增加task历史记录生成功能

project_evaluation_analysis/generate_task_static_history.py

@@ -0,0 +1,116 @@
+import sys
+import os
+
+import config
+
+curPath = os.path.abspath(os.path.dirname(__file__))
+rootPath = os.path.split(curPath)[0]
+sys.path.append(rootPath)
+
+import warnings
+from datetime import datetime,timedelta
+from utils import get_conn, get_month_date
+warnings.filterwarnings("ignore")
+
+def execute_sql(sql:str):
+    conn = get_conn()
+    curs = conn.cursor()
+    curs.execute(sql)
+    conn.commit()
+    conn.close()
+def record_task_history():
+    #每次调用此函数，默认记录当前日期的里程碑数据，然后回顾最近2个月的数据，若2个月之前的数据不存在，则以随机数的方式生成填充，并在is_test字段标记为true
+
+    # 查看当前历史表中已经有的数据
+    now = datetime.now()
+    today_str=now.strftime('%Y-%m-%d')
+    today=datetime.strptime(today_str,'%Y-%m-%d')
+    delta = timedelta(days=60)
+    begin_date=now-delta
+    begin_date_str=begin_date.strftime('%Y-%m-%d')
+    # 取出数据库中已有的最后一天的日期
+    last_date_sql=f'''SELECT max(date) as last_date from (
+SELECT date from ai_task_static_history where date>'{begin_date_str}' GROUP BY date order by date
+) as a'''
+
+    conn = get_conn()
+    curs = conn.cursor()
+    curs.execute(last_date_sql)
+    last_date=curs.fetchone()[0]
+    last_date_str=""
+    if last_date is None:
+        last_date_str=(today-timedelta(days=60)).strftime('%Y-%m-%d')
+    else:
+        last_date_str=last_date.strftime('%Y-%m-%d')
+    #last_date_str='2023-02-01'
+    if last_date_str==today_str:
+        config.logger.info("当前日期的task数据已存在，跳过生成")
+        return
+
+    last_date=datetime.strptime(last_date_str,'%Y-%m-%d')
+
+    delta=timedelta(days=1)
+    generate_day=last_date+delta
+    config.logger.info(f"开始生成{generate_day}到{today}的数据")
+    while generate_day<today:
+        insert_test_task(generate_day)
+        generate_day = generate_day + delta
+
+    insert_today_task(today)
+    config.logger.info("task历史数据记录完成")
+
+
+
+def insert_test_task(generate_day:datetime):
+    generate_day_str=generate_day.strftime('%Y-%m-%d')
+    generate_sql=f"""
+    insert into ai_task_static_history (repo_id,closed_count,open_count,date,is_test)
+
+SELECT rc.repo_id,
+case when rc.closed_count is null then ceil(random()*10)
+else ceil(rc.closed_count+random()*10)
+END as closed_count,
+case when o.open_count is null then ceil(random()*10)
+else ceil(random()*10+o.open_count)
+END as open_count,
+'{generate_day_str}' as date,true as is_test from
+(
+SELECT r.repo_id,c.closed_count from (
+SELECT repo_id from issue GROUP BY repo_id ) as r LEFT JOIN (
+SELECT repo_id,count(repo_id) as closed_count from issue  where is_closed=true and is_pull=false GROUP BY repo_id
+) as c on r.repo_id=c.repo_id
+
+) as rc left JOIN (
+SELECT repo_id,count(repo_id) as open_count from issue  where is_closed=false and is_pull=false GROUP BY repo_id
+) as o on rc.repo_id=o.repo_id ORDER BY rc.repo_id"""
+
+    execute_sql(generate_sql)
+
+def insert_today_task(generate_day:datetime):
+    generate_day_str = generate_day.strftime('%Y-%m-%d')
+    generate_sql = f"""
+    insert into ai_task_static_history (repo_id,closed_count,open_count,date,is_test)
+        SELECT rc.repo_id,
+case when rc.closed_count is null then 0
+else rc.closed_count
+END as closed_count,
+case when o.open_count is null then 0
+else o.open_count
+END as open_count,
+'{generate_day_str}' as date,false as is_test from
+(
+SELECT r.repo_id,c.closed_count from (
+SELECT repo_id from issue GROUP BY repo_id ) as r LEFT JOIN (
+SELECT repo_id,count(repo_id) as closed_count from issue  where is_closed=true and is_pull=false GROUP BY repo_id
+) as c on r.repo_id=c.repo_id
+
+) as rc left JOIN (
+SELECT repo_id,count(repo_id) as open_count from issue  where is_closed=false and is_pull=false GROUP BY repo_id
+) as o on rc.repo_id=o.repo_id ORDER BY rc.repo_id
+        """
+
+    execute_sql(generate_sql)
+
+if __name__ == '__main__':
+    # 生成测试数据
+    record_task_history()