Question

我正在使用ORACLE 11g DB，通过SQL Dev工具无法使用带有“Distinct”子句的Collect函数。在我的程序中使用时，它无法识别！

我的查询参考：

SELECT nvl(spicd.company_code, '') companycode
     , nvl(scc.company_description, '') companydesc
     , nvl(spicd.plant_code, '') plantcode
     , CAST(COLLECT(DISTINCT svh.haulier_code) AS varchar2_ntt) hauliercode
     , CAST(COLLECT(DISTINCT sh.hauier_name) AS varchar2_ntt) hauliername
  FROM saistb_company_code         scc
     , saistb_pve_indv_contact_det spicd

  LEFT OUTER JOIN saistb_vendor_haulier svh
    ON svh.company_code = spicd.company_code
   AND svh.plant_code = spicd.plant_code
   AND svh.vendor_code = spicd.vendor_code

  LEFT OUTER JOIN saistb_haulier sh
    ON sh.haulier_code = svh.haulier_code

 WHERE scc.company_code = spicd.company_code
   AND spicd.company_code LIKE <<companycode>>
   AND spicd.plant_code LIKE <<plantcode>>

 GROUP BY nvl(spicd.company_code, '')
        , nvl(scc.company_description, '')
        , nvl(spicd.plant_code, '')

此处varchar2_ntt是：

create or replace TYPE varchar2_ntt AS TABLE OF VARCHAR2(4000);

返回的错误是：

Error(49,6): PL/SQL: ORA-30482: DISTINCT option not allowed for this function

Answer 1

这是Oracle中一个老问题，其中SQL和PL / SQL引擎解析器不同，在这种情况下，PL / SQL不允许您在聚合函数中使用distinct。请参阅2003年this AskTom对此的回应。

他的建议是使用动态打开的引用游标。其他解决方案是使用外部查询（如Glenn的答案）包装您的查询，或者，我的偏好，将其转换为视图并从中进行选择。使用动态SQL也适用于这种情况。

Answer 2

如何将它包装在子查询中？

    SELECT companyCode, companyDesc, plantCode, 
           cast(COLLECT(haulierCode) as varchar2_ntt) haulierCode,
           cast(COLLECT(haulierName) as varchar2_ntt) haulierName
    FROM (

    SELECT
    nvl(spicd.COMPANY_CODE,'') companyCode,
    nvl(scc.COMPANY_DESCRIPTION,'') companyDesc,    
    nvl(spicd.PLANT_CODE,'') plantCode,
    nvl(sh.HAULIER_CODE, 'UNKNOWN HAULIER CODE') haulierCode,
    nvl(sh.haulier_name, 'UNKNOWN HAULIER NAME') haulierName
    from
    SAISTB_COMPANY_CODE scc,
    SAISTB_PVE_INDV_CONTACT_DET spicd

    left outer join SAISTB_VENDOR_HAULIER svh on
    svh.COMPANY_CODE = spicd.COMPANY_CODE and
    svh.PLANT_CODE = spicd.PLANT_CODE and
    svh.VENDOR_CODE = spicd.VENDOR_CODE

    left outer join SAISTB_HAULIER sh on
    sh.HAULIER_CODE = svh.HAULIER_CODE

    where
    scc.COMPANY_CODE = spicd.COMPANY_CODE
    and spicd.COMPANY_CODE like <<CompanyCode>>
    and spicd.PLANT_CODE like <<PlantCode>>

    group by
    nvl(spicd.COMPANY_CODE,''),         
    nvl(scc.COMPANY_DESCRIPTION,''),    
    nvl(spicd.PLANT_CODE,''),
    nvl(sh.HAULIER_CODE, 'UNKNOWN HAULIER CODE') haulierCode,
    nvl(sh.haulier_name, 'UNKNOWN HAULIER NAME') haulierName

    )

    GROUP BY companyCode, companyDesc, plantCode

内部组执行“不同”操作。

Oracle 11g和收集功能

2 个答案: