YouTube enrichment: legacy channels auto-upgrade to InnerTube (backwards compat)

Channels enriched before InnerTube existed (partial coverage, e.g. CaseyNeistat
36, Good Mythical MORE 25) were locked behind the 24h coverage gate and would
never reach the full ~240-date catalog on their own. Tag each enrichment row
with the source ('innertube'|'fallback'); legacy rows (method NULL after the
additive migration) bypass the gate ONCE so they re-enrich and upgrade, then
settle under the normal window. Non-destructive: existing follows, wished videos
(24), and cached dates (1176) are untouched — only coverage refreshes on next
open/sweep. Migration is additive (method TEXT); regression test added.

core/video/youtube_enrichment.py

@@ -205,7 +205,9 @@ class YoutubeDateEnricher:
                         filled += 1
         self._channels_done += 1
         self._dates_total += len(dates) + filled
-        db.mark_channel_dates_enriched(cid, len(dates) + filled)
+        # Tag the source so legacy (pre-InnerTube) rows are recognisable and upgrade.
+        method = "innertube" if dates else "fallback"
+        db.mark_channel_dates_enriched(cid, len(dates) + filled, method=method)
         # Terse per-channel summary (like the worker's "Synced full episode list…").
         logger.info("Dated %d videos for %s (%d bulk + %d per-video)",
                     len(dates) + filled, self._current, len(dates), filled)

database/video_database.py

@@ -109,6 +109,9 @@ _COLUMN_MIGRATIONS = [
     ("video_wishlist", "source", "TEXT NOT NULL DEFAULT 'tmdb'"),
     ("video_wishlist", "source_id", "TEXT"),
     ("video_wishlist", "parent_source_id", "TEXT"),   # owning channel youtube id (video rows)
+    # which source produced a channel's dates — NULL on legacy (pre-InnerTube) rows
+    # so they re-enrich once and upgrade to the full InnerTube catalog.
+    ("youtube_channel_enrichment", "method", "TEXT"),
 ]
 
 
@@ -2039,17 +2042,19 @@ class VideoDatabase:
         finally:
             conn.close()
 
-    def mark_channel_dates_enriched(self, channel_id, date_count=0) -> None:
-        """Record that the background enricher swept this channel (skip re-sweeps)."""
+    def mark_channel_dates_enriched(self, channel_id, date_count=0, method="innertube") -> None:
+        """Record that the background enricher swept this channel (skip re-sweeps).
+        ``method`` tags which source produced the dates; legacy rows have NULL and
+        get re-enriched once so they upgrade to the InnerTube catalog."""
         if not channel_id:
             return
         conn = self._get_connection()
         try:
             conn.execute(
-                "INSERT INTO youtube_channel_enrichment (channel_id, enriched_at, date_count) "
-                "VALUES (?, CURRENT_TIMESTAMP, ?) ON CONFLICT(channel_id) DO UPDATE SET "
-                "enriched_at=CURRENT_TIMESTAMP, date_count=excluded.date_count",
-                (str(channel_id), int(date_count or 0)))
+                "INSERT INTO youtube_channel_enrichment (channel_id, enriched_at, date_count, method) "
+                "VALUES (?, CURRENT_TIMESTAMP, ?, ?) ON CONFLICT(channel_id) DO UPDATE SET "
+                "enriched_at=CURRENT_TIMESTAMP, date_count=excluded.date_count, method=excluded.method",
+                (str(channel_id), int(date_count or 0), method or None))
             conn.commit()
         finally:
             conn.close()
@@ -2064,10 +2069,14 @@ class VideoDatabase:
         conn = self._get_connection()
         try:
             row = conn.execute(
-                "SELECT enriched_at, date_count FROM youtube_channel_enrichment WHERE channel_id=?",
+                "SELECT enriched_at, date_count, method FROM youtube_channel_enrichment WHERE channel_id=?",
                 (str(channel_id),)).fetchone()
             if not row or not row["enriched_at"]:
                 return False
+            # Legacy rows (enriched before InnerTube, method NULL) → always re-enrich
+            # once so they upgrade to the full catalog, then settle under the window.
+            if not row["method"]:
+                return False
             try:
                 when = datetime.strptime(row["enriched_at"], "%Y-%m-%d %H:%M:%S")
             except (ValueError, TypeError):

database/video_schema.sql

@@ -276,7 +276,8 @@ CREATE TABLE IF NOT EXISTS youtube_video_dates (
 CREATE TABLE IF NOT EXISTS youtube_channel_enrichment (
     channel_id  TEXT PRIMARY KEY,
     enriched_at TEXT NOT NULL DEFAULT CURRENT_TIMESTAMP,
-    date_count  INTEGER NOT NULL DEFAULT 0
+    date_count  INTEGER NOT NULL DEFAULT 0,
+    method      TEXT                        -- 'innertube' | 'fallback'; NULL = legacy → re-enrich once
 );
 
 -- ── Owned media files (the Library = content that has a file) ────────────────

tests/test_video_database.py

@@ -1182,3 +1182,13 @@ def test_channel_enrichment_tracking(db):
     db.mark_channel_dates_enriched("UCx", date_count=50)
     assert db.channel_dates_enriched_recently("UCx", within_hours=24) is True
     assert db.channel_dates_enriched_recently("UCx", within_hours=0) is False   # window respected
+
+
+def test_legacy_enrichment_rows_upgrade(db):
+    # A pre-InnerTube row (method NULL) with otherwise-good coverage must NOT be
+    # locked for 24h — it re-enriches once so it upgrades to the full catalog.
+    db.mark_channel_dates_enriched("UCleg", date_count=50, method=None)
+    assert db.channel_dates_enriched_recently("UCleg") is False
+    # After the InnerTube re-run, the normal window applies again (no churn).
+    db.mark_channel_dates_enriched("UCleg", date_count=240, method="innertube")
+    assert db.channel_dates_enriched_recently("UCleg") is True