Files
sac4cps-backend/microservices/data-ingestion-service/src/main.py
rafaeldpsilva aa07347604 Implement iterative FTP scan and skip logic with processed file cache
- Add iterative directory scanning to prevent infinite recursion - Cache
processed files in memory to avoid redundant database lookups - Skip
already processed files using cache and database fallback - Add tests
for skip logic and iterative scan behavior - Change logging for MongoDB
connection and file storage to debug level - Clean up FastAPI app and
remove redundant docstrings
2025-09-12 13:43:21 +01:00

123 lines
3.1 KiB
Python

from fastapi import FastAPI, HTTPException
from contextlib import asynccontextmanager
import asyncio
import logging
from datetime import datetime
from typing import Any
from ftp_monitor import FTPMonitor
from database import DatabaseManager
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
ftp_monitor = None
db_manager = None
@asynccontextmanager
async def lifespan(app: FastAPI):
global ftp_monitor, db_manager
logger.info("Starting SA4CPS Data Ingestion Service...")
db_manager = DatabaseManager()
await db_manager.connect()
ftp_monitor = FTPMonitor(db_manager)
monitoring_task = asyncio.create_task(ftp_monitor.start_monitoring())
logger.info("Service started successfully")
yield
logger.info("Shutting down service...")
monitoring_task.cancel()
await db_manager.close()
logger.info("Service shutdown complete")
app = FastAPI(
title="SA4CPS Data Ingestion Service",
description="Monitors FTP server for .sgl_v2 files and stores data in MongoDB",
version="1.0.0",
lifespan=lifespan
)
@app.get("/")
async def root():
return {
"service": "SA4CPS Data Ingestion Service",
"status": "running",
"timestamp": datetime.now().isoformat()
}
@app.get("/health")
async def health_check():
global ftp_monitor, db_manager
health_status = {
"service": "healthy",
"timestamp": datetime.now().isoformat(),
"database": "unknown",
"ftp_monitor": "unknown"
}
# Check database connection
if db_manager:
try:
await db_manager.ping()
health_status["database"] = "connected"
except Exception:
health_status["database"] = "disconnected"
health_status["service"] = "degraded"
# Check FTP monitor status
if ftp_monitor:
health_status["ftp_monitor"] = ftp_monitor.get_status()
health_status["last_check"] = ftp_monitor.get_last_check_time()
health_status["files_processed"] = ftp_monitor.get_processed_count()
return health_status
@app.get("/status")
async def get_status():
global ftp_monitor, db_manager
if not ftp_monitor:
raise HTTPException(status_code=503, detail="FTP monitor not initialized")
return {
"ftp_monitor": ftp_monitor.get_detailed_status(),
"database": await db_manager.get_stats() if db_manager else None,
"timestamp": datetime.now().isoformat()
}
@app.post("/trigger-check")
async def trigger_manual_check():
global ftp_monitor
if not ftp_monitor:
raise HTTPException(status_code=503, detail="FTP monitor not initialized")
try:
result = await ftp_monitor.check_for_new_files()
return {
"message": "Manual check completed",
"result": result,
"timestamp": datetime.now().isoformat()
}
except Exception as e:
logger.error(f"Manual check failed: {e}")
raise HTTPException(status_code=500, detail=f"Check failed: {str(e)}")
if __name__ == "__main__":
import uvicorn
uvicorn.run("main:app", host="0.0.0.0", port=8008, reload=True)