深圳积分商城网站建设,微信管理系统平台,本溪市城乡住房建设厅网站,app备案号查询平台官网Oracle - 多区间按权重取值逻辑 #xff0c;分时区-多层级-取配置方案https://blog.csdn.net/shijianduan1/article/details/133386281
某业务配置表#xff0c;按配置的时间区间及组织层级取方案#xff0c;形成报表展示出所有部门方案的取值#xff1b;
例如#xff0…Oracle - 多区间按权重取值逻辑 分时区-多层级-取配置方案https://blog.csdn.net/shijianduan1/article/details/133386281
某业务配置表按配置的时间区间及组织层级取方案形成报表展示出所有部门方案的取值
例如总公司配置20230101-20231231为方案3 分公司配置 20230301-20230731 为方案2部门配置20230601-20231031 为方案1配置优先级为 部门 分公司总公司 即啥都没配则使用总公司默认值 继上一篇拆分多区间后 再次进行优化目前有阶段性成果先记录一下等后续再有改进再发。
上一篇中 对日期区间的边界日期做了单独的拆分没有将边界日期和其临界的 日期合并。
本篇中会对其进行优化 单目前基础脚本只支持2个优先级如多个优先级则需要 增加多次查询
数据源 L:优先级23 , B1 开始日期, E1 结束日期
注意 这里默认数据源的 日期区间不会重复
拆分后结果 以下是代码 WITH S AS(SELECT 3 AS L, 20230101 AS B1, 20230821 AS E1FROM DUALUNION ALLSELECT 3 AS L, 20230822 AS B1, 20231231 AS E1FROM DUALUNION ALLSELECT 2 AS L, 20230301 AS B1, 20230731 AS E1FROM DUALUNION ALLSELECT 2 AS L, 20230901 AS B1, 20230931 AS E1FROM DUAL UNION ALLSELECT 2 AS L, 20230401 AS B1, 20230631 AS E1FROM DUAL)
,A AS(SELECT S.*, ROW_NUMBER() OVER(PARTITION BY L ORDER BY B1) AS N FROM S) --对优先级重新排序确保优先级差1
,A2 AS(SELECT A.*, NVL(PRE.E1, 0) B_PRE, NVL(NEX.B1, 20991231) AS E_NEXFROM ALEFT JOIN A PRE ON A.N PRE.N 1 AND PRE.L A.LLEFT JOIN A NEX ON A.N NEX.N - 1 AND NEX.L A.LWHERE A.L 2 ) , --同级别的优先级也是有边界的
A3 AS(SELECT * FROM A WHERE L 3)
SELECT A3.L, GREATEST(A3.B1, A2.B_PRE)as B2, A2.B1 as E2 FROM A2, A3 WHERE A3.B1 A2.B1 AND A3.E1 A2.B1
UNION ALL
SELECT A3.L, A2.E1 as B2, LEAST(A3.E1, A2.E_NEX) AS E2 FROM A2, A3 WHERE A3.B1 A2.E1 AND A3.E1 A2.E1
UNION ALL select L, B1, E1 from A2
order by L, B2;
