SQL查询

Question

我有一个查询，会删除重复assignment_name, effective_start_date, effective_end_date, effective_latest_change的重复行。员工的详细信息有两个表格

xx_as_table ：

Ass_nmber    eff_s_date             eff_e_date effective_l_change     special_ceiling         date_Start       employer_name         worker_type
-------------------------------------------------------------------------------------
10228        01-jan-1990           31-dec-4172             1                        0             01-feb-1989           XYZ Ltd           E
10228        03-feb-1989           31-dec-1989             2                        0             01-feb-1989           XYZ Ltd           E

xx_wrk_table

ass_number    date_Start       employer_name     worker_type
------------------------------------------------------------
10228         01-feb-1989      XYZ Ltd              E
10228         01-jan-1991      ABC Ltd              P 

现在，我想要一个检索上述数据的查询，如下所示：

Ass_nmber    eff_s_date          eff_e_date           eff_l_change     special_ceiling  date_Start       employer_name         worker_type
---------------------------------------------------------------------------------------------    
10228        01-jan-1990           31-dec-4172             1                        0             01-feb-1989           XYZ Ltd           E
10228        03-feb-1989           31-dec-1989             2                        0             01-feb-1989           XYZ Ltd           E

我想要xx_wrk_table中的date_start和min date_start以及针对min min_Start的详细信息。并且没有重复的行应该具有相同的assignment_name，effective_start_date，effective_end_date，effective_latest_change

为此，我写了一个查询：

select 
    ass_number,
    regexp_substr(ass_number, '[0-9]+') person_number,
    effective_start_date,
    effective_end_date,
    nvl((select min (wrk.date_start)
from   
    XX_WRK_STG
where
    WRK.PERSON_NUMBER = regexp_substr(assignment_number, '[0-9]+')), effective_start_date) D, t.effective_start_date),
        nvl(wrk.worker_type, 'E'),
        nvl(wrk.employer_name, 'XYZ'),
        effective_latest_change

      from  (SELECT xx_as_table.*,
               COUNT(*) OVER (PARTITION BY assignment_name, 
                                          effective_start_date,
                                          effective_end_date,
                                          effective_latest_change) AS c
        FROM   xx_as_table) T
 left outer join 
    XX_WRK_STG  wrk
on regexp_substr(t.assignment_name, '[0-9]+')=wrk.person_number
where c=1; 

但是这个查询不起作用，并且针对每个ass_number从wrk_table中检索到多个数据。如何修改查询以获得预期结果？

Answer 1

我无法完全理解您的表结构，因为在您的示例和书面查询中，您使用了不同的列名和一些其他列。 我还是试着帮忙。要查找具有最小日期的工作表中的行，您可以使用Rank function。

SELECT x_tab.Ass_nmber,  
       x_tab.eff_s_date,  
       x_tab.eff_e_date,  
       x_tab.effective_l_change,  
       x_tab.special_ceiling,  
       work_tab.date_Start,  
       work_tab.employer_name,  
       work_tab.worker_type  
  FROM (SELECT Ass_nmber,  
               eff_s_date,  
               eff_e_date,  
               effective_l_change,  
               special_ceiling,  
               COUNT(*) OVER (PARTITION BY Ass_nmber,  
                                           eff_s_date,  
                                           eff_e_date,  
                                           effective_l_change) cnt  
           FROM xx_as_table)x_tab,  
        (SELECT ass_number,  
                date_Start,  
                employer_name,  
                worker_type,  
                person_number,  
                RANK() OVER(PARTITION BY ass_number  
                                ORDER BY date_Start DESC) rnk  
           FROM xx_wrk_table) work_tab  
   WHERE s_tab.cnt = 1  
     AND work_tab.rnk = 1  
     AND regexp_substr(s_tab.Ass_nmber, '[0-9]+') = work_tab.person_number;  

如果它没有帮助，请提供正确的创建和插入语句。这将有助于其他人帮助你。感谢。