diff options
Diffstat (limited to 'packages/meshbay-node/src/meshbay_node/indexer/enrich_photo.py')
| -rw-r--r-- | packages/meshbay-node/src/meshbay_node/indexer/enrich_photo.py | 17 |
1 files changed, 17 insertions, 0 deletions
diff --git a/packages/meshbay-node/src/meshbay_node/indexer/enrich_photo.py b/packages/meshbay-node/src/meshbay_node/indexer/enrich_photo.py index 44f9ebb..93a68cc 100644 --- a/packages/meshbay-node/src/meshbay_node/indexer/enrich_photo.py +++ b/packages/meshbay-node/src/meshbay_node/indexer/enrich_photo.py @@ -143,6 +143,22 @@ class PhotoEnricher: on_done: Callable[[str, dict], Awaitable[None]], ) -> None: async with self._sem: + # entry.id is the file's own content hash — the same bytes + # produce the same thumbnail, so a hit here means this exact + # content was already decoded, resized and EXIF-read at some + # point (this run, an earlier one, even a previous daemon + # process — media_cache.db is the durable half of this). + # Without this check, every restart re-ran Pillow over every + # image in every configured photo_root from scratch — the + # in-memory GroupIndex enrichment fields don't survive a + # restart, but this cache does, and nothing was reading it + # before spawning the expensive work. + cached_hash = await self._media_cache.get_thumb_hash_by_file_id(entry.id) + cached_meta = await self._media_cache.get_photo_meta(entry.id) if cached_hash else None + if cached_hash and cached_meta: + await on_done(entry.id, {**cached_meta, "thumb_hash": cached_hash}) + return + fields: dict = {} try: thumb, width, height, taken_at, camera = await asyncio.wait_for( @@ -153,6 +169,7 @@ class PhotoEnricher: fields["camera"] = camera thumb_hash = blake3.blake3(thumb).hexdigest() await self._media_cache.put_thumb(thumb_hash, entry.id, thumb) + await self._media_cache.put_photo_meta(entry.id, width, height, taken_at, camera) fields["thumb_hash"] = thumb_hash except Exception as e: log.warning("Photo enrichment failed for %s: %s", file_path, e) |