The redirect proxy from 6ba65f1 still played the one-minute "watch on
eporner.com" clip away from home: checked from foreign networks
(check-host.net nodes in BR/DE/HU), a freshly signed link answers 302 ->
static.eporner.com/na.mp4, while the IP that resolved it gets the file. A
redirect therefore only works for a player behind the server's public IP.
/proxy/eporner/{alnum_id}/{height}.mp4 now streams the mp4: resolves the
link (cached 90 min, re-resolved when the CDN answers with a redirect),
forwards Range, and returns a sized body so 206 responses keep
Content-Length. It is no longer a redirect proxy, so app clients keep the
URLs in formats. Verified locally: HEAD, 0-1, mid-file and open-ended ranges
return 206 with Content-Length/Content-Range and valid MP4 data.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
Expired eporner mp4 links are answered with a 302 to static.eporner.com/
na.mp4, a one-minute "watch it on eporner.com" clip. The links expire
exactly two hours after the /xhr/video call, so anything played from a
favourite, history entry or a list left open past that hit the decoy.
New /proxy/eporner/{alnum_id}/{height} re-resolves the hash/xhr chain at
play time and 302s to the matching (else best) quality. The eporner
provider and melonstube's eporner destinations now emit these proxy URLs;
verified app clients get the best quality as video.url via the existing
redirect-proxy rule. Verified live: proxy -> fresh CDN link -> 206 video/mp4
for app and non-app clients.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
- Parser no longer requires a pagination block: single-page results
(small tags, short searches) used to parse to zero items.
- Search path is slugged like the site's own form: "2b/9s" 404'd, "%20"
is refused with 403 and "&" truncated the query.
- Probed archives need >= 4 videos to beat keyword search ("elf girl" was a
one-video tag); conclusive query -> target resolutions are remembered for 6h
so later pages and the prefetch skip the probes.
- Enriched details are cached per video URL for 2h (tokens last ~12h);
detail fetches go through get_probe so a lasting 429 stops the batch instead
of escalating each page to Jina/FlareSolverr. Archive failures serve stale
cache.
- requester: get_probe waits before retrying 429/5xx and compares redirect
paths percent-decoded (non-ASCII archives like /tags/pokémon-species/ were
treated as redirected away); get_with_headers backs off 1/2/4s on 429.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
decorate_channel ignored every provider-declared ytdlpCommand except
pimpbunny's. Send the declared command for freepornvideosxxx and missav,
where plain yt-dlp is verified to fail; other channels stay unchanged.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
Raw requester calls did not adopt the remembered FlareSolverr user agent,
so a RedGifs token minted under it was rejected on every gif lookup.
Tag/creator pages that mix non-RedGifs sources are refilled from later
cursor batches instead of coming back empty.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
The thumb proxy dropped the preview query string, rejected extensionless
getVideoPreview paths and warmed up through the slow fallback chain on
every image. check.py now verifies that a proxied video.url resolves
instead of failing it outright.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
xxdbx assets are now /id.jpg, not //host/..; prefixing https: produced
https:/id.jpg. cumlouder's unknown-tag 404 went through the requester's
mirror/FlareSolverr fallbacks (146s); read the raw response and return
an empty list at once while still retrying 5xx.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
Deleted lulustream files stayed listed and 502'd. myvidplay embeds lost
their host and were resolved against lulustream; route them through the
doodstream proxy (now allowing myvidplay/playmogo) with the host kept.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
The site serves one results page per searchgate POST, so later searches
and page 2 returned nothing on the shared cookie session. Use an isolated
session per (query, page). check.py now retries a refused HEAD (403/404/
405/501) with a ranged GET before reporting a URL as unreachable.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
Sponsored channel cards carry a stale data-video-id that 404s; use the
card link target. The signed pix-cdn77 thumbnail hosts 403 without a
Referer, so route them (tube8, ph-shorties) through a new host-restricted
aylo-thumb proxy. Also correct the pornhub-thumb route that sat under a
doubled /proxy prefix.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
The API ignores page= (every page 2+ repeated page 1), searches shared a
single cache entry, image-only posts were listed as unplayable videos, and
gzip/br bodies were not decoded because wreq had no decompression features.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
r.jina.ai now answers 403/429, so every request returned nothing. The
site serves plain HTML to the emulating client; parse cards, detail rows
and option lists from HTML, and return empty for no-hit searches (the
site renders a recommendation tray under the notice).
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
The site moved to article.video-card; the old front-video-card parser found
nothing and the unbounded detail fan-out timed out. The page URL now
resolves natively in yt-dlp, so it is returned directly.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi
