OpenRAG 0.5.0.dev30__py3-none-any.whl → 0.5.0.dev32__py3-none-any.whl

api/settings.py

@@ -1940,6 +1940,10 @@ async def rollback_onboarding(
         current_config.knowledge.embedding_model = ""
         current_config.onboarding.openrag_docs_ingested_version = None
         current_config.onboarding.openrag_docs_remote_signature = None
+        current_config.onboarding.assistant_message = None
+        current_config.onboarding.selected_nudge = None
+        current_config.onboarding.card_steps = None
+        current_config.onboarding.upload_steps = None
 
         embedding_only = body.embedding_only if body else False
 

connectors/google_drive/connector.py

@@ -3,9 +3,10 @@ import io
 import os
 import time
 from collections import deque
+from collections.abc import Iterable
 from dataclasses import dataclass
 from pathlib import Path
-from typing import Any, Dict, Iterable, List, Optional, Set
+from typing import Any
 
 from googleapiclient.errors import HttpError
 from googleapiclient.http import MediaIoBaseDownload
@@ -17,6 +18,7 @@ from .oauth import GoogleDriveOAuth
 
 logger = get_logger(__name__)
 
+
 # -------------------------
 # Config model
 # -------------------------
@@ -27,26 +29,26 @@ class GoogleDriveConfig:
     token_file: str
 
     # Selective sync
-    file_ids: Optional[List[str]] = None
-    folder_ids: Optional[List[str]] = None
+    file_ids: list[str] | None = None
+    folder_ids: list[str] | None = None
     recursive: bool = True
 
     # Shared Drives control
-    drive_id: Optional[str] = None  # when set, we use corpora='drive'
-    corpora: Optional[str] = None  # 'user' | 'drive' | 'domain'; auto-picked if None
+    drive_id: str | None = None  # when set, we use corpora='drive'
+    corpora: str | None = None  # 'user' | 'drive' | 'domain'; auto-picked if None
 
     # Optional filtering
-    include_mime_types: Optional[List[str]] = None
-    exclude_mime_types: Optional[List[str]] = None
+    include_mime_types: list[str] | None = None
+    exclude_mime_types: list[str] | None = None
 
     # Export overrides for Google-native types
-    export_format_overrides: Optional[Dict[str, str]] = None  # mime -> export-mime
+    export_format_overrides: dict[str, str] | None = None  # mime -> export-mime
 
     # Changes API state persistence (store these in your DB/kv if needed)
-    changes_page_token: Optional[str] = None
+    changes_page_token: str | None = None
 
     # Optional: resource_id for webhook cleanup
-    resource_id: Optional[str] = None
+    resource_id: str | None = None
 
 
 # -------------------------
@@ -89,7 +91,7 @@ class GoogleDriveConnector(BaseConnector):
         # Otherwise, implement your custom logic here.
         logger.debug(f"Emitting document: {doc.id} ({doc.filename})")
 
-    def __init__(self, config: Dict[str, Any]) -> None:
+    def __init__(self, config: dict[str, Any]) -> None:
         # Read from config OR env (backend env, not NEXT_PUBLIC_*):
         env_client_id = os.getenv(self.CLIENT_ID_ENV_VAR)
         env_client_secret = os.getenv(self.CLIENT_SECRET_ENV_VAR)
@@ -99,6 +101,7 @@ class GoogleDriveConnector(BaseConnector):
 
         # Token file default - use data directory for persistence
         from config.paths import get_data_file
+
         token_file = config.get("token_file") or get_data_file("google_drive_token.json")
         Path(token_file).parent.mkdir(parents=True, exist_ok=True)
 
@@ -114,9 +117,7 @@ class GoogleDriveConnector(BaseConnector):
             )
 
         # Normalize incoming IDs from any of the supported alias keys
-        def _first_present_list(
-            cfg: Dict[str, Any], keys: Iterable[str]
-        ) -> Optional[List[str]]:
+        def _first_present_list(cfg: dict[str, Any], keys: Iterable[str]) -> list[str] | None:
             for k in keys:
                 v = cfg.get(k)
                 if v:  # accept non-empty list
@@ -153,11 +154,15 @@ class GoogleDriveConnector(BaseConnector):
         # Drive client is built in authenticate()
         from google.oauth2.credentials import Credentials
 
-        self.creds: Optional[Credentials] = None
+        self.creds: Credentials | None = None
         self.service: Any = None
 
         # cache of resolved shortcutId -> target file metadata
-        self._shortcut_cache: Dict[str, Dict[str, Any]] = {}
+        self._shortcut_cache: dict[str, dict[str, Any]] = {}
+
+        import threading
+
+        self._lock = threading.Lock()
 
         # Authentication state
         self._authenticated: bool = False
@@ -170,20 +175,20 @@ class GoogleDriveConnector(BaseConnector):
         self._shortcut_cache.clear()
 
     @property
-    def _drives_get_flags(self) -> Dict[str, Any]:
+    def _drives_get_flags(self) -> dict[str, Any]:
         """
         Flags valid for GET-like calls (files.get, changes.getStartPageToken).
         """
         return {"supportsAllDrives": True}
 
     @property
-    def _drives_list_flags(self) -> Dict[str, Any]:
+    def _drives_list_flags(self) -> dict[str, Any]:
         """
         Flags valid for LIST-like calls (files.list, changes.list).
         """
         return {"supportsAllDrives": True, "includeItemsFromAllDrives": True}
 
-    def _pick_corpora_args(self) -> Dict[str, Any]:
+    def _pick_corpora_args(self) -> dict[str, Any]:
         """
         Decide corpora/driveId based on config.
 
@@ -197,7 +202,7 @@ class GoogleDriveConnector(BaseConnector):
         # Default to allDrives so Picker selections from Shared Drives work without explicit drive_id
         return {"corpora": "allDrives"}
 
-    def _resolve_shortcut(self, file_obj: Dict[str, Any]) -> Dict[str, Any]:
+    def _resolve_shortcut(self, file_obj: dict[str, Any]) -> dict[str, Any]:
         """
         If a file is a shortcut, fetch and return the real target metadata.
         """
@@ -234,7 +239,7 @@ class GoogleDriveConnector(BaseConnector):
             # shortcut target not accessible
             return file_obj
 
-    def _list_children(self, folder_id: str) -> List[Dict[str, Any]]:
+    def _list_children(self, folder_id: str) -> list[dict[str, Any]]:
         """
         List immediate children of a folder.
         """
@@ -245,7 +250,7 @@ class GoogleDriveConnector(BaseConnector):
 
         query = f"'{folder_id}' in parents and trashed = false"
         page_token = None
-        results: List[Dict[str, Any]] = []
+        results: list[dict[str, Any]] = []
 
         while True:
             resp = (
@@ -272,13 +277,13 @@ class GoogleDriveConnector(BaseConnector):
 
         return results
 
-    def _bfs_expand_folders(self, folder_ids: Iterable[str]) -> List[Dict[str, Any]]:
+    def _bfs_expand_folders(self, folder_ids: Iterable[str]) -> list[dict[str, Any]]:
         """
         Breadth-first traversal to expand folders to all descendant files (if recursive),
         or just immediate children (if not recursive). Folders themselves are returned
         as items too, but filtered later.
         """
-        out: List[Dict[str, Any]] = []
+        out: list[dict[str, Any]] = []
         queue = deque(folder_ids)
 
         while queue:
@@ -295,7 +300,7 @@ class GoogleDriveConnector(BaseConnector):
 
         return out
 
-    def _get_file_meta_by_id(self, file_id: str) -> Optional[Dict[str, Any]]:
+    def _get_file_meta_by_id(self, file_id: str) -> dict[str, Any] | None:
         """
         Fetch metadata for a file by ID (resolving shortcuts).
         """
@@ -304,30 +309,31 @@ class GoogleDriveConnector(BaseConnector):
                 "Google Drive service is not initialized. Please authenticate first."
             )
         try:
-            meta = (
-                self.service.files()
-                .get(
-                    fileId=file_id,
-                    fields=(
-                        "id, name, mimeType, modifiedTime, createdTime, size, "
-                        "webViewLink, parents, shortcutDetails, driveId"
-                    ),
-                    **self._drives_get_flags,
+            with self._lock:
+                meta = (
+                    self.service.files()
+                    .get(
+                        fileId=file_id,
+                        fields=(
+                            "id, name, mimeType, modifiedTime, createdTime, size, "
+                            "webViewLink, parents, shortcutDetails, driveId"
+                        ),
+                        **self._drives_get_flags,
+                    )
+                    .execute()
                 )
-                .execute()
-            )
             return self._resolve_shortcut(meta)
         except HttpError:
             return None
 
-    def _filter_by_mime(self, items: Iterable[Dict[str, Any]]) -> List[Dict[str, Any]]:
+    def _filter_by_mime(self, items: Iterable[dict[str, Any]]) -> list[dict[str, Any]]:
         """
         Apply include/exclude mime filters if configured.
         """
         include = set(self.cfg.include_mime_types or [])
         exclude = set(self.cfg.exclude_mime_types or [])
 
-        def keep(m: Dict[str, Any]) -> bool:
+        def keep(m: dict[str, Any]) -> bool:
             mt = m.get("mimeType")
             if exclude and mt in exclude:
                 return False
@@ -337,7 +343,7 @@ class GoogleDriveConnector(BaseConnector):
 
         return [m for m in items if keep(m)]
 
-    def _iter_selected_items(self) -> List[Dict[str, Any]]:
+    def _iter_selected_items(self) -> list[dict[str, Any]]:
         """
         Return a de-duplicated list of file metadata for the selected scope:
           - explicit file_ids (automatically expands folders to their contents)
@@ -351,16 +357,21 @@ class GoogleDriveConnector(BaseConnector):
         )
         self._clear_shortcut_cache()
 
-        seen: Set[str] = set()
-        items: List[Dict[str, Any]] = []
-        folders_to_expand: List[str] = []
+        seen: set[str] = set()
+        items: list[dict[str, Any]] = []
+        folders_to_expand: list[str] = []
 
         if self.cfg.file_ids:
-            logger.debug("[GoogleDrive] _iter_selected_items: processing %d file_id(s)", len(self.cfg.file_ids))
+            logger.debug(
+                "[GoogleDrive] _iter_selected_items: processing %d file_id(s)",
+                len(self.cfg.file_ids),
+            )
             for fid in self.cfg.file_ids:
                 meta = self._get_file_meta_by_id(fid)
                 if not meta:
-                    logger.debug("[GoogleDrive] _iter_selected_items: no metadata for file_id=%s", fid)
+                    logger.debug(
+                        "[GoogleDrive] _iter_selected_items: no metadata for file_id=%s", fid
+                    )
                     continue
 
                 if meta.get("mimeType") == "application/vnd.google-apps.folder":
@@ -378,7 +389,9 @@ class GoogleDriveConnector(BaseConnector):
             folders_to_expand.extend(self.cfg.folder_ids)
 
         if folders_to_expand:
-            logger.debug("[GoogleDrive] _iter_selected_items: expanding %d folder(s)", len(folders_to_expand))
+            logger.debug(
+                "[GoogleDrive] _iter_selected_items: expanding %d folder(s)", len(folders_to_expand)
+            )
             folder_children = self._bfs_expand_folders(folders_to_expand)
             for meta in folder_children:
                 meta = self._resolve_shortcut(meta)
@@ -394,11 +407,7 @@ class GoogleDriveConnector(BaseConnector):
             return []
 
         items = self._filter_by_mime(items)
-        items = [
-            m
-            for m in items
-            if m.get("mimeType") != "application/vnd.google-apps.folder"
-        ]
+        items = [m for m in items if m.get("mimeType") != "application/vnd.google-apps.folder"]
 
         if not items and (self.cfg.file_ids or self.cfg.folder_ids):
             logger.warning(
@@ -414,7 +423,7 @@ class GoogleDriveConnector(BaseConnector):
     # -------------------------
     # Download logic
     # -------------------------
-    def _pick_export_mime(self, source_mime: str) -> Optional[str]:
+    def _pick_export_mime(self, source_mime: str) -> str | None:
         """
         Choose export mime for Google-native docs if needed.
         """
@@ -437,7 +446,7 @@ class GoogleDriveConnector(BaseConnector):
         # Return None for non-Google-native or unsupported types
         return overrides.get(source_mime)
 
-    def _download_file_bytes(self, file_meta: Dict[str, Any]) -> bytes:
+    def _download_file_bytes(self, file_meta: dict[str, Any]) -> bytes:
         """
         Download bytes for a given file (exporting if Google-native).
         Raises ValueError if the item is a folder (folders cannot be downloaded).
@@ -478,57 +487,56 @@ class GoogleDriveConnector(BaseConnector):
             "application/vnd.google-apps.drawing",  # Google Drawings
         }
 
-        if mime_type in exportable_types:
-            # This is an exportable Google Workspace file - must use export_media
-            export_mime = self._pick_export_mime(mime_type)
-            if not export_mime:
-                # Default fallback for unsupported Google native types
-                export_mime = "application/pdf"
+        with self._lock:
+            if mime_type in exportable_types:
+                # This is an exportable Google Workspace file - must use export_media
+                export_mime = self._pick_export_mime(mime_type)
+                if not export_mime:
+                    # Default fallback for unsupported Google native types
+                    export_mime = "application/pdf"
 
-            logger.debug(
-                "[GoogleDrive] _download_file_bytes: using export_media (%s -> %s)",
-                mime_type,
-                export_mime,
-            )
-            # NOTE: export_media does not accept supportsAllDrives/includeItemsFromAllDrives
-            request = self.service.files().export_media(
-                fileId=file_id, mimeType=export_mime
-            )
-        else:
-            # This is a regular uploaded file (PDF, image, video, etc.) - use get_media
-            # Also handles non-exportable Google Apps files (Forms, Sites, Maps, etc.)
-            logger.debug("[GoogleDrive] _download_file_bytes: using get_media (%s)", mime_type)
-            # Binary download (get_media also doesn't accept the Drive flags)
-            request = self.service.files().get_media(fileId=file_id)
-
-        # Download the file with error handling for misclassified Google Docs
-        fh = io.BytesIO()
-        downloader = MediaIoBaseDownload(fh, request, chunksize=1024 * 1024)
-        done = False
-
-        try:
-            while not done:
-                status, done = downloader.next_chunk()
-                # Optional: you can log progress via status.progress()
-        except HttpError as e:
-            # If download fails with "fileNotDownloadable", it's a Docs Editor file
-            # that wasn't properly detected. Retry with export_media.
-            if "fileNotDownloadable" in str(e) and mime_type not in exportable_types:
-                logger.warning(
-                    f"Download failed for {file_id} ({mime_type}) with fileNotDownloadable error. "
-                    f"Retrying with export_media (file might be a Google Doc)"
-                )
-                export_mime = "application/pdf"
-                request = self.service.files().export_media(
-                    fileId=file_id, mimeType=export_mime
+                logger.debug(
+                    "[GoogleDrive] _download_file_bytes: using export_media (%s -> %s)",
+                    mime_type,
+                    export_mime,
                 )
-                fh = io.BytesIO()
-                downloader = MediaIoBaseDownload(fh, request, chunksize=1024 * 1024)
-                done = False
+                # NOTE: export_media does not accept supportsAllDrives/includeItemsFromAllDrives
+                request = self.service.files().export_media(fileId=file_id, mimeType=export_mime)
+            else:
+                # This is a regular uploaded file (PDF, image, video, etc.) - use get_media
+                # Also handles non-exportable Google Apps files (Forms, Sites, Maps, etc.)
+                logger.debug("[GoogleDrive] _download_file_bytes: using get_media (%s)", mime_type)
+                # Binary download (get_media also doesn't accept the Drive flags)
+                request = self.service.files().get_media(fileId=file_id)
+
+            # Download the file with error handling for misclassified Google Docs
+            fh = io.BytesIO()
+            downloader = MediaIoBaseDownload(fh, request, chunksize=1024 * 1024)
+            done = False
+
+            try:
                 while not done:
                     status, done = downloader.next_chunk()
-            else:
-                raise
+                    # Optional: you can log progress via status.progress()
+            except HttpError as e:
+                # If download fails with "fileNotDownloadable", it's a Docs Editor file
+                # that wasn't properly detected. Retry with export_media.
+                if "fileNotDownloadable" in str(e) and mime_type not in exportable_types:
+                    logger.warning(
+                        f"Download failed for {file_id} ({mime_type}) with fileNotDownloadable error. "
+                        f"Retrying with export_media (file might be a Google Doc)"
+                    )
+                    export_mime = "application/pdf"
+                    request = self.service.files().export_media(
+                        fileId=file_id, mimeType=export_mime
+                    )
+                    fh = io.BytesIO()
+                    downloader = MediaIoBaseDownload(fh, request, chunksize=1024 * 1024)
+                    done = False
+                    while not done:
+                        status, done = downloader.next_chunk()
+                else:
+                    raise
 
         data = fh.getvalue()
         logger.debug("[GoogleDrive] _download_file_bytes: done, %d bytes", len(data))
@@ -572,10 +580,10 @@ class GoogleDriveConnector(BaseConnector):
 
     async def list_files(
         self,
-        page_token: Optional[str] = None,
-        max_files: Optional[int] = None,
+        page_token: str | None = None,
+        max_files: int | None = None,
         **kwargs,
-    ) -> Dict[str, Any]:
+    ) -> dict[str, Any]:
         """
         List files in the currently selected scope (file_ids/folder_ids/recursive).
         Returns a dict with 'files' and 'next_page_token'.
@@ -589,7 +597,9 @@ class GoogleDriveConnector(BaseConnector):
                 "Google Drive service is not initialized. Please authenticate first."
             )
 
-        logger.debug("[GoogleDrive] list_files: entry (page_token=%s, max_files=%s)", page_token, max_files)
+        logger.debug(
+            "[GoogleDrive] list_files: entry (page_token=%s, max_files=%s)", page_token, max_files
+        )
 
         try:
             items = await asyncio.to_thread(self._iter_selected_items)
@@ -609,7 +619,7 @@ class GoogleDriveConnector(BaseConnector):
             logger.error("[GoogleDrive] list_files failed: %s", e, exc_info=True)
             raise
 
-    def _extract_google_drive_acl(self, file_meta: Dict) -> DocumentACL:
+    def _extract_google_drive_acl(self, file_meta: dict) -> DocumentACL:
         """
         Extract ACL from Google Drive file metadata.
 
@@ -624,10 +634,15 @@ class GoogleDriveConnector(BaseConnector):
         """
         try:
             # Fetch permissions (requires additional API call)
-            permissions_list = self.service.permissions().list(
-                fileId=file_meta["id"],
-                fields="permissions(emailAddress,role,type,deleted,displayName)"
-            ).execute()
+            with self._lock:
+                permissions_list = (
+                    self.service.permissions()
+                    .list(
+                        fileId=file_meta["id"],
+                        fields="permissions(emailAddress,role,type,deleted,displayName)",
+                    )
+                    .execute()
+                )
 
             allowed_users = []
             allowed_groups = []
@@ -730,12 +745,12 @@ class GoogleDriveConnector(BaseConnector):
             metadata={
                 "parents": meta.get("parents"),
                 "driveId": meta.get("driveId"),
-                "size": int(meta.get("size", 0))
-                if str(meta.get("size", "")).isdigit()
-                else None,
+                "size": int(meta.get("size", 0)) if str(meta.get("size", "")).isdigit() else None,
             },
         )
-        logger.debug("[GoogleDrive] get_file_content: done for file_id=%s (%d bytes)", file_id, len(blob))
+        logger.debug(
+            "[GoogleDrive] get_file_content: done for file_id=%s (%d bytes)", file_id, len(blob)
+        )
         return doc
 
     async def setup_subscription(self) -> str:
@@ -752,9 +767,7 @@ class GoogleDriveConnector(BaseConnector):
         # 1) Ensure we are authenticated and have a live Drive service
         ok = await self.authenticate()
         if not ok:
-            raise RuntimeError(
-                "GoogleDriveConnector.setup_subscription: not authenticated"
-            )
+            raise RuntimeError("GoogleDriveConnector.setup_subscription: not authenticated")
 
         # 2) Resolve webhook address (no param in ABC, so pull from config/env)
         webhook_address = getattr(self.cfg, "webhook_address", None) or os.getenv(
@@ -810,9 +823,7 @@ class GoogleDriveConnector(BaseConnector):
             }
 
             if not isinstance(channel_id, str) or not channel_id:
-                raise RuntimeError(
-                    f"Drive watch returned invalid channel id: {channel_id!r}"
-                )
+                raise RuntimeError(f"Drive watch returned invalid channel id: {channel_id!r}")
 
             return channel_id
 
@@ -850,13 +861,13 @@ class GoogleDriveConnector(BaseConnector):
 
         # Single-channel memory
         if getattr(self, "_active_channel", None):
-            ch = getattr(self, "_active_channel")
+            ch = self._active_channel
             if isinstance(ch, dict) and ch.get("channel_id") == subscription_id:
                 resource_id = ch.get("resource_id")
 
         # Multi-channel memory
         if resource_id is None and hasattr(self, "_subscriptions"):
-            subs = getattr(self, "_subscriptions")
+            subs = self._subscriptions
             if isinstance(subs, dict):
                 entry = subs.get(subscription_id)
                 if isinstance(entry, dict):
@@ -888,9 +899,7 @@ class GoogleDriveConnector(BaseConnector):
             ):
                 self._active_channel = {}
 
-            if hasattr(self, "_subscriptions") and isinstance(
-                self._subscriptions, dict
-            ):
+            if hasattr(self, "_subscriptions") and isinstance(self._subscriptions, dict):
                 self._subscriptions.pop(subscription_id, None)
 
             return True
@@ -902,7 +911,7 @@ class GoogleDriveConnector(BaseConnector):
                 pass
             return False
 
-    async def handle_webhook(self, payload: Dict[str, Any]) -> List[str]:
+    async def handle_webhook(self, payload: dict[str, Any]) -> list[str]:
         """
         Process a Google Drive Changes webhook.
         Drive push notifications do NOT include the changed files themselves; they merely tell us
@@ -915,7 +924,7 @@ class GoogleDriveConnector(BaseConnector):
         Returns:
             List[str]: unique list of affected file IDs (filtered to our selected scope).
         """
-        affected: List[str] = []
+        affected: list[str] = []
         try:
             # 1) Ensure we're authenticated / service ready
             ok = await self.authenticate()
@@ -941,9 +950,7 @@ class GoogleDriveConnector(BaseConnector):
             except Exception as e:
                 selected_ids = set()
                 try:
-                    logger.error(
-                        f"handle_webhook: scope build failed, proceeding unfiltered: {e}"
-                    )
+                    logger.error(f"handle_webhook: scope build failed, proceeding unfiltered: {e}")
                 except Exception:
                     pass
 
@@ -980,11 +987,7 @@ class GoogleDriveConnector(BaseConnector):
                     # Filter to our selected scope if we have one; otherwise accept all
                     if selected_ids and (rid not in selected_ids):
                         # Shortcut target might be in scope even if the shortcut isn't
-                        tgt = (
-                            fobj.get("shortcutDetails", {}).get("targetId")
-                            if fobj
-                            else None
-                        )
+                        tgt = fobj.get("shortcutDetails", {}).get("targetId") if fobj else None
                         if not (tgt and tgt in selected_ids):
                             continue
 
@@ -1007,7 +1010,7 @@ class GoogleDriveConnector(BaseConnector):
 
             # Deduplicate while preserving order
             seen = set()
-            deduped: List[str] = []
+            deduped: list[str] = []
             for x in affected:
                 if x not in seen:
                     seen.add(x)
@@ -1033,9 +1036,7 @@ class GoogleDriveConnector(BaseConnector):
                 blob = self._download_file_bytes(meta)
             except HttpError as e:
                 # Skip/record failures
-                logger.error(
-                    f"Failed to download {meta.get('name')} ({meta.get('id')}): {e}"
-                )
+                logger.error(f"Failed to download {meta.get('name')} ({meta.get('id')}): {e}")
                 continue
 
             from datetime import datetime
@@ -1081,12 +1082,10 @@ class GoogleDriveConnector(BaseConnector):
     # -------------------------
     def get_start_page_token(self) -> str:
         # getStartPageToken accepts supportsAllDrives (not includeItemsFromAllDrives)
-        resp = (
-            self.service.changes().getStartPageToken(**self._drives_get_flags).execute()
-        )
+        resp = self.service.changes().getStartPageToken(**self._drives_get_flags).execute()
         return resp["startPageToken"]
 
-    def poll_changes_and_sync(self) -> Optional[str]:
+    def poll_changes_and_sync(self) -> str | None:
         """
         Incrementally process changes since the last page token in cfg.changes_page_token.
 
@@ -1122,10 +1121,7 @@ class GoogleDriveConnector(BaseConnector):
                 # Match scope
                 if fid not in selected_ids:
                     # also consider shortcut target
-                    if (
-                        file_obj.get("mimeType")
-                        == "application/vnd.google-apps.shortcut"
-                    ):
+                    if file_obj.get("mimeType") == "application/vnd.google-apps.shortcut":
                         tgt = file_obj.get("shortcutDetails", {}).get("targetId")
                         if tgt and tgt in selected_ids:
                             pass
@@ -1189,8 +1185,8 @@ class GoogleDriveConnector(BaseConnector):
     # Optional: webhook stubs
     # -------------------------
     def build_watch_body(
-        self, webhook_address: str, channel_id: Optional[str] = None
-    ) -> Dict[str, Any]:
+        self, webhook_address: str, channel_id: str | None = None
+    ) -> dict[str, Any]:
         """
         Prepare the request body for changes.watch if you use webhooks.
         """
@@ -1200,7 +1196,7 @@ class GoogleDriveConnector(BaseConnector):
             "address": webhook_address,
         }
 
-    def start_watch(self, webhook_address: str) -> Dict[str, Any]:
+    def start_watch(self, webhook_address: str) -> dict[str, Any]:
         """
         Start a webhook watch on changes using the current page token.
         Persist the returned resourceId/expiration on your side.

connectors/langflow_connector_service.py

@@ -53,9 +53,13 @@ class LangflowConnectorService:
             filename=document.filename,
         )
 
+        import os
+
         from utils.file_utils import auto_cleanup_tempfile
 
-        suffix = get_file_extension(document.mimetype)
+        suffix = os.path.splitext(document.filename)[1]
+        if not suffix:
+            suffix = get_file_extension(document.mimetype)
 
         # Create temporary file from document content
         with auto_cleanup_tempfile(suffix=suffix) as tmp_path:

connectors/service.py

@@ -52,11 +52,15 @@ class ConnectorService:
         """Process a document from a connector using existing processing pipeline"""
 
         # Create temporary file from document content
+        import os
+
         from utils.file_utils import auto_cleanup_tempfile
 
-        with auto_cleanup_tempfile(
-            suffix=get_file_extension(document.mimetype)
-        ) as tmp_path:
+        suffix = os.path.splitext(document.filename)[1]
+        if not suffix:
+            suffix = get_file_extension(document.mimetype)
+
+        with auto_cleanup_tempfile(suffix=suffix) as tmp_path:
             # Write document content to temp file
             with open(tmp_path, "wb") as f:
                 f.write(document.content)
@@ -363,6 +367,7 @@ class ConnectorService:
             original_folder_ids = getattr(connector.cfg, "folder_ids", None)
 
         expanded_file_ids = file_ids  # Default to original IDs
+        expanded_files_info = []
 
         try:
             # Set the file_ids we want to sync in the connector's config
@@ -373,7 +378,12 @@ class ConnectorService:
             # Get the expanded list of file IDs (folders will be expanded to their contents)
             # This uses the connector's list_files() which calls _iter_selected_items()
             result = await connector.list_files()
-            expanded_file_ids = [f["id"] for f in result.get("files", [])]
+            expanded_files = result.get("files", [])
+            expanded_file_ids = [f["id"] for f in expanded_files]
+
+            # Save the expanded files info so we can set correct names in the task UI
+            for f in expanded_files:
+                expanded_files_info.append(f)
 
             if not expanded_file_ids:
                 logger.warning(
@@ -423,12 +433,15 @@ class ConnectorService:
 
         # Create custom task using TaskService
         original_filenames = {}
-        if file_infos:
+
+        # Combine file_infos and expanded_files_info
+        all_infos = (file_infos or []) + expanded_files_info
+        if all_infos:
             original_filenames = {
                 f["id"]: clean_connector_filename(
-                    f["name"], f.get("mimeType") or f.get("mimetype")
+                    f["name"], f.get("mimeType") or f.get("mimetype", "")
                 )
-                for f in file_infos
+                for f in all_infos
                 if "id" in f and "name" in f
             }
 

models/processors.py

@@ -257,15 +257,16 @@ class TaskProcessor:
                 file_hash=file_hash,
             )
             slim_doc = process_text_file(file_path)
-            # Override filename with original_filename if provided
-            if original_filename:
-                slim_doc["filename"] = original_filename
         else:
             from utils.docling_client import convert_file
 
             full_doc = await convert_file(file_path, httpx_client=clients.docling_http_client)
             slim_doc = extract_relevant(full_doc)
 
+        # Override filename with original_filename if provided
+        if original_filename:
+            slim_doc["filename"] = original_filename
+
         if chunk_size is not None:
             try:
                 cs = int(chunk_size)
@@ -527,10 +528,13 @@ class ConnectorFileProcessor(TaskProcessor):
             if not self.user_id:
                 raise ValueError("user_id not provided to ConnectorFileProcessor")
 
-            # Create temporary file from document content
             from utils.file_utils import auto_cleanup_tempfile
+            # Create temporary file from document content            import os
+
+            suffix = os.path.splitext(document.filename)[1]
+            if not suffix:
+                suffix = get_file_extension(document.mimetype)
 
-            suffix = get_file_extension(document.mimetype)
             with auto_cleanup_tempfile(suffix=suffix) as tmp_path:
                 # Write content to temp file
                 with open(tmp_path, 'wb') as f:
@@ -654,7 +658,10 @@ class LangflowConnectorFileProcessor(TaskProcessor):
             # Create temporary file and compute hash to check for duplicates
             from utils.file_utils import auto_cleanup_tempfile
 
-            suffix = get_file_extension(document.mimetype)
+            suffix = os.path.splitext(document.filename)[1]
+            if not suffix:
+                suffix = get_file_extension(document.mimetype)
+
             with auto_cleanup_tempfile(suffix=suffix) as tmp_path:
                 # Write content to temp file
                 with open(tmp_path, 'wb') as f:
@@ -740,7 +747,8 @@ class S3FileProcessor(TaskProcessor):
         from utils.hash_utils import hash_id
 
         try:
-            with auto_cleanup_tempfile() as tmp_path:
+            suffix = os.path.splitext(item)[1]
+            with auto_cleanup_tempfile(suffix=suffix) as tmp_path:
                 # Download object to temporary file
                 with open(tmp_path, 'wb') as tmp_file:
                     self.s3_client.download_fileobj(self.bucket, item, tmp_file)

{openrag-0.5.0.dev30.dist-info → openrag-0.5.0.dev32.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: openrag
-Version: 0.5.0.dev30
+Version: 0.5.0.dev32
 Summary: OpenRAG is a comprehensive Retrieval-Augmented Generation platform that enables intelligent document search and AI-powered conversations.
 Classifier: Development Status :: 4 - Beta
 Classifier: Environment :: Console

{openrag-0.5.0.dev30.dist-info → openrag-0.5.0.dev32.dist-info}/RECORD

@@ -22,7 +22,7 @@ api/provider_health.py,sha256=u4gNGhEAqsQY1ml6xWzWWAOfLHDrSVYPkw1NAmBCMsI,9690
 api/provider_validation.py,sha256=ik00rpXjL6pPCyJt3gEdeNhyoYlhIPGz_Kigmhf0zLw,31430
 api/router.py,sha256=nOusNdc-eOMRZo6a5eqNb91oOoleYNomHTGcKuPsMmU,6229
 api/search.py,sha256=z6M7PaO1oIdZtp_yU0eGkz0dTmDJItc7O4ULc67VIHk,1982
-api/settings.py,sha256=9ViBKRiyenlQdV_rZIbL_FQoZQEuMjR2-RxUH6bJl80,92480
+api/settings.py,sha256=UNvTkZSI1hHOKYzmNe1yPFBby-Mn22R0fDyeCwy1lyM,92701
 api/tasks.py,sha256=j3tAfpE_X4UxhiGPrgea5jkxILF6HZK9Hoy3QTFGG0g,1524
 api/upload.py,sha256=vh8pxlf7LnTIlRH8gsJhogTbFq93-SQ-1UNIQTtaer4,7288
 api/v1/__init__.py,sha256=A7dLY63RcJv6ii-Xh7WgREcwXixjzOXz2w5C0LGWTCc,190
@@ -43,8 +43,8 @@ config/settings.py,sha256=zhvJ3TAmdZwoWk55BPX4OuAAJiDTHlJg4_MoSWPhZMI,39298
 connectors/__init__.py,sha256=6ijOhF7jG-zJ3bIxQ4HIHZHqpT127cNvCDaaqdN_Pm8,390
 connectors/base.py,sha256=s9gBEIxl8_nZVtCeMKz7LrlOV-H0rCxvN_cFH8B9DMg,4532
 connectors/connection_manager.py,sha256=LrfiuoBB-hqNGjXhH-5x4uL6msU-OVOp7--4V302a14,28187
-connectors/langflow_connector_service.py,sha256=0455CdQMeFE320Le81e9fA40CduuPgIqodwkcisjI1U,17149
-connectors/service.py,sha256=1RQ1JnAJl0LeGwG9_NBjN1D3hae1j4NZRBdaS_bLoO8,17760
+connectors/langflow_connector_service.py,sha256=HjVgTmlpZCWsoR2msEuCouw3snX0HBHNv6Szm4A0Iik,17251
+connectors/service.py,sha256=_qyRDskMiNfPa2OyOan9gAmOgB5JjcLTidzzg-rhLWs,18229
 connectors/aws_s3/__init__.py,sha256=ZRXSKmYvYK7Kr7Rzcd0_YIMpqfAm5qjm06zF-YfPPhA,369
 connectors/aws_s3/api.py,sha256=CDpsdOZaOM7hT93Je4O_JD7GxJjPN0-_qgQPPSPiy40,6861
 connectors/aws_s3/auth.py,sha256=qvrjxNM5MvgK-I5F5Njl8B7pSJTrGIZ0wuZImFP9UVg,3060
@@ -52,7 +52,7 @@ connectors/aws_s3/connector.py,sha256=QZPFt0PycyIMp72ut2yxXrEYl45_N5MUloZXrw_xvf
 connectors/aws_s3/models.py,sha256=olSwBxzvVKqn6j3M5tENq7ksLQIuTfhGPGIEKmqxbR4,398
 connectors/aws_s3/support.py,sha256=vwsy6Q9EAssb2FeeT9NVwlJ9K_NWTulGh8HK4hxwmrg,1458
 connectors/google_drive/__init__.py,sha256=y3HqyvfVvNV8PYMgFF_MeM7djjWxkyDw3hJivLfhGXc,136
-connectors/google_drive/connector.py,sha256=QB4sxHN7FVqCKZlgbte5T4F3N1NKQAcxzMt09g4ZOcA,47544
+connectors/google_drive/connector.py,sha256=UCSpGxOtCGRtTtiF4p8cZLkWAEOvdymzGhRYLYztaMY,47609
 connectors/google_drive/oauth.py,sha256=7zRf5zuFNbrREGI516BJuOzuMpGEaqhAtyzENFyPp-U,8179
 connectors/ibm_cos/__init__.py,sha256=gp6L4nW7EIHsEYIlbZQUDE9o90JO-0CEojTUtEv5-Po,369
 connectors/ibm_cos/api.py,sha256=vsCGoJnEfUMWqrWAdw0znpJQUcex7Lwx1ODWZeuY7m8,8440
@@ -70,13 +70,13 @@ connectors/sharepoint/utils.py,sha256=OBNEbXb8uAfR-iR-ML0aOojGfGj03vJnKm9DbYf2hW
 mcp_http/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 mcp_http/server.py,sha256=JutP94fQRCizeCN69kKkdYDFE71L3DqLTCcb397T8Eo,7738
 models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-models/processors.py,sha256=e6qpstQq9DyHM4UMEUfjQqBoVO4xeLmSLkQcyt4oca0,35539
+models/processors.py,sha256=Vc6a1HIHpIe9t3Jx7HJ_lCRrSmWtzw63C5bzJpxTz8s,35793
 models/tasks.py,sha256=9iINXQZxA9-hcYzG8xC_baBfLa2w6gcRPJdQHsx_Kms,1643
 models/url.py,sha256=92rsrLLyIqUErYipYz5QwUNTdHwJ2jBRV7PIVpncTyo,5764
-openrag-0.5.0.dev30.dist-info/licenses/LICENSE,sha256=jWcnPXkUxEinX60Yn-_zjsY9i84vz2QSb0fi53djOkg,11334
+openrag-0.5.0.dev32.dist-info/licenses/LICENSE,sha256=jWcnPXkUxEinX60Yn-_zjsY9i84vz2QSb0fi53djOkg,11334
 services/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 services/api_key_service.py,sha256=NMz2ohJo0jcuHjDN3tHuzhCTVo6lsOqAkUhozzmWSv0,11720
-services/auth_service.py,sha256=nNxKuCHh0ziHuHU18EGEUnf242h1XYsfbrbfWTmCmQE,22125
+services/auth_service.py,sha256=nAKTvMyX-YUtiNfKF06Dq4E4iwHPBhfKR6nliqGpOSs,22090
 services/chat_service.py,sha256=Pg34Wg1krrRcZeVR0dGnZ_PU6mPB-CW4W39-2JwQAjQ,28366
 services/conversation_persistence_service.py,sha256=zfNMcgbD7GbQ_W8tB5jV_NDa_78l6fAmaIiiICg1YH8,6092
 services/document_service.py,sha256=-Lq5fYLnLz4mCsU4aS3tKXVTiDZ3P1jzyuNsxH-QdGw,7397
@@ -117,7 +117,7 @@ tui/screens/__init__.py,sha256=ZzB_lbAnSTwDqiG0ppafOE-sAiOUVaNy4V6x5AJvqyM,27
 tui/screens/config.py,sha256=_XcUisNNYR4-IVOhfXIZ9lCY1bPe8eaZY9JMh5nRi0g,27975
 tui/screens/diagnostics.py,sha256=ZMDD9KSY0GoXo67ft6j7x4RGWzI_Hy83_a3FovnstrM,24864
 tui/screens/logs.py,sha256=3fM6fAqjdFXVI_laiF5yl7lH8n0bSmWcdozMAWW3SLM,11853
-tui/screens/monitor.py,sha256=YOkrny2DdpDjqe6ChiFtjnygh4c8ECjy_9tHgEV4Gmk,46731
+tui/screens/monitor.py,sha256=kTDaaDjoJT5Lr4kGYwNeVV9jsOMLLYfapZnDcJjjbcE,46900
 tui/screens/welcome.py,sha256=YYotbQOVbiLZjnyvcZ5OJqYNrh0i_bI22T9WdXqMjHQ,28452
 tui/utils/__init__.py,sha256=MIui8sXgp8IUvEkaCNuI5kRME8XfwCnGsapBIldaY_s,29
 tui/utils/clipboard.py,sha256=tUoesmliVEVgrWarsNkEGWj3fHc2L_IsKEMIhwrgG9c,1901
@@ -159,8 +159,8 @@ utils/telemetry/__init__.py,sha256=xoStY8hoYWnVe0B_jNthSdzRb7AxCT0rhM8ba9Q0Obg,2
 utils/telemetry/category.py,sha256=wei3V5WyUaiadYjEE5s0ys9PY0TjnebIvTykQU0P_2k,1114
 utils/telemetry/client.py,sha256=RhjW1a27w7W6gHt9X7AxdRPdZGmCEQDPskkxj5nsuiQ,12500
 utils/telemetry/message_id.py,sha256=IAu0wBVNOGPcSlQ0bAzvqIYr1Ac_n8wSTSxXzq_tiV0,9634
-openrag-0.5.0.dev30.dist-info/METADATA,sha256=2p00OaaKrPEu6whS45WHl9IEz3Nc8VOrEcYpcM0DJeg,7819
-openrag-0.5.0.dev30.dist-info/WHEEL,sha256=aeYiig01lYGDzBgS8HxWXOg3uV61G9ijOsup-k9o1sk,91
-openrag-0.5.0.dev30.dist-info/entry_points.txt,sha256=ArKNHB63I3K8vynNkogvZOLmmQFpIsHDL7MvreCmju4,45
-openrag-0.5.0.dev30.dist-info/top_level.txt,sha256=5vfLBIkzuwBJrizhCSZiXV7NTzTRIMlNLsyzYDEeCF8,125
-openrag-0.5.0.dev30.dist-info/RECORD,,
+openrag-0.5.0.dev32.dist-info/METADATA,sha256=UfANP1bsEv8iz1SeZw2O9gH1qIOiUWpP_1YB8xttsE4,7819
+openrag-0.5.0.dev32.dist-info/WHEEL,sha256=aeYiig01lYGDzBgS8HxWXOg3uV61G9ijOsup-k9o1sk,91
+openrag-0.5.0.dev32.dist-info/entry_points.txt,sha256=ArKNHB63I3K8vynNkogvZOLmmQFpIsHDL7MvreCmju4,45
+openrag-0.5.0.dev32.dist-info/top_level.txt,sha256=5vfLBIkzuwBJrizhCSZiXV7NTzTRIMlNLsyzYDEeCF8,125
+openrag-0.5.0.dev32.dist-info/RECORD,,

services/auth_service.py

@@ -324,9 +324,7 @@ class AuthService:
 
             # Add expiry if provided
             if token_data.get("expires_in"):
-                expiry = datetime.now() + timedelta(
-                    seconds=int(token_data["expires_in"])
-                )
+                expiry = datetime.utcnow() + timedelta(seconds=int(token_data["expires_in"]))
                 token_file_data["expiry"] = expiry.isoformat()
 
             # Save tokens to file

tui/screens/monitor.py

@@ -3,29 +3,34 @@
 import asyncio
 import re
 import shutil
+from collections.abc import AsyncIterator
 from pathlib import Path
-from typing import Literal, Any, Optional, AsyncIterator
-
-# Define button variant type
-ButtonVariant = Literal["default", "primary", "success", "warning", "error"]
+from typing import Any, Literal, cast
 
+from rich.text import Text
 from textual.app import ComposeResult
-from textual.containers import Container, Vertical, Horizontal, ScrollableContainer
+from textual.containers import Horizontal, ScrollableContainer
+from textual.coordinate import Coordinate
 from textual.screen import Screen
-from textual.widgets import Header, Footer, Static, Button, DataTable
-from textual.timer import Timer
-from rich.text import Text
-from rich.table import Table
-
-from ..managers.container_manager import ContainerManager, ServiceStatus, ServiceInfo, format_port_conflict_message
+from textual.widgets import Button, DataTable, Footer, Static
+
+from ..managers.container_manager import (
+    ContainerManager,
+    ServiceInfo,
+    ServiceStatus,
+    format_port_conflict_message,
+)
 from ..managers.docling_manager import DoclingManager
 from ..utils.platform import RuntimeType
 from ..widgets.command_modal import CommandOutputModal
-from ..widgets.flow_backup_warning_modal import FlowBackupWarningModal
+from ..widgets.diagnostics_notification import notify_with_diagnostics
 from ..widgets.factory_reset_warning_modal import FactoryResetWarningModal
-from ..widgets.version_mismatch_warning_modal import VersionMismatchWarningModal
+from ..widgets.flow_backup_warning_modal import FlowBackupWarningModal
 from ..widgets.upgrade_instructions_modal import UpgradeInstructionsModal
-from ..widgets.diagnostics_notification import notify_with_diagnostics
+from ..widgets.version_mismatch_warning_modal import VersionMismatchWarningModal
+
+# Define button variant type
+ButtonVariant = Literal["default", "primary", "success", "warning", "error"]
 
 
 class MonitorScreen(Screen):
@@ -55,6 +60,8 @@ class MonitorScreen(Screen):
         self.refresh_timer = None
         self.operation_in_progress = False
         self._follow_task = None
+        self._follow_service: str | None = None
+        self._logs_buffer: list[str] = []
 
         # Track which table was last selected for mutual exclusion
         self._last_selected_table = None
@@ -63,10 +70,9 @@ class MonitorScreen(Screen):
     def container_manager(self) -> ContainerManager:
         """Get the shared container manager from the app."""
         if self._container_manager is None:
-            self._container_manager = self.app.container_manager
+            self._container_manager = cast(Any, self.app).container_manager
         return self._container_manager
 
-
     def compose(self) -> ComposeResult:
         """Create the monitoring screen layout."""
         # Just show the services content directly (no header, no tabs)
@@ -98,9 +104,7 @@ class MonitorScreen(Screen):
         yield Horizontal(id="services-controls", classes="button-row")
         # Create services table with image + digest info
         self.services_table = DataTable(id="services-table")
-        self.services_table.add_columns(
-            "Service", "Status", "Health", "Ports", "Image", "Digest"
-        )
+        self.services_table.add_columns("Service", "Status", "Health", "Ports", "Image", "Digest")
         yield self.services_table
         yield Static(" ")
 
@@ -118,12 +122,8 @@ class MonitorScreen(Screen):
         status_text = Text()
 
         if not self.container_manager.is_available():
-            status_text.append(
-                "WARNING: No container runtime available\n", style="bold red"
-            )
-            status_text.append(
-                "Please install Docker or Podman to continue.\n", style="dim"
-            )
+            status_text.append("WARNING: No container runtime available\n", style="bold red")
+            status_text.append("Please install Docker or Podman to continue.\n", style="dim")
             return status_text
 
         runtime_info = self.container_manager.get_runtime_info()
@@ -150,7 +150,7 @@ class MonitorScreen(Screen):
         """Initialize the screen when mounted."""
         await self._refresh_services()
         # Set up auto-refresh every 5 seconds
-        self.refresh_timer = self.set_interval(5.0, self._auto_refresh)
+        self.refresh_timer = self.set_interval(5.0, self._auto_refresh_services)
 
         self._focus_services_table()
 
@@ -186,9 +186,9 @@ class MonitorScreen(Screen):
                 images_set.add(img)
         # Ensure compose-declared images are also shown (e.g., langflow when stopped)
         try:
-            for img in (
-                self.container_manager._parse_compose_images()
-            ):  # best-effort, no YAML dep
+            for (
+                img
+            ) in await self.container_manager._parse_compose_images():  # best-effort, no YAML dep
                 if img:
                     images_set.add(img)
         except Exception:
@@ -205,7 +205,7 @@ class MonitorScreen(Screen):
             self.images_table.clear()
 
         # Add container service rows
-        for service_name, service_info in services.items():
+        for _service_name, service_info in services.items():
             status_style = self._get_status_style(service_info.status)
 
             self.services_table.add_row(
@@ -238,9 +238,7 @@ class MonitorScreen(Screen):
             if (docling_running or docling_starting)
             else "N/A"
         )
-        docling_pid = (
-            str(docling_status.get("pid")) if docling_status.get("pid") else "N/A"
-        )
+        docling_pid = str(docling_status.get("pid")) if docling_status.get("pid") else "N/A"
 
         if self.docling_table:
             self.docling_table.add_row(
@@ -275,7 +273,7 @@ class MonitorScreen(Screen):
         }
         return status_styles.get(status, "white")
 
-    async def _auto_refresh(self) -> None:
+    async def _auto_refresh_services(self) -> None:
         """Auto-refresh services if not in operation."""
         if not self.operation_in_progress:
             await self._refresh_services()
@@ -283,7 +281,6 @@ class MonitorScreen(Screen):
     def on_button_pressed(self, event: Button.Pressed) -> None:
         """Handle button presses."""
         button_id = event.button.id or ""
-        button_label = event.button.label or ""
 
         # Use button ID prefixes to determine action, ignoring any random suffix
         if button_id.startswith("start-btn"):
@@ -328,7 +325,7 @@ class MonitorScreen(Screen):
                 self.run_worker(self._show_logs(service_name))
                 self._start_follow(service_name)
 
-    async def _start_services(self, cpu_mode: Optional[bool] = None) -> None:
+    async def _start_services(self, cpu_mode: bool | None = None) -> None:
         """Start services with progress updates."""
         self.operation_in_progress = True
         try:
@@ -365,6 +362,7 @@ class MonitorScreen(Screen):
                 # This ensures docker compose reads the correct version
                 try:
                     from ..managers.env_manager import EnvManager
+
                     env_manager = EnvManager()
                     env_manager.ensure_openrag_version()
                     # Small delay to ensure .env file is written and flushed
@@ -433,7 +431,7 @@ class MonitorScreen(Screen):
                 # Show "this is the latest version" toast
                 self.notify(
                     f"You are running the latest version ({current_version}).",
-                    severity="success",
+                    severity="information",
                     timeout=5,
                 )
             else:
@@ -442,9 +440,7 @@ class MonitorScreen(Screen):
                     UpgradeInstructionsModal(current_version, latest_version)
                 )
         except Exception as e:
-            self.notify(
-                f"Error checking version: {str(e)}", severity="error", timeout=10
-            )
+            self.notify(f"Error checking version: {str(e)}", severity="error", timeout=10)
         finally:
             self.operation_in_progress = False
 
@@ -453,9 +449,7 @@ class MonitorScreen(Screen):
         self.operation_in_progress = True
         try:
             # Show factory reset warning modal first
-            should_continue = await self.app.push_screen_wait(
-                FactoryResetWarningModal()
-            )
+            should_continue = await self.app.push_screen_wait(FactoryResetWarningModal())
             if not should_continue:
                 self.notify("Factory reset cancelled", severity="information")
                 return
@@ -475,6 +469,7 @@ class MonitorScreen(Screen):
             try:
                 # Get paths from env config
                 from ..managers.env_manager import EnvManager
+
                 env_manager = EnvManager()
                 env_manager.load_existing_env()
 
@@ -487,7 +482,9 @@ class MonitorScreen(Screen):
 
                 if config_path.exists():
                     # Use container to handle files owned by container user
-                    success, msg = await self.container_manager.clear_directory_with_container(config_path)
+                    success, msg = await self.container_manager.clear_directory_with_container(
+                        config_path
+                    )
                     if not success:
                         # Fallback to regular rmtree if container method fails
                         shutil.rmtree(config_path)
@@ -497,18 +494,32 @@ class MonitorScreen(Screen):
                 # Also delete legacy TUI config folder if it exists (~/.openrag/tui/config/)
                 tui_config_path = expand_path(env_manager.config.openrag_tui_config_path_legacy)
                 if tui_config_path.exists():
-                    success, msg = await self.container_manager.clear_directory_with_container(tui_config_path)
+                    success, msg = await self.container_manager.clear_directory_with_container(
+                        tui_config_path
+                    )
                     if not success:
                         # Fallback to regular rmtree if container method fails
                         shutil.rmtree(tui_config_path)
                     # Recreate empty config directory
                     tui_config_path.mkdir(parents=True, exist_ok=True)
 
+                # Clear backend data directory (database, session ownership, conversations, oauth tokens)
+                data_path = expand_path(env_manager.config.openrag_data_path)
+                if data_path.exists():
+                    success, msg = await self.container_manager.clear_directory_with_container(
+                        data_path
+                    )
+                    if not success:
+                        shutil.rmtree(data_path)
+                    data_path.mkdir(parents=True, exist_ok=True)
+
                 # Delete flow backups only if user chose to (and they actually exist)
                 if self._check_flow_backups():
                     if delete_backups:
                         # Use container to handle files owned by container user
-                        success, msg = await self.container_manager.clear_directory_with_container(flows_backup_path)
+                        success, msg = await self.container_manager.clear_directory_with_container(
+                            flows_backup_path
+                        )
                         if not success:
                             # Fallback to regular rmtree if container method fails
                             shutil.rmtree(flows_backup_path)
@@ -516,8 +527,10 @@ class MonitorScreen(Screen):
                         flows_backup_path.mkdir(parents=True, exist_ok=True)
                         self.notify("Flow backups deleted", severity="information")
                     else:
-                        self.notify(f"Flow backups preserved in {flows_backup_path}", severity="information")
-                
+                        self.notify(
+                            f"Flow backups preserved in {flows_backup_path}", severity="information"
+                        )
+
             except Exception as e:
                 self.notify(
                     f"Error clearing config: {str(e)}",
@@ -548,12 +561,14 @@ class MonitorScreen(Screen):
 
         # Get data paths from env config
         from ..managers.env_manager import EnvManager
+
         env_manager = EnvManager()
         env_manager.load_existing_env()
 
         # Delete langflow-data directory (mirrors Makefile factory-reset behaviour)
         yield False, "Clearing Langflow data..."
         from tui.main import _resolve_langflow_data_path
+
         langflow_data_path = _resolve_langflow_data_path(Path.home() / ".openrag").resolve()
         home = Path.home().resolve()
         if not str(langflow_data_path).startswith(str(home) + "/"):
@@ -575,13 +590,13 @@ class MonitorScreen(Screen):
         try:
             # Show prune options modal
             from tui.widgets.prune_options_modal import PruneOptionsModal
-            
+
             prune_choice = await self.app.push_screen_wait(PruneOptionsModal())
-            
+
             if prune_choice == "cancel":
                 self.notify("Prune cancelled", severity="information")
                 return
-            
+
             # Choose the appropriate pruning method based on user choice
             if prune_choice == "all":
                 # Stop services and prune all images
@@ -591,7 +606,7 @@ class MonitorScreen(Screen):
                 # Prune only unused images (default)
                 command_generator = self.container_manager.prune_old_images()
                 modal_title = "Pruning Unused Images"
-            
+
             # Show command output in modal dialog
             modal = CommandOutputModal(
                 modal_title,
@@ -605,12 +620,15 @@ class MonitorScreen(Screen):
     def _check_flow_backups(self) -> bool:
         """Check if there are any flow backups in flows/backup directory."""
         from pathlib import Path
+
         from ..managers.env_manager import EnvManager
 
         # Get flows path from env config
         env_manager = EnvManager()
         env_manager.load_existing_env()
-        flows_path = Path(env_manager.config.openrag_flows_path.replace("$HOME", str(Path.home()))).expanduser()
+        flows_path = Path(
+            env_manager.config.openrag_flows_path.replace("$HOME", str(Path.home()))
+        ).expanduser()
         backup_dir = flows_path / "backup"
         if not backup_dir.exists():
             return False
@@ -651,9 +669,7 @@ class MonitorScreen(Screen):
             if success:
                 self.notify(message, severity="information")
             else:
-                self.notify(
-                    f"Failed to start docling serve: {message}", severity="error"
-                )
+                self.notify(f"Failed to start docling serve: {message}", severity="error")
             # Refresh again to show final status (running or stopped)
             await self._refresh_services()
         except Exception as e:
@@ -671,9 +687,7 @@ class MonitorScreen(Screen):
             if success:
                 self.notify(message, severity="information")
             else:
-                self.notify(
-                    f"Failed to stop docling serve: {message}", severity="error"
-                )
+                self.notify(f"Failed to stop docling serve: {message}", severity="error")
             # Refresh the services table to show updated status
             await self._refresh_services()
         except Exception as e:
@@ -689,9 +703,7 @@ class MonitorScreen(Screen):
             if success:
                 self.notify(message, severity="information")
             else:
-                self.notify(
-                    f"Failed to restart docling serve: {message}", severity="error"
-                )
+                self.notify(f"Failed to restart docling serve: {message}", severity="error")
             # Refresh the services table to show updated status
             await self._refresh_services()
         except Exception as e:
@@ -780,9 +792,7 @@ class MonitorScreen(Screen):
                 except Exception:
                     pass
         except Exception as e:
-            notify_with_diagnostics(
-                self.app, f"Error following logs: {e}", severity="error"
-            )
+            notify_with_diagnostics(self.app, f"Error following logs: {e}", severity="error")
 
     def action_refresh(self) -> None:
         """Refresh services manually."""
@@ -861,9 +871,7 @@ class MonitorScreen(Screen):
             self._update_mode_row()
             self.action_refresh()
         except Exception as e:
-            notify_with_diagnostics(
-                self.app, f"Failed to toggle mode: {e}", severity="error"
-            )
+            notify_with_diagnostics(self.app, f"Failed to toggle mode: {e}", severity="error")
 
     def _update_controls(self, services: list[ServiceInfo]) -> None:
         """Update control buttons based on running state."""
@@ -886,31 +894,19 @@ class MonitorScreen(Screen):
             # Add appropriate buttons based on service state
             if any_running:
                 # When services are running, show stop and restart
-                controls.mount(
-                    Button("Stop Services", variant="error", id=f"stop-btn{suffix}")
-                )
-                controls.mount(
-                    Button("Restart", variant="primary", id=f"restart-btn{suffix}")
-                )
+                controls.mount(Button("Stop Services", variant="error", id=f"stop-btn{suffix}"))
+                controls.mount(Button("Restart", variant="primary", id=f"restart-btn{suffix}"))
             else:
                 # When services are not running, show start
-                controls.mount(
-                    Button("Start Services", variant="success", id=f"start-btn{suffix}")
-                )
+                controls.mount(Button("Start Services", variant="success", id=f"start-btn{suffix}"))
 
             # Always show upgrade, prune, and reset buttons
-            controls.mount(
-                Button("Upgrade", variant="warning", id=f"upgrade-btn{suffix}")
-            )
-            controls.mount(
-                Button("Prune Images", variant="default", id=f"prune-btn{suffix}")
-            )
+            controls.mount(Button("Upgrade", variant="warning", id=f"upgrade-btn{suffix}"))
+            controls.mount(Button("Prune Images", variant="default", id=f"prune-btn{suffix}"))
             controls.mount(Button("Factory Reset", variant="error", id=f"reset-btn{suffix}"))
 
         except Exception as e:
-            notify_with_diagnostics(
-                self.app, f"Error updating controls: {e}", severity="error"
-            )
+            notify_with_diagnostics(self.app, f"Error updating controls: {e}", severity="error")
 
         # Update docling controls separately
         self._update_docling_controls()
@@ -940,9 +936,7 @@ class MonitorScreen(Screen):
                     Button("Stop", variant="error", id=f"docling-stop-btn{suffix}")
                 )
                 docling_controls.mount(
-                    Button(
-                        "Restart", variant="primary", id=f"docling-restart-btn{suffix}"
-                    )
+                    Button("Restart", variant="primary", id=f"docling-restart-btn{suffix}")
                 )
             elif docling_starting:
                 # Show disabled button or no button when starting
@@ -1102,7 +1096,7 @@ class MonitorScreen(Screen):
             return
 
         try:
-            table.cursor_coordinate = (row, 0)
+            table.cursor_coordinate = Coordinate(row, 0)
         except Exception:
             move_cursor = getattr(table, "move_cursor", None)
             if callable(move_cursor):
@@ -1111,9 +1105,7 @@ class MonitorScreen(Screen):
                 except Exception:
                     pass
 
-    def _focus_services_table(
-        self, row: str | None = None, set_last: bool = True
-    ) -> None:
+    def _focus_services_table(self, row: str | None = None, set_last: bool = True) -> None:
         """Focus the services table and update selection."""
         if not self.services_table:
             return