fix(node): make Photos/Video/Music enrichment survive node restarts

Only thumbnail bytes were ever durable in media_cache.db — every other
derived field (photo width/height/EXIF, video ffprobe duration/dims,
audio cover art) lived solely on the in-memory GroupIndex entry, so a
node restart re-decoded every photo through Pillow, re-ran ffprobe on
every video, and re-scanned for every album cover from scratch, even
though the answers already sat in the cache.

Adds photo_meta and video_meta tables (content-only fields, keyed by
file_id) and checks them before doing the expensive work. Audio gets no
new table: mutagen reads tags and duration in one inseparable call, so
caching duration alone buys nothing — instead cover-art extraction alone
is skipped via a new skip_cover flag when a cached cover already exists.

Deliberately excluded from all three caches: anything derived from the
filename or folder path (video display_title/season/episode via guessit,
audio artist/album folder-fallback) — those must keep being recomputed
fresh so a rename/move is still correctly re-derived by the existing
_reenrich_renamed_*_entries mechanisms, instead of silently handing back
a stale parse under the new name/location.

Regression tests prove cache reuse by deleting the source file (or cover)
between two enrichment runs, and prove rename/move correctness survives
the new cache by renaming/moving to a path that never exists on disk.

1 files changed, 28 insertions, 0 deletions

diff --git a/packages/meshbay-node/tests/test_enrich_photo.py b/packages/meshbay-node/tests/test_enrich_photo.py
index e0c1b73..877a71d 100644
--- a/packages/meshbay-node/tests/test_enrich_photo.py
+++ b/packages/meshbay-node/tests/test_enrich_photo.py
@@ -62,6 +62,34 @@ async def test_enricher_populates_dimensions_and_stores_thumbnail(tmp_path, medi
 
 
 @pytest.mark.asyncio
+async def test_enricher_reuses_cached_meta_without_touching_the_file_again(tmp_path, media_cache):
+    """
+    The gap this closes: enrichment fields only ever lived in the in-memory
+    GroupIndex, so every daemon restart re-ran Pillow over every photo in
+    every configured root from scratch, even though media_cache.db (the
+    thumbnail bytes) already had the answer. A second PhotoEnricher sharing
+    the same media_cache — standing in for "the daemon restarted" — must
+    reuse it instead. Proven strongly: the source file is deleted between
+    the two runs, so a second real decode attempt would fail outright
+    rather than merely being redundant.
+    """
+    img = tmp_path / "reused.jpg"
+    _save_jpeg(img, size=(300, 200))
+    entry = IndexEntry(id="fileid_reuse", name=img.name, path=img.name,
+                        size=img.stat().st_size, type="image", added_at=0)
+
+    first_enricher = PhotoEnricher(media_cache)
+    _, first_fields = await _run(first_enricher, entry, img)
+    assert first_fields.get("thumb_hash")
+
+    img.unlink()  # a real second decode would now raise, not just be wasteful
+    second_enricher = PhotoEnricher(media_cache)
+    _, second_fields = await _run(second_enricher, entry, img)
+
+    assert second_fields == first_fields
+
+
+@pytest.mark.asyncio
 async def test_enricher_no_exif_degrades_gracefully(tmp_path, media_cache):
     """A screenshot or a re-saved image with no EXIF block at all is the
     ordinary case, not an error — must not raise and must leave taken_at/