Files
aliyunApsSkill/aliyun-sync/aps-aliyun-sync/aps_scheduler.py
2026-04-13 18:09:52 +08:00

131 lines
4.5 KiB
Python

"""
Long-running scheduler — watches for new JSON files and syncs to MySQL.
Modes:
--watch : File watcher, syncs immediately when new JSON appears (default)
--cron : One-shot sync, meant to be called by system cron/launchd
--daemon : Combined: runs initial sync + watches for changes
"""
import time
import sys
import signal
import argparse
import logging
from pathlib import Path
from datetime import datetime
from aps_db_sync import APSSyncer, DB_CONFIG, JSON_DIR
LOG_FORMAT = "%(asctime)s [%(levelname)s] %(message)s"
logging.basicConfig(level=logging.INFO, format=LOG_FORMAT)
logger = logging.getLogger("aps_scheduler")
WATCH_INTERVAL_SECONDS = 30
PROCESSED_MARKER_DIR = JSON_DIR / ".aps_sync_processed"
def _update_watch_interval(value: int):
global WATCH_INTERVAL_SECONDS
WATCH_INTERVAL_SECONDS = value
class SyncScheduler:
def __init__(self, db_config: dict = None):
self.db_config = db_config or DB_CONFIG
self.running = True
PROCESSED_MARKER_DIR.mkdir(exist_ok=True)
signal.signal(signal.SIGINT, self._shutdown)
signal.signal(signal.SIGTERM, self._shutdown)
def _shutdown(self, signum, frame):
logger.info("Shutdown signal received, stopping...")
self.running = False
def _marker_path(self, json_path: Path) -> Path:
return PROCESSED_MARKER_DIR / f"{json_path.stem}.synced"
def _is_processed(self, json_path: Path) -> bool:
marker = self._marker_path(json_path)
if not marker.exists():
return False
marker_mtime = marker.stat().st_mtime
json_mtime = json_path.stat().st_mtime
return marker_mtime >= json_mtime
def _mark_processed(self, json_path: Path):
marker = self._marker_path(json_path)
marker.write_text(datetime.now().isoformat())
def find_unprocessed_files(self) -> list[Path]:
pattern = "aps_aliyun_customers_with_bills_*.json"
all_files = sorted(JSON_DIR.glob(pattern), key=lambda p: p.stat().st_mtime)
return [f for f in all_files if not self._is_processed(f)]
def sync_file(self, json_path: Path) -> bool:
logger.info("Syncing: %s", json_path.name)
try:
syncer = APSSyncer(db_config=self.db_config)
syncer.sync_from_json(str(json_path))
self._mark_processed(json_path)
return True
except Exception as e:
logger.error("Sync failed for %s: %s", json_path.name, e)
return False
def run_once(self):
unprocessed = self.find_unprocessed_files()
if not unprocessed:
logger.info("No unprocessed JSON files found")
return 0
count = 0
for f in unprocessed:
if self.sync_file(f):
count += 1
logger.info("Processed %d/%d files", count, len(unprocessed))
return count
def run_watch(self):
logger.info("Watching %s for new JSON files (interval=%ds)", JSON_DIR, WATCH_INTERVAL_SECONDS)
self.run_once()
while self.running:
time.sleep(WATCH_INTERVAL_SECONDS)
unprocessed = self.find_unprocessed_files()
for f in unprocessed:
if not self.running:
break
self.sync_file(f)
logger.info("Watcher stopped")
def main():
parser = argparse.ArgumentParser(description="APS Sync Scheduler")
parser.add_argument("--mode", choices=["watch", "cron", "daemon"], default="watch",
help="watch=file watcher, cron=one-shot, daemon=watch with initial sync")
parser.add_argument("--host", default=DB_CONFIG["host"])
parser.add_argument("--port", type=int, default=DB_CONFIG["port"])
parser.add_argument("--user", default=DB_CONFIG["user"])
parser.add_argument("--password", default=DB_CONFIG["password"])
parser.add_argument("--database", default=DB_CONFIG["database"])
parser.add_argument("--interval", type=int, default=WATCH_INTERVAL_SECONDS,
help="Watch interval in seconds")
args = parser.parse_args()
_update_watch_interval(args.interval)
config = {
"host": args.host, "port": args.port, "user": args.user,
"password": args.password, "database": args.database, "charset": "utf8mb4",
}
scheduler = SyncScheduler(db_config=config)
if args.mode == "cron":
count = scheduler.run_once()
sys.exit(0 if count >= 0 else 1)
else:
scheduler.run_watch()
if __name__ == "__main__":
main()