knowledge/structure/worker.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
单线程结构化处理 Worker
通过 Docker 的副本数实现并发，不再使用应用内多线程调度器。
"""

import os
import sys
import time
import signal

sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))

from utils.logging_config import get_logger
from structure.structure_processor import StructureProcessor


class StructureWorker:
    def __init__(self,
                 interval_seconds: int = 120,
                 idle_sleep_seconds: int = 10):
        self.interval_seconds = interval_seconds
        self.idle_sleep_seconds = idle_sleep_seconds
        self.running = True
        self.logger = get_logger('StructureWorker')
        self.processor = StructureProcessor()

        signal.signal(signal.SIGINT, self._handle_signal)
        signal.signal(signal.SIGTERM, self._handle_signal)

    def _handle_signal(self, signum, frame):
        self.logger.info(f"收到信号 {signum}，准备优雅退出…")
        self.running = False

    def run_forever(self):
        self.logger.info(f"启动 StructureWorker，间隔: {self.interval_seconds}s，空闲等待: {self.idle_sleep_seconds}s")
        while self.running:
            start_time = time.time()
            try:
                processed = self.processor.process_single_record()
                if not processed:
                    time.sleep(self.idle_sleep_seconds)
                else:
                    elapsed = time.time() - start_time
                    wait_time = max(0, self.interval_seconds - elapsed)
                    if wait_time > 0:
                        time.sleep(wait_time)
            except Exception as exc:
                self.logger.error(f"处理异常: {exc}", exc_info=True)
                time.sleep(self.idle_sleep_seconds)


def main():
    interval_seconds = int(os.getenv('INTERVAL_SECONDS', '120'))
    idle_sleep_seconds = int(os.getenv('IDLE_SLEEP_SECONDS', '10'))
    worker = StructureWorker(interval_seconds=interval_seconds,
                             idle_sleep_seconds=idle_sleep_seconds)
    worker.run_forever()


if __name__ == '__main__':
    main()