重构现代化 FastAPI 后端项目框架

scripts/auto_realtime.py

@@ -0,0 +1,156 @@
+from logging.handlers import TimedRotatingFileHandler
+import influxdb_api
+import os
+import logging
+import globals
+from datetime import datetime, timedelta, timezone
+import schedule
+import time
+import shutil
+from influxdb_client import InfluxDBClient, BucketsApi, WriteApi, OrganizationsApi, Point, QueryApi
+import simulation
+import influxdb_info
+import project_info
+
+def setup_logger():
+    # 创建日志目录
+    log_dir = "logs"
+    os.makedirs(log_dir, exist_ok=True)
+
+    # 配置基础日志格式
+    log_format = "%(asctime)s - %(levelname)s - %(message)s"
+    formatter = logging.Formatter(log_format)
+
+    # 创建主 Logger
+    logger = logging.getLogger()
+    logger.setLevel(logging.INFO)  # 全局日志级别
+
+    # --- 1. 按日期分割的日志文件 Handler ---
+    log_file = os.path.join(log_dir, "simulation.log")
+    file_handler = TimedRotatingFileHandler(
+        filename=log_file,
+        when="midnight",  # 每天午夜轮转
+        interval=1,
+        backupCount=7,
+        encoding="utf-8"
+    )
+    file_handler.suffix = "simulation-%Y-%m-%d.log"  # 文件名格式
+    file_handler.setFormatter(formatter)
+    file_handler.setLevel(logging.INFO)        # 文件记录所有级别日志
+
+    # --- 2. 控制台实时输出 Handler ---
+    console_handler = logging.StreamHandler()  # 默认输出到 sys.stderr (控制台)
+    console_handler.setFormatter(formatter)
+    console_handler.setLevel(logging.INFO)     # 控制台仅显示 INFO 及以上级别
+
+    # 将 Handler 添加到 Logger
+    logger.addHandler(file_handler)
+    #logger.addHandler(console_handler)
+
+    return logger
+
+logger = setup_logger()
+
+# 2025/02/01
+def get_next_time() -> str:
+    """
+    获取下一个1分钟时间点，返回格式为字符串'YYYY-MM-DDTHH:MM:00+08:00'
+    :return: 返回字符串格式的时间，表示下一个1分钟的时间点
+    """
+    # 获取当前时间，并设定为北京时间
+    now = datetime.now() # now 类型为 datetime，表示当前本地时间
+    # 获取当前的分钟，并且将秒和微秒置为零
+    current_time = now.replace(second=0, microsecond=0) # current_time 类型为 datetime，时间的秒和微秒部分被清除
+    return current_time.strftime('%Y-%m-%dT%H:%M:%S+08:00')
+
+
+# 2025/02/06
+def store_realtime_SCADA_data_job() -> None:
+    """
+    定义的任务1，每分钟执行1次，每次执行时，更新get_real_value_time并调用store_realtime_SCADA_data_to_influxdb函数
+    :return: None
+    """
+    # 获取当前时间并更新get_real_value_time，转换为字符串格式
+    get_real_value_time: str = get_next_time()  # get_real_value_time 类型为 str，格式为'2025-02-01T18:45:00+08:00'
+    # 调用函数执行任务
+    influxdb_api.store_realtime_SCADA_data_to_influxdb(get_real_value_time)
+    logger.info('{} -- Successfully store realtime SCADA data.'.format(datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
+
+
+# 2025/02/06
+def get_next_15minute_time() -> str:
+    """
+    获取下一个15分钟的时间点，返回格式为字符串'YYYY-MM-DDTHH:MM:00+08:00'
+    :return: 返回字符串格式的时间，表示下一个15分钟执行时间点
+    """
+    now = datetime.now()
+    # 向上舍入到下一个15分钟
+    next_15minute = (now.minute // 15 + 1) * 15 - 15
+    if next_15minute == 60:
+        next_15minute = 0
+        now = now + timedelta(hours=1)
+    next_time = now.replace(minute=next_15minute, second=0, microsecond=0)
+    return next_time.strftime('%Y-%m-%dT%H:%M:%S+08:00')
+
+
+# 2025/02/07
+def run_simulation_job() -> None:
+    """
+    定义的任务3，每15分钟执行一次在store_realtime_SCADA_data_to_influxdb之后执行run_simulation。
+    :return: None
+    """
+    # 获取当前时间，并检查是否是整点15分钟
+    current_time = datetime.now()
+    if current_time.minute % 15 == 0:
+        print(f"{current_time.strftime('%Y-%m-%d %H:%M:%S')} -- Start simulation task.")
+        # 计算前，获取scada_info中的信息，按照设定的方法修改pg数据库
+        simulation.query_corresponding_element_id_and_query_id(project_info.name)
+        simulation.query_corresponding_pattern_id_and_query_id(project_info.name)
+        region_result = simulation.query_non_realtime_region(project_info.name)
+        globals.source_outflow_region_id = simulation.get_source_outflow_region_id(project_info.name, region_result)
+        globals.realtime_region_pipe_flow_and_demand_id = simulation.query_realtime_region_pipe_flow_and_demand_id(project_info.name, region_result)
+        globals.pipe_flow_region_patterns = simulation.query_pipe_flow_region_patterns(project_info.name)
+        globals.non_realtime_region_patterns = simulation.query_non_realtime_region_patterns(project_info.name, region_result)
+        globals.source_outflow_region_patterns, realtime_region_pipe_flow_and_demand_patterns = simulation.get_realtime_region_patterns(project_info.name,
+                                                                                                globals.source_outflow_region_id,
+                                                                                                globals.realtime_region_pipe_flow_and_demand_id)
+        modify_pattern_start_time: str = get_next_15minute_time()  # 获取下一个15分钟时间点
+        # print(modify_pattern_start_time)
+        simulation.run_simulation(name=project_info.name, simulation_type="realtime", modify_pattern_start_time=modify_pattern_start_time)
+
+        logger.info('{} -- Successfully run simulation and store realtime simulation result.'.format(datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
+    else:
+        logger.info(f"{current_time.strftime('%Y-%m-%d %H:%M:%S')} -- Skipping the simulation task.")
+
+
+# 2025/02/06
+def realtime_task() -> None:
+    """
+    定时执行任务1和，使用schedule库每1分钟执行一次store_realtime_SCADA_data_job函数。
+    该任务会一直运行，定期调用store_realtime_SCADA_data_job获取SCADA数据。
+    :return:
+    """
+    # 等待到整分对齐
+    now = datetime.now()
+    wait_seconds = 60 - now.second
+    time.sleep(wait_seconds)
+    # 使用 .at(":00") 指定在每分钟的第0秒执行
+    schedule.every(1).minute.at(":00").do(store_realtime_SCADA_data_job)
+    # 每15分钟执行一次run_simulation_job
+    schedule.every(1).minute.at(":00").do(run_simulation_job)
+    # 持续执行任务，检查是否有待执行的任务
+    while True:
+        schedule.run_pending()  # 执行所有待处理的定时任务
+        time.sleep(1)  # 暂停1秒，避免过于频繁的任务检查
+
+
+if __name__ == "__main__":
+    url = influxdb_info.url
+    token = influxdb_info.token
+    org_name = influxdb_info.org
+
+    client = InfluxDBClient(url=url, token=token)
+    # step2: 先查询pg数据库中scada_info的信息，然后存储SCADA数据到SCADA_data这个bucket里
+    influxdb_api.query_pg_scada_info_realtime(project_info.name)
+    # 自动执行
+    realtime_task()