Duplicates, delete & storage, users page with J-ID avatars

Backend:
- Perceptual hashes (aHash/dHash/pHash/wHash via imagehash, no imgdd)
  stored on items, computed on upload/download and by the new
  compute_visual_hashes command
- Duplicates API: exact duplicates (multi-location items), visual matches
  for one item, union-find similarity groups with pagination
- Delete API with ownership/staff checks, per-item and per-copy deletion,
  watched-folder path validation; storage overview and temp cleanup;
  file list accepts j_ids batches
- Staged uploads are flagged visual_match with their library matches
  (threshold via VISUAL_MATCH_THRESHOLD)
- Staff users API: list with upload counts, set role and avatar by J-ID;
  User.avatar FK with signed avatar URLs
- Download threads close their DB connection and stale tasks are reaped,
  keeping behaviour Gunicorn-friendly

Frontend:
- /duplicates: exact duplicate groups with per-copy delete, visual
  similarity controls, search similar to a J-ID, paginated groups with
  selection, bulk delete and dismiss
- /delete: storage cards, delete by J-ID with preview grid, temp cleanup
- /users: staff directory with role selects and avatar J-ID inputs
- Nav + command palette entries; top-bar avatar; upload cards and the
  metadata modal show library visual matches

backend/apps/library/uploads.py

@@ -28,10 +28,40 @@ from . import services
 from .models import MediaItem, TempUpload
 from .permissions import CanUpload
 from .serializers import TempUploadSerializer
+from .tools import HASH_FIELDS, hashes_similarity
 
 logger = logging.getLogger(__name__)
 
 
+def find_library_matches(path, limit=10):
+    """Library items visually similar to a staged file."""
+    hashes = services.compute_visual_hashes(path)
+    if not hashes:
+        return []
+    algorithms = list(HASH_FIELDS)
+    threshold = settings.VISUAL_MATCH_THRESHOLD
+    matches = []
+    for item in MediaItem.objects.prefetch_related("locations"):
+        similarity = hashes_similarity(
+            hashes,
+            {field: getattr(item, field, "") for field in algorithms},
+            algorithms,
+            threshold,
+        )
+        if similarity is None:
+            continue
+        location = item.locations.first()
+        matches.append(
+            {
+                "j_id": f"J-{item.id}",
+                "filename": Path(location.rel_path).name if location else item.md5,
+                "similarity": round(similarity * 100, 1),
+            }
+        )
+    matches.sort(key=lambda entry: entry["similarity"], reverse=True)
+    return matches[:limit]
+
+
 def complete_temp_upload(temp, download_url=None):
     """Index the upload into the library.
 
@@ -57,6 +87,7 @@ def complete_temp_upload(temp, download_url=None):
             raise
         item, _, location, _ = services.index_file(destination, folder)
         services.rename_location_to_j_id(item, location)
+        services.ensure_visual_hashes(item)
         if temp.file:
             temp.file.delete(save=False)
     else:
@@ -69,6 +100,7 @@ def complete_temp_upload(temp, download_url=None):
             shutil.copyfileobj(source, target)
         item, _, location, _ = services.index_file(destination, folder)
         services.rename_location_to_j_id(item, location)
+        services.ensure_visual_hashes(item)
         temp.file.delete(save=False)
 
     temp.library_item = item
@@ -143,6 +175,11 @@ class TempUploadViewSet(
             temp.resolution = TempUpload.RESOLUTION_DUPLICATE
             temp.library_item = existing
             temp.file.delete(save=False)
+        else:
+            matches = find_library_matches(temp.file.path)
+            if matches:
+                temp.visual_matches = matches
+                temp.status = TempUpload.STATUS_VISUAL_MATCH
         temp.save()
         return Response(
             self.get_serializer(temp).data, status=status.HTTP_201_CREATED