Enhancement (beta): Switch the AI vector store to LanceDB (#12944)

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: shamoon <shamoon@users.noreply.github.com>

src/paperless/settings/__init__.py

@@ -97,8 +97,7 @@ MODEL_FILE = get_path_from_env(
     DATA_DIR / "classification_model.pickle",
 )
 LLM_INDEX_DIR = DATA_DIR / "llm_index"
-LLM_INDEX_LOCK = DATA_DIR / "locks" / "llm_index.lock"
-(DATA_DIR / "locks").mkdir(parents=True, exist_ok=True)
+LLM_INDEX_LOCK = LLM_INDEX_DIR / "index.lock"
 
 LOGGING_DIR = get_path_from_env("PAPERLESS_LOGGING_DIR", DATA_DIR / "log")
 
@@ -644,6 +643,7 @@ LOGGING = {
         "kombu": {"handlers": ["file_celery"], "level": "DEBUG"},
         "_granian": {"handlers": ["file_paperless"], "level": "DEBUG"},
         "granian.access": {"handlers": ["file_paperless"], "level": "DEBUG"},
+        "httpx": {"level": "WARNING"},
     },
 }
 

src/paperless/views.py

@@ -49,7 +49,7 @@ from paperless.serialisers import GroupSerializer
 from paperless.serialisers import PaperlessAuthTokenSerializer
 from paperless.serialisers import ProfileSerializer
 from paperless.serialisers import UserSerializer
-from paperless_ai.indexing import vector_store_file_exists
+from paperless_ai.indexing import llm_index_exists
 
 
 class PaperlessObtainAuthTokenView(ObtainAuthToken):
@@ -467,7 +467,7 @@ class ApplicationConfigurationViewSet(ModelViewSet[ApplicationConfiguration]):
             or old_llm_context_size != new_llm_context_size
         )
         rebuild_needed = new_ai_index_enabled and (
-            not vector_store_file_exists() or embedding_config_changed
+            not llm_index_exists() or embedding_config_changed
         )
 
         if rebuild_needed: