共计 5454 个字符,预计需要花费 14 分钟才能阅读完成。
这篇文章主要讲解了“SQL Server 中怎么改写内联表值函数”,文中的讲解内容简单清晰,易于学习与理解,下面请大家跟着丸趣 TV 小编的思路慢慢深入,一起来研究和学习“SQL Server 中怎么改写内联表值函数”吧!
问题 SQL:
SELECT TOP 1001 ha.HuntApplicationID ,
ha.PartyNumber ,
mht.Name AS MasterHuntTypeName ,
htly.LicenseYear ,
lStatus.[Status] AS DrawTicketStatus ,
isnull(dbo.udf_GetHuntApplicationPartyCount(ha.HuntApplicationID), 0) AS MemberCount ,
count( won.DrawTicketLicenseID) AS DrawnMemberCount ,
won.drawticketid ,
dt.PreDrawNonResidentMemberCount AS NRMemberCount ,
dbo.udf_GetAvgPreferencePoints(dt.DrawTicketID) AS PreferencePointAverage ,
CASE
WHEN ha.Quantity 1 THEN NULL
ELSE dt.PreDrawRandomNumber
END AS PreDrawRandomNumber ,
dsm.Name AS DrawSelectionMethodName ,
dt.DrawnSequence ,
dt.PreferencePointRank ,
dt.DrawID ,
dt.RandomRank
FROM dbo.HuntApplication ha
JOIN dbo.HuntTypeLicenseYear htly ON ha.HuntTypeLicenseYearID = htly.HuntTypeLicenseYearID
JOIN dbo.MasterHuntType mht ON htly.MasterHuntTypeID = mht.MasterHuntTypeID
LEFT JOIN dbo.HuntApplicationLicense hal ON ha.HuntApplicationID = hal.HuntApplicationID
LEFT JOIN dbo.DrawTicket dt ON ha.HuntApplicationID = dt.HuntApplicationID
LEFT JOIN dbo.DrawTicketLicense won ON dt.DrawTicketID = won.DrawTicketID
AND won.WasDrawn = 1
LEFT JOIN dbo.DrawSelectionMethod dsm ON dt.DrawSelectionMethodID = dsm.DrawSelectionMethodID
LEFT JOIN dbo.StatusCode lStatus ON dt.StatusCodeID = lStatus.StatusCodeID
JOIN dbo.DrawTicketHuntChoice dthc ON dt.DrawTicketID = dthc.DrawTicketID
CROSS APPLY dbo.tvf_GetHuntApplicationPartyCount(ha.HuntApplicationID) hapc
CROSS APPLY dbo.tvf_GetAvgPreferencePoints(dt.DrawTicketID) app
WHERE 1 = 1
AND htly.MasterHuntTypeID = @iMasterHuntTypeID
AND htly.LicenseYear = @iLicenseYear
AND dt.StatusCodeID = @iDrawTicketStatusCodeID
AND dthc.WasDrawn = @iHuntChoiceWasDrawn
GROUP BY ha.HuntApplicationID,
ha.PartyNumber,
mht.[Name],
htly.LicenseYear,
lStatus.[Status],
isnull(dbo.udf_GetHuntApplicationPartyCount(ha.HuntApplicationID), 0),
won.DrawTicketID,
dt.PreDrawNonResidentMemberCount,
dbo.udf_GetAvgPreferencePoints(dt.DrawTicketID),
CASE
WHEN ha.Quantity 1 THEN NULL
ELSE dt.PreDrawRandomNumber
END,
dsm.[Name],
dt.DrawnSequence,
dt.PreferencePointRank,
dt.DrawID,
dt.RandomRank
ORDER BY htly.LicenseYear DESC,
mht.Name,
lStatus.[Status],
dt.DrawID,
PreferencePointAverage DESC,
PreDrawRandomNumber,
ha.PartyNumber
静态函数:
CREATE FUNCTION [dbo].[udf_GetAvgPreferencePoints]
(@DrawTicketID INT)
RETURNS NUMERIC (18, 3)
BEGIN
RETURN
SELECT TOP 1
CONVERT(DECIMAL, dt.PreDrawPreferencePointTotal) / NULLIF(CONVERT(DECIMAL, dt.PreDrawMemberCount),0)
FROM dbo.DrawTicket dt
WHERE dt.DrawTicketID = @DrawTicketID
END
执行时间 40s
这是典型可以进行静态函数改写内联表值函数的 sql:
改写后:
SELECT TOP 1001 ha.HuntApplicationID ,
ha.PartyNumber ,
mht.Name AS MasterHuntTypeName ,
htly.LicenseYear ,
lStatus.[Status] AS DrawTicketStatus ,
--isnull(dbo.udf_GetHuntApplicationPartyCount(ha.HuntApplicationID), 0) AS MemberCount ,
isnull(hapc.MemberCount, 0) AS MemberCount,
count( won.DrawTicketLicenseID) AS DrawnMemberCount ,
won.drawticketid ,
dt.PreDrawNonResidentMemberCount AS NRMemberCount ,
--dbo.udf_GetAvgPreferencePoints(dt.DrawTicketID) AS PreferencePointAverage ,
app.PreferencePointAverage PreferencePointAverage,
CASE
WHEN ha.Quantity 1 THEN NULL
ELSE dt.PreDrawRandomNumber
END AS PreDrawRandomNumber ,
dsm.Name AS DrawSelectionMethodName ,
dt.DrawnSequence ,
dt.PreferencePointRank ,
dt.DrawID ,
dt.RandomRank
FROM dbo.HuntApplication ha
JOIN dbo.HuntTypeLicenseYear htly ON ha.HuntTypeLicenseYearID = htly.HuntTypeLicenseYearID
JOIN dbo.MasterHuntType mht ON htly.MasterHuntTypeID = mht.MasterHuntTypeID
LEFT JOIN dbo.HuntApplicationLicense hal ON ha.HuntApplicationID = hal.HuntApplicationID
LEFT JOIN dbo.DrawTicket dt ON ha.HuntApplicationID = dt.HuntApplicationID
LEFT JOIN dbo.DrawTicketLicense won ON dt.DrawTicketID = won.DrawTicketID
AND won.WasDrawn = 1
LEFT JOIN dbo.DrawSelectionMethod dsm ON dt.DrawSelectionMethodID = dsm.DrawSelectionMethodID
LEFT JOIN dbo.StatusCode lStatus ON dt.StatusCodeID = lStatus.StatusCodeID
JOIN dbo.DrawTicketHuntChoice dthc ON dt.DrawTicketID = dthc.DrawTicketID
CROSS APPLY dbo.tvf_GetHuntApplicationPartyCount(ha.HuntApplicationID) hapc
CROSS APPLY dbo.tvf_GetAvgPreferencePoints(dt.DrawTicketID) app
WHERE 1 = 1
AND htly.MasterHuntTypeID = @iMasterHuntTypeID
AND htly.LicenseYear = @iLicenseYear
AND dt.StatusCodeID = @iDrawTicketStatusCodeID
AND dthc.WasDrawn = @iHuntChoiceWasDrawn
GROUP BY ha.HuntApplicationID,
ha.PartyNumber,
mht.[Name],
htly.LicenseYear,
lStatus.[Status],
--isnull(dbo.udf_GetHuntApplicationPartyCount(ha.HuntApplicationID), 0),
isnull(hapc.MemberCount, 0),
won.DrawTicketID,
dt.PreDrawNonResidentMemberCount,
--dbo.udf_GetAvgPreferencePoints(dt.DrawTicketID),
app.PreferencePointAverage,
CASE
WHEN ha.Quantity 1 THEN NULL
ELSE dt.PreDrawRandomNumber
END,
dsm.[Name],
dt.DrawnSequence,
dt.PreferencePointRank,
dt.DrawID,
dt.RandomRank
ORDER BY htly.LicenseYear DESC,
mht.Name,
lStatus.[Status],
dt.DrawID,
PreferencePointAverage DESC,
PreDrawRandomNumber,
ha.PartyNumber
对应的表值函数:
CREATE FUNCTION [dbo].[tvf_GetAvgPreferencePoints]
(@DrawTicketID INT)
RETURNS TABLE WITH SCHEMABINDING
RETURN
SELECT TOP 1
CONVERT(DECIMAL, dt.PreDrawPreferencePointTotal) / NULLIF(CONVERT(DECIMAL, dt.PreDrawMemberCount),0) as PreferencePointAverage
FROM dbo.DrawTicket dt
WHERE dt.DrawTicketID = @DrawTicketID
GO
改写后执行时间从 40s 降低到 16s,对于倾斜列的优化速度更为明显
感谢各位的阅读,以上就是“SQL Server 中怎么改写内联表值函数”的内容了,经过本文的学习后,相信大家对 SQL Server 中怎么改写内联表值函数这一问题有了更深刻的体会,具体使用情况还需要大家实践验证。这里是丸趣 TV,丸趣 TV 小编将为大家推送更多相关知识点的文章,欢迎关注!
正文完