soundcloud: prefer the progressive mp3 stream (HLS exchange 404s anonymously)

Live testing showed SoundCloud only serves the progressive+audio/mpeg
transcoding to anonymous clients: its media exchange returns 200 with a direct,
range-streamable mp3 URL (cf-media.sndcdn.com, 206, audio/mpeg), while the plain
hls+audio/mpeg exchange 404s for every track (streamable or not). The provider
picked HLS, so every queued track failed to resolve a URL and was skipped.

pick_stream_url now prefers progressive, falling back to hls. A progressive URL
is a plain mp3 the player streams on its normal windowed-HTTP path (no .m3u8, so
HlsStream is bypassed); HlsStream stays the fallback for HLS-only tracks. A
genuinely restricted track (Go+/label preview, geo-blocked) still 404s the
exchange and is skipped, not crashed. Note: SoundCloud login does not help here
- public streaming is client_id-only.

Verified live: search a streamable track -> resolve -> 206 range GET returns
audio/mpeg with an mp3 frame-sync header. 19 unit tests pass.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

architecture/soundcloud-provider.md

@@ -135,14 +135,24 @@ Track ids and playlist ids are both numeric, so leaf/container paths use a
 Terms/URLs are percent-encoded into one segment (`encode_segment`/
 `decode_segment`); numeric ids are already URL-safe.
 
-### D4 — Playback via a new HLS source in `audio-player`
+### D4 — Playback: progressive mp3 (HLS source retained as fallback)
 
-- `get_urls_for_track(/soundcloud/track/<id>)`: `GET /tracks/<id>` (or use a
-  cached transcoding), pick the `hls + audio/mpeg` transcoding, `GET
-  <transcoding.url>?client_id=…` → the `.m3u8` media URL, and **return that URL
-  as `urls[0]`** (the player consumes only the first). One API round-trip —
-  unlike abs's pure string-building — because the media URL is signed and
-  ephemeral.
+**Revised after live testing (2026-07-24).** The original plan chose HLS mp3,
+but live probing showed SoundCloud's plain `hls + audio/mpeg` transcoding
+exchange **404s for anonymous clients** (every track, streamable or not), while
+the **`progressive + audio/mpeg`** transcoding returns 200 with a direct,
+range-streamable mp3 URL (`cf-media.sndcdn.com`, `206`, `audio/mpeg`). So the
+provider now **prefers progressive**, which the player streams on its normal
+windowed-HTTP path — no HLS needed for the common case. The `HlsStream` built
+for the original plan is kept as a fallback for any track that offers only HLS.
+
+- `get_urls_for_track(/soundcloud/track/<id>)`: `GET /tracks/<id>`, pick the
+  best mp3 transcoding (`progressive` first, then `hls`), `GET
+  <transcoding.url>?client_id=…` → the media URL, and **return it as `urls[0]`**
+  (the player consumes only the first). One API round-trip — unlike abs's pure
+  string-building — because the media URL is signed and ephemeral. A track
+  whose exchange 404s (Go+/label preview, geo-blocked) surfaces
+  `NotStreamable`/`NotFound` and is skipped, never a crash.
 - **New `audio-player` component `HlsStream`** — a `stream-download`
   `SourceStream`, sibling to `WindowedHttpStream`: on create it fetches the
   `.m3u8` (a media playlist), parses `#EXTINF`/segment URIs (resolving relative

soundclouddy/src/api.rs

@@ -122,10 +122,11 @@ pub trait Sc: Debug + Send + Sync {
     /// follows `ids`.
     async fn hydrate_tracks(&self, ids: &[String]) -> Result<Vec<ScTrack>, FetchError>;
 
-    /// Resolves a track id to a **playable HLS `.m3u8` media URL**: picks the
-    /// `hls + audio/mpeg` transcoding and exchanges its API URL for the signed
-    /// media URL. The returned URL is ephemeral and secret — never logged.
-    /// [`FetchError::NotStreamable`] when no such transcoding exists.
+    /// Resolves a track id to a **playable, signed media URL**: picks the best
+    /// mp3 transcoding (progressive preferred — it streams anonymously; HLS
+    /// `.m3u8` as fallback) and exchanges its API URL for the media URL. The
+    /// returned URL is ephemeral and secret — never logged.
+    /// [`FetchError::NotStreamable`] when no mp3 transcoding exists.
     async fn resolve_stream_url(&self, track_id: &str) -> Result<String, FetchError>;
 
     /// The signed-in user's liked tracks (at most `limit`). Requires OAuth.
@@ -385,7 +386,7 @@ impl Sc for ScApi {
             .get_json(&format!("{API_BASE}/tracks/{track_id}"), &[])
             .await?;
         let transcoding_url =
-            pick_hls_mp3(&dto.media.transcodings).ok_or(FetchError::NotStreamable)?;
+            pick_stream_url(&dto.media.transcodings).ok_or(FetchError::NotStreamable)?;
         let media: MediaUrl = self.get_json(&transcoding_url, &[]).await?;
         if media.url.is_empty() {
             return Err(FetchError::NotStreamable);
@@ -430,11 +431,23 @@ impl Sc for ScApi {
     }
 }
 
-/// Picks the `hls + audio/mpeg` transcoding's API url (mp3-HLS, D4).
-fn pick_hls_mp3(transcodings: &[TranscodingDto]) -> Option<String> {
+/// Picks the best playable mp3 transcoding's API url. **Prefers `progressive`
+/// over `hls`**: SoundCloud anonymously serves the progressive mp3 stream (a
+/// direct, range-streamable file the player decodes on the normal path), while
+/// the plain `hls` mp3 exchange currently 404s for anonymous clients (verified
+/// live 2026-07-24). HLS stays a fallback for the rare track that only offers
+/// it (then the player's `HlsStream` kicks in on the `.m3u8`). The
+/// `*-encrypted-hls` variants are AES/DRM and deliberately ignored.
+fn pick_stream_url(transcodings: &[TranscodingDto]) -> Option<String> {
+    let is_mp3 = |t: &&TranscodingDto| t.format.mime_type.starts_with("audio/mpeg");
     transcodings
         .iter()
-        .find(|t| t.format.protocol == "hls" && t.format.mime_type.starts_with("audio/mpeg"))
+        .find(|t| t.format.protocol == "progressive" && is_mp3(t))
+        .or_else(|| {
+            transcodings
+                .iter()
+                .find(|t| t.format.protocol == "hls" && is_mp3(t))
+        })
         .map(|t| t.url.clone())
 }
 
@@ -524,7 +537,7 @@ struct PublisherDto {
     album_title: Option<String>,
 }
 
-#[derive(Default, Deserialize)]
+#[derive(Default, Clone, Deserialize)]
 struct FormatDto {
     #[serde(default)]
     protocol: String,
@@ -532,7 +545,7 @@ struct FormatDto {
     mime_type: String,
 }
 
-#[derive(Default, Deserialize)]
+#[derive(Default, Clone, Deserialize)]
 struct TranscodingDto {
     #[serde(default)]
     url: String,
@@ -717,34 +730,39 @@ mod api_tests {
     }
 
     #[test]
-    fn picks_hls_mp3_transcoding() {
-        let transcodings = vec![
-            TranscodingDto {
+    fn prefers_progressive_then_hls_mp3() {
+        let opus_hls = TranscodingDto {
             url: "https://api/opus".into(),
             format: FormatDto {
                 protocol: "hls".into(),
                 mime_type: "audio/ogg; codecs=\"opus\"".into(),
             },
-            },
-            TranscodingDto {
-                url: "https://api/mp3".into(),
+        };
+        let mp3_hls = TranscodingDto {
+            url: "https://api/mp3-hls".into(),
             format: FormatDto {
                 protocol: "hls".into(),
                 mime_type: "audio/mpeg".into(),
             },
-            },
-            TranscodingDto {
-                url: "https://api/prog".into(),
+        };
+        let mp3_prog = TranscodingDto {
+            url: "https://api/mp3-prog".into(),
             format: FormatDto {
                 protocol: "progressive".into(),
                 mime_type: "audio/mpeg".into(),
             },
-            },
-        ];
+        };
+        // Progressive wins when present (it is what streams anonymously).
+        let all = vec![opus_hls, mp3_hls.clone(), mp3_prog];
         assert_eq!(
-            pick_hls_mp3(&transcodings),
-            Some("https://api/mp3".to_string())
+            pick_stream_url(&all),
+            Some("https://api/mp3-prog".to_string())
         );
-        assert_eq!(pick_hls_mp3(&[]), None);
+        // HLS mp3 is the fallback when there is no progressive.
+        assert_eq!(
+            pick_stream_url(&[mp3_hls]),
+            Some("https://api/mp3-hls".to_string())
+        );
+        assert_eq!(pick_stream_url(&[]), None);
     }
 }

soundclouddy/tests/live.rs

@@ -38,10 +38,10 @@ async fn live_scrape_search_and_stream() {
         api.cached_client_id().await.is_some()
     );
 
-    let tracks = api
-        .search_tracks("boards of canada", 5)
-        .await
-        .expect("search decodes");
+    // "lofi" returns freely-streamable (policy=MONETIZE) tracks; a label
+    // artist like Boards of Canada would be Go+/preview-only and 404 the media
+    // exchange (that restriction is expected, not a provider bug).
+    let tracks = api.search_tracks("lofi", 5).await.expect("search decodes");
     assert!(!tracks.is_empty(), "search returns tracks");
     let first = &tracks[0];
     println!(
@@ -49,16 +49,13 @@ async fn live_scrape_search_and_stream() {
         first.artist, first.title, first.id
     );
 
-    // Resolve a playable HLS media URL for it.
+    // Resolve a playable media URL (progressive mp3, or an HLS m3u8 fallback).
     let url = api
         .resolve_stream_url(&first.id)
         .await
         .expect("stream url resolves");
-    assert!(
-        url.contains(".m3u8"),
-        "stream URL is an HLS playlist: {url}"
-    );
-    println!("stream url resolves to an m3u8 (len {})", url.len());
+    assert!(url.starts_with("http"), "stream URL is absolute: {url}");
+    println!("stream url resolves (len {})", url.len());
 }
 
 #[tokio::test]