引用WHERE子句中的列别名

Question

SELECT logcount, logUserID, maxlogtm
   , DATEDIFF(day, maxlogtm, GETDATE()) AS daysdiff
FROM statslogsummary
WHERE daysdiff > 120

我明白了

"无效的列名称daysdiff".

Maxlogtm是一个日期时间字段.这是让我疯狂的小东西.

Answer 1

SELECT
   logcount, logUserID, maxlogtm,
   DATEDIFF(day, maxlogtm, GETDATE()) AS daysdiff
FROM statslogsummary
WHERE ( DATEDIFF(day, maxlogtm, GETDATE() > 120)

通常,您不能在WHERE子句中引用字段别名.(将其视为整个SELECT包含别名,在WHERE子句之后应用.)

但是,正如其他答案中所提到的,您可以强制SQL SELECT在WHERE子句之前处理.这通常用括号来强制执行逻辑操作顺序或使用公用表表达式(CTE):

括号/子选择:

SELECT
   *
FROM
(
   SELECT
      logcount, logUserID, maxlogtm,
      DATEDIFF(day, maxlogtm, GETDATE()) AS daysdiff
   FROM statslogsummary   
) as innerTable
WHERE daysdiff > 120

或者看看Adam对CTE版本的回答.

Answer 2

如果要在WHERE子句中使用别名,则需要将其包装在子选择或CTE中:

WITH LogDateDiff AS
(
   SELECT logcount, logUserID, maxlogtm
      , DATEDIFF(day, maxlogtm, GETDATE()) AS daysdiff
   FROM statslogsummary
)
SELECT logCount, logUserId, maxlogtm, daysdiff
FROM LogDateDiff
WHERE daysdiff > 120

Answer 3

如果您不想在CTE中列出所有列,另一种方法是使用outer apply:

select
    s.logcount, s.logUserID, s.maxlogtm,
    a.daysdiff
from statslogsummary as s
    outer apply (select datediff(day, s.maxlogtm, getdate()) as daysdiff) as a
where a.daysdiff > 120

Answer 4

无需重复执行代码的最有效方法是使用HAVING而不是WHERE

SELECT logcount, logUserID, maxlogtm
   , DATEDIFF(day, maxlogtm, GETDATE()) AS daysdiff
FROM statslogsummary
HAVING daysdiff > 120

Answer 5

怎么样使用子查询（这在Mysql中对我有用）？

SELECT * from (SELECT logcount, logUserID, maxlogtm
   , DATEDIFF(day, maxlogtm, GETDATE()) AS daysdiff
FROM statslogsummary) as 'your_alias'
WHERE daysdiff > 120

Answer 6

根据文档，HAVING 在 MySQL 中工作：

在HAVING子句被添加到SQL，因为关键字无法与合计函数使用的地方。

Answer 7

您可以引用列别名，但需要使用以下方式定义它CROSS/OUTER APPLY：

SELECT s.logcount, s.logUserID, s.maxlogtm, c.daysdiff
FROM statslogsummary s
CROSS APPLY (SELECT DATEDIFF(day, s.maxlogtm, GETDATE()) AS daysdiff) c
WHERE c.daysdiff > 120;

DBFiddle 演示

优点：

• 表达式的单一定义（更易于维护/无需复制粘贴）
• 无需使用 CTE/outerquery 包装整个查询
• 参考的可能性WHERE/GROUP BY/ORDER BY
• 可能更好的性能（单次执行）