Playlist count: show the TRUE total + load ~200 (was capped at ~100 via yt-dlp)

A big playlist read '97' because yt-dlp flat caps playlist extraction at ~100 and
its playlist_count is often unset, so we showed the FETCHED count. Now the detail
endpoint overlays the InnerTube playlist browse (browseId VL<id>): curator-ordered
videos up to ~200 (vs yt-dlp's ~100) AND the real count from the header's
numVideosText (e.g. '512 videos'). The billboard shows the true total; when the
list is partial it says 'Showing the first N of TOTAL videos (YouTube limits
playlist browsing).' Live-validated (Veritasium uploads: total 512, 200 loaded).
103 tests green.

Honest cap: YouTube's browse only exposes ~200 of a playlist (page 2 returns no
continuation), so very large playlists show the right COUNT but list ~200.

api/video/youtube.py

@@ -378,10 +378,19 @@ def register_routes(bp):
             limit = 200
         try:
             db = get_video_db()
-            pl = yt.resolve_playlist("https://www.youtube.com/playlist?list=" + playlist_id,
-                                     limit=max(1, min(500, limit)))
+            pl = yt.resolve_playlist("https://www.youtube.com/playlist?list=" + playlist_id, limit=60)
             if not pl:
                 return jsonify({"success": False, "error": "Playlist not found"}), 404
+            # yt-dlp flat caps playlists at ~100 and its count is often unset. Overlay
+            # the InnerTube catalog (curator order, up to ~200) + the TRUE header count.
+            try:
+                cat = yt.innertube_playlist_catalog(playlist_id)
+                if cat.get("videos"):
+                    pl["videos"] = cat["videos"]
+                if cat.get("total"):
+                    pl["video_count"] = cat["total"]
+            except Exception:
+                pass
             vids = pl.get("videos") or []
             ids = [v.get("youtube_id") for v in vids]
             wished = db.youtube_video_wish_state(ids)

core/video/youtube.py

@@ -614,6 +614,70 @@ def innertube_channel_videos_page(channel_id, continuation=None, now=None, post=
     return {"videos": videos, "continuation": innertube_continuation(j)}
 
 
+def _innertube_playlist_total(obj):
+    """The playlist's TRUE video count from its browse header ('512 videos'), or None.
+    Reliable even when yt-dlp's playlist_count isn't populated."""
+    for t in _json_find_all(obj, "numVideosText", []):
+        m = re.search(r"([\d,]+)", "".join(_json_all_strings(t, [])))
+        if m:
+            try:
+                return int(m.group(1).replace(",", ""))
+            except ValueError:
+                pass
+    return None
+
+
+def innertube_playlist_page(playlist_id, continuation=None, now=None, post=None):
+    """One InnerTube page of a PLAYLIST's videos (browseId VL<id>) in the curator's
+    order: {"videos": [...], "continuation": token|None, "total": N|None}. Fixes
+    yt-dlp flat's ~100-item cap (browse pages to ~200) and exposes the real count.
+    Same item shape as innertube_channel_videos_page."""
+    pid = str(playlist_id or "").strip()
+    if not continuation and not pid:
+        return {"videos": [], "continuation": None, "total": None}
+    if now is None:
+        now = datetime.now(timezone.utc).date()
+    payload = ({"context": _INNERTUBE_CTX, "continuation": continuation} if continuation
+               else {"context": _INNERTUBE_CTX, "browseId": "VL" + pid})
+    try:
+        j = _innertube_post(payload, post)
+    except Exception:
+        return {"videos": [], "continuation": None, "total": None}
+    if not isinstance(j, dict):
+        return {"videos": [], "continuation": None, "total": None}
+    videos = []
+    for it in innertube_parse_video_items(j):
+        it["published_at"] = relative_to_date(it.pop("relative", None), now)
+        videos.append(it)
+    return {"videos": videos, "continuation": innertube_continuation(j), "total": _innertube_playlist_total(j)}
+
+
+def innertube_playlist_catalog(playlist_id, pages=40, now=None, post=None):
+    """A playlist's video list (curator order) + TRUE total via InnerTube:
+    {"videos": [...], "total": N|None}. yt-dlp flat caps at ~100; the browse pages
+    to ~200 and the header carries the real count. {} on failure."""
+    pid = str(playlist_id or "").strip()
+    if not pid:
+        return {"videos": [], "total": None}
+    if now is None:
+        now = datetime.now(timezone.utc).date()
+    out, seen, token, total = [], set(), None, None
+    for _ in range(max(1, pages)):
+        page = innertube_playlist_page(pid, continuation=token, now=now, post=post)
+        if total is None:
+            total = page.get("total")
+        for v in page.get("videos") or []:
+            if v.get("youtube_id") and v["youtube_id"] not in seen:
+                seen.add(v["youtube_id"])
+                out.append(v)
+        token = page.get("continuation")
+        if not token:
+            break
+        if post is None:
+            time.sleep(_INNERTUBE_DELAY)
+    return {"videos": out, "total": total}
+
+
 def innertube_channel_catalog(channel_id, pages=_INNERTUBE_PAGES, now=None, post=None):
     """A channel's video catalog (list + approximate dates) via InnerTube, up to
     ``pages`` pages: [{youtube_id, title, thumbnail_url, published_at}]. Like

tests/test_video_api.py

@@ -698,6 +698,7 @@ def test_youtube_playlist_follow_detail_and_watchlist(tmp_path, monkeypatch):
            "thumbnail_url": "t", "videos": [{"youtube_id": "a", "title": "A"}, {"youtube_id": "b", "title": "B"}]}
     monkeypatch.setattr(ytmod, "parse_playlist_id", lambda u: "PLx" if "list=" in (u or "") else None)
     monkeypatch.setattr(ytmod, "resolve_playlist", lambda *a, **k: dict(_PL))
+    monkeypatch.setattr(ytmod, "innertube_playlist_catalog", lambda *a, **k: {"videos": [], "total": None})
     # resolve detects a playlist link → returns a playlist (not a channel)
     r = client.get("/api/video/youtube/resolve?url=https://youtube.com/playlist?list=PLx").get_json()
     assert r["success"] and r["playlist"]["playlist_id"] == "PLx" and r["following"] is False
@@ -778,6 +779,7 @@ def test_youtube_playlists_and_playlist_videos(tmp_path, monkeypatch):
                         lambda *a, **k: {"playlist_id": "PL1", "title": "Trailers",
                                          "videos": [{"youtube_id": "a", "title": "A"},
                                                     {"youtube_id": "b", "title": "B"}]})
+    monkeypatch.setattr(ytmod, "innertube_playlist_catalog", lambda *a, **k: {"videos": [], "total": None})
     # 'a' is wished → should hydrate wished=True
     videoapi._video_db.add_videos_to_wishlist({"youtube_id": "UCx", "title": "X"}, [{"youtube_id": "a", "title": "A"}])
     pv = client.get("/api/video/youtube/playlist/PL1").get_json()

tests/test_video_youtube.py

@@ -561,3 +561,22 @@ def test_innertube_channel_catalog_pages_until_no_token():
     assert cat[0]["title"] == "Title v1" and cat[2]["published_at"] == "2026-06-12"
     assert seq["n"] == 2                                                 # stopped when the token ran out
     assert yt.innertube_channel_catalog("not-uc", post=lambda p: None) == []
+
+
+def test_innertube_playlist_page_keeps_order_and_true_count():
+    now = date(2026, 6, 17)
+    page = _page([_lk_item("a", "1 day ago"), _lk_item("b", "2 days ago")], token="TOK")
+    page["header"] = {"numVideosText": {"runs": [{"text": "512"}, {"text": " videos"}]}}   # the real total
+    got = yt.innertube_playlist_page("PLx", now=now, post=lambda p: page)
+    assert [v["youtube_id"] for v in got["videos"]] == ["a", "b"]        # curator order preserved
+    assert got["total"] == 512 and got["continuation"] == "TOK"
+    # catalog pages until no token, surfacing the header count
+    pages = [page, _page([_lk_item("c", "3 days ago")])]   # 2nd page: no token → stop
+    seq = {"n": 0}
+
+    def post(p):
+        i = seq["n"]; seq["n"] += 1; return pages[i]
+
+    cat = yt.innertube_playlist_catalog("PLx", now=now, post=post)
+    assert [v["youtube_id"] for v in cat["videos"]] == ["a", "b", "c"] and cat["total"] == 512
+    assert yt.innertube_playlist_catalog("", post=lambda p: None) == {"videos": [], "total": None}

webui/static/video/video-detail.js

@@ -1570,10 +1570,14 @@
         var vids = pl.videos || [];
         ytVideoMap = {};
         vids.forEach(function (v) { ytVideoMap[v.youtube_id] = v; });
+        var total = pl.video_count || vids.length;
+        // YouTube only exposes ~200 of a playlist via browsing — be honest when partial.
+        var note = total > vids.length
+            ? 'Showing the first ' + vids.length + ' of ' + total + ' videos (YouTube limits playlist browsing).' : '';
         var season = { season_number: 1, title: 'Videos', poster_url: ytProx(pl.thumbnail_url),
             episode_owned: 0, episode_total: vids.length, episodes: vids.map(ytEpisodeOf) };
         return { kind: 'playlist', source: 'youtube', id: pl.playlist_id, title: pl.title || 'Playlist',
-            overview: '', poster_url: ytProx(pl.thumbnail_url), has_poster: !!pl.thumbnail_url,
+            overview: note, poster_url: ytProx(pl.thumbnail_url), has_poster: !!pl.thumbnail_url,
             backdrop_url: ytProx(pl.thumbnail_url), has_backdrop: !!pl.thumbnail_url,
             genres: pl.channel_title ? [pl.channel_title] : [], handle: null,
             subscriber_count: null, view_count: null, video_count: pl.video_count,