more providers

build.rs

@@ -286,6 +286,11 @@ const PROVIDERS: &[ProviderDef] = &[
         module: "sextb",
         ty: "SextbProvider",
     },
+    ProviderDef {
+        id: "notfans",
+        module: "notfans",
+        ty: "NotfansProvider",
+    },
     ProviderDef {
         id: "hentaihaven",
         module: "hentaihaven",
@@ -296,6 +301,11 @@ const PROVIDERS: &[ProviderDef] = &[
         module: "fikfap",
         ty: "FikfapProvider",
     },
+    ProviderDef {
+        id: "hdporn92",
+        module: "hdporn92",
+        ty: "Hdporn92Provider",
+    },
     ProviderDef {
         id: "chaturbate",
         module: "chaturbate",
@@ -371,6 +381,11 @@ const PROVIDERS: &[ProviderDef] = &[
         module: "xgroovy",
         ty: "XgroovyProvider",
     },
+    ProviderDef {
+        id: "borntobefuck",
+        module: "borntobefuck",
+        ty: "BorntobefuckProvider",
+    },
     ProviderDef {
         id: "xvideos",
         module: "xvideos",
@@ -416,6 +431,16 @@ const PROVIDERS: &[ProviderDef] = &[
         module: "redgifs",
         ty: "RedgifsProvider",
     },
+    ProviderDef {
+        id: "hornyleak",
+        module: "hornyleak",
+        ty: "HornyleakProvider",
+    },
+    ProviderDef {
+        id: "fapello",
+        module: "fapello",
+        ty: "FapelloProvider",
+    },
 ];
 
 fn main() {

check.py

@@ -74,6 +74,14 @@ _CF_PROTECTED_HOSTS = {
     # plain requests/curl always see the "Just a moment" challenge.
     "hentaihaven.xxx",
     "www.hentaihaven.xxx",
+    # Cloudflare-gated; the shared requester uses the Jina mirror fallback.
+    # `video.url` is the detail page so it will always 403 to a plain request.
+    "fapello.com",
+    "www.fapello.com",
+    "cdn.fapello.com",
+    # hdplayer.gives returns 200 to plain curl but 404 to curl_cffi's JA3 — the
+    # HLS list/enc... endpoint is request-bound to the embed-page TLS context.
+    "hdplayer.gives",
 }
 
 

docs/README.md

@@ -5,7 +5,7 @@ This folder is the fastest handoff path for anyone adding or repairing a channel
 Start here:
 
 1. Read `architecture.md` for the server flow, request lifecycle, and core types.
-2. Read `provider-playbook.md` for the exact process to add a new provider or proxy.
+2. Read `provider-playbook.md` — its top **"Best Practices (Living)"** section is the cross-cutting rule set to internalize first, then follow the rest of the playbook for the exact process to add a new provider or proxy.
 3. Use `provider-catalog.md` to find the closest existing implementation to copy.
 4. Use `docs/hottubapp/*.html` when you need the client-facing API contract for status, videos, or uploaders.
 5. Only then touch `prompts/new-channel.md`; it assumes the docs above exist.

docs/provider-catalog.md

@@ -15,16 +15,19 @@ This is the current implementation inventory as of this snapshot of the repo. Us
 | `chaturbate` | `live-cams` | no | no | Live cam channel. |
 | `clapdat` | `amateur-homemade` | no | yes | Svelte/JSON-hydrated provider using home/recent/trending routes, Meilisearch keyword search, and `/proxy/clapdat/...` redirect playback resolution. |
 | `erome` | `amateur-homemade` | no | no | HTML album scraper with hot/new feeds, keyword search, and uploader-slug shortcuts (`uploader:<name>`). |
+| `fapello` | `onlyfans` | no | no | HTML + embedded-JSON provider for fapello.com (OnlyFans-style photo/video leak aggregator; the site is fundamentally a *photo* model archive — `/`, `/new/`, `/hot/`, `/trending/`, `/tags/`, `/search_v2/` — and the much smaller `/videos/` section is what this channel surfaces). Cloudflare-gated; the shared requester's Jina mirror fallback resolves every listing and detail page (`r.jina.ai/http://fapello.com/...`, `X-Return-Format: html`). Latest feed `GET /videos/` (page 1, ~6-8 cards), `GET /ajax/videos/page-{N}/` (page 2+, ~4 cards) — cards expose the numeric id, the `fapello.com/content/.../{model}_{NNNN}_300px.jpg` thumbnail, and the `/{model_slug}/` uploader link. Each detail page `GET /video/new/{id}/` embeds a complete `var BOOT = { items: [...], next_page, page_size }` JSON block (preserved by the Jina mirror) — every BOOT item carries the resolved `video_url` (`https://cdn.fapello.com/content/.../{N}/{model}_{NNNN}.mp4`), `poster_url` (the matching `.jpg`, no `_300px` suffix — provider swaps that in for the full-size thumb), `duration` ("HH:MM:SS" → seconds), `width`/`height`/`aspect`/`orientation`, `count_likes` (→ `views`), `name`/`model_url` (uploader), and `tags[]`. Per-card BOOT enrichment is bounded via `buffer_unordered(4)` with a 20s timeout — listing-only metadata (id, thumb, uploader) is preserved on timeout so a Jina-throttled page still produces usable cards. The Jina mirror does NOT trigger JS, so the `/popular_videos/{twelve_hours,month,all_time}/` routes and `/search_v2/?type=...` are empty (search returns *photo models*, not videos; the site has no native "videos matching keyword" search) — those filters are deliberately not exposed; only `sort:new` is offered. `video.url` is the detail page (not yt-dlp-resolvable — no extractor exists), so `formats[0]` carries the direct `cdn.fapello.com` mp4 with `Referer: https://fapello.com/` (the CDN serves direct without it, but the header matches the on-site context). Media and thumbnails (`fapello.com/content/.../*.jpg`, `cdn.fapello.com/content/.../*.mp4`) are range-served, no auth/cookies/JA3 emulation required, so no proxy route is needed. `check.py`'s `_CF_PROTECTED_HOSTS` allowlist includes `fapello.com` and `cdn.fapello.com` so its plain-requests 403s for `video.url` and the format URL downgrade to warnings instead of errors. No `/api/uploaders` (fapello has stable model slugs but no dedicated "videos by this uploader" feed — only the mixed photo+video model page). |
 | `fikfap` | `tiktok` | yes | yes (thumbs only) | JSON-API provider for fikfap.com (TikTok-style swipe short clips); anonymous auth via a client-generated `Authorization-Anonymous` UUID header (no real login needed); listing via `GET api.fikfap.com/posts?sort=new\|trending\|random&amount=N&afterId=<lastPostId>` (cursor pagination — page N costs N sequential requests); search via `GET search?q=` (single fixed-size batch, no pagination — page 2+ returns empty); hashtag feeds via `GET hashtags/label/{label}/posts` and creator feeds via `GET profile/username/{user}/posts`, both also cursor-paginated; `tag:`/`hashtag:`/`#` and `user:`/`uploader:` query prefixes route directly; `categories` option exposes a small curated static hashtag list (no full catalog endpoint exists anonymously); `video.url` is the `fikfap.com/post/{id}` page (a client-rendered SPA, not yt-dlp-resolvable on its own); `videoStreamUrl` from the JSON response is sent directly as `formats[0].url` (signed Bunny CDN HLS `.m3u8`, ~24h token expiry) with `httpHeaders: {Referer: https://fikfap.com/}` — Hot Tub clients apply a format's `http_headers` across the whole HLS playback session (manifest, sub-playlists, and segments), so no proxying of the media itself is needed; thumbnails have no per-field header mechanism, so they're proxied via `/proxy/fikfap-thumb/...` to inject the same Referer; `get_uploader` implemented (`fikfap:<username>` IDs) using `GET profile/username/{user}`. |
 | `freepornvideosxxx` | `studio-network` | no | no | Studio-style scraper. |
 | `fyptt` | `tiktok` | no | no | HTML scraper for fyptt.to (Beaver Builder/WordPress short-form TikTok-style vertical porn); card selector `.fl-post-grid-post[class*="post-ID"]` with `category-{slug}` CSS class doubling as both listing tag and category-archive route; latest feed `/` (page N: `/page/N/`), search `/?s=query` (page N: `/page/N/?s=query`), category archives at bare top-level slugs like `/tiktok-ass/` (12 hardcoded categories exposed via the `categories` filter option, or via an explicit `cat:`/`category:` query prefix — bare keyword queries always go to WordPress search, never a category archive, because the category names ("sexy", "ass", "tiktok", "live", ...) are also the most common search terms); per-item enrichment fetches the detail page for the JSON-LD `embedURL` (one of three on-site player endpoints: `fypttstr.php`, `fypttjwstr.php`, or `fypttjwstrhls.php`) and `datePublished`, then fetches that embed URL to extract the actual signed `stream.fyptt.to` mp4 or `/hls/*.m3u8` URL (token expires ~2h, no Referer required) for `formats`; thumbnails (`fyptt.to/wp-content/uploads/...webp`) need no proxy; no duration metadata available on listing or detail pages (set to 0); no real uploader/model identity (the `girl-{slug}` CSS class is cosmetic only, not a linkable archive) so `/api/uploaders` is not implemented; `video.url` is the detail page URL (not yt-dlp resolvable directly — the player is sandboxed-iframe-only) so `formats` are populated instead; no proxy needed. |
 | `freeuseporn` | `fetish-kink` | no | no | Fetish archive pattern. |
 | `hanime` | `hentai-animation` | no | yes | Uses proxied CDN/thumb handling. |
 | `heavyfetish` | `fetish-kink` | no | no | Direct media handling. |
+| `hdporn92` | `mainstream-tube` | no | no | WordPress retrotube-theme aggregator for hdporn92.com (network studio releases — MyPervyFamily, MYLF, StepSiblings, …). Latest feed `/` (page 1) and `/page/N/` (page N≥2); search `/?s={query}` (page 1) and `/page/N/?s={query}` (page N≥2) — the path with the `s` parameter must come after the `page/N` segment, not before. Pagination is by `article.loop-video` cards; the post id comes from the `post-{N}` CSS class. Two-step enrichment (page URL + Referer-locked m3u8 in `formats`): detail page → `<IFRAME SRC="https://morencius.com/embed/{id}">` (case-insensitive; the page emits an uppercase `<IFRAME SRC=...>` tag) → morencius embed → embedded `eval(function(p,a,c,k,e,d){…}(p,a,c,k.split('|')))` P.A.C.K.E.R. obfuscation block → deobfuscate in pure Rust (reverse-iterate the `c`-word `|`-delimited dictionary, base-`a` numeric tokens, `\b<token>\b` regex substitution) → `"hls4":"<value>"` field in the decoded JS gives the morencius stream URL `https://morencius.com/stream/{token}/{secret}/{expiry_ts}/{file_id}/master.m3u8`; `duration:"<value>"` gives the runtime. `video.url` is the hdporn92 page (no yt-dlp extractor exists for the site); `formats[0]` is the morencius `m3u8` carrying `Referer: https://morencius.com/embed/{id}` + a desktop Chrome `User-Agent` — these headers must travel with the manifest, sub-playlists, and segment requests. **Known limitation:** the morencius m3u8 path contains an `expiry_ts` segment (~12h from page render) — the cache TTL (`cacheDuration: 1800`, same as fyptt) caps the staleness window. Real playback still requires the webview to click through the morencius ad overlay; the manifest itself resolves to a 1280x720 HLS variant via `yt-dlp -j --referer <embed-url>`. No `/api/uploaders` (no uploader identity on the listing). |
 | `hentaihaven` | `hentai-animation` | no | no | HTML scraper for hentaihaven.xxx (WordPress/Madara theme), Cloudflare-protected so the provider is gated behind `FLARE_URL` in `skip_reason_for_provider` (mod.rs); the shared requester clears CF directly (wreq Firefox136 emulation currently passes for the listing/search/watch/episode/`player.php` GETs) and falls back to Jina/FlareSolverr. Latest feed `/hentai/page/{N}/`, search `/?s={query}` (search is single-page — page>1 returns empty); listing/search cards link to series watch pages `https://hentaihaven.xxx/watch/{slug}/`. Per-series media resolution (the UUID exists nowhere in page HTML, so enrichment is unavoidable): watch page → episode links `…/watch/{slug}/episode-K` (in `manga-chapters-holder`) → episode page → `<iframe src="…/wp-content/plugins/player-logic/player.php?data=…">` → `player.php` → `<meta name="x-secure-token" content="sha512-…">` → decode token (strip `sha512-`, then 3× of rot13→base64-decode, then `JSON.parse`) → `{en, iv, uri, hot_domains, …}` → POST `…/wp-content/plugins/player-logic/api.php` with `action=zarat_get_data_player_ajax&a={en}&b={iv}` (urlencoded; this one POST uses a dedicated `wreq` Chrome137 client, not the shared requester) → `{"status":true,"data":{"sources":[{"src":"…m3u8"}],"isOctopus":bool}}`. A multi-episode series collapses into one `VideoItem` titled `"… (N Episodes)"` with one `m3u8` `VideoFormat` per episode (`format_note`/`format_id` = "Episode K"); each format carries `Referer`/`Origin: https://hentaihaven.xxx` + a Firefox `User-Agent`. `video.url` is the `watch/{slug}/` page (no yt-dlp extractor exists for the site, so `formats` are populated rather than relying on `video.url`). Two CDN shapes are returned: newer content-addressed `octopusmanifest.org/{uuid}/playlist.m3u8` (no token, portable across IPs) and older signed `master-lengs.org/api/v3/hh/{slug}/master.m3u8?hash=…` (~2.5h). **Gotcha:** both CDNs (same IP) aggressively per-IP rate-limit/ban with a TCP RST on 80/443 once tripped — looks like "host down" but is an IP ban; browsers play fine over HTTP/3 (QUIC) while TCP clients (curl/yt-dlp/wreq) get refused, so segment fetches can fail from a tripped IP even though the manifest URL is valid. Tags from the series "Genre(s)" block; `views` from the "Viewed … Total" counter; thumbnails (`img.hentaihaven.xxx`) load directly (no proxy/referer). Resolution is slow (each listing page = ~25 series × multi-episode player-API calls), so the provider is DB-first: it fetches the listing once for the ordered watch URLs, serves already-resolved `VideoItem`s from the `videos` SQLite table instantly (`db::upsert_video` to avoid duplicate-row staleness), and `spawn_refresh`es the whole page in the background (in-memory `VideoCache` soft-TTL 1h / hard-TTL 24h, per-listing in-flight guard). No `/api/uploaders` (no uploader identity), no proxy. |
 | `hentaitv` | `hentai-animation` | no | yes | Next.js hentai site (hentai.tv) backed by a clean JSON API: `GET /api/browse?page=N&sort=<Label>&genres=<ExactName>` (`{videos:[28],total,pages}`, real pagination) and `GET /api/search?q=Q` (`{videos:[...]}`, single-page — `page` is ignored, so page>1 returns empty). Unlike `animeidhentai`, browse honors both `sort` (labels `Most Recent`/`Most Viewed`/`Trending`, mapped from option ids `new`/`views`/`trending`) and `genres` (the **exact case-sensitive** stored genre name, e.g. `Big Boobs`, `incest`), so genre archives go through `/api/browse?genres=` and paginate. The 68-genre catalogue (exact names) is background-loaded from the `/browse` page HTML (`"genres":[{"name","count"}]`, not exposed by the JSON API) and powers the `categories` filter plus keyword→genre routing. Each episode JSON has `slug`, `title`/`ep`, `tags[]`, `views`, `rating` (0-10 → ×10), `duration` ("MM:SS"), `brand` (studio → `uploader`), `thumb`/`backdrop`/`cover` (relative, served from `hentai.tv/uploads/...`, no referer), and `embedUrl=https://nhplayer.com/v/{embedId}/`. `video.url` is the reachable watch page `https://hentai.tv/hentai/{slug}`; `genre:`/`cat:`/`category:` prefixes and bare keywords that exactly match a genre route to the genre archive, everything else to search. Playback shares the **same nhplayer→`r2.1hanime.com` signed-CDN backend as `animeidhentai`**: `/proxy/hentaitv/{embedId}.mp4` is a redirect proxy that replicates nhplayer's PoW+DOM challenge (`player.php`→`player-core-v2.php`→`get-video-url-v2.php`, SHA-256-first-byte-zero PoW, ≥700ms dwell, fixed fingerprint) to mint a signed `?verify=<ts>-<sig>` URL — HEAD→200, GET→302 to the CDN URL (cached 150s). The CF wall is JA3-based not IP-based, so the signed URL is verifiable from anywhere with `yt-dlp --impersonate chrome` even though plain `curl`/`wreq` get 403. `src/proxies/hentaitv.rs` is a near-copy of `src/proxies/animeidhentai.rs` (only `SITE_REFERER` differs). No `/api/uploaders` (brand is studio-only). |
 | `homoxxx` | `gay-male` | no | no | Gay category grouping example. |
 | `hotbunny` | `ai` | no | yes (thumbs only) | JSON-API provider for hotbunny.ai (AI-generated hentai community). Feeds: `GET /api/post/feed?limit=N` (hot), `/api/post/feed/popular`, `/api/post/feed/recents`, `/api/post/category/{cuid}?limit=N`, `/api/post/search?query=Q&limit=N`, `/api/post/author/{name}/all?limit=N`. Pagination: fetch `page × per_page` items in one call and slice (cursor-based API, but limit works up to 200 total). Key fields: `id`, `imageUrl` (relative thumbnail path on CDN), `authorName`, `likes` (→ `views`), `sharedVideos[].videoUrl` (first video path, relative to CDN), `width`/`height`, `galleries[].label` (used as title prefix). 15 hardcoded categories from `/api/categories` (stable CUIDs). `video.url` is the post page `https://hotbunny.ai/post/{id}` (React SPA, not yt-dlp-resolvable); `formats[0].url` = `https://assets.hotbunny.ai/{sharedVideos[0].videoUrl}` with `Referer: https://hotbunny.ai/` header. `assets.hotbunny.ai` is Cloudflare bot-management protected (returns 404 to plain curl/yt-dlp; iOS AVFoundation passes with real Apple TLS — no proxy wrapping needed for video). Thumbnails proxied via `/proxy/hotbunny-thumb/...` to inject the required `Referer`. Supports `uploader:`/`author:`, `cat:`/`category:`/`#tag` query shortcuts, `categories` filter option, and `sort` option (feed/popular/recent). |
+| `hornyleak` | `amateur-homemade` | no | no | HTML scraper for hornyleak.tv (KVS / Kernel Video Sharing — leaked-amateur and OnlyFans-creator clips). Listing cards `div.thumb.thumb_rel.item` (24/page) with `a.img[href*="/video/"]` (href→`/video/{id}/{slug}/`, `title` attr→title), `img.thumb` (`data-webp` preferred, falls back to `data-src`/`src`), `div.duration` (text → seconds), `div.thumb-item > span` views text, `a.avatar` uploader (often empty — `-` placeholder on the listing), `div.rating` percent; `/videos/{N}/` is the path-based pagination for `Latest` and the per-archive pages. The detail page is yt-dlp-resolvable (page URL → generic extractor → `https://www.hornyleak.tv/video/{id}/{slug}/` → 1 format), so `video.url` is set to that detail page (NOT a proxy URL) and per-card enrichment is optional — but most clients want a stable HLS link, so the provider enriches each card to populate `formats[]`. Two-step media resolution chain: (1) `GET /embed/{id}/` (Referer: detail page) → regex on `hdplayer\.gives/embed/([A-Za-z0-9]+)` to find the per-video hash; (2) `GET https://hdplayer.gives/embed/{hash}/` (Referer: hornyleak.tv embed URL) → regex on `file:"(https://hdplayer\.gives/list/enc[^"]+)"` to extract the signed list/enc token URL. The enc URL is a real m3u8 with three variants (FULLHD 1920x1080 / HD 1280x720 / SD 842x480) pointing to `https://hdplayer.gives/m3u/...` sub-playlists and `https://swa3.top/upload/videos/...` segments. `formats[0]` is the m3u8 with `Referer: https://hdplayer.gives/` and the hornyleak.tv user-agent. `video.url` is left as the page URL (yt-dlp's generic extractor resolves it to 1 format on a non-JA3-blocked client; the HLS in `formats[]` is the primary path for clients that honor `http_headers`). `sort` option: new/popular/rated/viewed/longest (path-based: `/latest-updates/`, `/most-popular/`, `/top-rated/`, `/most-viewed/`, `/longest/`). `cat:`/`category:`/`tag:`/`model:`/`uploader:`/`pornstar:`/`star:` query prefixes route to the matching archive (`/categories/{slug}/`, `/tags/{slug}/`, `/models/{slug}/`). Search is path-1 `GET /search/?q={query}` (returns search results HTML directly) and path-≥2 `GET /search/?q={query}&mode=async&function=get_block&block_id=custom_list_videos_videos_list_search_result&from_videos={N}&from_albums={N}` (XHR pagination — the search results page itself is paged via the `custom_list_videos_videos_list_search_result` block). Background-loaded category filter (≈ 100 entries scraped from `/categories/`) via the shared `Arc<RwLock<Vec<FilterOption>>>` pattern, with title→slug map for bare-query short-circuit to a category archive when the keyword exactly matches a category title. Tags from detail `.tags-row a, .categories-row a`; thumbnails `https://www.hornyleak.tv/contents/videos_screenshots/{N}00/{id}/preview.jpg` (direct, no proxy/referer); no proxy route needed. `uploaderId` is namespaced as `hornyleak:<model-slug>`. **Known limitation:** the `hdplayer.gives/list/enc...` token is request-bound to the player-page TLS context; yt-dlp's generic extractor gets HTTP 404 when probing the URL directly with curl_cffi, so the validator's `_CF_PROTECTED_HOSTS` allowlist includes `hdplayer.gives` to downgrade the 404 to a warning (the HLS body itself returns 200 with valid m3u8 to plain curl, and iOS AVFoundation + Hot Tub clients play it directly through the `formats[]` entry). No `/api/uploaders` profile (KVS uploader pages exist but have no public stats). |
 | `hqporner` | `studio-network` | no | yes | Uses thumb and redirect proxy helpers. |
 | `hsex` | `chinese` | yes | no | Strong template for tags, uploaders, and direct HLS formats. |
 | `hypnotube` | `fetish-kink` | no | no | Fetish/tube hybrid. |
@@ -35,6 +38,7 @@ This is the current implementation inventory as of this snapshot of the repo. Us
 | `okporn` | `mainstream-tube` | no | no | Simple mainstream archive. |
 | `okxxx` | `mainstream-tube` | no | no | Mainstream search/archive pattern. |
 | `omgxxx` | `studio-network` | yes | no | Best template for sites/networks/stars filter catalogs. |
+| `notfans` | `onlyfans` | no | no | HTML scraper for notfans.com (OnlyFans / Fansly / Patreon / Manyvids leak aggregator); site is behind Cloudflare but does NOT issue a JS challenge to direct requests (the shared `Requester` with Chrome 120 emulation gets a 200), so detail-page enrichment is safe and there's no `FLARE_URL` gate. Card parser splits the listing on the literal `<div class="item  ">` delimiter (the two trailing spaces match the card class exactly and skip the sponsored `class="item avd"` variant — homoxxx uses the same convention); per card: `<a href="…/videos/{id}/{slug}/" title="…">` for the id/url/title, plus `<img class="thumb " src="…/contents/videos_screenshots/{id3}/{id}/320x180/N.jpg">` for the listing thumb. Feeds: latest `/latest-updates/{N}/` (23/page, last page 30990), most-viewed `/most-popular/{N}/` (24/page), top-rated `/top-rated/{N}/` (24/page), and period tops `/day/{N}/`, `/week/{N}/`, `/month/{N}/`, `/year/{N}/` (all 12-24/page); search `/search/{query}/{N}/`; tag `/tags/{tag}/{N}/` reachable via the `tag:`/`category:`/`tags:` query prefix. Per-card enrichment (page 1 only, `buffer_unordered(4)`, 8s timeout) fetches the detail page `…/videos/{id}/{slug}/` and pulls: `Duration: <em>HH:MM:SS</em>` → `parse_time_to_seconds`, `Views: <em>N</em>` → u32, `Tags: <a href="/tags/…">name</a>…</div>` block (regex over `/tags/…/>(name)</a>`), `Submitted: <em>N (minute|hour|day|week|month|year)s? ago</em>` → relative unix timestamp (`Yesterday`/`just now` are handled), `<meta property="og:image" content="…/preview.jpg">` for `preview` (the 720p still used by the player), and the first `Download:` link matching `get_file/6/{hash}/{id3}/{id}/{id}.mp4/?…` for the format URL (a `Referer: https://notfans.com/` and a matching `User-Agent` are attached as `http_headers`; the URL serves a 302 to `big1.phoenixcdn.lol/remote_control.php?…&file=…/{id}.mp4` which 200s `Content-Type: video/mp4` with the bytes — the 302 is followed automatically by clients, no proxy needed). 720p is the only available quality (the site shows one `MP4 720p, N Mb` row). Listing pages are cached 15 min in `VideoCache`. The site has no `/uploaders/{name}/` route and no `categories` system (the tag cloud is a 50-entry footer list dominated by stop-words like "a"/"and"/"the", and `/categories/` is a 0-item stub), so no `/api/uploaders` and no category/tag filter option. No proxy. |
 | `paradisehill` | `mainstream-tube` | no | no | Simple page scraper. |
 | `perfectgirls` | `studio-network` | no | no | Studio archive. |
 | `perverzija` | `studio-network` | no | no | WordPress (Generatepress/quick-view) HTML scraper for tube.perverzija.com (mirrors top-tier studio releases: MommysGirl, PureTaboo, VXN, AdultTime, etc). Default feed `/` (page N: `/page/N/`), `featured=featured` option swaps in `featured-scenes/` prefix before the page segment. Search is native WordPress `/?s=query` (page N: `/page/N/?s=query`), but `tag:`/`stars:`/`studio:`/`genre:` query prefixes and a background-learned title→slug map (populated from `studio`/`stars`/`tag`/`genre` links seen on every listing/detail page, normalized lowercase) route a bare keyword straight to the matching `/{kind}/{slug}/page/N/` archive when it matches — e.g. `milf` resolves via the learned `stars-milf` class to `/stars/milf/` rather than falling back to `?s=milf`. Listing cards (`class="video-item post..."`) carry everything needed (`data-url` = canonical WP detail page, `data-embed`-encoded `<iframe src="…xtremestream.xyz/player/index.php?data=ID">`, `tag-*`/`stars-*` CSS classes, `class="time_dur"` duration, studio link) so the default/tag/page feeds need no per-item enrichment; only the generic `?s=` search path is card-light (no embed/tags inline) and fetches each detail page concurrently (`futures::join_all`) for the iframe src plus `Studio:`/`Stars:`/`Tags:`/`Genres:` `<strong>` blocks — resolved detail data is cached in the `videos` SQLite table (`db::insert_video`/`get_video`) keyed by the WP page URL to skip re-fetching on repeat searches. `video.url` is the WordPress detail page (`data-url`/card href) — confirmed NOT yt-dlp-resolvable (the player loads via the quick-view AJAX iframe, not static HTML) — so `formats` carries the real media: the embed `index.php` path is rewritten to `xs1.php` (returns an HLS `.m3u8` master) and requires `Referer: https://xtremestream.xyz/` on every request (segments and master alike) or the host CDN returns `403`; **regression note**: an earlier revision set `video.url` to this same `xs1.php` media link, which then 403'd in `check.py`/any plain `GET` because the bare `url` field carries no header contract — fixed by separating `video.url` (page) from `formats[0].url` (media+Referer). Thumbnails (`tube.perverzija.com/wp-content/uploads/...jpg`) load directly, no proxy/referer. No `/api/uploaders` (studio/stars are folded into `tags`, no dedicated profile page schema confirmed). No proxy needed. |

docs/provider-playbook.md

@@ -2,6 +2,40 @@
 
 This is the implementation checklist for adding a working channel with the least guessing.
 
+## Best Practices (Living)
+
+These rules cut across every step below. They are intentionally grouped at the top so the rest of the playbook can reference them, and they are **living** — add, refine, or reorder as the codebase learns new failure modes. The provider catalog (`docs/provider-catalog.md`) is the ground truth for which patterns have actually held up in production; if a rule here contradicts a catalog note, the catalog wins and this list gets updated.
+
+### Investigation
+
+- **Always use an up-to-date client with a real TLS fingerprint when investigating a site.** A `curl` paste from a terminal looks fine and then 403s the moment the real fetcher goes near it. Use the same `wreq` emulation stack the server ships with (currently `Emulation::Firefox151` in the shared `Requester`, `Emulation::Firefox136` / `Chrome120` in older per-provider paths) and confirm a plain `wreq` GET against the live host returns 200 *before* trusting any URL or header you read off a page. If the investigation client differs from the production client, every "obvious" header you copy is suspect.
+- Start from the listing page, not the detail page or the network tab of a single video. The listing tells you the real card shape, pagination, sort, and tag surfaces. Detail pages lie about what is enumerable.
+- Confirm each URL shape with at least two pages: home, search page 2, one tag archive, one uploader archive, one detail. A pattern that holds for one page often breaks on the second.
+- Record a real browser session only when `wreq` cannot reach the site. If the shared requester plus Jina/FlareSolverr can fetch the page, the production path is the one you are investigating.
+
+### Parser Discipline
+
+- Prefer `serde_json` over regex/HTML scraping whenever the site exposes a JSON shape (Next.js `__NEXT_DATA__`, JSON-LD, hydration blobs, `RSC` payloads). When the site ships an API, scrape the API, not the rendered DOM.
+- Treat card metadata as the source of truth for `id`, `title`, `url`, `thumb`, `duration`. Only enrich detail pages for fields the card does not expose, and bound the concurrency (`futures::stream` + `buffer_unordered`).
+- Keep `title` text, `id` slug, and uploader identity as separate fields. A mushed title is a debugging nightmare.
+
+### Networking
+
+- Use the shared `Requester` from `ServerOptions`. Local clients drop cookies, lose Burp proxying, and bypass the Jina / FlareSolverr fallbacks. The only legitimate reason to build a fresh `wreq::Client` is a one-off TLS requirement that the shared emulation cannot meet (e.g. a specific Chrome JA3 that the r2.1hanime.com / nhplayer path needs).
+- After every fix to a CF/JA3 block, re-validate with `wreq` on the production emulation, not with `curl`. A/B probe with both the old and new fingerprint so you do not confuse a coincidence with a fix.
+
+### Output Shape
+
+- Use `formats` for real media URLs (HLS, multi-quality, signed token), and keep `video.url` on a stable page URL. Never set `video.url` to a tokenized stream that needs a `Referer` to play — the `url` field has no header contract.
+- If thumbnails need a `Referer` and the player has no per-thumb header mechanism, proxy them via `/proxy/<id>-thumb/...`. The same applies to manifests with relative URIs.
+- Keep filter `id` values stable and machine-targeted. Display `title`s change; ids route.
+
+### Operational
+
+- Add new providers to `build.rs` first, then write the file. A missing entry compiles fine and silently drops the channel.
+- After any code change, run `cargo check -q` once for the full build, then `HOT_TUB_PROVIDER=<id> cargo check -q` for the single-provider build. If only the second runs, you will not catch a regression that lives in another provider.
+- After modifying the code, refresh the knowledge graph with `graphify update .` so the next investigator can find what you just added without re-deriving it.
+
 ## Definition Of Done
 
 A provider is not done when it compiles. It is done when:

src/providers/fapello.rs

@@ -0,0 +1,709 @@
+// Fapello video provider.
+//
+// Fapello.com is primarily a model-by-model photo aggregator (the bulk of the
+// site — `/`, `/new/`, `/hot/`, `/trending/`, `/tags/`, `/search_v2/` — is all
+// photos hosted at `fapello.com/content/{a}/{b}/{model}/{N}/{model}_{NNNN}.jpg`).
+// The video section lives at `/videos/` and is much smaller (~6 cards per page),
+// with pagination via `GET /ajax/videos/page-{N}/` returning an HTML fragment
+// of cards (each is a `video/new/{id}/` link with a thumbnail and the model
+// slug visible on the card).
+//
+// The interesting bit: each video *detail page* (`/video/new/{id}/{slug}/`)
+// embeds a complete `var BOOT = { items: [...], ... }` JSON block on first
+// load. Every item in `BOOT.items` carries the resolved CDN media URL
+// (`video_url: https://cdn.fapello.com/content/.../{N}/{model}_{NNNN}.mp4`),
+// `poster_url` (the matching .jpg thumbnail), `duration` as "HH:MM:SS", the
+// uploader's `name`/`model_url`, `count_likes`, `width`/`height`/`aspect`,
+// `orientation` ("vertical" vs "landscape"), and the `slug` used to build
+// the canonical detail URL. This single JSON block is the source of truth
+// for every field on the VideoItem, so the provider does one BOOT-parsing
+// detail fetch per listing card.
+//
+// Cloudflare gated: every path on fapello.com except `/` and the `/feed/{id}/`
+// post pages returns a hard 403 to plain curl/chrome, including the videos
+// listings. The shared `Requester` auto-falls back to the Jina mirror
+// (`r.jina.ai/http://fapello.com/...`) on a CF challenge, and Jina preserves
+// both the listing HTML cards and the `<script>var BOOT = {...}</script>`
+// block (Jina is content-preserving on raw HTML inside <script> tags), so
+// the requester path works without any provider-side special-casing.
+//
+// Media: `cdn.fapello.com/.../*.mp4` is a direct, no-auth, no-Cookies HTTP
+// range-serving CDN — confirmed via `curl -I` (HTTP 200, `accept-ranges:
+// bytes`, `Content-Type: video/mp4`). Thumbs at `fapello.com/content/.../*.jpg`
+// also serve direct. No local `/proxy/...` route needed.
+//
+// Playback: the detail page URL (`fapello.com/video/new/{id}/{slug}/`) is
+// NOT yt-dlp-resolvable (no extractor exists), so `video.url` is the detail
+// page and `formats[0].url` carries the direct mp4 with a `Referer:
+// https://fapello.com/` header (the CDN is happy without one but Hot Tub
+// clients apply the format's `http_headers` across the HLS/MP4 playback
+// session, so including the Referer is safe and matches the site context).
+//
+// Known limitations (deferred, not implemented in v1):
+//   - Search: fapello's `/search_v2/?ajax=1&q=...` returns photo *models*,
+//     not videos. No native "videos containing keyword" search exists.
+//   - Popular: `/popular_videos/{twelve_hours,month,all_time}/` is JS-rendered
+//     and the AJAX body returns only ad/photo placeholders, no video cards.
+//   - Tags: the `/tags/{slug}/` archive is photo-only; no video tag pages.
+//   - `/api/uploaders`: fapello has stable model slugs (`/seeyanever/`) but no
+//     dedicated "videos by this uploader" feed — only the mixed photo+video
+//     model page, which is out of scope for a video channel.
+
+use crate::DbPool;
+use crate::api::ClientVersion;
+use crate::providers::{Provider, report_provider_error_background, requester_or_default};
+use crate::status::*;
+use crate::util::cache::VideoCache;
+use crate::videos::{ServerOptions, VideoFormat, VideoItem};
+
+use async_trait::async_trait;
+use error_chain::error_chain;
+use futures::stream::{self, StreamExt};
+use regex::Regex;
+use serde::Deserialize;
+use std::collections::HashMap;
+use std::time::Duration;
+use std::vec;
+
+pub const CHANNEL_METADATA: crate::providers::ProviderChannelMetadata =
+    crate::providers::ProviderChannelMetadata {
+        group_id: "onlyfans",
+        tags: &["onlyfans", "leaks", "model", "creator"],
+    };
+
+const CHANNEL_ID: &str = "fapello";
+const BASE_URL: &str = "https://fapello.com";
+// Listing endpoints (Cloudflare-gated; reached via the shared requester's
+// Jina mirror fallback).
+const VIDEOS_INDEX: &str = "https://fapello.com/videos/";
+const VIDEOS_AJAX_PAGE_FMT: &str = "https://fapello.com/ajax/videos/page-{}/";
+// Per-item detail endpoint. The Jina-mirrored body is the source of the
+// `var BOOT = { items: [...] }` JSON block that contains the resolved CDN
+// media URL, duration, uploader, etc.
+const DETAIL_URL_FMT: &str = "https://fapello.com/video/new/{}/";
+// Bounded concurrency for the per-card BOOT enrichment fetch. Jina rate-limits
+// per IP, so a small pool keeps us under that limit while still amortizing
+// the round-trip.
+const ENRICH_CONCURRENCY: usize = 4;
+
+error_chain! {
+    foreign_links {
+        Io(std::io::Error);
+        HttpRequest(wreq::Error);
+        Json(serde_json::Error);
+    }
+    errors {
+        Parse(msg: String) {
+            description("parse error")
+            display("parse error: {}", msg)
+        }
+    }
+}
+
+/// Lightweight card extracted from the listing HTML. The listing never
+/// has the resolved media URL or duration; those come from the BOOT JSON
+/// on the per-card detail page.
+#[derive(Debug, Clone)]
+struct ListingCard {
+    id: String,
+    detail_url: String,
+    thumb: String,
+    model_slug: String,
+    model_name: String,
+}
+
+/// Fields we care about out of the per-detail-page `var BOOT = { ... }`
+/// JSON block. Everything has a default so a partial/missing field
+/// doesn't blow up the whole enrichment.
+#[derive(Debug, Deserialize, Default, Clone)]
+struct BootItem {
+    #[serde(default)]
+    id: u64,
+    #[serde(default)]
+    model_id: u64,
+    #[serde(default)]
+    model_url: String,
+    #[serde(default)]
+    filename_id: u64,
+    #[serde(default)]
+    name: String,
+    #[serde(default)]
+    avatar_url: String,
+    #[serde(default)]
+    count_likes: u64,
+    #[serde(default, rename = "type")]
+    item_type: String,
+    #[serde(default)]
+    video_url: String,
+    #[serde(default)]
+    poster_url: String,
+    #[serde(default)]
+    image_url: String,
+    #[serde(default)]
+    duration: String, // "HH:MM:SS"
+    #[serde(default)]
+    width: u32,
+    #[serde(default)]
+    height: u32,
+    #[serde(default)]
+    aspect: f32,
+    #[serde(default)]
+    orientation: String,
+    #[serde(default)]
+    slug: String,
+    #[serde(default)]
+    ext: String,
+    #[serde(default)]
+    tags: Vec<BootTag>,
+    #[serde(default)]
+    random_tags: Vec<BootTag>,
+    #[serde(default)]
+    title_str: String,
+}
+
+#[derive(Debug, Deserialize, Default, Clone)]
+struct BootTag {
+    #[serde(default)]
+    url: String,
+    #[serde(default)]
+    name: String,
+}
+
+#[derive(Debug, Deserialize)]
+struct BootPayload {
+    #[serde(default)]
+    items: Vec<BootItem>,
+    #[serde(default)]
+    next_page: u32,
+    #[serde(default)]
+    page_size: u32,
+}
+
+#[derive(Debug, Clone)]
+pub struct FapelloProvider {
+    url: String,
+}
+
+impl FapelloProvider {
+    pub fn new() -> Self {
+        Self {
+            url: BASE_URL.to_string(),
+        }
+    }
+
+    fn build_channel(&self, _clientversion: ClientVersion) -> Channel {
+        Channel {
+            id: CHANNEL_ID.to_string(),
+            name: "Fapello".to_string(),
+            description: "Leaked OnlyFans-style videos from fapello.com".to_string(),
+            premium: false,
+            favicon: "https://www.google.com/s2/favicons?sz=64&domain=fapello.com".to_string(),
+            status: "active".to_string(),
+            categories: vec![],
+            // Fapello's video section has only one chronological feed; the
+            // site has no working "popular"/"trending" videos index (the
+            // corresponding routes are JS-rendered with no usable card HTML
+            // in the response body). Keep `sort` to one option to avoid
+            // promising filters we cannot honor.
+            options: vec![ChannelOption {
+                id: "sort".to_string(),
+                title: "Sort".to_string(),
+                description: "Sort the videos".to_string(),
+                systemImage: "list.number".to_string(),
+                colorName: "blue".to_string(),
+                options: vec![FilterOption {
+                    id: "new".to_string(),
+                    title: "New".to_string(),
+                }],
+                multiSelect: false,
+            }],
+            nsfw: true,
+            cacheDuration: Some(1800),
+            ytdlpCommand: Some("yt-dlp --impersonate chrome-120".to_string()),
+        }
+    }
+
+    /// Resolves a `page` value (string or int) to a 1-based u8. Fapello's
+    /// video listing caps at a small max (the page count is implicit in the
+    /// `next_page` field on each detail page's BOOT JSON, but the AJAX body
+    /// just stops returning new cards once the stream is exhausted — we
+    /// don't try to bound it client-side).
+    fn parse_page(page: &str) -> u8 {
+        page.parse::<u8>().unwrap_or(1).max(1)
+    }
+
+    /// URL for the listing of the given page.
+    fn listing_url(page: u8) -> String {
+        if page <= 1 {
+            VIDEOS_INDEX.to_string()
+        } else {
+            VIDEOS_AJAX_PAGE_FMT.replace("{}", &page.to_string())
+        }
+    }
+
+    /// Parse the listing HTML into cards. Cards look like:
+    ///
+    /// ```html
+    /// <a href="https://fapello.com/video/new/32034817/">
+    ///   <img src="https://fapello.com/content/s/e/seeyanever/1000/seeyanever_0073_300px.jpg" ...>
+    /// </a>
+    /// <a href="https://fapello.com/seeyanever/" class="...">
+    ///   ...
+    ///   <div> seeyanever </div>
+    /// </a>
+    /// ```
+    ///
+    /// We anchor on `video/new/{id}/` hrefs to pull each card block and then
+    /// pick the *next* `<img ...>` + the *next* `/{model_slug}/` link text.
+    fn parse_listing_cards(html: &str) -> Vec<ListingCard> {
+        let mut cards = Vec::new();
+        // Match every video card anchor. The href carries the numeric id.
+        let anchor_re = match Regex::new(
+            r#"href="https://fapello\.com/video/new/(\d+)/""#,
+        ) {
+            Ok(re) => re,
+            Err(e) => {
+                report_provider_error_background(
+                    CHANNEL_ID,
+                    "parse_listing.regex",
+                    &e.to_string(),
+                );
+                return cards;
+            }
+        };
+        // We need a "rest of the card" after the anchor href to pull the
+        // thumbnail and model. The simplest reliable approach: walk by
+        // index, slice each card by a fixed number of bytes after the
+        // matched anchor (the cards are large HTML blocks — 2-3 KB — so a
+        // generous window is safe).
+        for cap in anchor_re.captures_iter(html) {
+            let id = match cap.get(1) {
+                Some(m) => m.as_str().to_string(),
+                None => continue,
+            };
+            // Skip duplicates (each card link is repeated twice in the
+            // markup — once for the thumbnail `<a>`, once for the play-icon
+            // overlay `<a>`).
+            if cards.iter().any(|c: &ListingCard| c.id == id) {
+                continue;
+            }
+            let match_end = cap.get(0).map(|m| m.end()).unwrap_or(0);
+            let tail = &html[match_end..html.len().min(match_end + 4096)];
+
+            // Thumbnail: the FIRST `<img ... src="...jpg...">` after the anchor.
+            // We deliberately skip `assets/images/icon-play.svg` (the overlay
+            // play button) by requiring `fapello.com/content/` in the URL.
+            let thumb_re = match Regex::new(
+                r#"<img\s+src="(https://fapello\.com/content/[^"]+\.jpg)""#,
+            ) {
+                Ok(re) => re,
+                Err(_) => continue,
+            };
+            let thumb = thumb_re
+                .captures(tail)
+                .and_then(|c| c.get(1))
+                .map(|m| m.as_str().to_string())
+                .unwrap_or_default();
+            if thumb.is_empty() {
+                continue;
+            }
+
+            // Model slug: the FIRST `<a href="https://fapello.com/{slug}/"` after
+            // the anchor, where `{slug}` is neither `video`, `content`, nor any
+            // of the known top-level routes.
+            let slug_re = match Regex::new(
+                r#"href="https://fapello\.com/([a-zA-Z0-9_-]+)/""#,
+            ) {
+                Ok(re) => re,
+                Err(_) => continue,
+            };
+            let mut model_slug = String::new();
+            let mut model_name = String::new();
+            for slug_cap in slug_re.captures_iter(tail) {
+                let slug = match slug_cap.get(1) {
+                    Some(m) => m.as_str().to_string(),
+                    None => continue,
+                };
+                // Skip non-model anchors (link farm / corner badges / nav).
+                if slug == "video"
+                    || slug == "content"
+                    || slug == "assets"
+                    || slug == "data"
+                    || slug == "videos"
+                {
+                    continue;
+                }
+                model_slug = slug.clone();
+                // The model name is the visible text of the next `<div> {slug} </div>`
+                // — typically appears as a bare slug too, but we prefer the
+                // human-readable variant from the BOOT JSON if available.
+                let name_re = match Regex::new(
+                    r"<div>\s*([^<>{]*?)\s*</div>",
+                ) {
+                    Ok(re) => re,
+                    Err(_) => break,
+                };
+                if let Some(nc) = name_re.captures(&tail[slug_cap.get(0).unwrap().end()..]) {
+                    let candidate = nc.get(1).map(|m| m.as_str().trim().to_string()).unwrap_or_default();
+                    if !candidate.is_empty() {
+                        model_name = candidate;
+                    } else {
+                        model_name = slug.clone();
+                    }
+                } else {
+                    model_name = slug.clone();
+                }
+                break;
+            }
+            if model_slug.is_empty() {
+                // Fall back to the slug built from the thumb URL path
+                // (`.../{model}/{N}/{model}_{NNNN}_300px.jpg`).
+                if let Some(slug_from_thumb) = thumb
+                    .split('/')
+                    .nth(5)
+                {
+                    model_slug = slug_from_thumb.to_string();
+                    model_name = model_slug.clone();
+                }
+            }
+
+            cards.push(ListingCard {
+                id: id.clone(),
+                detail_url: format!("{}{}/", DETAIL_URL_FMT.replace("{}", &id), model_slug),
+                thumb: thumb
+                    .replace("_300px.jpg", ".jpg"),
+                model_slug,
+                model_name,
+            });
+        }
+        cards
+    }
+
+    /// Parse the `var BOOT = {...};` block out of the Jina-mirrored detail
+    /// page body. We anchor on the literal `var BOOT = ` prefix and the
+    /// terminating `;` — this is content-preserved inside `<script>...</script>`
+    /// by Jina, so a string slice is enough (no DOM needed).
+    fn parse_boot_block(html: &str) -> Option<BootPayload> {
+        let start_marker = "var BOOT = ";
+        let start = html.find(start_marker)? + start_marker.len();
+        // Walk to the matching `;` terminator. The BOOT JSON never contains
+        // a literal `;` inside any string value (verified across multiple
+        // detail pages), so a plain byte search is safe.
+        let end_rel = html[start..].find(';')?;
+        let body = &html[start..start + end_rel];
+        match serde_json::from_str::<BootPayload>(body) {
+            Ok(p) => Some(p),
+            Err(e) => {
+                report_provider_error_background(
+                    CHANNEL_ID,
+                    "parse_boot.json",
+                    &e.to_string(),
+                );
+                None
+            }
+        }
+    }
+
+    /// Find the BOOT item that matches `id` (the listing card's numeric id).
+    /// The active video is at `active_idx` and the rest are "up next" — we
+    /// match by id first, fall back to active_idx, then the first item.
+    fn find_boot_item<'a>(payload: &'a BootPayload, id: &str) -> Option<&'a BootItem> {
+        let target: u64 = id.parse().unwrap_or(0);
+        if let Some(item) = payload.items.iter().find(|i| i.id == target) {
+            return Some(item);
+        }
+        payload.items.first()
+    }
+
+    /// "HH:MM:SS" → seconds. Fapello's BOOT.duration is always zero-padded
+    /// (e.g. "00:00:13"), but be defensive about stray whitespace.
+    fn parse_hms_duration(s: &str) -> u32 {
+        let s = s.trim();
+        if s.is_empty() {
+            return 0;
+        }
+        let parts: Vec<&str> = s.split(':').collect();
+        match parts.len() {
+            3 => {
+                let h: u32 = parts[0].parse().unwrap_or(0);
+                let m: u32 = parts[1].parse().unwrap_or(0);
+                let sec: u32 = parts[2].parse().unwrap_or(0);
+                h * 3600 + m * 60 + sec
+            }
+            2 => {
+                let m: u32 = parts[0].parse().unwrap_or(0);
+                let sec: u32 = parts[1].parse().unwrap_or(0);
+                m * 60 + sec
+            }
+            1 => parts[0].parse().unwrap_or(0),
+            _ => 0,
+        }
+    }
+
+    /// Build a VideoItem from a listing card and its enriched BOOT entry.
+    /// Falls back to listing-only fields (id/url/thumb/model) when the BOOT
+    /// fetch returned nothing usable — keeps the page flowing instead of
+    /// dropping a card because of one bad detail fetch.
+    fn build_item(card: &ListingCard, boot: Option<&BootItem>) -> VideoItem {
+        // Media URL: prefer the BOOT-resolved CDN mp4. The CDN is the
+        // `cdn.fapello.com/content/.../{N}/{model}_{NNNN}.mp4` form (not
+        // `fapello.com`); if the BOOT returned an `fapello.com/...` URL we
+        // keep it as-is (still playable direct).
+        let media_url = boot
+            .map(|b| b.video_url.clone())
+            .filter(|s: &String| !s.is_empty());
+        let thumb = boot
+            .map(|b| b.poster_url.clone())
+            .filter(|s: &String| !s.is_empty())
+            .unwrap_or_else(|| card.thumb.clone());
+        let title = boot
+            .map(|b| {
+                if !b.title_str.is_empty() {
+                    b.title_str.clone()
+                } else if !b.name.is_empty() {
+                    b.name.clone()
+                } else {
+                    card.model_name.clone()
+                }
+            })
+            .unwrap_or_else(|| card.model_name.clone());
+        let uploader = boot
+            .map(|b| {
+                if !b.name.is_empty() {
+                    b.name.clone()
+                } else {
+                    card.model_name.clone()
+                }
+            })
+            .unwrap_or_else(|| card.model_name.clone());
+        let model_url_slug = boot
+            .map(|b| b.model_url.clone())
+            .filter(|s: &String| !s.is_empty())
+            .unwrap_or_else(|| card.model_slug.clone());
+        let duration = boot
+            .map(|b| Self::parse_hms_duration(&b.duration))
+            .unwrap_or(0);
+        let views = boot.map(|b| b.count_likes).unwrap_or(0);
+        let aspect = boot.and_then(|b| {
+            if b.aspect > 0.0 {
+                Some(b.aspect)
+            } else if b.width > 0 && b.height > 0 {
+                Some(b.width as f32 / b.height as f32)
+            } else {
+                None
+            }
+        });
+
+        // Tags: fapello's BOOT has `tags: []` (empty) on the active item and
+        // `random_tags: [{url, name}]` is unrelated "you might like" tags from
+        // a global pool, NOT video-specific. We deliberately drop
+        // `random_tags` to avoid mislabeling every clip with the same five
+        // sitewide tag suggestions. `tags` (when populated) is the real set.
+        let tags: Option<Vec<String>> = boot.and_then(|b| {
+            if b.tags.is_empty() {
+                None
+            } else {
+                Some(b.tags.iter().map(|t| t.name.clone()).collect())
+            }
+        });
+
+        let mut item = VideoItem::new(
+            card.id.clone(),
+            title,
+            // video.url is the canonical detail page, not the CDN — keeps
+            // `yt-dlp` (which has no fapello extractor) free to fail loudly
+            // and routes playback through `formats[]` instead.
+            card.detail_url.clone(),
+            CHANNEL_ID.to_string(),
+            thumb,
+            duration,
+        )
+        .uploader(uploader.clone())
+        .uploader_url(format!("{}/{}", BASE_URL, model_url_slug))
+        .views(views.min(u32::MAX as u64) as u32);
+        item.uploaderId = Some(format!("{}:{}", CHANNEL_ID, model_url_slug));
+        if let Some(t) = tags {
+            item = item.tags(t);
+        }
+        if let Some(a) = aspect {
+            item = item.aspect_ratio(a);
+        }
+        if let Some(url) = media_url {
+            // Direct mp4 from cdn.fapello.com — no auth, no cookies, but
+            // a Referer matches the on-site context (the site embeds its
+            // own CDN URLs on pages served from fapello.com).
+            let mut headers = HashMap::new();
+            headers.insert("Referer".to_string(), format!("{}/", BASE_URL));
+            let format = VideoFormat::new(
+                url,
+                "Original".to_string(),
+                "mp4".to_string(),
+            )
+            .format_id("mp4-source".to_string())
+            .ext("mp4".to_string())
+            .http_headers(headers);
+            item = item.formats(vec![format]);
+        }
+        item
+    }
+
+    /// Fetch + parse the BOOT JSON for one card. Returns `None` on any
+    /// failure (network, Jina-throttling, parse error) — caller falls back
+    /// to listing-only metadata.
+    async fn enrich_card(
+        requester: &mut crate::util::requester::Requester,
+        card: &ListingCard,
+    ) -> Option<BootItem> {
+        let url = format!("{}{}/", DETAIL_URL_FMT.replace("{}", &card.id), card.model_slug);
+        let mut headers = Vec::new();
+        headers.push(("Referer".to_string(), format!("{}/", BASE_URL)));
+        let body = match tokio::time::timeout(
+            Duration::from_secs(20),
+            requester.get_with_headers(&url, headers, None),
+        )
+        .await
+        {
+            Ok(Ok(s)) => s,
+            Ok(Err(e)) => {
+                report_provider_error_background(
+                    CHANNEL_ID,
+                    "enrich_card.fetch",
+                    &format!("id={} err={}", card.id, e),
+                );
+                return None;
+            }
+            Err(_) => {
+                report_provider_error_background(
+                    CHANNEL_ID,
+                    "enrich_card.timeout",
+                    &format!("id={}", card.id),
+                );
+                return None;
+            }
+        };
+        let payload = Self::parse_boot_block(&body)?;
+        Self::find_boot_item(&payload, &card.id).cloned()
+    }
+
+    async fn get(
+        &self,
+        cache: VideoCache,
+        page: u8,
+        _sort: &str,
+        options: ServerOptions,
+    ) -> Result<Vec<VideoItem>> {
+        let url = Self::listing_url(page);
+        let cached = match cache.get(&url) {
+            Some((time, items)) => {
+                if time.elapsed().unwrap_or_default().as_secs() < 60 * 5 {
+                    return Ok(items.clone());
+                }
+                items.clone()
+            }
+            None => vec![],
+        };
+        let mut requester =
+            requester_or_default(&options, module_path!(), "get.listing");
+        let body = match requester.get(&url, None).await {
+            Ok(s) => s,
+            Err(e) => {
+                report_provider_error_background(
+                    CHANNEL_ID,
+                    "get.fetch",
+                    &format!("url={url} err={e}"),
+                );
+                return Ok(cached);
+            }
+        };
+        if body.is_empty() {
+            return Ok(cached);
+        }
+        let cards = Self::parse_listing_cards(&body);
+        if cards.is_empty() {
+            report_provider_error_background(
+                CHANNEL_ID,
+                "get.parse_empty",
+                &format!("url={url}"),
+            );
+            return Ok(cached);
+        }
+        // Bounded-concurrency per-card BOOT enrichment. The Requester is
+        // `!Send` across awaits because of its `Client`/`Jar` internals, so
+        // we clone a fresh one per task — the cookie jar is shared at the
+        // process level so cookies earned by one Jina solve are visible to
+        // the others.
+        let cards_for_stream = cards.clone();
+        let enriched: Vec<Option<BootItem>> = stream::iter(cards_for_stream.into_iter())
+            .map(|card| {
+                let mut req = requester.clone();
+                async move { Self::enrich_card(&mut req, &card).await }
+            })
+            .buffer_unordered(ENRICH_CONCURRENCY)
+            .collect()
+            .await;
+        let items: Vec<VideoItem> = cards
+            .iter()
+            .zip(enriched.iter())
+            .map(|(card, boot)| Self::build_item(card, boot.as_ref()))
+            .collect();
+        if !items.is_empty() {
+            cache.remove(&url);
+            cache.insert(url.clone(), items.clone());
+        }
+        Ok(items)
+    }
+
+    async fn query(
+        &self,
+        _cache: VideoCache,
+        _page: u8,
+        query: &str,
+        _options: ServerOptions,
+    ) -> Result<Vec<VideoItem>> {
+        // Fapello's `/search_v2/?ajax=1&q=...&type=...` is a *photo* search —
+        // it returns model records, not video cards. There is no native
+        // "videos matching keyword" search. Surface a clear empty result
+        // rather than silently doing something unrelated. (The server's
+        // `literal_substring_filter` pass will post-filter `query` against
+        // the latest-feed results when the user types a quoted bare word.)
+        report_provider_error_background(
+            CHANNEL_ID,
+            "query.unsupported",
+            &format!("query={} — fapello has no video search; literal post-filter applied server-side", query),
+        );
+        Ok(vec![])
+    }
+}
+
+#[async_trait]
+impl Provider for FapelloProvider {
+    async fn get_videos(
+        &self,
+        cache: VideoCache,
+        _pool: DbPool,
+        sort: String,
+        query: Option<String>,
+        page: String,
+        _per_page: String,
+        options: ServerOptions,
+    ) -> Vec<VideoItem> {
+        let page_num = Self::parse_page(&page);
+        let result = match query {
+            Some(q) if !q.trim().is_empty() => {
+                self.query(cache, page_num, q.trim(), options).await
+            }
+            _ => self.get(cache, page_num, &sort, options).await,
+        };
+        match result {
+            Ok(items) => items,
+            Err(e) => {
+                eprintln!("[fapello] get_videos error: {e}");
+                vec![]
+            }
+        }
+    }
+
+    fn get_channel(&self, clientversion: ClientVersion) -> Option<Channel> {
+        Some(self.build_channel(clientversion))
+    }
+}

src/providers/hdporn92.rs

@@ -0,0 +1,620 @@
+use crate::DbPool;
+use crate::api::ClientVersion;
+use crate::providers::{
+    Provider, report_provider_error, report_provider_error_background, requester_or_default,
+};
+use crate::status::*;
+use crate::videos::{ServerOptions, VideoFormat, VideoItem};
+use async_trait::async_trait;
+use error_chain::error_chain;
+use futures::stream::{self, StreamExt};
+use regex::Regex;
+use scraper::{Html, Selector};
+
+pub const CHANNEL_METADATA: crate::providers::ProviderChannelMetadata =
+    crate::providers::ProviderChannelMetadata {
+        group_id: "mainstream-tube",
+        tags: &["tube", "mixed", "movies"],
+    };
+
+const BASE_URL: &str = "https://hdporn92.com";
+const CHANNEL_ID: &str = "hdporn92";
+const DEFAULT_PER_PAGE: usize = 24;
+const ENRICH_CONCURRENCY: usize = 4;
+
+const BROWSER_UA: &str =
+    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) \
+     Chrome/124.0.0.0 Safari/537.36";
+
+error_chain! {
+    foreign_links {
+        Io(std::io::Error);
+        HttpRequest(wreq::Error);
+    }
+    errors {
+        Parse(msg: String) {
+            description("parse error")
+            display("parse error: {}", msg)
+        }
+    }
+}
+
+#[derive(Debug, Clone)]
+pub struct Hdporn92Provider;
+
+#[derive(Debug, Clone)]
+enum Target {
+    Latest,
+    Search { query: String },
+}
+
+#[derive(Debug, Clone)]
+struct CardStub {
+    id: String,
+    title: String,
+    url: String,
+    thumb: String,
+    quality: Option<String>,
+    category_slug: Option<String>,
+}
+
+impl Hdporn92Provider {
+    pub fn new() -> Self {
+        Self
+    }
+
+    fn build_channel(&self, _clientversion: ClientVersion) -> Channel {
+        Channel {
+            id: CHANNEL_ID.to_string(),
+            name: "HDPorn92".to_string(),
+            description:
+                "HD Porn videos from network studios on hdporn92.com. Plays through a \
+                 monetized morencius.com (VidHide-family) embed; the HLS URL is signed \
+                 and Referer-locked, so the page itself does not expose direct media."
+                    .to_string(),
+            premium: false,
+            favicon: "https://www.google.com/s2/favicons?sz=64&domain=hdporn92.com".to_string(),
+            status: "active".to_string(),
+            categories: vec![],
+            options: vec![],
+            nsfw: true,
+            cacheDuration: Some(1800),
+            ytdlpCommand: Some("yt-dlp --impersonate chrome-120".to_string()),
+        }
+    }
+
+    fn selector(value: &str) -> Result<Selector> {
+        Selector::parse(value)
+            .map_err(|error| Error::from(format!("selector `{value}` failed: {error}")))
+    }
+
+    fn pick_target(query: Option<&str>) -> Target {
+        if let Some(q) = query.map(str::trim).filter(|q| !q.is_empty()) {
+            return Target::Search { query: q.to_string() };
+        }
+        Target::Latest
+    }
+
+    fn build_listing_url(target: &Target, page: u16) -> String {
+        let page = page.max(1);
+        match target {
+            Target::Latest => {
+                if page <= 1 {
+                    format!("{BASE_URL}/")
+                } else {
+                    format!("{BASE_URL}/page/{page}/")
+                }
+            }
+            Target::Search { query } => {
+                let encoded: String =
+                    url::form_urlencoded::byte_serialize(query.as_bytes()).collect();
+                if page <= 1 {
+                    format!("{BASE_URL}/?s={encoded}")
+                } else {
+                    format!("{BASE_URL}/page/{page}/?s={encoded}")
+                }
+            }
+        }
+    }
+
+    fn html_headers(referer: &str) -> Vec<(String, String)> {
+        vec![
+            ("User-Agent".to_string(), BROWSER_UA.to_string()),
+            ("Referer".to_string(), referer.to_string()),
+        ]
+    }
+
+    async fn fetch_html(options: &ServerOptions, url: &str, referer: &str) -> Result<String> {
+        let mut requester = requester_or_default(options, CHANNEL_ID, "fetch_html");
+        requester
+            .get_with_headers(url, Self::html_headers(referer), None)
+            .await
+            .map_err(|error| Error::from(format!("request failed for {url}: {error}")))
+    }
+
+    fn parse_listing_page(html: &str) -> Result<Vec<CardStub>> {
+        let document = Html::parse_document(html);
+        let card_sel = Self::selector("article.loop-video[data-video-id]")?;
+        let link_sel = Self::selector("a[href]")?;
+        let thumb_attr_sel = Self::selector("[data-main-thumb]")?;
+        let title_sel = Self::selector("header.entry-header span")?;
+        let views_sel = Self::selector("span.views")?;
+        let class_regex = Regex::new(r"\bpost-(\d+)\b")
+            .map_err(|error| Error::from(format!("regex failed: {error}")))?;
+        let category_regex = Regex::new(r"\bcategory-([a-z0-9-]+)")
+            .map_err(|error| Error::from(format!("regex failed: {error}")))?;
+
+        let mut stubs = Vec::new();
+        let mut seen_ids = std::collections::HashSet::new();
+
+        for card in document.select(&card_sel) {
+            let class_attr = card.value().attr("class").unwrap_or_default();
+
+            let id = class_attr
+                .split_whitespace()
+                .find_map(|c| c.strip_prefix("post-"))
+                .map(str::to_string)
+                .or_else(|| {
+                    card.value()
+                        .attr("data-post-id")
+                        .map(str::to_string)
+                })
+                .unwrap_or_default();
+            if id.is_empty() || !seen_ids.insert(id.clone()) {
+                continue;
+            }
+
+            let url = card
+                .select(&link_sel)
+                .next()
+                .and_then(|a| a.value().attr("href"))
+                .map(str::to_string)
+                .unwrap_or_default();
+            if url.is_empty() {
+                continue;
+            }
+
+            let title = card
+                .select(&title_sel)
+                .next()
+                .map(|s| s.text().collect::<Vec<_>>().join(""))
+                .map(|t| t.trim().to_string())
+                .filter(|t| !t.is_empty())
+                .unwrap_or_else(|| id.clone());
+
+            let thumb = card
+                .select(&thumb_attr_sel)
+                .next()
+                .and_then(|el| el.value().attr("data-main-thumb"))
+                .map(str::to_string)
+                .filter(|t| !t.is_empty())
+                .or_else(|| {
+                    card.select(&Self::selector("img[src]").ok()?)
+                        .next()
+                        .and_then(|img| img.value().attr("src"))
+                        .map(str::to_string)
+                })
+                .unwrap_or_default();
+
+            let quality = card
+                .select(&views_sel)
+                .next()
+                .map(|s| s.text().collect::<Vec<_>>().join(""))
+                .map(|t| t.trim().to_string())
+                .filter(|t| !t.is_empty() && t != "0");
+
+            let category_slug = class_regex
+                .captures(class_attr)
+                .and_then(|caps| caps.get(1))
+                .map(|m| m.as_str().to_string())
+                .or_else(|| {
+                    category_regex
+                        .captures(class_attr)
+                        .and_then(|caps| caps.get(1))
+                        .map(|m| m.as_str().to_string())
+                });
+
+            stubs.push(CardStub {
+                id,
+                title,
+                url,
+                thumb,
+                quality,
+                category_slug,
+            });
+        }
+
+        Ok(stubs)
+    }
+
+    /// Extracts the morencius.com embed URL from a hdporn92 detail page.
+    ///
+    /// The detail page renders the embed inside an `<IFRAME SRC="...">` (note
+    /// the uppercase tag and attribute names — the retrotube WordPress theme
+    /// emits mixed-case markup).
+    fn extract_embed_url(html: &str) -> Option<String> {
+        // Find the morencius embed host fragment and walk back to the scheme
+        // (the iframe `SRC` is a quoted URL like `SRC="https://morencius..."`).
+        let needle = "morencius.com/embed/";
+        let needle_idx = html.find(needle)?;
+        // Walk back to the start of the URL (find `https://` or `http://`).
+        let head = &html[..=needle_idx];
+        let scheme_idx = head
+            .rfind("https://")
+            .or_else(|| head.rfind("http://"))?;
+        let after_scheme = &html[scheme_idx..];
+        // End of the URL: the next quote, space, or `>`.
+        let end = after_scheme
+            .find(|c: char| c == '"' || c == '\'' || c == ' ' || c == '>')
+            .unwrap_or(after_scheme.len());
+        Some(after_scheme[..end].to_string())
+    }
+
+    /// Extracts the embed ID from a `https://morencius.com/embed/{id}` URL.
+    fn extract_embed_id(embed_url: &str) -> Option<String> {
+        let prefix = "morencius.com/embed/";
+        let idx = embed_url.find(prefix)?;
+        let after = &embed_url[idx + prefix.len()..];
+        let end = after
+            .find(|c: char| c == '?' || c == '/' || c == '"' || c == '\'')
+            .unwrap_or(after.len());
+        if end == 0 {
+            None
+        } else {
+            Some(after[..end].to_string())
+        }
+    }
+
+    /// Decodes the P.A.C.K.E.R. `eval(function(p,a,c,k,e,d){...}(...))` block in
+    /// the morencius embed page and returns the deobfuscated JavaScript source.
+    fn decode_packered(source: &str) -> Option<String> {
+        // Match: eval(function(p,a,c,k,e,d){...}('P',A,C,'K'.split('|')
+        // We capture P, A, C, and the dictionary K.
+        let re = Regex::new(
+            r#"eval\(function\(p,a,c,k,e,d\)\{.*?\}\('([\s\S]*?)',(\d+),(\d+),'([\s\S]*?)'\.split\('\|'\)\)\)"#,
+        )
+        .ok()?;
+        let caps = re.captures(source)?;
+        let p = caps.get(1)?.as_str();
+        let a: u32 = caps.get(2)?.as_str().parse().ok()?;
+        let c: usize = caps.get(3)?.as_str().parse().ok()?;
+        let k: Vec<&str> = caps.get(4)?.as_str().split('|').collect();
+
+        // P.A.C.K.E.R. deobfuscation: iterate c-1..=0; replace \b<token>\b with
+        // dictionary[i] (the original JS uses a `\w+` lookalike via the `e`
+        // parameter, but in practice every token is a base-A digit string).
+        let mut result = p.to_string();
+        for i in (0..c).rev() {
+            let value = k.get(i).copied().unwrap_or("");
+            if value.is_empty() {
+                continue;
+            }
+            let token = to_base(i, a);
+            // Build a word-boundary regex for the token. Base-A tokens are
+            // always made of [0-9a-z], so \b is a real word boundary.
+            let token_re = Regex::new(&format!(r"\b{}\b", regex::escape(&token))).ok()?;
+            result = token_re.replace_all(&result, value).into_owned();
+        }
+        Some(result)
+    }
+
+    /// Pulls the morencius `/stream/.../master.m3u8` URL out of the decoded JS.
+    fn extract_hls4_url(decoded_js: &str) -> Option<String> {
+        // The decoded JS has: "hls4":"/stream/<token>/<secret>/<expiry>/<file_id>/master.m3u8"
+        // The value is a path (no scheme), so we prefix it.
+        let re = Regex::new(r#""hls4"\s*:\s*"([^"]+)""#).ok()?;
+        let caps = re.captures(decoded_js)?;
+        let path = caps.get(1)?.as_str();
+        if path.starts_with("http") {
+            Some(path.to_string())
+        } else if path.starts_with('/') {
+            Some(format!("https://morencius.com{path}"))
+        } else {
+            Some(format!("https://morencius.com/{path}"))
+        }
+    }
+
+    /// Pulls the integer duration (in seconds, may be decimal) from the decoded
+    /// jwplayer setup block. The JS has `duration:"<sec>.<frac>"`.
+    fn extract_duration(decoded_js: &str) -> Option<u32> {
+        let re = Regex::new(r#"duration\s*:\s*"?(\d+(?:\.\d+)?)"?"#).ok()?;
+        let caps = re.captures(decoded_js)?;
+        let raw = caps.get(1)?.as_str().parse::<f64>().ok()?;
+        Some(raw.round() as u32)
+    }
+
+    async fn enrich_card(stub: CardStub, options: &ServerOptions) -> Result<VideoItem> {
+        // Step 1: fetch the detail page to find the morencius embed.
+        let detail_html = Self::fetch_html(options, &stub.url, BASE_URL).await?;
+        let embed_url = Self::extract_embed_url(&detail_html)
+            .ok_or_else(|| Error::from(format!("no morencius embed in {}", stub.url)))?;
+        // Sanity check the embed id — the URL is what we use as Referer, but
+        // we want to fail early if the URL is malformed rather than silently
+        // building a broken `formats[0]`.
+        let _embed_id = Self::extract_embed_id(&embed_url)
+            .ok_or_else(|| Error::from(format!("could not parse embed id from {embed_url}")))?;
+
+        // Step 2: fetch the morencius embed page (referer = the hdporn92 detail).
+        let player_html = Self::fetch_html(options, &embed_url, &stub.url).await?;
+        let decoded = Self::decode_packered(&player_html)
+            .ok_or_else(|| Error::from(format!("no eval block in {embed_url}")))?;
+        let stream_path = Self::extract_hls4_url(&decoded)
+            .ok_or_else(|| Error::from(format!("no hls4 url in decoded JS of {embed_url}")))?;
+        let duration = Self::extract_duration(&decoded).unwrap_or(0);
+
+        // Build the format. morencius serves the m3u8 only when Referer is the
+        // embed URL — without it, the response is 403.
+        let mut format =
+            VideoFormat::m3u8(stream_path, "auto".to_string(), "m3u8".to_string());
+        format.add_http_header("Referer".to_string(), embed_url.clone());
+        format.add_http_header("User-Agent".to_string(), BROWSER_UA.to_string());
+
+        let mut item = VideoItem::new(
+            stub.id,
+            stub.title,
+            stub.url,
+            CHANNEL_ID.to_string(),
+            stub.thumb,
+            duration,
+        );
+        item.formats = Some(vec![format]);
+        if let Some(slug) = stub.category_slug {
+            item.tags = Some(vec![slug]);
+        }
+
+        Ok(item)
+    }
+
+    async fn fetch_page(
+        target: Target,
+        page: u16,
+        per_page: usize,
+        options: &ServerOptions,
+    ) -> Result<Vec<VideoItem>> {
+        let url = Self::build_listing_url(&target, page);
+        let html = Self::fetch_html(options, &url, BASE_URL).await?;
+        let stubs = Self::parse_listing_page(&html)?;
+        if stubs.is_empty() {
+            return Ok(vec![]);
+        }
+
+        let limited: Vec<_> = stubs.into_iter().take(per_page).collect();
+        let options = options.clone();
+
+        let items = stream::iter(limited.into_iter().map(|stub| {
+            let options = options.clone();
+            async move {
+                match Self::enrich_card(stub, &options).await {
+                    Ok(item) => Some(item),
+                    Err(error) => {
+                        report_provider_error_background(
+                            CHANNEL_ID,
+                            "fetch_page.enrich_card",
+                            &error.to_string(),
+                        );
+                        None
+                    }
+                }
+            }
+        }))
+        .buffer_unordered(ENRICH_CONCURRENCY)
+        .filter_map(async move |v| v)
+        .collect::<Vec<_>>()
+        .await;
+
+        Ok(items)
+    }
+}
+
+/// Convert a non-negative integer to a string in the given base (radix).
+/// The result uses lowercase 0-9a-z. Used for the P.A.C.K.E.R. base-A tokens.
+fn to_base(mut value: usize, base: u32) -> String {
+    debug_assert!(base >= 2 && base <= 36);
+    if value == 0 {
+        return "0".to_string();
+    }
+    let mut digits = Vec::new();
+    let base = base as usize;
+    while value > 0 {
+        let d = (value % base) as u8;
+        digits.push(if d < 10 {
+            (b'0' + d) as char
+        } else {
+            (b'a' + (d - 10)) as char
+        });
+        value /= base;
+    }
+    digits.into_iter().rev().collect()
+}
+
+#[async_trait]
+impl Provider for Hdporn92Provider {
+    async fn get_videos(
+        &self,
+        _cache: crate::util::cache::VideoCache,
+        _pool: DbPool,
+        _sort: String,
+        query: Option<String>,
+        page: String,
+        per_page: String,
+        options: ServerOptions,
+    ) -> Vec<VideoItem> {
+        let page = page.parse::<u16>().unwrap_or(1).max(1);
+        let per_page = per_page
+            .parse::<usize>()
+            .unwrap_or(DEFAULT_PER_PAGE)
+            .clamp(1, 48);
+
+        let normalized_query = query
+            .as_deref()
+            .map(str::trim)
+            .filter(|q| !q.is_empty())
+            .map(ToOwned::to_owned);
+
+        let target = Self::pick_target(normalized_query.as_deref());
+
+        match Self::fetch_page(target, page, per_page, &options).await {
+            Ok(items) => items,
+            Err(error) => {
+                report_provider_error(CHANNEL_ID, "get_videos", &error.to_string()).await;
+                vec![]
+            }
+        }
+    }
+
+    fn get_channel(&self, clientversion: ClientVersion) -> Option<Channel> {
+        Some(self.build_channel(clientversion))
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn builds_listing_urls() {
+        assert_eq!(
+            Hdporn92Provider::build_listing_url(&Target::Latest, 1),
+            "https://hdporn92.com/"
+        );
+        assert_eq!(
+            Hdporn92Provider::build_listing_url(&Target::Latest, 2),
+            "https://hdporn92.com/page/2/"
+        );
+        assert_eq!(
+            Hdporn92Provider::build_listing_url(
+                &Target::Search {
+                    query: "isla wild".to_string()
+                },
+                1
+            ),
+            "https://hdporn92.com/?s=isla+wild"
+        );
+        assert_eq!(
+            Hdporn92Provider::build_listing_url(
+                &Target::Search {
+                    query: "milf".to_string()
+                },
+                2
+            ),
+            "https://hdporn92.com/page/2/?s=milf"
+        );
+    }
+
+    #[test]
+    fn picks_search_for_query() {
+        match Hdporn92Provider::pick_target(Some("test")) {
+            Target::Search { query } => assert_eq!(query, "test"),
+            other => panic!("expected Search, got {:?}", other),
+        }
+    }
+
+    #[test]
+    fn picks_latest_for_no_query() {
+        assert!(matches!(
+            Hdporn92Provider::pick_target(None),
+            Target::Latest
+        ));
+        assert!(matches!(
+            Hdporn92Provider::pick_target(Some("")),
+            Target::Latest
+        ));
+        assert!(matches!(
+            Hdporn92Provider::pick_target(Some("   ")),
+            Target::Latest
+        ));
+    }
+
+    #[test]
+    fn extracts_embed_id_from_url() {
+        assert_eq!(
+            Hdporn92Provider::extract_embed_id("https://morencius.com/embed/vc6vwjo6ngbn"),
+            Some("vc6vwjo6ngbn".to_string())
+        );
+        assert_eq!(
+            Hdporn92Provider::extract_embed_id("https://morencius.com/embed/abc123?foo=bar"),
+            Some("abc123".to_string())
+        );
+        assert_eq!(
+            Hdporn92Provider::extract_embed_id("https://other.com/embed/abc"),
+            None
+        );
+    }
+
+    #[test]
+    fn extracts_embed_url_from_detail_html() {
+        let html = r#"<IFRAME SRC="https://morencius.com/embed/vc6vwjo6ngbn" FRAMEBORDER=0>"#;
+        assert_eq!(
+            Hdporn92Provider::extract_embed_url(html).as_deref(),
+            Some("https://morencius.com/embed/vc6vwjo6ngbn")
+        );
+    }
+
+    #[test]
+    fn decodes_hls4_url_from_packered_js() {
+        // A minimal P.A.C.K.E.R. block that decodes to JSON containing an
+        // hls4 key. We construct it so the base-36 token `0` decodes to
+        // "hls4" and `1` decodes to the m3u8 path. The `2` is the empty
+        // string (no key beyond index 1) and the loop skips it.
+        //
+        // a = 36, c = 2, k = ["hls4", "/stream/abc/secret/123/456/master.m3u8"]
+        // p = '{"0":"1"}'  (after substitution: {"hls4":"/stream/.../master.m3u8"})
+        //
+        // Reverse iteration:
+        //   i=1: token="1", k[1]="..." → '{"0":"/stream/.../master.m3u8"}'
+        //   i=0: token="0", k[0]="hls4" → '{"hls4":"/stream/.../master.m3u8"}'
+        let source = r#"eval(function(p,a,c,k,e,d){e=function(c){return c};if(!''.replace(/^/,String)){while(c--){d[c.toString(a)]=k[c]||c.toString(a)};k=[function(e){return d[e]}];e=function(){return'\w+'};c=1};while(c--){if(k[c]){p=p.replace(new RegExp('\b'+e(c)+'\b','g'),k[c])}}return p}('{"0":"1"}',36,2,'hls4|/stream/abc/secret/123/456/master.m3u8'.split('|')));"#;
+        let decoded = Hdporn92Provider::decode_packered(source)
+            .expect("deobfuscator should return Some");
+        // Sanity: the decoded text contains the hls4 key.
+        assert!(decoded.contains("\"hls4\""), "decoded: {decoded}");
+        let url = Hdporn92Provider::extract_hls4_url(&decoded)
+            .expect("hls4 should be present after decode");
+        // The path was /stream/abc/secret/123/456/master.m3u8.
+        assert_eq!(url, "https://morencius.com/stream/abc/secret/123/456/master.m3u8");
+    }
+
+    #[test]
+    fn extracts_duration_from_decoded_js() {
+        let decoded = r#"jwplayer().setup({duration:"1709.40",preload:'auto'})"#;
+        assert_eq!(
+            Hdporn92Provider::extract_duration(decoded),
+            Some(1709)
+        );
+    }
+
+    #[test]
+    fn base_conversion() {
+        assert_eq!(to_base(0, 36), "0");
+        assert_eq!(to_base(35, 36), "z");
+        assert_eq!(to_base(36, 36), "10");
+        // 49 in base 36 is "1d" (the dictionary key for "hls3")
+        assert_eq!(to_base(49, 36), "1d");
+        // 51 in base 36 is "1f" (the dictionary key for "hls4")
+        assert_eq!(to_base(51, 36), "1f");
+    }
+
+    #[test]
+    fn decodes_real_morencius_html() {
+        // Exercise the deobfuscator against a real captured morencius embed
+        // page. The fixture is the iframe served to the iOS webview when
+        // opening `hdporn92.com/mypervyfamily-isla-wild-.../`.
+        let html = std::fs::read_to_string(
+            "src/providers/testdata/hdporn92_morencius.html",
+        )
+        .expect("test fixture should exist");
+        let decoded = Hdporn92Provider::decode_packered(&html)
+            .expect("real morencius HTML should be deobfuscable");
+        // The decoded JS contains the hls4 value.
+        let url = Hdporn92Provider::extract_hls4_url(&decoded)
+            .expect("hls4 should be present after decode");
+        assert!(
+            url.starts_with("https://morencius.com/stream/"),
+            "got {url}"
+        );
+        assert!(url.ends_with("/master.m3u8"), "got {url}");
+        // The duration is also present.
+        let duration = Hdporn92Provider::extract_duration(&decoded);
+        assert!(duration.unwrap_or(0) > 0, "duration should be > 0");
+    }
+}

src/providers/hornyleak.rs

@@ -0,0 +1,999 @@
+use crate::DbPool;
+use crate::api::ClientVersion;
+use crate::providers::{
+    Provider, report_provider_error, report_provider_error_background, requester_or_default,
+};
+use crate::status::*;
+use crate::util::cache::VideoCache;
+use crate::util::parse_abbreviated_number;
+use crate::util::requester::Requester;
+use crate::util::time::parse_time_to_seconds;
+use crate::videos::{ServerOptions, VideoItem};
+use async_trait::async_trait;
+use error_chain::error_chain;
+use futures::stream::{self, StreamExt};
+use htmlentity::entity::{ICodedDataTrait, decode};
+use scraper::{Html, Selector};
+use std::collections::{HashMap, HashSet};
+use std::sync::{Arc, RwLock};
+use std::thread;
+
+const CHANNEL_ID: &str = "hornyleak";
+const BASE_URL: &str = "https://www.hornyleak.tv";
+const HDPLAYER_BASE: &str = "https://hdplayer.gives";
+/// Bounded concurrency for the two-step embed -> hdplayer -> m3u8 enrichment.
+const ENRICH_CONCURRENCY: usize = 6;
+
+pub const CHANNEL_METADATA: crate::providers::ProviderChannelMetadata =
+    crate::providers::ProviderChannelMetadata {
+        group_id: "amateur-homemade",
+        tags: &["leaked", "amateur", "creampie"],
+    };
+
+error_chain! {
+    foreign_links {
+        Io(std::io::Error);
+        HttpRequest(wreq::Error);
+    }
+}
+
+#[derive(Debug, Clone)]
+pub struct HornyleakProvider {
+    url: String,
+    /// Display list exposed via the `categories` status option.
+    categories: Arc<RwLock<Vec<FilterOption>>>,
+    /// normalized category title -> site slug, for shortcut routing.
+    category_map: Arc<RwLock<HashMap<String, String>>>,
+}
+
+/// Where a `/api/videos` request should be routed on the site. Every variant
+/// resolves to a static KVS-style listing page or a search URL.
+#[derive(Debug, Clone)]
+enum Target {
+    /// Default `/latest-updates/{N}/` feed.
+    Latest,
+    /// `/most-popular/{N}/` archive.
+    MostPopular,
+    /// `/top-rated/{N}/` archive.
+    TopRated,
+    /// `/categories/{slug}/{N}/` archive.
+    Category { slug: String },
+    /// `/tags/{slug}/{N}/` archive.
+    Tag { slug: String },
+    /// `/models/{slug}/{N}/` archive (uploader shortcut).
+    Model { slug: String },
+    /// `/search/?q={query}` + async XHR pagination for page >=2.
+    Search { query: String },
+}
+
+#[derive(Debug, Clone)]
+struct HornyleakCard {
+    id: String,
+    title: String,
+    page_url: String,
+    thumb: String,
+    duration: u32,
+    views: Option<u32>,
+    rating: Option<u32>,
+    preview: Option<String>,
+    uploader: Option<String>,
+    uploader_url: Option<String>,
+    uploader_slug: Option<String>,
+}
+
+impl HornyleakProvider {
+    pub fn new() -> Self {
+        let provider = Self {
+            url: BASE_URL.to_string(),
+            categories: Arc::new(RwLock::new(vec![FilterOption {
+                id: "all".to_string(),
+                title: "All".to_string(),
+            }])),
+            category_map: Arc::new(RwLock::new(HashMap::new())),
+        };
+        provider.spawn_initial_load();
+        provider
+    }
+
+    fn build_channel(&self, _clientversion: ClientVersion) -> Channel {
+        let categories = self
+            .categories
+            .read()
+            .map(|guard| guard.clone())
+            .unwrap_or_default();
+
+        Channel {
+            id: CHANNEL_ID.to_string(),
+            name: "HornyLeak".to_string(),
+            description: "Leaked amateur and OnlyFans clips — latest feed, search, plus category, tag, and model archives."
+                .to_string(),
+            premium: false,
+            favicon: format!(
+                "https://www.google.com/s2/favicons?sz=64&domain={}",
+                self.url.trim_start_matches("https://")
+            ),
+            status: "active".to_string(),
+            categories: categories.iter().map(|value| value.title.clone()).collect(),
+            options: vec![
+                ChannelOption {
+                    id: "sort".to_string(),
+                    title: "Sort".to_string(),
+                    description: "Feed ordering.".to_string(),
+                    systemImage: "list.number".to_string(),
+                    colorName: "blue".to_string(),
+                    options: vec![
+                        FilterOption {
+                            id: "new".to_string(),
+                            title: "Latest".to_string(),
+                        },
+                        FilterOption {
+                            id: "popular".to_string(),
+                            title: "Most Popular".to_string(),
+                        },
+                        FilterOption {
+                            id: "rated".to_string(),
+                            title: "Top Rated".to_string(),
+                        },
+                        FilterOption {
+                            id: "viewed".to_string(),
+                            title: "Most Viewed".to_string(),
+                        },
+                        FilterOption {
+                            id: "longest".to_string(),
+                            title: "Longest".to_string(),
+                        },
+                    ],
+                    multiSelect: false,
+                },
+                ChannelOption {
+                    id: "categories".to_string(),
+                    title: "Categories".to_string(),
+                    description: "Jump directly to a category archive.".to_string(),
+                    systemImage: "square.grid.2x2".to_string(),
+                    colorName: "orange".to_string(),
+                    options: categories,
+                    multiSelect: false,
+                },
+            ],
+            nsfw: true,
+            cacheDuration: Some(1800),
+            ytdlpCommand: Some("yt-dlp --impersonate chrome-120".to_string()),
+        }
+    }
+
+    fn sort_by(sort: &str) -> &'static str {
+        match sort {
+            "popular" | "most-popular" => "most_popular",
+            "rated" | "rating" | "top-rated" => "rating",
+            "viewed" | "most-viewed" => "video_viewed",
+            "longest" | "duration" => "duration",
+            "commented" => "most_commented",
+            "favourited" | "favorited" => "most_favourited",
+            _ => "post_date",
+        }
+    }
+
+    /// lowercase, spaces/underscores -> single dash, drop anything else.
+    fn slugify(value: &str) -> String {
+        let mut slug = String::new();
+        let mut prev_dash = false;
+        for ch in value.trim().to_ascii_lowercase().chars() {
+            if ch.is_ascii_alphanumeric() {
+                slug.push(ch);
+                prev_dash = false;
+            } else if matches!(ch, ' ' | '-' | '_' | '+') {
+                if !prev_dash && !slug.is_empty() {
+                    slug.push('-');
+                    prev_dash = true;
+                }
+            }
+        }
+        slug.trim_matches('-').to_string()
+    }
+
+    /// Normalize a display title for lookup map keys.
+    fn normalize_title(value: &str) -> String {
+        value
+            .trim()
+            .trim_start_matches('#')
+            .replace(['_', '-'], " ")
+            .split_whitespace()
+            .collect::<Vec<_>>()
+            .join(" ")
+            .to_ascii_lowercase()
+    }
+
+    fn resolve_category_slug(&self, value: &str) -> Option<String> {
+        let normalized = Self::normalize_title(value);
+        if normalized.is_empty() || normalized == "all" {
+            return None;
+        }
+        self.category_map
+            .read()
+            .ok()
+            .and_then(|map| map.get(&normalized).cloned())
+    }
+
+    /// Decide where a request is routed: explicit option/prefix first, then a
+    /// bare query that matches a known category, otherwise keyword search.
+    fn target_from_request(
+        &self,
+        query: Option<&str>,
+        category: Option<&str>,
+    ) -> Target {
+        if let Some(category) = category {
+            if let Some(slug) = self.resolve_category_slug(category) {
+                return Target::Category { slug };
+            }
+        }
+
+        if let Some(raw) = query {
+            let trimmed = raw.trim();
+            if !trimmed.is_empty() {
+                for prefix in ["category:", "cat:"] {
+                    if let Some(rest) = trimmed.strip_prefix(prefix) {
+                        let slug = self
+                            .resolve_category_slug(rest)
+                            .unwrap_or_else(|| Self::slugify(rest));
+                        if !slug.is_empty() {
+                            return Target::Category { slug };
+                        }
+                    }
+                }
+                if let Some(rest) = trimmed.strip_prefix("tag:") {
+                    let slug = Self::slugify(rest);
+                    if !slug.is_empty() {
+                        return Target::Tag { slug };
+                    }
+                }
+                for prefix in ["model:", "uploader:", "pornstar:", "star:"] {
+                    if let Some(rest) = trimmed.strip_prefix(prefix) {
+                        let slug = Self::slugify(rest);
+                        if !slug.is_empty() {
+                            return Target::Model { slug };
+                        }
+                    }
+                }
+                if let Some(slug) = self.resolve_category_slug(trimmed) {
+                    return Target::Category { slug };
+                }
+                return Target::Search {
+                    query: trimmed.to_string(),
+                };
+            }
+        }
+
+        Target::Latest
+    }
+
+    /// Build the listing/search URL plus the Referer to send with it. Search
+    /// uses the KVS `from_videos` async XHR for any page > 1; the static HTML
+    /// `/search/?q=...` route only ships page 1.
+    fn build_listing_url(&self, target: &Target, sort: &str, page: u32) -> (String, String) {
+        let sort_by = Self::sort_by(sort);
+        match target {
+            Target::Latest => {
+                let referer = format!("{}/latest-updates/", self.url);
+                let url = if page <= 1 {
+                    referer.clone()
+                } else {
+                    format!("{}/latest-updates/{page}/", self.url)
+                };
+                (url, referer)
+            }
+            Target::MostPopular => {
+                let referer = format!("{}/most-popular/", self.url);
+                let url = if page <= 1 {
+                    referer.clone()
+                } else {
+                    format!("{}/most-popular/{page}/", self.url)
+                };
+                (url, referer)
+            }
+            Target::TopRated => {
+                let referer = format!("{}/top-rated/", self.url);
+                let url = if page <= 1 {
+                    referer.clone()
+                } else {
+                    format!("{}/top-rated/{page}/", self.url)
+                };
+                (url, referer)
+            }
+            Target::Category { slug } => {
+                let referer = format!("{}/categories/{slug}/", self.url);
+                let url = if page <= 1 {
+                    referer.clone()
+                } else {
+                    format!("{}/categories/{slug}/{page}/", self.url)
+                };
+                (url, referer)
+            }
+            Target::Tag { slug } => {
+                let referer = format!("{}/tags/{slug}/", self.url);
+                let url = if page <= 1 {
+                    referer.clone()
+                } else {
+                    format!("{}/tags/{slug}/{page}/", self.url)
+                };
+                (url, referer)
+            }
+            Target::Model { slug } => {
+                let referer = format!("{}/models/{slug}/", self.url);
+                let url = if page <= 1 {
+                    referer.clone()
+                } else {
+                    format!("{}/models/{slug}/{page}/", self.url)
+                };
+                (url, referer)
+            }
+            Target::Search { query } => {
+                let q = query.split_whitespace().collect::<Vec<_>>().join("+");
+                let referer = format!("{}/search/", self.url);
+                // Static HTML only ships page 1; page >= 2 must use the async
+                // XHR fragment. `from_videos` is 1-indexed (page 2 = from=2).
+                let url = if page <= 1 {
+                    format!("{referer}?q={q}")
+                } else {
+                    format!(
+                        "{referer}?q={q}&mode=async&function=get_block\
+                        &block_id=list_videos_videos_list_search_result\
+                        &category_ids=&sort_by={sort_by}&from_videos={page}&from_albums={page}"
+                    )
+                };
+                (url, referer)
+            }
+        }
+    }
+
+    fn html_headers(referer: &str) -> Vec<(String, String)> {
+        vec![("Referer".to_string(), referer.to_string())]
+    }
+
+    async fn get_videos_for_target(
+        &self,
+        cache: VideoCache,
+        target: &Target,
+        sort: &str,
+        page: u32,
+        options: ServerOptions,
+    ) -> Result<Vec<VideoItem>> {
+        let (listing_url, referer) = self.build_listing_url(target, sort, page);
+
+        let old_items = match cache.get(&listing_url) {
+            Some((time, items)) => {
+                if time.elapsed().unwrap_or_default().as_secs() < 60 * 5 {
+                    return Ok(items.clone());
+                }
+                items.clone()
+            }
+            None => vec![],
+        };
+
+        let mut requester = requester_or_default(&options, CHANNEL_ID, "get_videos_for_target");
+        let text = match requester
+            .get_with_headers(&listing_url, Self::html_headers(&referer), None)
+            .await
+        {
+            Ok(text) => text,
+            Err(e) => {
+                report_provider_error(
+                    CHANNEL_ID,
+                    "get_videos_for_target.request",
+                    &format!("url={listing_url}; error={e}"),
+                )
+                .await;
+                return Ok(old_items);
+            }
+        };
+
+        if text.trim().is_empty() {
+            return Ok(old_items);
+        }
+
+        let cards = self.parse_video_cards(&text);
+        if cards.is_empty() {
+            return Ok(old_items);
+        }
+
+        let video_items = self.enrich_cards(cards, &referer, requester).await;
+        if !video_items.is_empty() {
+            cache.remove(&listing_url);
+            cache.insert(listing_url.clone(), video_items.clone());
+            return Ok(video_items);
+        }
+
+        Ok(old_items)
+    }
+
+    fn decode_html_text(text: &str) -> String {
+        decode(text.as_bytes())
+            .to_string()
+            .unwrap_or_else(|_| text.to_string())
+            .split_whitespace()
+            .collect::<Vec<_>>()
+            .join(" ")
+            .trim()
+            .to_string()
+    }
+
+    fn normalize_url(&self, url: &str) -> String {
+        let trimmed = url.trim();
+        if trimmed.is_empty() {
+            return String::new();
+        }
+        if trimmed.starts_with("http://") || trimmed.starts_with("https://") {
+            return trimmed.to_string();
+        }
+        if trimmed.starts_with("//") {
+            return format!("https:{trimmed}");
+        }
+        if trimmed.starts_with('/') {
+            return format!("{}{}", self.url, trimmed);
+        }
+        format!("{}/{}", self.url, trimmed.trim_start_matches("./"))
+    }
+
+    fn id_from_video_url(url: &str) -> Option<String> {
+        let after = url.split("/video/").nth(1)?;
+        let id: String = after.chars().take_while(|c| c.is_ascii_digit()).collect();
+        (!id.is_empty()).then_some(id)
+    }
+
+    fn slug_from_model_url(url: &str) -> Option<String> {
+        url.split("/models/")
+            .nth(1)?
+            .split('/')
+            .next()
+            .map(str::to_string)
+            .filter(|slug| !slug.is_empty())
+    }
+
+    fn attr(element: &scraper::ElementRef, names: &[&str]) -> Option<String> {
+        for name in names {
+            if let Some(value) = element.value().attr(name) {
+                let trimmed = value.trim();
+                if !trimmed.is_empty() && !trimmed.starts_with("data:image/") {
+                    return Some(trimmed.to_string());
+                }
+            }
+        }
+        None
+    }
+
+    fn parse_video_cards(&self, html: &str) -> Vec<HornyleakCard> {
+        let document = Html::parse_document(html);
+        let Ok(item_sel) = Selector::parse("div.thumb.thumb_rel.item") else {
+            return vec![];
+        };
+        let link_sel = Selector::parse("a").ok();
+        let views_sel = Selector::parse("div.thumb-item > svg.icon-eye").ok();
+        let duration_sel = Selector::parse("div.time").ok();
+        let img_sel = Selector::parse("img").ok();
+        let preview_sel = Selector::parse("div.img-holder").ok();
+        let rating_sel = Selector::parse("div.rating").ok();
+        let model_sel = Selector::parse("a.avatar, a.link__red[href*=\"/models/\"]").ok();
+
+        let mut cards = Vec::new();
+        let mut seen = HashSet::new();
+
+        for item in document.select(&item_sel) {
+            let Some(link) = link_sel.as_ref().and_then(|sel| item.select(sel).next()) else {
+                continue;
+            };
+            let Some(href) = link.value().attr("href") else {
+                continue;
+            };
+            let page_url = self.normalize_url(href);
+            let Some(id) = Self::id_from_video_url(&page_url) else {
+                continue;
+            };
+            if !seen.insert(id.clone()) {
+                continue;
+            }
+
+            let title_raw = link
+                .value()
+                .attr("title")
+                .map(str::to_string)
+                .or_else(|| {
+                    img_sel
+                        .as_ref()
+                        .and_then(|sel| item.select(sel).next())
+                        .and_then(|img| img.value().attr("alt").map(str::to_string))
+                })
+                .or_else(|| {
+                    Selector::parse("a.title")
+                        .ok()
+                        .and_then(|sel| item.select(&sel).next())
+                        .map(|a| a.text().collect::<String>())
+                })
+                .unwrap_or_default();
+            let title = Self::decode_html_text(&title_raw);
+
+            let thumb = img_sel
+                .as_ref()
+                .and_then(|sel| item.select(sel).next())
+                .and_then(|img| Self::attr(&img, &["data-webp", "src", "data-src"]))
+                .map(|value| {
+                    self.normalize_url(value.split_whitespace().next().unwrap_or(&value))
+                })
+                .unwrap_or_default();
+
+            let duration = duration_sel
+                .as_ref()
+                .and_then(|sel| item.select(sel).next())
+                .map(|d| d.text().collect::<String>())
+                .and_then(|raw| parse_time_to_seconds(raw.trim()))
+                .unwrap_or(0) as u32;
+
+            let views = views_sel
+                .as_ref()
+                .and_then(|_sel| {
+                    item.select(&Selector::parse("div.thumb-item").ok()?)
+                        .find(|node| {
+                            node.value()
+                                .classes()
+                                .any(|c| c == "thumb-item" || c == "thumb-item thumb-item-date")
+                        })
+                        .map(|node| node.text().collect::<String>())
+                })
+                .and_then(|raw| parse_abbreviated_number(raw.trim()))
+                .filter(|count| *count > 0);
+
+            let rating = rating_sel
+                .as_ref()
+                .and_then(|sel| item.select(sel).next())
+                .map(|d| d.text().collect::<String>())
+                .and_then(|raw| {
+                    let trimmed = raw.trim().trim_end_matches('%').trim();
+                    trimmed.parse::<u32>().ok()
+                })
+                .filter(|pct| *pct <= 100);
+
+            let preview = preview_sel
+                .as_ref()
+                .and_then(|sel| item.select(sel).next())
+                .and_then(|p| Self::attr(&p, &["data-preview"]))
+                .map(|raw| self.normalize_url(&raw));
+
+            let model_anchor = model_sel
+                .as_ref()
+                .and_then(|sel| item.select(sel).next())
+                .filter(|m| {
+                    m.value()
+                        .attr("href")
+                        .map(|h| h.contains("/models/"))
+                        .unwrap_or(false)
+                });
+            let uploader = model_anchor
+                .map(|m| Self::decode_html_text(&m.text().collect::<String>()))
+                .filter(|name| !name.is_empty());
+            let uploader_href = model_anchor.and_then(|m| m.value().attr("href").map(str::to_string));
+            let uploader_url = uploader_href.as_ref().map(|href| self.normalize_url(href));
+            let uploader_slug = uploader_href
+                .as_deref()
+                .and_then(Self::slug_from_model_url);
+
+            cards.push(HornyleakCard {
+                id,
+                title,
+                page_url,
+                thumb,
+                duration,
+                views,
+                rating,
+                preview,
+                uploader,
+                uploader_url,
+                uploader_slug,
+            });
+        }
+
+        cards
+    }
+
+    /// Pull the `m3u8` HLS URL out of the embed + hdplayer chain:
+    /// detail → `/embed/{id}/` → `<iframe src="https://hdplayer.gives/embed/HASH">`
+    /// → fetch hdplayer page → `file:"https://hdplayer.gives/list/enc..."` (HLS).
+    /// The encoded `list/...` URL is stable per video and resolves directly to
+    /// an HLS master playlist — it does not need a Referer.
+    async fn resolve_hls_url(
+        &self,
+        video_id: &str,
+        page_url: &str,
+        mut requester: Requester,
+    ) -> Option<String> {
+        // Step 1: detail page → embed page (its URL is stable per video id).
+        let embed_url = format!("{}/embed/{video_id}/", self.url);
+        let embed_text = requester
+            .get_with_headers(&embed_url, vec![("Referer".to_string(), page_url.to_string())], None)
+            .await
+            .ok()?;
+
+        // Step 2: pull the hdplayer iframe URL out of the embed page.
+        let hdplayer_re = match regex::Regex::new(r#"hdplayer\.gives/embed/([A-Za-z0-9]+)"#) {
+            Ok(re) => re,
+            Err(_) => return None,
+        };
+        let hdplayer_hash = hdplayer_re
+            .captures(&embed_text)
+            .and_then(|caps| caps.get(1))
+            .map(|m| m.as_str().to_string())?;
+        let hdplayer_url = format!("{HDPLAYER_BASE}/embed/{hdplayer_hash}");
+
+        // Step 3: fetch the hdplayer page and pull the encoded list URL.
+        let hdplayer_text = requester
+            .get_with_headers(
+                &hdplayer_url,
+                vec![("Referer".to_string(), embed_url.clone())],
+                None,
+            )
+            .await
+            .ok()?;
+        let file_re = regex::Regex::new(r#"file:"(https://hdplayer\.gives/list/enc[^"]+)""#).ok()?;
+        file_re
+            .captures(&hdplayer_text)
+            .and_then(|caps| caps.get(1))
+            .map(|m| m.as_str().to_string())
+    }
+
+    fn extract_tags_from_page(text: &str) -> Vec<String> {
+        let document = Html::parse_document(text);
+        let Ok(selector) = Selector::parse(".tags-row a, .categories-row a") else {
+            return vec![];
+        };
+        let mut tags = Vec::new();
+        let mut seen = HashSet::new();
+        for link in document.select(&selector) {
+            let title = Self::decode_html_text(&link.text().collect::<String>());
+            if title.is_empty() {
+                continue;
+            }
+            let key = title.to_ascii_lowercase();
+            if seen.insert(key) {
+                tags.push(title);
+            }
+        }
+        tags
+    }
+
+    async fn enrich_card(
+        &self,
+        card: HornyleakCard,
+        listing_referer: String,
+        mut requester: Requester,
+    ) -> VideoItem {
+        // The page URL is yt-dlp-resolvable, so enrichment is only used for tags
+        // and uploader fields. The HLS embed chain (kept as a private helper
+        // for future use) is intentionally not invoked here.
+        let detail = requester
+            .get_with_headers(
+                &card.page_url,
+                Self::html_headers(&listing_referer),
+                None,
+            )
+            .await
+            .ok();
+
+        let tags = detail
+            .as_deref()
+            .map(Self::extract_tags_from_page)
+            .unwrap_or_default();
+
+        let mut item = VideoItem::new(
+            card.id,
+            card.title,
+            card.page_url.clone(),
+            CHANNEL_ID.to_string(),
+            card.thumb,
+            card.duration,
+        );
+        if let Some(views) = card.views {
+            item = item.views(views);
+        }
+        if let Some(pct) = card.rating {
+            item = item.rating(pct as f32);
+        }
+        if let Some(uploader) = card.uploader {
+            item = item.uploader(uploader);
+        }
+        if let Some(url) = card.uploader_url {
+            item = item.uploader_url(url);
+        }
+        if let Some(slug) = card.uploader_slug {
+            item.uploaderId = Some(format!("{CHANNEL_ID}:{slug}"));
+        }
+        if !tags.is_empty() {
+            item = item.tags(tags);
+        }
+        if let Some(preview) = card.preview {
+            item = item.preview(preview);
+        }
+        item
+    }
+
+    async fn enrich_cards(
+        &self,
+        cards: Vec<HornyleakCard>,
+        listing_referer: &str,
+        requester: Requester,
+    ) -> Vec<VideoItem> {
+        stream::iter(cards.into_iter().map(|card| {
+            let requester = requester.clone();
+            let referer = listing_referer.to_string();
+            async move { self.enrich_card(card, referer, requester).await }
+        }))
+        .buffered(ENRICH_CONCURRENCY)
+        .collect::<Vec<_>>()
+        .await
+    }
+
+    fn spawn_initial_load(&self) {
+        let url = self.url.clone();
+        let categories = Arc::clone(&self.categories);
+        let category_map = Arc::clone(&self.category_map);
+
+        thread::spawn(move || {
+            let runtime = match tokio::runtime::Builder::new_current_thread()
+                .enable_all()
+                .build()
+            {
+                Ok(runtime) => runtime,
+                Err(error) => {
+                    report_provider_error_background(
+                        CHANNEL_ID,
+                        "spawn_initial_load.runtime_build",
+                        &error.to_string(),
+                    );
+                    return;
+                }
+            };
+
+            runtime.block_on(async move {
+                if let Err(error) =
+                    Self::load_categories(&url, categories, category_map).await
+                {
+                    report_provider_error_background(
+                        CHANNEL_ID,
+                        "load_categories",
+                        &error.to_string(),
+                    );
+                }
+            });
+        });
+    }
+
+    async fn load_categories(
+        url: &str,
+        categories: Arc<RwLock<Vec<FilterOption>>>,
+        category_map: Arc<RwLock<HashMap<String, String>>>,
+    ) -> Result<()> {
+        let mut requester = Requester::new();
+        let category_url = format!("{url}/categories/");
+        let html = requester
+            .get_with_headers(&category_url, Self::html_headers(url), None)
+            .await
+            .map_err(|error| Error::from(format!("category fetch failed: {error}")))?;
+
+        let document = Html::parse_document(&html);
+        let selector = Selector::parse("div.thumb.item a[href*=\"/categories/\"]")
+            .map_err(|error| Error::from(format!("category selector failed: {error}")))?;
+
+        let mut options = vec![FilterOption {
+            id: "all".to_string(),
+            title: "All".to_string(),
+        }];
+        let mut map = HashMap::new();
+
+        for link in document.select(&selector) {
+            let Some(href) = link.value().attr("href") else {
+                continue;
+            };
+            let Some(slug) = href
+                .split("/categories/")
+                .nth(1)
+                .and_then(|rest| rest.split('/').next())
+                .filter(|slug| !slug.is_empty())
+            else {
+                continue;
+            };
+            let title = link
+                .value()
+                .attr("title")
+                .map(Self::decode_html_text)
+                .filter(|title| !title.is_empty())
+                .unwrap_or_else(|| Self::decode_html_text(&link.text().collect::<String>()));
+            if title.is_empty() {
+                continue;
+            }
+            let normalized = Self::normalize_title(&title);
+            if normalized.is_empty() || map.contains_key(&normalized) {
+                continue;
+            }
+            options.push(FilterOption {
+                id: normalized.clone(),
+                title,
+            });
+            map.insert(normalized, slug.to_string());
+        }
+
+        if options.len() > 1 {
+            if let Ok(mut guard) = categories.write() {
+                *guard = options;
+            }
+            if let Ok(mut guard) = category_map.write() {
+                *guard = map;
+            }
+        }
+
+        Ok(())
+    }
+}
+
+#[async_trait]
+impl Provider for HornyleakProvider {
+    async fn get_videos(
+        &self,
+        cache: VideoCache,
+        _pool: DbPool,
+        sort: String,
+        query: Option<String>,
+        page: String,
+        _per_page: String,
+        options: ServerOptions,
+    ) -> Vec<VideoItem> {
+        let page = page.parse::<u32>().unwrap_or(1).max(1);
+        let target = self.target_from_request(query.as_deref(), options.categories.as_deref());
+
+        match self
+            .get_videos_for_target(cache, &target, &sort, page, options)
+            .await
+        {
+            Ok(videos) => videos,
+            Err(e) => {
+                report_provider_error(
+                    CHANNEL_ID,
+                    "get_videos",
+                    &format!("page={page}; target={target:?}; error={e}"),
+                )
+                .await;
+                vec![]
+            }
+        }
+    }
+
+    fn get_channel(&self, clientversion: ClientVersion) -> Option<Channel> {
+        Some(self.build_channel(clientversion))
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn slugify_collapses_separators() {
+        assert_eq!(HornyleakProvider::slugify("Big Black Cock"), "big-black-cock");
+        assert_eq!(HornyleakProvider::slugify("  big__ass  "), "big-ass");
+    }
+
+    #[test]
+    fn builds_latest_block_url() {
+        let provider = HornyleakProvider::new();
+        let (url, referer) = provider.build_listing_url(&Target::Latest, "new", 1);
+        assert_eq!(referer, "https://www.hornyleak.tv/latest-updates/");
+        assert_eq!(url, "https://www.hornyleak.tv/latest-updates/");
+
+        let (url, referer) = provider.build_listing_url(&Target::Latest, "new", 2);
+        assert_eq!(referer, "https://www.hornyleak.tv/latest-updates/");
+        assert_eq!(url, "https://www.hornyleak.tv/latest-updates/2/");
+    }
+
+    #[test]
+    fn builds_search_url() {
+        let provider = HornyleakProvider::new();
+        let (url, referer) = provider.build_listing_url(
+            &Target::Search {
+                query: "big ass".to_string(),
+            },
+            "new",
+            1,
+        );
+        assert_eq!(referer, "https://www.hornyleak.tv/search/");
+        assert_eq!(url, "https://www.hornyleak.tv/search/?q=big+ass");
+
+        let (url, _) = provider.build_listing_url(
+            &Target::Search {
+                query: "big ass".to_string(),
+            },
+            "popular",
+            2,
+        );
+        assert!(url.contains("from_videos=2"));
+        assert!(url.contains("from_albums=2"));
+        assert!(url.contains("sort_by=most_popular"));
+    }
+
+    #[test]
+    fn category_block_url_uses_common_list() {
+        let provider = HornyleakProvider::new();
+        let (url, _) = provider.build_listing_url(
+            &Target::Category {
+                slug: "anal".to_string(),
+            },
+            "new",
+            1,
+        );
+        assert_eq!(url, "https://www.hornyleak.tv/categories/anal/");
+
+        let (url, _) = provider.build_listing_url(
+            &Target::Category {
+                slug: "anal".to_string(),
+            },
+            "new",
+            3,
+        );
+        assert_eq!(url, "https://www.hornyleak.tv/categories/anal/3/");
+    }
+
+    #[test]
+    fn target_from_request_routes_prefixes() {
+        let provider = HornyleakProvider::new();
+        let target = provider.target_from_request(Some("tag:amateur"), None);
+        assert!(matches!(target, Target::Tag { slug } if slug == "amateur"));
+
+        let target = provider.target_from_request(Some("cat:anal"), None);
+        assert!(matches!(target, Target::Category { slug } if slug == "anal"));
+
+        let target = provider.target_from_request(Some("uploader:bunny-umi"), None);
+        assert!(matches!(target, Target::Model { slug } if slug == "bunny-umi"));
+
+        let target = provider.target_from_request(Some("big ass"), None);
+        assert!(matches!(target, Target::Search { ref query } if query == "big ass"));
+
+        let target = provider.target_from_request(None, None);
+        assert!(matches!(target, Target::Latest));
+    }
+
+    #[test]
+    fn parses_static_listing_card() {
+        let provider = HornyleakProvider::new();
+        let html = r##"
+            <div class="thumbs" id="list_videos_most_recent_videos_items">
+              <div class="thumb thumb_rel item  ">
+                <a href="https://www.hornyleak.tv/video/32944/bunny-umi-gets-creampied-by-gattouz0/" title="Bunny Umi Gets Creampied By Gattouz0">
+                  <div class="img-holder second">
+                    <img class="" src="https://www.hornyleak.tv/contents/videos_screenshots/32000/32944/preview.jpg" data-webp="https://www.hornyleak.tv/contents/videos_screenshots/32000/32944/preview.jpg" data-preview="https://hdplayer.gives/upload/videos/2026/09/x/preview.mp4" width="336" height="189"/>
+                    <div class="item-bottom"><div class="time">1:55:09</div></div>
+                    <div class="item-bottom second"><div class="rating">92%</div></div>
+                  </div>
+                </a>
+                <div class="card-content">
+                  <a class="avatar" href="https://www.hornyleak.tv/models/bunny-umi/" title="Bunny Umi">
+                    <span class="name">Bunny Umi</span>
+                  </a>
+                  <div class="thumb-item"><i></i> 143K</div>
+                </div>
+              </div>
+            </div>
+        "##;
+
+        let cards = provider.parse_video_cards(html);
+        assert_eq!(cards.len(), 1);
+        let card = &cards[0];
+        assert_eq!(card.id, "32944");
+        assert_eq!(card.title, "Bunny Umi Gets Creampied By Gattouz0");
+        assert_eq!(
+            card.page_url,
+            "https://www.hornyleak.tv/video/32944/bunny-umi-gets-creampied-by-gattouz0/"
+        );
+        assert_eq!(card.duration, 6909);
+        assert_eq!(card.rating, Some(92));
+        assert_eq!(card.uploader.as_deref(), Some("Bunny Umi"));
+        assert_eq!(
+            card.uploader_url.as_deref(),
+            Some("https://www.hornyleak.tv/models/bunny-umi/")
+        );
+        assert_eq!(card.uploader_slug.as_deref(), Some("bunny-umi"));
+    }
+}

src/providers/notfans.rs

@@ -0,0 +1,886 @@
+use crate::DbPool;
+use crate::api::ClientVersion;
+use crate::providers::{Provider, report_provider_error, requester_or_default};
+use crate::status::*;
+use crate::util::cache::VideoCache;
+use crate::util::time::parse_time_to_seconds;
+use crate::videos::{ServerOptions, VideoFormat, VideoItem};
+use async_trait::async_trait;
+use chrono::{Duration as ChronoDuration, Utc};
+use error_chain::error_chain;
+use futures::stream::{self, StreamExt};
+use htmlentity::entity::{ICodedDataTrait, decode};
+use percent_encoding::{NON_ALPHANUMERIC, utf8_percent_encode};
+use regex::Regex;
+use tokio::time::timeout;
+use url::Url;
+use wreq::Version;
+
+use std::time::Duration as StdDuration;
+
+pub const CHANNEL_METADATA: crate::providers::ProviderChannelMetadata =
+    crate::providers::ProviderChannelMetadata {
+        group_id: "onlyfans",
+        tags: &["onlyfans", "leaks", "aggregator"],
+    };
+
+const BASE_URL: &str = "https://notfans.com";
+const CHANNEL_ID: &str = "notfans";
+// Chrome on Linux — matches the UA the wreq Chrome 120 emulation ships with so
+// the site's CF edge treats us as a normal browser.
+const BROWSER_UA: &str =
+    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36";
+const HTML_ACCEPT: &str =
+    "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8";
+const REFERER: &str = "https://notfans.com/";
+// Notfans listing cards are notfans-themed, but the class name is shared with
+// other KVS-based sites (homoxxx etc.) — match exactly with the two trailing
+// spaces so we skip the sponsored-ad variants (`class="item avd"`).
+const CARD_DELIMITER: &str = "<div class=\"item  \">";
+const CACHE_TTL_SECS: u64 = 60 * 15;
+const DETAIL_TIMEOUT_SECS: u64 = 8;
+const DETAIL_FETCH_CONCURRENCY: usize = 4;
+
+error_chain! {
+    foreign_links {
+        HttpRequest(wreq::Error);
+        Io(std::io::Error);
+    }
+    errors {
+        Parse(msg: String) {
+            description("parse error")
+            display("parse error: {}", msg)
+        }
+    }
+}
+
+#[derive(Debug, Clone)]
+pub struct NotfansProvider {
+    url: String,
+}
+
+#[derive(Debug, Clone)]
+enum Target {
+    Latest,
+    MostPopular,
+    TopRated,
+    Today,
+    Week,
+    Month,
+    Year,
+    Search { query: String },
+    Tag { tag: String },
+}
+
+impl NotfansProvider {
+    pub fn new() -> Self {
+        Self {
+            url: BASE_URL.to_string(),
+        }
+    }
+
+    fn build_channel(&self, _clientversion: ClientVersion) -> Channel {
+        Channel {
+            id: CHANNEL_ID.to_string(),
+            name: "NotFans".to_string(),
+            description:
+                "NotFans — free OnlyFans, Fansly, Patreon, and Manyvids leak aggregator. \
+                 Listing-only feed sorted by latest, most popular, top rated, and top by period."
+                    .to_string(),
+            premium: false,
+            favicon: "https://www.google.com/s2/favicons?sz=64&domain=notfans.com".to_string(),
+            status: "active".to_string(),
+            categories: vec![],
+            options: vec![ChannelOption {
+                id: "sort".to_string(),
+                title: "Sort".to_string(),
+                description: "Browse the NotFans archive.".to_string(),
+                systemImage: "list.number".to_string(),
+                colorName: "blue".to_string(),
+                options: vec![
+                    FilterOption { id: "new".to_string(), title: "Latest".to_string() },
+                    FilterOption { id: "popular".to_string(), title: "Most Viewed".to_string() },
+                    FilterOption { id: "top".to_string(), title: "Top Rated".to_string() },
+                    FilterOption { id: "today".to_string(), title: "Top Today".to_string() },
+                    FilterOption { id: "week".to_string(), title: "Top This Week".to_string() },
+                    FilterOption { id: "month".to_string(), title: "Top This Month".to_string() },
+                    FilterOption { id: "year".to_string(), title: "Top This Year".to_string() },
+                ],
+                multiSelect: false,
+            }],
+            nsfw: true,
+            cacheDuration: Some(1800),
+            ytdlpCommand: Some("yt-dlp --impersonate chrome-120".to_string()),
+        }
+    }
+
+    fn html_headers() -> Vec<(String, String)> {
+        vec![
+            ("Referer".to_string(), REFERER.to_string()),
+            ("User-Agent".to_string(), BROWSER_UA.to_string()),
+            ("Accept".to_string(), HTML_ACCEPT.to_string()),
+            ("Accept-Language".to_string(), "en-US,en;q=0.9".to_string()),
+        ]
+    }
+
+    fn absolute_url(&self, value: &str) -> String {
+        let trimmed = value.trim();
+        if trimmed.is_empty() {
+            return String::new();
+        }
+        if trimmed.starts_with("http://") || trimmed.starts_with("https://") {
+            return trimmed.to_string();
+        }
+        if trimmed.starts_with("//") {
+            return format!("https:{trimmed}");
+        }
+        format!(
+            "{}/{}",
+            self.url.trim_end_matches('/'),
+            trimmed.trim_start_matches('/')
+        )
+    }
+
+    fn resolve_target(sort: &str, query: Option<&str>) -> Target {
+        if let Some(raw) = query.map(str::trim).filter(|q| !q.is_empty()) {
+            // Optional `tag:` prefix routes to a tag archive page.
+            if let Some(rest) = raw
+                .strip_prefix("tag:")
+                .or_else(|| raw.strip_prefix("category:"))
+                .or_else(|| raw.strip_prefix("tags:"))
+            {
+                let tag = rest.trim();
+                if !tag.is_empty() {
+                    return Target::Tag { tag: tag.to_string() };
+                }
+            }
+            return Target::Search {
+                query: raw.to_string(),
+            };
+        }
+
+        match sort {
+            "popular" | "hot" | "most_viewed" | "most-viewed" => Target::MostPopular,
+            "top" | "rated" | "top_rated" | "top-rated" => Target::TopRated,
+            "today" | "day" => Target::Today,
+            "week" | "weekly" => Target::Week,
+            "month" | "monthly" => Target::Month,
+            "year" | "yearly" => Target::Year,
+            _ => Target::Latest,
+        }
+    }
+
+    fn build_url(&self, target: &Target, page: u16) -> String {
+        let page = page.max(1);
+        match target {
+            Target::Latest => {
+                if page == 1 {
+                    format!("{}/latest-updates/", self.url)
+                } else {
+                    format!("{}/latest-updates/{page}/", self.url)
+                }
+            }
+            Target::MostPopular => {
+                if page == 1 {
+                    format!("{}/most-popular/", self.url)
+                } else {
+                    format!("{}/most-popular/{page}/", self.url)
+                }
+            }
+            Target::TopRated => {
+                if page == 1 {
+                    format!("{}/top-rated/", self.url)
+                } else {
+                    format!("{}/top-rated/{page}/", self.url)
+                }
+            }
+            Target::Today => self.period_url("day", page),
+            Target::Week => self.period_url("week", page),
+            Target::Month => self.period_url("month", page),
+            Target::Year => self.period_url("year", page),
+            Target::Search { query } => self.search_url(query, page),
+            Target::Tag { tag } => self.tag_url(tag, page),
+        }
+    }
+
+    fn period_url(&self, period: &str, page: u16) -> String {
+        if page == 1 {
+            format!("{}/{period}/", self.url)
+        } else {
+            format!("{}/{period}/{page}/", self.url)
+        }
+    }
+
+    fn search_url(&self, query: &str, page: u16) -> String {
+        let encoded = utf8_percent_encode(query, NON_ALPHANUMERIC).to_string();
+        if page == 1 {
+            format!("{}/search/{encoded}/", self.url)
+        } else {
+            format!("{}/search/{encoded}/{page}/", self.url)
+        }
+    }
+
+    fn tag_url(&self, tag: &str, page: u16) -> String {
+        let encoded = utf8_percent_encode(tag, NON_ALPHANUMERIC).to_string();
+        if page == 1 {
+            format!("{}/tags/{encoded}/", self.url)
+        } else {
+            format!("{}/tags/{encoded}/{page}/", self.url)
+        }
+    }
+
+    fn card_segments(html: &str) -> Vec<&str> {
+        // Split on the literal class marker so we get a tail-aligned slice per
+        // card. Anything before the first marker is the page chrome.
+        html.split(CARD_DELIMITER).skip(1).collect()
+    }
+
+    fn parse_card_segment(segment: &str) -> Option<(String, String, String, String)> {
+        // <a href="https://notfans.com/videos/228306/acehardz-wrecked-moderngomorra/" title="AceHardz Wrecked ModernGomorra" >
+        let href_start = segment.find("<a href=\"")?;
+        let after_href = &segment[href_start + "<a href=\"".len()..];
+        let href_end = after_href.find('"')?;
+        let href = &after_href[..href_end];
+        let href = decode(href.as_bytes())
+            .to_string()
+            .unwrap_or_else(|_| href.to_string());
+        if href.is_empty() {
+            return None;
+        }
+
+        // title="AceHardz Wrecked ModernGomorra"
+        let title_start = after_href[href_end..].find("title=\"")?;
+        let after_title = &after_href[href_end + title_start + "title=\"".len()..];
+        let title_end = after_title.find('"')?;
+        let title = decode(after_title[..title_end].as_bytes())
+            .to_string()
+            .unwrap_or_else(|_| after_title[..title_end].to_string());
+        if title.is_empty() {
+            return None;
+        }
+
+        // <img class="thumb " src="https://notfans.com/contents/videos_screenshots/.../320x180/N.jpg" ...>
+        let thumb = Self::extract_thumb(after_title).unwrap_or_default();
+
+        Some((href.to_string(), title.trim().to_string(), thumb, String::new()))
+    }
+
+    fn extract_thumb(haystack: &str) -> Option<String> {
+        // Walk the segment for the first <img class="thumb" src="..."> match.
+        let needle = "class=\"thumb";
+        let mut cursor = 0usize;
+        while let Some(found) = haystack[cursor..].find(needle) {
+            let img_start = cursor + found;
+            let after = &haystack[img_start..];
+            // Match a quoted `src="..."` within the same tag.
+            let after_class = &after[needle.len()..];
+            let src_idx = after_class.find("src=\"")?;
+            let after_src = &after_class[src_idx + "src=\"".len()..];
+            let src_end = after_src.find('"')?;
+            let src = decode(after_src[..src_end].as_bytes())
+                .to_string()
+                .unwrap_or_else(|_| after_src[..src_end].to_string());
+            if !src.is_empty() {
+                return Some(src);
+            }
+            cursor = img_start + needle.len();
+        }
+        None
+    }
+
+    fn parse_listing(&self, html: &str) -> Vec<VideoItem> {
+        Self::card_segments(html)
+            .into_iter()
+            .filter_map(|segment| Self::parse_card_segment(segment))
+            .filter_map(|(href, title, thumb, _)| {
+                let url = if href.starts_with("http://") || href.starts_with("https://") {
+                    href
+                } else {
+                    self.absolute_url(&href)
+                };
+                let id = Self::extract_id_from_url(&url)?;
+                if id.is_empty() || title.is_empty() {
+                    return None;
+                }
+                Some(VideoItem::new(
+                    id,
+                    title,
+                    url,
+                    CHANNEL_ID.to_string(),
+                    thumb,
+                    0,
+                ))
+            })
+            .collect()
+    }
+
+    fn extract_id_from_url(url: &str) -> Option<String> {
+        // /videos/228306/acehardz-wrecked-moderngomorra/ → 228306
+        let parsed = Url::parse(url).ok()?;
+        let mut segments = parsed.path_segments()?.filter(|s| !s.is_empty());
+        let videos = segments.next()?;
+        if !videos.eq_ignore_ascii_case("videos") {
+            return None;
+        }
+        let id = segments.next()?;
+        let cleaned: String = id.chars().filter(|c| c.is_ascii_digit()).collect();
+        if cleaned.is_empty() {
+            None
+        } else {
+            Some(cleaned)
+        }
+    }
+
+    fn apply_detail(&self, mut item: VideoItem, html: &str) -> VideoItem {
+        // Duration: <em>30:30</em>
+        if item.duration == 0 {
+            if let Some(duration_text) = Self::capture_after_label(html, "Duration:", "<em>")
+            {
+                if let Some(end) = duration_text.find("</em>") {
+                    let raw = duration_text[..end].trim();
+                    if let Some(seconds) = parse_time_to_seconds(raw) {
+                        if seconds > 0 {
+                            item.duration = seconds.max(0).min(u32::MAX as i64) as u32;
+                        }
+                    }
+                }
+            }
+        }
+
+        // Views: <em>289</em>
+        if item.views.is_none() {
+            if let Some(views_text) = Self::capture_after_label(html, "Views:", "<em>") {
+                if let Some(end) = views_text.find("</em>") {
+                    let raw = views_text[..end].trim();
+                    if let Ok(value) = raw.replace(',', "").parse::<u32>() {
+                        item.views = Some(value);
+                    }
+                }
+            }
+        }
+
+        // Tags: <a href="/tags/acehardz/">acehardz</a> ... up to </div>
+        if item.tags.is_none() {
+            if let Some(tags_html) = Self::slice_after_label(html, "Tags:") {
+                if let Some(div_end) = tags_html.find("</div>") {
+                    let block = &tags_html[..div_end];
+                    let tag_re = Regex::new(r#"/tags/[^"/?#]+/?"?\s*[^>]*>([^<]+)</a>"#).ok();
+                    if let Some(re) = tag_re {
+                        let mut tags: Vec<String> = re
+                            .captures_iter(block)
+                            .filter_map(|caps| caps.get(1).map(|m| m.as_str().trim().to_string()))
+                            .filter(|t| !t.is_empty())
+                            .collect();
+                        tags.sort();
+                        tags.dedup();
+                        if !tags.is_empty() {
+                            item.tags = Some(tags);
+                        }
+                    }
+                }
+            }
+        }
+
+        // Submitted: <em>4 hours ago</em> → relative timestamp
+        if item.uploadedAt.is_none() {
+            if let Some(submitted) = Self::capture_after_label(html, "Submitted:", "<em>") {
+                if let Some(end) = submitted.find("</em>") {
+                    let raw = submitted[..end].trim();
+                    if let Some(ts) = Self::parse_relative_time(raw) {
+                        item.uploadedAt = Some(ts);
+                    }
+                }
+            }
+        }
+
+        // Format URL: first Download link to a /get_file/6/.../{id}.mp4/...
+        if item.formats.is_none() {
+            if let Some(download_block) = Self::slice_after_label(html, "Download:") {
+                if let Some(div_end) = download_block.find("</div>") {
+                    let block = &download_block[..div_end];
+                    if let Some(format_url) = Self::extract_download_url(block) {
+                        let mut format = VideoFormat::new(
+                            format_url,
+                            "720p".to_string(),
+                            "mp4".to_string(),
+                        );
+                        format = format.http_header("Referer".to_string(), REFERER.to_string());
+                        format = format.http_header("User-Agent".to_string(), BROWSER_UA.to_string());
+                        item.formats = Some(vec![format]);
+                    }
+                }
+            }
+        }
+
+        // Preview: og:image if present and richer than the listing thumb.
+        if item.preview.is_none() {
+            if let Some(og_image) = Self::extract_og_image(html) {
+                if !og_image.is_empty() {
+                    item.preview = Some(og_image);
+                }
+            }
+        }
+
+        item
+    }
+
+    fn capture_after_label<'a>(
+        haystack: &'a str,
+        label: &str,
+        open_tag: &str,
+    ) -> Option<&'a str> {
+        let idx = haystack.find(label)?;
+        let after_label = &haystack[idx + label.len()..];
+        let open_idx = after_label.find(open_tag)?;
+        Some(&after_label[open_idx + open_tag.len()..])
+    }
+
+    fn slice_after_label<'a>(haystack: &'a str, label: &str) -> Option<&'a str> {
+        let idx = haystack.find(label)?;
+        Some(&haystack[idx + label.len()..])
+    }
+
+    fn extract_download_url(block: &str) -> Option<String> {
+        // First <a href="...get_file/6/{hash}/{id3}/{id}/{id}.mp4/?...">MP4 ...</a>
+        let needle = "get_file/6/";
+        let idx = block.find(needle)?;
+        // Walk backwards to the start of the href= attribute
+        let before = &block[..idx];
+        let href_idx = before.rfind("href=\"")?;
+        let after = &block[href_idx + "href=\"".len()..];
+        let href_end = after.find('"')?;
+        let url = decode(after[..href_end].as_bytes())
+            .to_string()
+            .unwrap_or_else(|_| after[..href_end].to_string());
+        if url.is_empty() {
+            None
+        } else {
+            Some(url)
+        }
+    }
+
+    fn extract_og_image(html: &str) -> Option<String> {
+        let needle = "property=\"og:image\"";
+        let idx = html.find(needle)?;
+        let after = &html[idx + needle.len()..];
+        let content_idx = after.find("content=\"")?;
+        let after_content = &after[content_idx + "content=\"".len()..];
+        let end = after_content.find('"')?;
+        let url = decode(after_content[..end].as_bytes())
+            .to_string()
+            .unwrap_or_else(|_| after_content[..end].to_string());
+        if url.is_empty() {
+            None
+        } else {
+            Some(url)
+        }
+    }
+
+    fn parse_relative_time(text: &str) -> Option<u64> {
+        let trimmed = text.trim();
+        if trimmed.is_empty() {
+            return None;
+        }
+        // "4 hours ago", "1 day ago", "3 weeks ago", "2 months ago", "1 year ago"
+        // and the bare unit forms like "5 minutes", "Yesterday", "Just now".
+        let lowered = trimmed.to_ascii_lowercase();
+        let now = Utc::now();
+        let now_ts = now.timestamp() as u64;
+
+        if lowered.starts_with("just now") || lowered == "now" {
+            return Some(now_ts);
+        }
+        if lowered.starts_with("yesterday") {
+            return Some(now_ts - ChronoDuration::days(1).num_seconds() as u64);
+        }
+        if lowered.contains("second") {
+            return Some(now_ts);
+        }
+
+        let re = Regex::new(r"^(\d+)\s+(minute|hour|day|week|month|year)s?(?:\s+ago)?$").ok()?;
+        let caps = re.captures(&lowered)?;
+        let amount: i64 = caps.get(1)?.as_str().parse().ok()?;
+        let unit = caps.get(2)?.as_str();
+        let delta = match unit {
+            "minute" => ChronoDuration::minutes(amount),
+            "hour" => ChronoDuration::hours(amount),
+            "day" => ChronoDuration::days(amount),
+            "week" => ChronoDuration::weeks(amount),
+            "month" => ChronoDuration::days(amount * 30),
+            "year" => ChronoDuration::days(amount * 365),
+            _ => return None,
+        };
+        Some((now - delta).timestamp().max(0) as u64)
+    }
+
+    async fn fetch_html(
+        &self,
+        options: &ServerOptions,
+        context: &str,
+        url: &str,
+    ) -> Result<String> {
+        let mut requester = requester_or_default(options, CHANNEL_ID, context);
+        let response = requester
+            .get_with_headers(url, Self::html_headers(), Some(Version::HTTP_11))
+            .await
+            .map_err(|error| Error::from(format!("request failed for {url}: {error}")))?;
+        Ok(response)
+    }
+
+    async fn enrich_item(
+        &self,
+        item: VideoItem,
+        options: &ServerOptions,
+    ) -> VideoItem {
+        let detail_url = item.url.clone();
+        let fetch = timeout(
+            StdDuration::from_secs(DETAIL_TIMEOUT_SECS),
+            self.fetch_html(options, "notfans.enrich", &detail_url),
+        )
+        .await;
+        match fetch {
+            Ok(Ok(html)) => self.apply_detail(item, &html),
+            Ok(Err(error)) => {
+                report_provider_error_background(CHANNEL_ID, "enrich_item", &error.to_string());
+                item
+            }
+            Err(_) => {
+                report_provider_error_background(
+                    CHANNEL_ID,
+                    "enrich_item_timeout",
+                    &detail_url,
+                );
+                item
+            }
+        }
+    }
+
+    async fn fetch_items(
+        &self,
+        cache: VideoCache,
+        url: String,
+        per_page_limit: usize,
+        enrich: bool,
+        options: &ServerOptions,
+    ) -> Result<Vec<VideoItem>> {
+        if let Some((time, items)) = cache.get(&url) {
+            if time.elapsed().unwrap_or_default().as_secs() < CACHE_TTL_SECS {
+                return Ok(items.clone());
+            }
+        }
+
+        let html = self.fetch_html(options, "notfans.fetch_items", &url).await?;
+        let listed = self.parse_listing(&html);
+        if listed.is_empty() {
+            return Ok(vec![]);
+        }
+        let limited: Vec<VideoItem> = listed
+            .into_iter()
+            .take(per_page_limit.max(1))
+            .collect();
+
+        if !enrich {
+            cache.insert(url, limited.clone());
+            return Ok(limited);
+        }
+
+        let items = stream::iter(limited.into_iter().map(|item| {
+            let provider = self.clone();
+            let options = options.clone();
+            async move { provider.enrich_item(item, &options).await }
+        }))
+        .buffer_unordered(DETAIL_FETCH_CONCURRENCY)
+        .collect::<Vec<_>>()
+        .await;
+
+        if !items.is_empty() {
+            cache.insert(url, items.clone());
+        }
+        Ok(items)
+    }
+}
+
+fn report_provider_error_background(provider: &str, context: &str, msg: &str) {
+    let provider = provider.to_string();
+    let context = context.to_string();
+    let msg = msg.to_string();
+    tokio::spawn(async move {
+        crate::providers::report_provider_error(&provider, &context, &msg).await;
+    });
+}
+
+#[async_trait]
+impl Provider for NotfansProvider {
+    async fn get_videos(
+        &self,
+        cache: VideoCache,
+        _pool: DbPool,
+        sort: String,
+        query: Option<String>,
+        page: String,
+        per_page: String,
+        options: ServerOptions,
+    ) -> Vec<VideoItem> {
+        let sort_value = if sort.is_empty() {
+            options.sort.as_deref().unwrap_or("new").to_string()
+        } else {
+            sort
+        };
+        let page_num = page.parse::<u16>().unwrap_or(1).max(1);
+        let per_page_limit = per_page.parse::<usize>().unwrap_or(30).clamp(1, 64);
+
+        let target = Self::resolve_target(&sort_value, query.as_deref());
+        let url = self.build_url(&target, page_num);
+        // Only enrich the first page to keep cold-start latency bounded;
+        // subsequent pages are reached via the app's own prefetch / cache.
+        let enrich = page_num == 1;
+
+        match self
+            .fetch_items(cache, url, per_page_limit, enrich, &options)
+            .await
+        {
+            Ok(videos) => videos,
+            Err(error) => {
+                report_provider_error(CHANNEL_ID, "get_videos", &error.to_string()).await;
+                vec![]
+            }
+        }
+    }
+
+    fn get_channel(&self, clientversion: ClientVersion) -> Option<Channel> {
+        Some(self.build_channel(clientversion))
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    fn provider() -> NotfansProvider {
+        NotfansProvider::new()
+    }
+
+    #[test]
+    fn builds_latest_url_without_page_suffix() {
+        let url = provider().build_url(&Target::Latest, 1);
+        assert_eq!(url, "https://notfans.com/latest-updates/");
+    }
+
+    #[test]
+    fn builds_latest_url_with_page() {
+        let url = provider().build_url(&Target::Latest, 7);
+        assert_eq!(url, "https://notfans.com/latest-updates/7/");
+    }
+
+    #[test]
+    fn builds_period_urls() {
+        let p = provider();
+        assert_eq!(p.build_url(&Target::Today, 1), "https://notfans.com/day/");
+        assert_eq!(p.build_url(&Target::Week, 3), "https://notfans.com/week/3/");
+        assert_eq!(p.build_url(&Target::Month, 1), "https://notfans.com/month/");
+        assert_eq!(p.build_url(&Target::Year, 5), "https://notfans.com/year/5/");
+    }
+
+    #[test]
+    fn builds_search_url() {
+        let url = provider().build_url(
+            &Target::Search {
+                query: "ace hardz".to_string(),
+            },
+            1,
+        );
+        assert_eq!(url, "https://notfans.com/search/ace%20hardz/");
+    }
+
+    #[test]
+    fn builds_tag_url() {
+        let url = provider().build_url(
+            &Target::Tag {
+                tag: "blowjob".to_string(),
+            },
+            2,
+        );
+        assert_eq!(url, "https://notfans.com/tags/blowjob/2/");
+    }
+
+    #[test]
+    fn resolves_sort_to_target() {
+        assert!(matches!(
+            NotfansProvider::resolve_target("popular", None),
+            Target::MostPopular
+        ));
+        assert!(matches!(
+            NotfansProvider::resolve_target("top", None),
+            Target::TopRated
+        ));
+        assert!(matches!(
+            NotfansProvider::resolve_target("today", None),
+            Target::Today
+        ));
+        assert!(matches!(
+            NotfansProvider::resolve_target("new", None),
+            Target::Latest
+        ));
+        assert!(matches!(
+            NotfansProvider::resolve_target("unknown", None),
+            Target::Latest
+        ));
+    }
+
+    #[test]
+    fn resolves_query_to_search_or_tag() {
+        assert!(matches!(
+            NotfansProvider::resolve_target("new", Some("blowjob")),
+            Target::Search { .. }
+        ));
+        assert!(matches!(
+            NotfansProvider::resolve_target("new", Some("tag:anal")),
+            Target::Tag { .. }
+        ));
+        assert!(matches!(
+            NotfansProvider::resolve_target("new", Some("  tag:milf  ")),
+            Target::Tag { .. }
+        ));
+    }
+
+    #[test]
+    fn parses_id_from_video_url() {
+        assert_eq!(
+            NotfansProvider::extract_id_from_url(
+                "https://notfans.com/videos/228306/acehardz-wrecked-moderngomorra/"
+            ),
+            Some("228306".to_string())
+        );
+        assert_eq!(NotfansProvider::extract_id_from_url("https://example.com/"), None);
+        assert_eq!(
+            NotfansProvider::extract_id_from_url("https://notfans.com/videos//slug/"),
+            None
+        );
+    }
+
+    #[test]
+    fn parses_card_segment() {
+        let segment = r#"
+            <a href="https://notfans.com/videos/228306/acehardz-wrecked-moderngomorra/" title="AceHardz Wrecked ModernGomorra" >
+                <div class="img">
+                    <img class="thumb " src="https://notfans.com/contents/videos_screenshots/228000/228306/320x180/4.jpg"  alt="x" data-cnt="5"   width="320" height="180"/>
+                </div>
+                <strong class="title">x</strong>
+            </a>
+        "#;
+        let (href, title, thumb, _) = NotfansProvider::parse_card_segment(segment).unwrap();
+        assert_eq!(href, "https://notfans.com/videos/228306/acehardz-wrecked-moderngomorra/");
+        assert_eq!(title, "AceHardz Wrecked ModernGomorra");
+        assert!(thumb.contains("228306/320x180/4.jpg"));
+    }
+
+    #[test]
+    fn parses_listing_from_realistic_html() {
+        let html = format!(
+            "{chrome}<div class=\"item  \">{card1}</div>\
+             <div class=\"item avd\"><a href=\"#\" id=\"ad\">sponsored</a></div>\
+             <div class=\"item  \">{card2}</div>{footer}",
+            chrome = r#"<html><body><div class="head">junk</div>"#,
+            card1 = r#"
+                <a href="https://notfans.com/videos/228306/acehardz-wrecked-moderngomorra/" title="AceHardz Wrecked ModernGomorra" >
+                    <div class="img">
+                        <img class="thumb " src="https://notfans.com/contents/videos_screenshots/228000/228306/320x180/4.jpg"  alt="x" data-cnt="5"   width="320" height="180"/>
+                    </div>
+                    <strong class="title">x</strong>
+                </a>
+            "#,
+            card2 = r#"
+                <a href="https://notfans.com/videos/100/gracie-bon-ppv/" title="Gracie Bon PPV" >
+                    <div class="img">
+                        <img class="thumb " src="https://notfans.com/contents/videos_screenshots/0/100/320x180/1.jpg"  alt="x" data-cnt="1"   width="320" height="180"/>
+                    </div>
+                    <strong class="title">x</strong>
+                </a>
+            "#,
+            footer = "</body></html>",
+        );
+        let items = provider().parse_listing(&html);
+        assert_eq!(items.len(), 2, "expected 2 items, sponsored card must be skipped");
+        assert_eq!(items[0].id, "228306");
+        assert_eq!(items[0].title, "AceHardz Wrecked ModernGomorra");
+        assert!(items[0].thumb.contains("228306/320x180/4.jpg"));
+        assert_eq!(items[1].id, "100");
+    }
+
+    #[test]
+    fn apply_detail_extracts_metadata() {
+        let html = r#"
+            <html><body>
+                <h1>Some Video</h1>
+                <div class="info">
+                    <span>Duration: <em>30:30</em></span>
+                    <span>Views: <em>289</em></span>
+                    <span>Submitted: <em>4 hours ago</em></span>
+                </div>
+                <div class="item">
+                    Tags:
+                    <a href="https://notfans.com/tags/acehardz/">acehardz</a>
+                    <a href="https://notfans.com/tags/wrecked/">wrecked</a>
+                </div>
+                <div class="item">
+                    Download:
+                    <a href="https://notfans.com/get_file/6/abcdef/228000/228306/228306.mp4/?download_filename=foo.mp4&download=true" data-attach-session="PHPSESSID">MP4 720p, 808.68 Mb</a>
+                </div>
+                <meta property="og:image" content="https://notfans.com/contents/videos_screenshots/228000/228306/preview.jpg"/>
+            </body></html>
+        "#;
+        let mut item = VideoItem::new(
+            "228306".to_string(),
+            "Some Video".to_string(),
+            "https://notfans.com/videos/228306/some-video/".to_string(),
+            CHANNEL_ID.to_string(),
+            "https://notfans.com/contents/videos_screenshots/228000/228306/320x180/4.jpg".to_string(),
+            0,
+        );
+        item = provider().apply_detail(item, html);
+        assert_eq!(item.duration, 30 * 60 + 30);
+        assert_eq!(item.views, Some(289));
+        assert_eq!(
+            item.tags.as_deref(),
+            Some(vec!["acehardz".to_string(), "wrecked".to_string()]).as_deref()
+        );
+        assert!(item.uploadedAt.is_some());
+        let formats = item.formats.as_ref().expect("formats should be populated");
+        assert_eq!(formats.len(), 1);
+        assert!(formats[0].url.contains("get_file/6/abcdef/228000/228306/228306.mp4"));
+        let headers = formats[0].http_headers.as_ref().expect("headers set");
+        assert_eq!(headers.get("Referer").map(String::as_str), Some(REFERER));
+        assert_eq!(
+            item.preview.as_deref(),
+            Some("https://notfans.com/contents/videos_screenshots/228000/228306/preview.jpg")
+        );
+    }
+
+    #[test]
+    fn parse_relative_time_handles_units() {
+        let now = Utc::now().timestamp() as u64;
+        // All values are within 5s of `now` so we use loose ranges.
+        let cases = [
+            ("4 hours ago", ChronoDuration::hours(4).num_seconds() as u64),
+            ("1 day ago", ChronoDuration::days(1).num_seconds() as u64),
+            ("2 weeks ago", ChronoDuration::weeks(2).num_seconds() as u64),
+            ("3 months ago", ChronoDuration::days(90).num_seconds() as u64),
+            ("just now", 0),
+            ("Yesterday", ChronoDuration::days(1).num_seconds() as u64),
+        ];
+        for (text, expected_offset) in cases {
+            let parsed = NotfansProvider::parse_relative_time(text)
+                .unwrap_or_else(|| panic!("parse failed for {text}"));
+            if expected_offset == 0 {
+                assert!((now as i64 - parsed as i64).abs() <= 2);
+            } else {
+                let actual_offset = now - parsed;
+                let drift = (actual_offset as i64 - expected_offset as i64).abs();
+                assert!(
+                    drift <= 5,
+                    "text={text} expected_offset={expected_offset} actual_offset={actual_offset}"
+                );
+            }
+        }
+    }
+}