Question

我致力于自动将CSV数据导入数据库。这时我的代码会生成一个名为#temp1的表，如下所示：

CompetitionID   DateKickOff                TimeKickOff         TeamIDHome  TeamIDAway   ScoreHome   ScoreAway
--------------- -------------------------- ------------------- ----------- ------------ ----------- ------------
2               2013-03-08 00:00:00.000    14:02:00.0000000    21          43           0           4

此时我需要检查数据库中是否存在相同的DateKickOff和TeamIDHome值组合，以避免导入重复数据。如果这个组合都准备好存在于表Match的一行上，那么我需要有效地跳过以下代码：

INSERT INTO Match
SELECT *
FROM #temp1

--Add MatchID column to Temp Data file and fill it with the most recent match ID
ALTER TABLE #CSVTest_Data
ADD MatchID INT

UPDATE #CSVTest_Data
SET MatchID = (SELECT TOP 1 MatchID
               FROM Match
               ORDER BY MatchID DESC)

INSERT INTO Data (MatchID,
                  OddsFirstTimeTaken, 
                  OddsLastTimeTaken, 
                  MarketName, 
                  Outcome, 
                  Odds, 
                  NumberOfBets, 
                  VolumeMatched, 
                  InPlay)
SELECT MatchID,
       FirstTimeTaken, 
       LatestTimeTaken, 
       Market, 
       Outcome, 
       Odds, 
       NumberOfBets, 
       VolumeMatched, 
       InPlay
FROM #CSVTest_Data

显然，如果数据不是重复条目，则需要运行上面的代码。我真的很感激这方面的一些帮助。

编辑：只是为了澄清，数据的比较需要在“插入匹配”之前进行。代码发生。如果数据不重复，SQL Server将在Match表中增加MatchID的主键。然后，在写入我的数据＆＃39;之前，我将获得这个新的MatchID值并将其写入我的第二个临时表。表。如果没有新的条目添加到匹配＆＃39;表，那么没有数据必须写入＆＃39;数据＆＃39;表

Answer 1

您可以使用EXCEPT关键字：

WITH NewData AS (
    SELECT FirstTimeTaken
         , LatestTimeTaken
         , Market
         , Outcome
         , Odds
         , NumberOfBets
         , VolumeMatched
         , InPlay
    FROM #CSVTest_Data -- Coming data

    EXCEPT --Minus

    SELECT FirstTimeTaken
         , LatestTimeTaken
         , Market
         , Outcome
         , Odds
         , NumberOfBets
         , VolumeMatched
         , InPlay
    FROM Data --Existing Data
)

INSERT INTO Data (OddsFirstTimeTaken, OddsLastTimeTaken, MarketName, Outcome, Odds, NumberOfBets, VolumeMatched, InPlay)
SELECT FirstTimeTaken, LatestTimeTaken, Market, Outcome, Odds, NumberOfBets, VolumeMatched, InPlay
FROM NewData --Insert New Data only

编辑：

如果您拥有主要身份密钥，即使在插入语句中也不需要使用它们。 SQL Server将弄清楚如何处理它们。

Answer 2

试试这个，

INSERT INTO Data (MatchID,
                  OddsFirstTimeTaken, 
                  OddsLastTimeTaken, 
                  MarketName, 
                  Outcome, 
                  Odds, 
                  NumberOfBets, 
                  VolumeMatched, 
                  InPlay)
SELECT t.MatchID,
       t.FirstTimeTaken, 
       t.LatestTimeTaken, 
       t.Market, 
       t.Outcome, 
       t.Odds, 
       t.NumberOfBets, 
       t.VolumeMatched, 
       t.InPlay
FROM #CSVTest_Data as t
left join Data as d on t.DateKickOff = d.DateKickOff
            and t.TeamIDHome = d.TeamIDHome
where d.DateKickOff is null and d.TeamIDHome is null

Answer 3

这可以通过对每个表格的重复数据进行单独检查来实现：

INSERT INTO  Match    
SELECT *
FROM #temp1
EXCEPT
SELECT CompetitionID, DateKickOff, TimeKickOff, TeamIDHome, TeamIDAway, ScoreHome, ScoreAway
FROM Match

DELETE #CSVTest_Data
  FROM #CSVTest_Data d
WHERE EXISTS( SELECT * from Data d2 WHERE
    d.FirstTimeTaken = d2.OddsFirstTimeTaken AND
    d.LatestTimeTaken = d2.OddsLastTimeTaken AND
    d.Market = d2.MarketName AND
    d.Outcome = d2.Outcome AND
    d.Odds = d2.Odds AND
    d.NumberOfBets = d2.NumberOfBets AND
    d.VolumeMatched = d2.VolumeMatched AND
    d.InPlay = d2.InPlay)

--Add MatchID column to Temp Data file and fill it with the most recent match ID
ALTER TABLE #CSVTest_Data ADD MatchID INT
update #CSVTest_Data
     Set MatchID = (SELECT TOP 1 MatchID FROM BetfairFootballDB..Match
                   ORDER BY MatchID DESC)

INSERT INTO  BetfairFootballDB..Data (MatchID, OddsFirstTimeTaken, OddsLastTimeTaken, MarketName, Outcome, Odds, NumberOfBets, VolumeMatched, InPlay)
SELECT MatchID, FirstTimeTaken, LatestTimeTaken, Market, Outcome, Odds, NumberOfBets, VolumeMatched, InPlay
FROM #CSVTest_Data

在运行某些SQL Server代码之前，请检查数据库中是否存在数据（避免重复）

3 个答案: