Move document scanning trigger by command line to background task

- Added background task management
- Prevented concurrent scanning
- Tracked scanning progress
- Improved startup performance
- Enhanced error handling
This commit is contained in:
yangdx
2025-02-14 01:12:39 +08:00
parent 47268c2ab1
commit 08364e99fb

View File

@@ -6,7 +6,7 @@ from fastapi import (
Form, Form,
BackgroundTasks, BackgroundTasks,
) )
import asyncio
import threading import threading
import os import os
import json import json
@@ -730,6 +730,8 @@ def create_app(args):
postgres_db = None postgres_db = None
oracle_db = None oracle_db = None
tidb_db = None tidb_db = None
# Store background tasks
app.state.background_tasks = set()
try: try:
# Check if PostgreSQL is needed # Check if PostgreSQL is needed
@@ -794,20 +796,19 @@ def create_app(args):
# Auto scan documents if enabled # Auto scan documents if enabled
if args.auto_scan_at_startup: if args.auto_scan_at_startup:
try: # Start scanning in background
new_files = doc_manager.scan_directory_for_new_files() with progress_lock:
for file_path in new_files: if not scan_progress["is_scanning"]:
try: scan_progress["is_scanning"] = True
await index_file(file_path) scan_progress["indexed_count"] = 0
except Exception as e: scan_progress["progress"] = 0
trace_exception(e) # Create background task
logging.error(f"Error indexing file {file_path}: {str(e)}") task = asyncio.create_task(run_scanning_process())
app.state.background_tasks.add(task)
ASCIIColors.info( task.add_done_callback(app.state.background_tasks.discard)
f"Indexed {len(new_files)} documents from {args.input_dir}" ASCIIColors.info(f"Started background scanning of documents from {args.input_dir}")
) else:
except Exception as e: ASCIIColors.info("Skip document scanning cause anohter scanning is active")
logging.error(f"Error during startup indexing: {str(e)}")
yield yield