双WHILE循环？

时间：2015-08-12 18:46:29

标签： sql sql-server tsql recursive-query

总体而言，我需要提取一堆不同的人数据并按站点汇总。最终结果看起来像这样：

Site1
   Person1 general info
   Person1 sales info
   Person1 customer info
   Person1 vendor info
   Person2 general info
   Person2 sales info
   Person2 customer info
   Person2 vendor info
   Person3 general info
   .
   .
   .
Site2
   Person1 general info
   Person1 sales info
   Person1 customer info
   Person1 vendor info
   Person2 general info
   Person2 sales info
   Person2 customer info
   Person2 vendor info
   Person3 general info
   .
   .
   .
Site 4
   Person1 general info
  ...etc

每个网站可能有也可能没有相同的人。这些网站都在一张桌子上。 Person信息的4个类别中的每一个都在单独的表中。一般信息适用于表1中的所有人，并包含站点编号。所有人的销售信息都在表2中，还包括每个人的站点编号。客户和供应商信息都在单独的表中，但仅引用员工的personID（而不是站点）。

此外，为了bcp我的结果，我将从该查询中获取结果并将其放入临时表中，然后将每个结果附加到最终文本文件中。

我在这里的代码使结果如下所示：

Site1
Person1 general info
Person1 sales info
Person1 customer info
Person1 vendor info
Site4
Person2 general info
Person2 sales info
Person2 customer info
Person2 vendor info
Site2
Person3 general info
...etc

这里只有一个网站的代码，一个人：

DECLARE @salespersonid INT
DECLARE @siteNum VARCHAR(14)
DECLARE @count INT
DECLARE @rID INT
DECLARE @SQLCmd as VARCHAR(500)
DECLARE @fileName VARCHAR(50)

set @count = (select count(*) from tmp_salesSDrec)
set @rID = 1 --because siteid is sequential and unique in tmp_salesSDrec

WHILE @count > 0
BEGIN

set @siteNum = (select siteNumber from tmp_salesSDrec where rID = @rID)

set @salespersonid = (select salespersonid from tmp_sales01rec where rID = @rID)

--drop the tmp_salestemp table
IF EXISTS (SELECT * FROM sys.objects WHERE object_id = OBJECT_ID(N'[dbo].[tmp_salestemp]') AND TYPE IN (N'U'))
DROP TABLE [dbo].[tmp_salestemp]

SELECT field1 into tmp_salestemp
FROM
(
     --SDrec
          SELECT CONCAT(recordID,',',CAST(LayoutVersion as VARCHAR(3)),',',submissionType,',',salesYear,',',siteType,',',',',',',',',',',','
                              ,CAST(siteCode as VARCHAR(6)),',',',',
                         TPSCode,',',physicalsiteName,',',siteAddr,',',siteCity,',',sicustate,',',siteZip,',',',',',',adminContFirstName,',',adminContLastName,',',
                         adminContPhone,',',adminContEmail,',',techContFirstName,',',techContLastName,',',techContPhone,',',techContEmail,',',CAST(numreports as VARCHAR(4))
                         ,',',CAST(ConversionFactor as VARCHAR(6)),',',',',calTypeCY,',',',',',',calTypeCY1,',',',',',',calTypeCY2,',',',',',',calTypeCY3,',',',',',',',',',',',',','
                              ,CAST(MatchingName as VARCHAR(2)),',',',',
                         extractDate,',',eor)
          as field1, 1 as sortOrder, recordID
          from tmp_salesSDrec
          where siteNumber = @siteNum
  UNION ALL
     --01rec
          SELECT CONCAT(recordID,',',',',CAST(siteATPCode as VARCHAR(6)),',',empNumber,',',stateNumber,',',',',',',',',firstName,',',',',lastName
                         ,',',address_1,',',',',city,',',[state],',',zip,',',',',CONVERT(VARCHAR(8),dob,112),',',gender,',',salesgrade,',',',',',',salesgraddate,',',',',',',salesrank,',',',',',',',',',',
                         partnerYouth,',',siteConsent,',',eor)
          as field1, 2 as sortOrder, recordID
          from tmp_sales01rec
          where salespersonid = @salespersonid and siteNumber = @siteNum
  UNION ALL
     --02rec
          select CONCAT(recordID,',',salesgrade,',',siteAttended,',',siteATPCode,',',',',salesYear,',',',',term,',',blockSales,',',workInProgress,',',
                         CAST(sectorID as VARCHAR(25)),',',sectorName,',',',',',',',',PrepInd,',',salesAtmpt,',',salesEarned,',',sectorsalesgrade,',',LOWER(area),',',eor)
          as field1, 3 as sortOrder, recordID
          from tmp_sales02rec
          where salespersonid = @salespersonid and siteNumber = @siteNum
    UNION ALL
     --03rec
          select CONCAT(recordID,',',custType,',',custDate,',',subcustName,',',custScore,',',eor)
          as field1, 4 as sortOrder, recordID
          from tmp_sales03rec
          where salespersonid = @salespersonid
    UNION ALL
     --04rec
          select CONCAT(recordID,',',vendorType,',',otherVendorType,',',vendorDate,',',vendorStatus,',',eor)
          as field1, 5 as sortOrder, recordID
          from tmp_sales04rec
          where salespersonid = @salespersonid
) sq
order by sortOrder, recordID;

SELECT @SQLCmd = 'bcp "select * from [sales].dbo.tmp_salestemp" queryout "c:\bcpTemp.csv" -w -t -T -S && type c:\bcpTemp.csv >> c:\sales_2015.csv' ;
EXEC master..xp_cmdshell @SQLCmd;

     set @rID = @rID + 1;

     set @count = @count - 1;



END

我在想我需要为每个salespersonID添加第二个while循环，然后为这些站点设置原始循环。

有没有人有任何建议或更好的方法来实现这一目标？

1 个答案:

答案 0 :(得分：0)

你应该能够在没有任何while循环的情况下编写它。使用基于纯集的查询会快得多。将您的查询更改为如下所示，并将其全部导出为一大批。

SELECT field1 into tmp_salestemp
FROM
(
     --SDrec
          SELECT CONCAT(recordID,',',CAST(LayoutVersion as VARCHAR(3)),',',submissionType,',',salesYear,',',siteType,',',',',',',',',',',','
                              ,CAST(siteCode as VARCHAR(6)),',',',',
                         TPSCode,',',physicalsiteName,',',siteAddr,',',siteCity,',',sicustate,',',siteZip,',',',',',',adminContFirstName,',',adminContLastName,',',
                         adminContPhone,',',adminContEmail,',',techContFirstName,',',techContLastName,',',techContPhone,',',techContEmail,',',CAST(numreports as VARCHAR(4))
                         ,',',CAST(ConversionFactor as VARCHAR(6)),',',',',calTypeCY,',',',',',',calTypeCY1,',',',',',',calTypeCY2,',',',',',',calTypeCY3,',',',',',',',',',',',',','
                              ,CAST(MatchingName as VARCHAR(2)),',',',',
                         extractDate,',',eor)
          as field1, 1 as sortOrder, recordID, tmp_salesSDrec.siteNumber, convert(int,null) salespersonid
          from tmp_salesSDrec
  UNION ALL
     --01rec
          SELECT CONCAT(recordID,',',',',CAST(siteATPCode as VARCHAR(6)),',',empNumber,',',stateNumber,',',',',',',',',firstName,',',',',lastName
                         ,',',address_1,',',',',city,',',[state],',',zip,',',',',CONVERT(VARCHAR(8),dob,112),',',gender,',',salesgrade,',',',',',',salesgraddate,',',',',',',salesrank,',',',',',',',',',',
                         partnerYouth,',',siteConsent,',',eor)
          as field1, 2 as sortOrder, recordID, tmp_salesSDrec.siteNumber, salespersonid
          from tmp_sales01rec 
          join tmp_salesSDrec 
          on tmp_sales01rec.siteNumber = tmp_salesSDrec.siteNumber
  UNION ALL
     --02rec
          select CONCAT(recordID,',',salesgrade,',',siteAttended,',',siteATPCode,',',',',salesYear,',',',',term,',',blockSales,',',workInProgress,',',
                         CAST(sectorID as VARCHAR(25)),',',sectorName,',',',',',',',',PrepInd,',',salesAtmpt,',',salesEarned,',',sectorsalesgrade,',',LOWER(area),',',eor)
          as field1, 3 as sortOrder, recordID, tmp_salesSDrec.siteNumber, salespersonid
          from tmp_sales01rec 
          join tmp_salesSDrec 
          on tmp_sales01rec.siteNumber = tmp_salesSDrec.siteNumber
    UNION ALL
     --03rec
          select CONCAT(recordID,',',custType,',',custDate,',',subcustName,',',custScore,',',eor)
          as field1, 4 as sortOrder, recordID, tmp_salesSDrec.siteNumber, salespersonid
          from tmp_sales01rec 
          join tmp_salesSDrec 
          on tmp_sales01rec.siteNumber = tmp_salesSDrec.siteNumber
    UNION ALL
     --04rec
          select CONCAT(recordID,',',vendorType,',',otherVendorType,',',vendorDate,',',vendorStatus,',',eor)
          as field1, 5 as sortOrder, recordID, tmp_salesSDrec.siteNumber, salespersonid
          from tmp_sales01rec 
          join tmp_salesSDrec 
          on tmp_sales01rec.siteNumber = tmp_salesSDrec.siteNumber
) sq
order by tmp_salesSDrec.siteNumber, salespersonid, sortOrder, recordID;

我没有您的任何数据，因此我无法对此进行测试，因此可能需要进行一些调整。

相关问题
最新问题