|
|
@@ -0,0 +1,151 @@
|
|
|
+# -*- coding: utf-8 -*-
|
|
|
+from odps import ODPS
|
|
|
+from threading import Timer
|
|
|
+import threading
|
|
|
+from utils import RedisHelper, execute_sql_from_odps
|
|
|
+from config import set_config
|
|
|
+from log import Log
|
|
|
+import json
|
|
|
+from datetime import datetime
|
|
|
+from queue import Queue
|
|
|
+from tqdm import tqdm
|
|
|
+import sys
|
|
|
+
|
|
|
+config_, _ = set_config()
|
|
|
+log_ = Log()
|
|
|
+redis_helper = RedisHelper()
|
|
|
+
|
|
|
+REDIS_PREFIX = "redis:vid_title_tags:"
|
|
|
+EXPIRE_TIME = 6 * 3600
|
|
|
+
|
|
|
+def worker(queue, executor):
|
|
|
+ while True:
|
|
|
+ row = queue.get()
|
|
|
+ if row is None: # 结束信号
|
|
|
+ queue.task_done()
|
|
|
+ break
|
|
|
+ executor(row)
|
|
|
+ queue.task_done()
|
|
|
+def records_process_for_list(records, executor, max_size=50, num_workers=10):
|
|
|
+ # 创建一个线程安全的队列
|
|
|
+ queue = Queue(maxsize=max_size) # 可以调整 maxsize 以控制内存使用
|
|
|
+ # 设置线程池大小
|
|
|
+ num_workers = num_workers
|
|
|
+ # 启动工作线程
|
|
|
+ threads = []
|
|
|
+ for _ in range(num_workers):
|
|
|
+ t = threading.Thread(target=worker, args=(queue, executor))
|
|
|
+ t.start()
|
|
|
+ threads.append(t)
|
|
|
+ # 读取数据并放入队列
|
|
|
+ for row in tqdm(records):
|
|
|
+ queue.put(row)
|
|
|
+ # 发送结束信号
|
|
|
+ for _ in range(num_workers):
|
|
|
+ queue.put(None)
|
|
|
+ # 等待所有任务完成
|
|
|
+ queue.join()
|
|
|
+ # 等待所有工作线程结束
|
|
|
+ for t in threads:
|
|
|
+ t.join()
|
|
|
+def process_and_store(row):
|
|
|
+ video_id, json_str = row
|
|
|
+ key = REDIS_PREFIX + str(video_id)
|
|
|
+ expire_time = EXPIRE_TIME
|
|
|
+ redis_helper.set_data_to_redis(key, json_str, expire_time)
|
|
|
+
|
|
|
+def check_data(project, table, date, hour) -> int:
|
|
|
+ """检查数据是否准备好,输出数据条数"""
|
|
|
+ odps = ODPS(
|
|
|
+ access_id=config_.ODPS_CONFIG['ACCESSID'],
|
|
|
+ secret_access_key=config_.ODPS_CONFIG['ACCESSKEY'],
|
|
|
+ project=project,
|
|
|
+ endpoint=config_.ODPS_CONFIG['ENDPOINT'],
|
|
|
+ connect_timeout=3000,
|
|
|
+ read_timeout=500000,
|
|
|
+ pool_maxsize=1000,
|
|
|
+ pool_connections=1000
|
|
|
+ )
|
|
|
+ try:
|
|
|
+ t = odps.get_table(name=table)
|
|
|
+ log_.info(f"检查分区是否存在-【 dt={date} hh={hour}】")
|
|
|
+ check_res = t.exist_partition(partition_spec=f'dt={date}/hh={hour}')
|
|
|
+ if check_res:
|
|
|
+ sql = f'select * from {project}.{table} where dt = {date} and hh = {hour}'
|
|
|
+ log_.info(sql)
|
|
|
+ with odps.execute_sql(sql=sql).open_reader() as reader:
|
|
|
+ data_count = reader.count
|
|
|
+ else:
|
|
|
+ log_.info("表{}分区{}/{}不存在".format(table, date, hour))
|
|
|
+ data_count = 0
|
|
|
+ except Exception as e:
|
|
|
+ log_.error("table:{},date:{},hour:{} no data. return data_count=0,报错原因是:{}".format(table, date, hour, e))
|
|
|
+ data_count = 0
|
|
|
+ return data_count
|
|
|
+
|
|
|
+def get_sql(project, table, date, hour):
|
|
|
+ sql = '''
|
|
|
+ SELECT vid
|
|
|
+ ,tags
|
|
|
+ ,title
|
|
|
+ FROM {}.{}
|
|
|
+ WHERE dt = '{}'
|
|
|
+ and hh = '{}'
|
|
|
+ '''.format(
|
|
|
+ project, table, date, hour
|
|
|
+ )
|
|
|
+ print("sql:" + sql)
|
|
|
+ records = execute_sql_from_odps(project=project, sql=sql)
|
|
|
+ video_list = []
|
|
|
+ with records.open_reader() as reader:
|
|
|
+ for record in reader:
|
|
|
+ video_id = record['vid']
|
|
|
+ m = dict()
|
|
|
+ try:
|
|
|
+ m["tags"] = record['tags']
|
|
|
+ except Exception as e:
|
|
|
+ log_.error(e)
|
|
|
+ try:
|
|
|
+ m["title"] = record['title']
|
|
|
+ except Exception as e:
|
|
|
+ log_.error(e)
|
|
|
+
|
|
|
+ json_str = json.dumps(m)
|
|
|
+ video_list.append([video_id, json_str])
|
|
|
+ return video_list
|
|
|
+
|
|
|
+
|
|
|
+def main():
|
|
|
+ try:
|
|
|
+ date = sys.argv[1]
|
|
|
+ hour = sys.argv[2]
|
|
|
+ except Exception as e:
|
|
|
+ date = datetime.now().strftime('%Y%m%d')
|
|
|
+ hour = datetime.now().hour
|
|
|
+ log_.info("没有读取到参数，采用系统时间：{}".format(e))
|
|
|
+ log_.info("使用时间参数-日期：{}，小时：{}".format(date, str(hour)))
|
|
|
+ # 1 判断上游数据表是否生产完成
|
|
|
+ project = "loghubods"
|
|
|
+ table = "alg_recsys_feature_title_tags"
|
|
|
+ table_data_cnt = check_data(project, table, date, hour)
|
|
|
+ if table_data_cnt == 0:
|
|
|
+ log_.info("上游数据{}未就绪{}/{}，等待...".format(table, date, hour))
|
|
|
+ Timer(60, main).start()
|
|
|
+ else:
|
|
|
+ log_.info("上游数据就绪，count={}，开始读取数据表".format(table_data_cnt))
|
|
|
+ # 2 读取数据表 处理特征
|
|
|
+ video_list = get_sql(project, table, date, hour)
|
|
|
+ # 3 写入redis
|
|
|
+ log_.info("video的数据量:{}".format(len(video_list)))
|
|
|
+ records_process_for_list(video_list, process_and_store, max_size=100, num_workers=8)
|
|
|
+
|
|
|
+if __name__ == '__main__':
|
|
|
+ log_.info("开始执行:" + datetime.now().strftime("%Y-%m-%d %H:%M:%S"))
|
|
|
+ main()
|
|
|
+ log_.info("完成执行:" + datetime.now().strftime("%Y-%m-%d %H:%M:%S"))
|
|
|
+
|
|
|
+
|
|
|
+
|
|
|
+# previous_date_str = (datetime.strptime(date, "%Y%m%d") - timedelta(days=1)).strftime("%Y%m%d")
|
|
|
+# cd /root/zhangbo/rov-offline
|
|
|
+# python alg_recsys_recall_01_vid2titletags_redis.py 20240508 13
|
zhangbo