Question

我有一个查询，检查三个表A，B和C中是否至少存在一行，我想检查theDate = myDate。我希望能够查看表是否包含特定日期的数据（myDate）。我希望用户在运行查询时选择变量myDate。我怎么做？仅供参考：我正在使用Teradata：

查询：

SELECT 'A' AS "Table", CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END AS "True?"
WHERE EXISTS (SELECT * FROM A WHERE theDate=myDate) 

UNION ALL

SELECT 'B', CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END
WHERE EXISTS (SELECT * FROM B WHERE theDate=myDate) 

UNION ALL

SELECT 'C', CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END
WHERE EXISTS (SELECT * FROM C WHERE theDate=myDate)

Answer 1

我认为动态过程/宏将最符合您的要求。编译以下程序：

REPLACE PROCEDURE your_proc (IN in_mydate DATE)
READS SQL DATA
DYNAMIC RESULT SETS 1
BEGIN
DECLARE stmt_str VARCHAR(60000);
DECLARE RESPONSE CURSOR WITH RETURN ONLY TO CLIENT FOR STMT;

-----------------------------------------------------------------------------------------------
-------------------------------- PROGRAM LOGIC ------------------------------------------------
-----------------------------------------------------------------------------------------------

        SET stmt_str =          'SELECT ''A'' AS "Table", CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END AS "True?"'
                                ||'WHERE EXISTS (SELECT * FROM A WHERE theDate='''||in_mydate||''')' 
                                ||'UNION ALL'
                                ||'SELECT ''B'', CASE WHEN COUNT(*) = 1 THEN ''Y'' ELSE ''N'' END'
                                ||'WHERE EXISTS (SELECT * FROM B WHERE theDate='''||in_mydate||''') '
                                ||'UNION ALL'
                                ||'SELECT ''C'', CASE WHEN COUNT(*) = 1 THEN ''Y'' ELSE ''N'' END'
                                ||'WHERE EXISTS (SELECT * FROM C WHERE theDate='''||in_mydate||''');';



PREPARE STMT FROM stmt_str;
OPEN RESPONSE;
DEALLOCATE PREPARE STMT;

END;

在此之后调用语句只需输入您的日期（作为输入参数）。输出将是您想要的选择查询的结果。

如果你使用宏，下面的代码会给你输出：

        REPLACE MACRO your_macro (mydate DATE)
        AS (
        SELECT 'A' AS "Table", CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END         AS "True?"
        WHERE EXISTS (SELECT * FROM A WHERE theDate=:myDate) 

        UNION ALL

        SELECT 'B', CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END
        WHERE EXISTS (SELECT * FROM B WHERE theDate=:myDate) 

        UNION ALL

        SELECT 'C', CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END
        WHERE EXISTS (SELECT * FROM C WHERE theDate=:myDate);
       );

由于

Answer 2

您可以将“变量”放在单行易失性表中，然后在查询中加入该表：

create volatile table varTable
as (select <your date value> as thedate)
with data
on commit preserve rows
;

SELECT 'A' AS "Table", 
CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END AS "True?"
from
a inner join vartable
  on a.<date column> = vartable.thedate

Answer 3

我发现在我看来简单的解决方案非常有效。？theDate是在运行查询之前由用户设置的变量。（对我来说，其他解决方案的问题是我无权创建存储过程或宏）

SELECT 'A' AS "Table", CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END AS "True?"
WHERE EXISTS (SELECT * FROM A WHERE theDate=?theDate) 

UNION ALL

SELECT 'B', CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END
WHERE EXISTS (SELECT * FROM B WHERE theDate=?theDate) 

UNION ALL

SELECT 'C', CASE WHEN COUNT(*) = 1 THEN 'Y' ELSE 'N' END
WHERE EXISTS (SELECT * FROM C WHERE theDate=?theDate)

用户在运行查询时设置的变量Teradata

3 个答案: