Fix noodlemagazine thumb proxy; check.py verifies proxied urls

The thumb proxy dropped the preview query string, rejected extensionless
getVideoPreview paths and warmed up through the slow fallback chain on
every image. check.py now verifies that a proxied video.url resolves
instead of failing it outright.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MUYdCh2nCVsQYgJAT7gwSi

check.py

@@ -354,7 +354,29 @@ def check_video(video: dict, channel_id: str, results: Results, run_ytdlp: bool)
     if not vurl:
         results.err(channel_id, f"{label}: missing url")
     elif "127.0.0.1" in vurl or "localhost" in vurl:
-        results.err(channel_id, f"{label}: url points to hottub server: {vurl}")
+        # A proxied video.url is legitimate when the page itself cannot be
+        # played directly (see the hottub skill: "if a video URL requires a
+        # server-side proxy, set it as the video URL"). Verify that it
+        # actually resolves instead of rejecting it outright.
+        resolved_vurl = follow_proxy_redirect(vurl)
+        if resolved_vurl and resolved_vurl != vurl:
+            ok, code = http_ok(resolved_vurl)
+            target = resolved_vurl
+        else:
+            ok, code = http_ok(vurl)
+            target = vurl
+        if ok:
+            results.info(channel_id, f"{label}: proxied url OK (HTTP {code})")
+        elif _is_cf_protected(target):
+            results.warn(
+                channel_id,
+                f"{label}: proxied url target unreachable HTTP={code} (CF-protected host, expected): {vurl}",
+            )
+        else:
+            results.err(
+                channel_id,
+                f"{label}: proxied url does not resolve (HTTP={code}, target {target[:120]}): {vurl}",
+            )
     else:
         ok, code = http_ok(vurl)
         if not ok:

docs/provider-fix-todo.md

@@ -37,7 +37,7 @@ Status: `[ ]` open, `[~]` in progress, `[x]` fixed and verified, `[-]` not repro
 |---|---|---|---|
 | [x] | `xxdbx` | malformed thumbs `https:/7004….jpg` | assets are now site-relative (`/id.jpg`) rather than `//host/..`; new `absolute_url` resolves all three forms. 36 items, p2/search OK, check.py passes |
 | [x] | `ph-shorties` | thumbs return 403 | same Referer-gated `pix-cdn77.phncdn.com` thumbs; routed through `aylo-thumb` proxy. check.py passes |
-| [ ] | `noodlemagazine` | thumbs unreachable (-1 / 400) in check.py | thumb proxy |
+| [x] | `noodlemagazine` | thumbs unreachable (-1 / 400) in check.py | thumb proxy dropped the preview's query string (`?id&tkn..`), rejected extensionless `getVideoPreview` paths, and warmed up via noodlemagazine.com through the fallback chain on every image (10-28s). All 24 thumbs now 200 in ~0.08s. Titles fix from earlier commit stands |
 
 ## 4. Search
 
@@ -62,7 +62,7 @@ Status: `[ ]` open, `[~]` in progress, `[x]` fixed and verified, `[-]` not repro
 
 | | Item | Notes |
 |---|---|---|
-| [ ] | `check.py` "url points to hottub server" | contradicts the hottub skill (proxy-needing videos use the proxy as `item.url`). Affects spankbang, hqporner, pornhd3x, pimpbunny, sxyprn, noodlemagazine. Decide: exempt proxy URLs that resolve, rather than hard-fail. |
+| [x] | `check.py` "url points to hottub server" | contradicted the hottub skill | proxied `video.url` is now verified to resolve (redirect target reachable, CF hosts -> warning) instead of hard-failing. Affected: spankbang, hqporner, pornhd3x, pimpbunny, sxyprn, noodlemagazine |
 
 ## Open decision (needs the owner)
 

src/proxies/noodlemagazine.rs

@@ -105,9 +105,11 @@ impl NoodlemagazineProxy {
 
     fn has_allowed_image_extension(path: &str) -> bool {
         let path = path.to_ascii_lowercase();
-        [".jpg", ".jpeg", ".png", ".webp", ".avif", ".gif"]
+        // `.../i.mycdn.me/getVideoPreview?id=..` previews have no extension.
-            .iter()
+        path.ends_with("/getvideopreview")
-            .any(|ext| path.ends_with(ext))
+            || [".jpg", ".jpeg", ".png", ".webp", ".avif", ".gif"]
+                .iter()
+                .any(|ext| path.ends_with(ext))
     }
 
     fn is_disallowed_thumb_host(host: &str) -> bool {
@@ -292,20 +294,23 @@ pub async fn get_image(
     requester: web::types::State<Requester>,
 ) -> Result<impl web::Responder, web::Error> {
     let endpoint = req.match_info().query("endpoint").to_string();
-    let image_url = NoodlemagazineProxy::normalize_image_url(&endpoint);
+    let mut image_url = NoodlemagazineProxy::normalize_image_url(&endpoint);
+    // The preview's own query (`id`, `idx`, `type`, `tkn`, `fn`) arrives as the
+    // request query string, not as part of the matched path.
+    let query = req.query_string();
+    if !query.is_empty() && !image_url.contains('?') {
+        image_url.push('?');
+        image_url.push_str(query);
+    }
 
     if !NoodlemagazineProxy::is_allowed_thumb_url(&image_url) {
         return Ok(web::HttpResponse::BadRequest().finish());
     }
 
+    // No site warm-up first: the image CDN answers directly (~0.1s). Warming
+    // via noodlemagazine.com cost 10-30s through the requester's fallbacks;
+    // the retry below still warms up when the first fetch fails.
     let mut requester = requester.get_ref().clone();
-    let _ = requester
-        .get_with_headers(
-            NoodlemagazineProxy::root_referer(),
-            NoodlemagazineProxy::root_html_headers(),
-            Some(Version::HTTP_11),
-        )
-        .await;
 
     let mut headers = NoodlemagazineProxy::image_headers(&requester, image_url.as_str());
     let mut upstream = requester
@@ -409,6 +414,9 @@ mod tests {
 
     #[test]
     fn allows_https_image_thumbs_but_rejects_local_or_non_images() {
+        assert!(NoodlemagazineProxy::is_allowed_thumb_url(
+            "https://img.pvvstream.pro/preview/a/-1_2/240/i.mycdn.me/getVideoPreview?id=1&idx=0&type=39&tkn=x"
+        ));
         assert!(NoodlemagazineProxy::is_allowed_thumb_url(
             "https://noodlemagazine.com/thumbs/example.webp"
         ));