Question

我终于在最后对数据进行分组时写了一篇文章。基本上，我已经设计了查询，现在我正在获取我想要查看的所有数据，但现在我需要按清单编号（man_no列）对结果进行分组。基本上，如果结果只显示唯一清单编号的第一个实例，我会很高兴，因为后续实例实际上是相同的数据。

有什么建议吗？正如您从数据中看到的那样，这些记录实际上是相同的数据，但在＆＃34;托运＆＃34;专栏结果......我可能只是写得不好？

QUERY：

    SELECT landing_jobs.man_no, ltrim(rtrim(brun.veh_code)) as Run_Vehicle,
    ltrim(rtrim(landing_jobs.col_zone)) + ' - ' + ltrim(rtrim(landing_jobs.del_zone))         as coldel,
    CASE
    WHEN con2.long_no <> ' ' and landing_jh.est_time<=job_history2.est_time and         jmaster.col_state='SA' THEN dateadd(ss,1800,job_history2.est_time)
    WHEN (con2.long_no <> ' ' and landing_jh.est_time>job_history2.est_time and jmaster.col_state='SA') OR (con2.long_no IS null and jmaster.col_state='SA') then dateadd(ss,1800,Landing_JH.est_time)
    WHEN con2.long_no <> ' ' and landing_jh.est_time<=job_history2.est_time and jmaster.col_state='WA' THEN dateadd(ss,7200,job_history2.est_time)
    WHEN (con2.long_no <> ' ' and landing_jh.est_time>job_history2.est_time and jmaster.col_state='WA') OR (con2.long_no IS null and jmaster.col_state='WA') then dateadd(ss,7200,Landing_JH.est_time)
    WHEN con2.long_no <> ' ' and landing_jh.est_time<=job_history2.est_time and jmaster.col_state='QLD' THEN dateadd(ss,0,job_history2.est_time)
    WHEN (con2.long_no <> ' ' and landing_jh.est_time>job_history2.est_time and jmaster.col_state='QLD') OR (con2.long_no IS null and jmaster.col_state='QLD') then dateadd(ss,0,Landing_JH.est_time)
    ELSE landing_jh.est_time
    END AS dep_mel_time,

    CASE 
    WHEN con2.del_code<>' ' and con2.del_code<>landing_jobs.del_code then         LTRIM(rtrim(landing_jobs.del_code)) + ' / ' + LTRIM(rtrim(con2.del_code)) 
    ELSE ltrim(rtrim(landing_jobs.del_code))
    END AS deliveries,

    CASE
    WHEN con3.long_no<>' ' THEN LTRIM(RTRIM(landing_jobs.long_no)) + '-' +         LTRIM(rtrim(landing_jobs.cus_code)) + ' and ' + LTRIM(rtrim(con2.long_no)) + '-' + LTRIM(rtrim(con2.cus_code)) + ' and ' + LTRIM(rtrim(con3.long_no)) + '-' + LTRIM(rtrim(con3.cus_code))
    WHEN con2.long_no<>' ' THEN LTRIM(RTRIM(landing_jobs.long_no)) + '-' + LTRIM(rtrim(landing_jobs.cus_code)) + ' and ' + LTRIM(rtrim(con2.long_no)) + '-' + LTRIM(rtrim(con2.cus_code))
    ELSE LTRIM(rtrim(landing_jobs.long_no)) + '-' + ltrim(rtrim(Landing_Jobs.cus_code))
    END AS consignments,

    ltrim(rtrim(landing_jobs.trl_code1)) as con1trl1,
    ltrim(rtrim(landing_jobs.trl_code2)) as con1trl2,
    ltrim(rtrim(con2.trl_code1)) as con2trl1,
    ltrim(rtrim(con2.trl_code2)) as con2trl2,
    ltrim(rtrim(driver2.name))

    FROM landing_jobs left outer join landing_runs brun on landing_jobs.man_no=brun.job_no left outer join landing_driver driver2 on brun.drv_code=driver2.code left outer join landing_equipment on landing_equipment.code=brun.veh_code left outer join landing_driver on landing_driver.code=brun.drv_code left outer join landing_JH on landing_jobs.job_no=landing_JH.job_no and landing_JH.job_status='C' left outer join landing_jobs con2 on landing_jobs.man_no=con2.man_no and landing_jobs.job_no<>con2.job_no left outer join landing_JH job_history2 on job_history2.job_no=con2.job_no and job_history2.job_no<>landing_JH.job_no and job_history2.job_status='C' left outer join landing_jobs con3 on landing_jobs.man_no=con3.man_no and landing_jobs.job_no<>con3.job_no and con3.job_no<>con2.job_no LEFT join landing_jobs jmaster on landing_jobs.mas_no=jmaster.job_no
    WHERE landing_jobs.type='JL' and brun.work_code<>'LOCAL' and brun.veh_code<>'SC' and brun.veh_code<>' ' AND landing_jobs.col_zone<>landing_jobs.del_zone AND jmaster.srv_code<>'B/T' and landing_jobs.leg_item=1
    AND landing_jh.est_time >= '2013-07-01'


    order by dep_mel_time

返回值：

 man_no     Run_Vehicle     coldel     dep_mel_time     deliveries        consignments                        con1trl1     con1trl2     con2trl1     con2trl2     Driver
 349802     091V            ADE - WAGG 7/05/2014 18:00  WAGGA             349835JL#1-COPE and 349832JL#1-BFL     445       2541                                   EY Peter
 349802     091V            ADE - WAGG 7/05/2014 18:00  WAGGA             349832JL#1-BFL and 349835JL#1-COPE     2541                   445                       EY Peter
 349792     010V            ADE - MEL  7/05/2014 18:00  COPEMEL           349808JL#1-BFL and 349806JL#1-COPE     2005                   430                       THIELE Laura
 349792     010V            ADE - MEL  7/05/2014 18:00  COPEMEL           349806JL#1-COPE and 349808JL#1-BFL     430       2005                                   THIELE Laura
 349799     117V            ADE - MEL  7/05/2014 18:00  BFLMEL / POLYDAN  349825JL#1-BFL and 349823JL#1-POLYAIRE 2611                   5580                      DESTRO Bill
 349799     117V            ADE - MEL  7/05/2014 18:00  POLYDAN / BFLMEL  349823JL#1-POLYAIRE and 349825JL#1-BFL 5580      2611                                   DESTRO Bill
 349516     259             MEL - SYD  7/05/2014 19:00  CAMSYD            349455JL#1-CAMERON                     2590      688          NULL        NULL          WAIN Sarah
 349794     113V            ADE - NHIL 7/05/2014 19:00  NHILL             349820JL#1-WA FR                       2608      643          NULL        NULL          STEPNELL William
 349766     047K            SYD - MEL  7/05/2014 19:00  COSRIN            349686JL#1-COSTCO                      CC2622    CC5504       NULL        NULL          WINTERTON Mike
 349753     039K            SYD - MEL  7/05/2014 19:00  WAFGMELB          349661JL#1-WA FR                       2619      716          NULL        NULL          SIDHU Marpreet

Answer 1

您可以使用row_number()执行所需操作。我建议：

with cte as (<your query here (without the `order by`)>)
select t.*
from (select cte.*, row_number() over (partition by man_no order by dep_mel_time desc) as seqnum
      from cte
     ) t
where seqnum = 1;

作为一个注释，给定man_no的记录对我来说看起来不像重复。我看到列中有很多不同之处。

SQL - 分组数据

1 个答案: