sql - 具有变量的普遍 SQL 子查询

问题描述

为了在不到几分钟的执行时间内获得我需要的数据，我有 2 个单独的查询：

select
    ' ' as POHI_ACD,
    LTRIM(RTRIM(OePo_Header.PO_ID)) as POHI_PONUM,
    ' ' as POHI_POTYP,
    ' ' as POHI_VND,
    LTRIM(RTRIM(OePo_Header.CLIENT_ID)) as POHI_CO,
    'B1' as POHI_WA,
    'n/a' as PODI_COMMNT, -- Comment filled in separate query
    ' ' as POHI_CUSER,
    ' ' as POHI_TSCRT,
    ' ' as POHI_PGMUP,
    ' ' as POHI_USRUP,
    '01/01/0001 00:00:00' as POHI_TSLUP,
    ' ' as POHI_FILEPATH
from (
        OePo_Header
        inner join OePo_Item 
        on OePo_Header.PO_ID = OePo_Item.PO_ID
    )
    left join OePo_Remarks
    on OePo_Header.WHSE_INST_ID = OePo_Remarks.AUTOKEY
    and OePo_Header.PO_ID = OePo_Remarks.PO_ID
where (
        (OePo_Item.UNITQTY_ORDER = 1)
        and (OePo_Header.SHIPTO_ID like '%*B1')
        and(OePo_Item.CLOSED_FLG = 'N')
        and(OePo_Item.INVOICED_COUNT = 0)
        and(OePo_Item.ITEM_TYPE = '2')
        and(OePo_Item.VENDOR_ID <> 'WHSE')
    )
group by
    LTRIM(RTRIM(OePo_Header.PO_ID)),
    LTRIM(RTRIM(OePo_Header.CLIENT_ID)),
    'B1',
    if(TEXT = Null, '', TEXT),
    '',
    '',
    OEPO_HEADER.PO_ID
order by
    OEPO_HEADER.PO_ID;

和

select
    OePo_Remarks.TEXT as POHI_COMMNT
from (
    OePo_Remarks
    inner join OePo_Header
    on OePo_Remarks.AUTOKEY = OePo_Header.WHSE_INST_ID
    and OePo_Remarks.PO_ID = OePo_Header.PO_ID
    )
where
    OePo_Remarks.PO_ID like '%208672%' -- fill in with appropriate PO num from outer code, 208672 is just an example. (PO_IDs can be padded, hence the %s)
    OePo_Remarks.zflags like '%w%'

我将如何设置一个子查询，它使用来自外部查询的数据来获取我正在寻找的内容，而不是组合调用这两个查询的代码中的数据？就像是：

select
    ' ' as POHI_ACD,
    LTRIM(RTRIM(OePo_Header.PO_ID)) as POHI_PONUM,
    ' ' as POHI_POTYP,
    ' ' as POHI_VND,
    LTRIM(RTRIM(OePo_Header.CLIENT_ID)) as POHI_CO,
    'B1' as POHI_WA,
    -- subquery start
    ( select
        OePo_Remarks.TEXT
    from (
        OePo_Remarks
        inner join OePo_Header
        on OePo_Remarks.AUTOKEY = OePo_Header.WHSE_INST_ID
        and OePo_Remarks.PO_ID = OePo_Header.PO_ID
        )
    where (
        OePo_Remarks.PO_ID like CONCAT('%', POHI_PONUM, '%') -- Here is where I need the correct PONUM from the outer query.
        and OePo_Remarks.zflags like '%w%'
        )
    ) as POHI_COMMNT, --subquery end
    ' ' as POHI_CUSER,
    ' ' as POHI_TSCRT,
    ' ' as POHI_PGMUP,
    ' ' as POHI_USRUP,
    '01/01/0001 00:00:00' as POHI_TSLUP,
    ' ' as POHI_FILEPATH
from (
        OePo_Header
        inner join OePo_Item 
        on OePo_Header.PO_ID = OePo_Item.PO_ID
    )
    left join OePo_Remarks
    on OePo_Header.WHSE_INST_ID = OePo_Remarks.AUTOKEY
    and OePo_Header.PO_ID = OePo_Remarks.PO_ID
where (
        (OePo_Item.UNITQTY_ORDER = 1)
        and (OePo_Header.SHIPTO_ID like '%*B1')
        and(OePo_Item.CLOSED_FLG = 'N')
        and(OePo_Item.INVOICED_COUNT = 0)
        and(OePo_Item.ITEM_TYPE = '2')
        and(OePo_Item.VENDOR_ID <> 'WHSE')
    )
group by
    LTRIM(RTRIM(OePo_Header.PO_ID)),
    LTRIM(RTRIM(OePo_Header.CLIENT_ID)),
    'B1',
    if(OePo_Remarks.TEXT = Null, '', OePo_Remarks.TEXT),
    '',
    '',
    OePo_Header.PO_ID
order by
    OePo_Header.PO_ID;

使用效率低得多的单个查询，我可以在单个查询中获取所需的数据，但在此查询中执行需要几分钟，并且在寻找相似（但不相同）的单独查询中可能需要更长的数量级）数据。有没有办法让这个工作与变量一起工作？（除非有一种join组合可以让我在几分钟内收集到正确的数据？）

标签： sqlvariablessubquerypervasive

sql - 具有变量的普遍 SQL 子查询

问题描述

解决方案

推荐阅读