Question

我有3张桌子：联想（ID，姓名）种族（ID，姓名）统计信息（ID，日期，associationId，TYPEID，数量）

我的目标是获得所有关联的列表，以及所有关联的列表，类型列表，数量，先排序，关联总数，然后是每种类型。为此，我有这个问题：

SELECT association.name AS associationName,
        race.name AS raceName,
        SUM(report.quantity) AS quantity,
        subrequest.totalquantity as totalquantity
    FROM stats report
    JOIN association ON association.id = report.associationId
    JOIN race ON race.id = report.raceId
    JOIN (
        SELECT SUM(report.quantity) AS totalquantity,
            association.id AS subId
        FROM stats report
        JOIN association ON association.id = report.associationId
        WHERE date LIKE "2013-11-%"
        GROUP BY association.id
        ORDER BY totalquantity DESC
    ) subrequest ON subrequest.subId = association.id
    WHERE date LIKE "2013-11-%"
    GROUP BY association.id, race.id
        ORDER BY totalquantity DESC, quantity DESC

它工作得非常好！但现在，我希望每个协会只有两种最大的类型。我试过这个：

set @num := 0, @association := '';
select associationName, raceName, quantity, totalquantity,
@num := if(@association = associationName, @num + 1, 1) as row_number,
        @association := associationName as dummy
FROM (
    SELECT association.name AS associationName,
        race.name AS raceName,
        SUM(report.quantity) AS quantity,
        subrequest.totalquantity as totalquantity
    FROM stats report
    JOIN association ON association.id = report.associationId
    JOIN race ON race.id = report.raceId
    JOIN (
        SELECT SUM(report.quantity) AS totalquantity,
            association.id AS subId
        FROM stats report
        JOIN association ON association.id = report.associationId
        WHERE date LIKE "2013-11-%"
        GROUP BY association.id
        ORDER BY totalquantity DESC
    ) subrequest ON subrequest.subId = association.id
    WHERE date LIKE "2013-11-%"
    GROUP BY association.id, race.id
    ORDER BY totalquantity DESC, quantity DESC
) x
WHERE row_number <= 2

有了这个，我有一个错误：＃1054 - 'where子句'中的未知列'row_number'如果我删除WHERE子句，它工作正常，我有预期的结果。我还有一个列row_number，其中包含正确的数字。

我必须这样做才能使它发挥作用：

set @num := 0, @association := '';
select associationName, raceName, quantity, totalquantity
FROM (
    select associationName, raceName, quantity, totalquantity,
    @num := if(@association = associationName, @num + 1, 1) as row_number,
            @association := associationName as dummy
    FROM (
        SELECT association.name AS associationName,
            race.name AS raceName,
            SUM(report.quantity) AS quantity,
            subrequest.totalquantity as totalquantity
        FROM association_race_report report
        JOIN association ON association.id = report.associationId
        JOIN race ON race.id = report.raceId
        JOIN (
            SELECT SUM(report.quantity) AS totalquantity,
                association.id AS subId
            FROM association_race_report report
            JOIN association ON association.id = report.associationId
            WHERE date LIKE "2013-11-%"
            GROUP BY association.id
            ORDER BY totalquantity DESC
        ) subrequest ON subrequest.subId = association.id
        WHERE date LIKE "2013-11-%"
        GROUP BY association.id, race.id
        ORDER BY totalquantity DESC, quantity DESC
    ) x
)y
WHERE row_number <= 2

这很好用！但我不明白为什么我需要添加另一级别的请求才能使WHERE子句起作用？添加无用的查询级别真让我烦恼。有没有办法让它发挥作用？为什么MySQL会引发这个错误？

谢谢！

Answer 1

MySQL不喜欢列引用，因为它在查询的select部分中定义 - 它还没有用于连接和where子句。

我看到你发现的解决方案，都是用定义它的表达式替换别名，或者创建一个中间子查询以避免重写整个事物。（查询优化器将nromally删除中间查询。）

为什么MySQL告诉我该列不存在？

1 个答案: