refac

backend/open_webui/utils/access_control/files.py

@@ -87,3 +87,38 @@ async def has_access_to_file(
                 return True
 
     return False
+
+
+async def get_accessible_folder_files(
+    entries: list[dict] | None,
+    user: UserModel,
+    db: AsyncSession | None = None,
+) -> list[dict]:
+    """Filter folder.data['files'] entries to those the caller can read.
+
+    Each entry is expected to have 'type' ('file' or 'collection') and 'id'.
+    Admins bypass all checks. Unknown types are kept as-is.
+    """
+    if not entries:
+        return []
+    if user.role == 'admin':
+        return list(entries)
+
+    accessible: list[dict] = []
+    for entry in entries:
+        if not isinstance(entry, dict):
+            continue
+        entry_type = entry.get('type')
+        entry_id = entry.get('id')
+        if not entry_id:
+            accessible.append(entry)
+            continue
+        if entry_type == 'file':
+            if await has_access_to_file(entry_id, 'read', user, db=db):
+                accessible.append(entry)
+        elif entry_type == 'collection':
+            if await Knowledges.check_access_by_user_id(entry_id, user.id, 'read', db=db):
+                accessible.append(entry)
+        else:
+            accessible.append(entry)
+    return accessible

backend/open_webui/utils/middleware.py

@@ -106,6 +106,7 @@ from open_webui.utils.tools import (
     get_terminal_tools,
 )
 from open_webui.utils.access_control import has_connection_access
+from open_webui.utils.access_control.files import get_accessible_folder_files
 from open_webui.utils.plugin import load_function_module_by_id
 from open_webui.utils.filter import (
     get_sorted_filter_ids,
@@ -2407,15 +2408,17 @@ async def process_chat_payload(request, form_data, user, metadata, model):
             if 'system_prompt' in folder.data:
                 form_data = await apply_system_prompt_to_body(folder.data['system_prompt'], form_data, metadata, user)
             if 'files' in folder.data:
+                # Defensive: filter to entries the caller can still read.
+                allowed_files = await get_accessible_folder_files(folder.data['files'], user)
                 if metadata.get('params', {}).get('function_calling') != 'native':
                     form_data['files'] = [
-                        *folder.data['files'],
+                        *allowed_files,
                         *form_data.get('files', []),
                     ]
                 else:
                     # Native FC: skip RAG injection, builtin tools
                     # will read folder knowledge from metadata.
-                    metadata['folder_knowledge'] = folder.data['files']
+                    metadata['folder_knowledge'] = allowed_files
 
     # Model "Knowledge" handling
     user_message = get_last_user_message(form_data['messages'])
@@ -2615,7 +2618,7 @@ async def process_chat_payload(request, form_data, user, metadata, model):
                     folder = await Folders.get_folder_by_id_and_user_id(folder_id, user.id)
                     if folder and folder.data and 'files' in folder.data:
                         files = [f for f in files if f.get('id', None) != folder_id]
-                        files = [*files, *folder.data['files']]
+                        files = [*files, *await get_accessible_folder_files(folder.data['files'], user)]
 
         # files = [*files, *[{"type": "url", "url": url, "name": url} for url in urls]]
         # Remove duplicate files based on their content