream tubes — yt-dlp
|
|
# NB: 2026-05-18 cross-IP test potwierdził że xvideos/xnxx/pornhub/youporn/redtube
|
|
# CDN URLs są **time-bound** (nie IP-bound) — mobile_direct_ok auto-detect w
|
|
# playback.py daje mobile direct fetch, zero VPS bandwidth.
|
|
# pornhub — 2026-06-02: yt-dlp z VPS dostaje HTTP 403 (Pornhub blokuje Hetzner IP;
|
|
# yt-dlp aktualny, inne yt-dlp tuby działają → blok specyficzny dla PH). WebView
|
|
# fallback gra z residential IP telefonu (jak xhamster). Wcześniej `_ytdlp.extract`
|
|
# zwracał 0 źródeł → "nie działa odtwarzanie".
|
|
"pornhubcom": _vps_blocked_fallback.extract,
|
|
"redtubecom": _ytdlp.extract,
|
|
"xvideoscom": _ytdlp.extract,
|
|
"xnxxcom": _ytdlp.extract,
|
|
"youporncom": _ytdlp.extract,
|
|
# porntrex KVS — 2026-05-22 VPS znów dociera (HTTP 200). Dedykowany extractor:
|
|
# flashvars `video_url` → `get_file` 302 → CDN time-bound signed URL
|
|
# (`expires`+`md5`, NIE IP-bound) → mobile gra direct, zero VPS bandwidth.
|
|
"porntrexcom": porntrex.extract,
|
|
# VPS-blocked tubes — KVS / Cloudflare blokuje Hetzner IP, ale działają z residential
|
|
# IP (potwierdzone Chrome DevTools MCP 2026-05-15). Mobile WebView + INJECTED_JS
|
|
# (PlayerScreen.tsx:805) skanuje <video>.src + XHR — łapie URL po decode-ie player JS.
|
|
"xhamstercom": _vps_blocked_fallback.extract,
|
|
# pornditt — KVS jak yespornvip (function/0 + license). VPS dociera → resolve
|
|
# server-side (decode + follow 302 → portable twa.tgprn.com CDN). Wcześniej WebView
|
|
# fallback łapał VAST preroll (trafostatic) zamiast contentu. Patrz pornditt.py/_kvs.py.
|
|
"porndittcom": pornditt.extract,
|
|
# fpoxxx — KVS, plain get_file + license. 2026-06-01 (task #20): get_file 302 →
|
|
# `videos3.fpo.xxx/remote_control.php?acctoken=<base64>` — zdekodowany acctoken
|
|
# zawiera WBITY IP serwera-resolvera → definitywnie IP-bound. WebView only.
|
|
"fpoxxx": _vps_blocked_fallback.extract,
|
|
"sxylandcom": _vps_blocked_fallback.extract,
|
|
# Aggregator tubes — generic embed-iframe → hoster unpacker
|
|
"latestpornvideocom": latestpornvideo.extract,
|
|
"xmoviesforyoucom": _embed_iframe.extract,
|
|
"watchporn": _embed_iframe.extract,
|
|
"siskavideo": _embed_iframe.extract,
|
|
"porn4dayspw": _embed_iframe.extract,
|
|
"porndishcom": _embed_iframe.extract,
|
|
# xxxfreewatch — DELISTED 2026-05-18. 790 solo-orphan scen, 0% match, CF-walled z VPS.
|
|
"latestleaksco": _embed_iframe.extract,
|
|
"mypornerleakcom": _embed_iframe.extract,
|
|
# PornHat — dedicated extractor: tylko `<source>` z player area (skip sidebar
|
|
# trailer URLs `_preview*.mp4`), dedupe po filename. Get_file 302 → CDN, proxy
|
|
# follow_redirects=True wymagane (fix w stream_proxy.py).
|
|
"pornhatcom": pornhat.extract,
|
|
# Freshporno KVS (function/0 + license). 2026-06-04 DevTools + cross-IP re-test
|
|
# NAPRAWIA błąd z #20: finalny cdn4.freshporno.org/remote_control.php jest PORTABLE
|
|
# (token time-bound nie IP-bound — VPS odtworzył token z residential → 206) ale
|
|
# wymaga browser-TLS (curl_cffi chrome/ExoPlayer → 206; plain curl → 000). W #20
|
|
# testowałem plain-curl-em poza sesją → 000 → błędnie „nieosiągalny" → WebView.
|
|
# Teraz backend-resolve jak yespornvip/pornditt (_kvs używa curl_cffi chrome).
|
|
# Native, multi-quality, zero proxy/WebView. (zweryfikowane na emulatorze przed deploy)
|
|
"freshpornoorg": freshporno.extract,
|
|
# porn00 — KVS (plain get_file + license). 2026-06-04 DevTools + cross-IP re-test
|
|
# NAPRAWIA błąd z #20: finalny fe.porn00.org/...?token=&expires= jest PORTABLE
|
|
# (token time-bound nie IP-bound — Bright Data residential proxy z innego IP → 206)
|
|
# ale wymaga browser-TLS (curl_cffi chrome → 206; plain curl → 403). W #20
|
|
# testowałem finalny URL plain-curl-em → 403 → błędnie „IP-bound" → WebView.
|
|
# Teraz backend-resolve przez _kvs (curl_cffi chrome), native multi-quality,
|
|
# ZERO proxy (wcześniej force_proxy łamał no-proxy). Same mechanizm co freshporno.
|
|
"porn00org": porn00.extract,
|
|
# pornxp — `<source> //sr.porn-xp.com/<token>/.../720.mp4` (redirect → xpxp.eu).
|
|
# 2026-06-01 (task #20): 403 cross-IP → token w path IP-bound. WebView only.
|
|
"pornxpph": _vps_blocked_fallback.extract,
|
|
# yesporn.vip — KVS engine. VPS znów dociera (HTTP 200, odblokowane jak porntrex),
|
|
# więc resolvujemy SERVER-SIDE: dekoduj flashvars `video_url`/alt/alt2 (function/0/ +
|
|
# license_code, algo kt_player) → follow get_file 302 → portable cdn5 url (time-bound,
|
|
# NIE IP/cookie-bound, zweryfikowane cross-IP 2026-05-31). Mobile gra direct natywnie,
|
|
# multi-quality, ZERO WebView/reklam/preroll. Wcześniej WebView fallback pokazywał
|
|
# ad-heavy stronę a scrape łapał preroll-reklamę (bkcdn) zamiast wideo.
|
|
"yespornvip": yespornvip.extract,
|
|
# Direct-scraping tubes (mają też search scraper w connectors/direct_scrapers/)
|
|
# — używają identycznego embed-iframe pattern dla streamingu.
|
|
# hdporn92com — DELISTED 2026-05-18. Scene pages to SEO shell bez player iframe,
|
|
# JS hijackuje kliki na popunder. Wszystkie playback_sources mass-marked dead.
|
|
# 0dayxx wraps watchporn.to embed. watchporn.to/get_file/ token IP-bound (302→410
|
|
# cross-IP). Switch na WebView fallback. ~5k scen.
|
|
"0dayxxcom": _vps_blocked_fallback.extract,
|
|
# CF-protected tube — curl_cffi w fetch_tube_html bypassa JA3, embed-iframe pattern.
|
|
"perverzijacom": _embed_iframe.extract,
|
|
# Special: WebView-only (Yii2 session-bound player).
|
|
"paradisehillcc": paradisehill.extract,
|
|
# PornDoe — dołączony 2026-05-21 (theporndude audit). Stream URL nie inline w
|
|
# SSR HTML (player JS init po Play click), więc WebView fallback: mobile pobiera
|
|
# /watch/<id> z phone IP, player JS dekoduje video.src, INJECTED_JS scrape.
|
|
# 0 VPS bandwidth — zgodne z pre-public bandwidth/anonimowość priorytet.
|
|
"porndoecom": _vps_blocked_fallback.extract,
|
|
# fullmovies.xxx + hdporn.gg — BRAKOWAŁO extractora (try_extract→None→"no stream";
|
|
# fullmovies.xxx + hdporn.gg — ta sama platforma (`<source>/get_file/8512/`).
|
|
# 2026-06-04 (DevTools + cross-IP, naprawia „loading forever" + bug 19866e9e):
|
|
# get_file binduje fpvcdn do IP FETCHERA + jest stateless + ważny ≥90s, więc
|
|
# oddajemy get_file NIEZRESOLWOWANY (mobile_direct) — telefon follow-uje 302 →
|
|
# fpvcdn z IP telefonu → gra. POMIJAMY 4K (time-out 30s na fpvcdn = przyczyna
|
|
# „loading forever"; 720/480p gra ~1s). Native, multi-quality, ZERO proxy/WebView.
|
|
# (#19866e9e wcześniej źle: założyłem „get_file 403 IP-bound" testem plain-curl.)
|
|
"fullmoviesxxx": fullmovies.extract,
|
|
"hdporngg": hdporngg.extract,
|
|
}
|
|
|
|
|
|
def try_extract(sitetag: str, page_url: str) -> list[StreamSource] | None:
|
|
"""Próbuje rozwiązać stream URL dla danego tube'a + page_url.
|
|
|
|
Zwraca listę StreamSource (różne quality/kontener) lub None gdy:
|
|
- brak extractora dla tego sitetag
|
|
- extractor zwrócił None / nie znalazł URL'a
|
|
|
|
Raises HosterDead gdy embed page wprost mówi że video deleted/not found —
|
|
caller (playback.py) łapie i oznacza playback_source.dead_at.
|
|
"""
|
|
extractor = _REGISTRY.get(sitetag)
|
|
if extractor is None:
|
|
return None
|
|
try:
|
|
return extractor(page_url)
|
|
except (HosterDead, TubePageError):
|
|
raise
|
|
except Exception as e:
|
|
log.warning("extractor for %s failed on %s: %s", sitetag, page_url, e)
|
|
return None
|
|
|
|
|
|
def supported_sitetags() -> tuple[str, ...]:
|
|
"""Zwraca listę sitetag-ów które mają zarejestrowany extractor."""
|
|
return tuple(_REGISTRY.keys())
|
|
|
|
|
|
def is_vps_blocked_fallback(sitetag: str) -> bool:
|
|
"""True gdy sitetag resolvuje się TYLKO przez WebView fallback (IP-bound CDN /
|
|
ad-heavy / CAPTCHA — np. fpoxxx, pornxpph, pornhubcom). Takie źródła dają gorszy
|
|
UX (reklamy, czarny ekran) niż natywny KVS/direct resolve, więc UI powinien je
|
|
rankować NIŻEJ gdy scena ma też natywne źródło (bug-report 2026-06-07: scena
|
|
pokazywała fpoxxx-WebView przed działającym freshporno bo sort był alfabetyczny)."""
|
|
return _REGISTRY.get(sitetag) is _vps_blocked_fallback.extract
|
|
|
|
|
|
__all__ = [
|
|
"try_extract",
|
|
"supported_sitetags",
|
|
"StreamSource",
|
|
"HosterDead",
|
|
"TubePageError",
|
|
"extract_stream_from_hoster",
|
|
"unpack_packer",
|
|
"fetch_tube_html",
|
|
"browser_get",
|
|
]
|