fix(backfill): don't skip the tail of the recent/deep boundary page

When recent backfill stopped partway through a page (age cutoff or count cap),
it stored the *next* page token as the deep-backfill cursor, so the older items
remaining on that same page were collected by neither pass — a channel could
report backfill_done while silently missing a band of videos right around the
365-day cutoff. Now resume deep from the page we stopped on (re-fetched; inserts
are idempotent). Verified: Pánczél went 200 -> 209 stored (= full uploads
playlist) after re-backfill.

backend/app/sync/videos.py

@@ -107,25 +107,34 @@ def backfill_channel_recent(db: Session, yt: YouTubeClient, channel: Channel) ->
     next_cursor: str | None = None
     done = False
 
-    while len(collected) < settings.backfill_recent_max_videos:
+    while True:
+        page_start = page_token  # token that fetched THIS page (None = first page)
         data = yt.get_playlist_items(channel.uploads_playlist_id, page_token)
-        reached_cutoff = False
+        stopped_mid_page = False
         for item in data.get("items", []):
             stub = _stub_from_playlist_item(item, channel.id)
             if stub is None:
                 continue
             if stub["published_at"] and stub["published_at"] < cutoff:
-                reached_cutoff = True
+                stopped_mid_page = True
                 break
             collected.append(stub)
             if len(collected) >= settings.backfill_recent_max_videos:
+                stopped_mid_page = True
                 break
         page_token = data.get("nextPageToken")
-        if reached_cutoff or not page_token:
-            next_cursor = page_token
-            done = not page_token
+        if stopped_mid_page:
+            # We stopped partway through this page (age cutoff or count cap), so the older
+            # items still on it weren't collected. Resume the deep backfill FROM THIS page
+            # (re-fetch it; inserts are idempotent) so that tail isn't silently skipped.
+            next_cursor = page_start
+            done = False
             break
-        next_cursor = page_token
+        if not page_token:
+            next_cursor = None
+            done = True
+            break
+        next_cursor = page_token  # whole page consumed, more pages remain
 
     channel.recent_synced_at = _now()
     channel.backfill_cursor = next_cursor