Question

我有一个存储过程。

SELECT ADDRESS.ADDRESSID, NAME, DOORNO, STREET, STATE, COUNT(SUB.ADDRESSID) AS FREQUENCY
FROM ADDRESS 
JOIN
(
(SELECT ADDRESSID FROM ADDRESS WHERE KEYWORD LIKE '%' + @name + '%')
UNION ALL
(SELECT ADDRESSID FROM ADDRESS WHERE KEYWORD LIKE '%' + @doorno + '%')
UNION ALL
(SELECT ADDRESSID FROM ADDRESS WHERE KEYWORD LIKE '%' + @street + '%')
UNION ALL
(SELECT ADDRESSID FROM ADDRESS WHERE KEYWORD LIKE '%' + @state + '%')
) 
SUB
ON SUB.ADDRESSID=ADDRESS.ADDRESSID
GROUP BY ADDRESS.ADDRESSID, NAME, DOORNO, STREET, STATE
ORDER BY FREQUENCY DESC

当我手动执行此操作时，我得到了正确的结果。让我们说这些是结果

╔════╦══════╦════════════════╦═════════════════════╦═══════╦═══════════╗
║ ID ║ NAME ║     DOORNO     ║       STREET        ║ STATE ║ FREQUENCY ║
╠════╬══════╬════════════════╬═════════════════════╬═══════╬═══════════╣
║  4 ║ John ║ 29             ║ Police Station      ║ DC    ║         3 ║
║  1 ║ John ║ Door 13-9-16/A ║ patimeeda           ║ NJ    ║         2 ║
║  3 ║ Paul ║ Plot 304       ║ Near Police Station ║ NJ    ║         1 ║
╚════╩══════╩════════════════╩═════════════════════╩═══════╩═══════════╝

但是当我使用Sqldataadapter和数据集从C＃代码调用此SP时，我总是得到频率为1 如下所示。

╔════╦══════╦════════════════╦═════════════════════╦═══════╦═══════════╗
║ ID ║ NAME ║     DOORNO     ║       STREET        ║ STATE ║ FREQUENCY ║
╠════╬══════╬════════════════╬═════════════════════╬═══════╬═══════════╣
║  1 ║ John ║ Door 13-9-16/A ║ patimeeda           ║ NJ    ║         1 ║
║  3 ║ Paul ║ Plot 304       ║ Near Police Station ║ NJ    ║         1 ║
║  4 ║ John ║ 29             ║ Police Station      ║ DC    ║         1 ║
╚════╩══════╩════════════════╩═════════════════════╩═══════╩═══════════╝

排序对我来说非常重要。知道我失去这个的原因吗？

C＃.NET代码

SqlCommand cmd = new SqlCommand("Add_GetAddressSearchResults", con);
        cmd.CommandType=CommandType.StoredProcedure;
        cmd.Parameters.Add("@name", SqlDbType.VarChar, 50).Value=addressModel.Name;
        cmd.Parameters.Add("@doorno", SqlDbType.VarChar, 50).Value=addressModel.DoorNo;
        cmd.Parameters.Add("@street", SqlDbType.VarChar, 50).Value=addressModel.Street;
        cmd.Parameters.Add("@state", SqlDbType.VarChar, 50).Value=addressModel.State;

        SqlDataAdapter da = new SqlDataAdapter(cmd);
        DataSet ds= new DataSet();
        da.Fill(ds);
        ds.WriteXml("c:/ram.xls"); // to check the data

Answer 1

如果你已经消除了其他明显的原因（不同的数据库，不同的查询......），我强烈怀疑你在ADDRESSID列中有NULL值，ANSI_NULLS配置不同。

我认为C＃连接的默认值与SQL Server Management Studio中的默认值不同。

要对此进行测试，请尝试在手动执行查询时打开和关闭ANSI_NULLS，和/或尝试修改查询以按照您希望的方式显式处理NULL，例如类似的东西：

ON ISNULL(SUB.ADDRESSID,0)=ISNULL(ADDRESS.ADDRESSID,0)

您还可以尝试在存储过程中明确设置此（以及其他类似设置，如ANSI_PADDING）。

Answer 2

我可能错了，但是您从子查询中提取的SUB再次包含ADDRESS的记录，ADDRESSID上的内部联接可能会将频率设为1，除非SUB包含的内容超过了ADDRESS中的内容，您的正确结果将是因为传递的参数在ADDRESS中为您提供了SUB的超集

数据集数据与存储过程不同

2 个答案: