如何使用“a”字段按“b”字段的最大值合并两个生成的表？

Question

有下一个查询：

select `privmsgs_id`, `contact`, `privmsgs_date`, `contentType` from (
    (
        select max(`privmsgs_id`) as `privmsgs_id`, `contact`, max(`privmsgs_date`) as `privmsgs_date`, `contentType` from (
            (
                select `privmsgs_from_userid` as `contact`, `privmsgs_id`, `privmsgs_date`, 'message' as `contentType`
                    from `privmsgs_table`
                where `privmsgs_to_userid` = 305026
            ) union (
                select `privmsgs_to_userid` as `contact`, `privmsgs_id`, `privmsgs_date`, 'message' as `contentType`
                    from `privmsgs_table`
                where `privmsgs_from_userid` = 305026
            )
        ) as `tmp` group by `contact` order by `privmsgs_date` desc
    ) union (
        select max(`privmsgs_id`) as `privmsgs_id`, `contact`, max(`privmsgs_date`) as `privmsgs_date`, `contentType` from (
            (
                select `from_userid` as `contact`, `id` as `privmsgs_id`, `date` as `privmsgs_date`, 'postcard' as `contentType`
                    from `postcards_table`
                where `to_userid` = 305026
            ) union (
                select `to_userid` as `contact`, `id` as `privmsgs_id`, `date` as `privmsgs_date`, 'postcard' as `contentType`
                    from `postcards_table`
                where `from_userid` = 305026
            )
        ) as `tmp1` group by `contact` order by `privmsgs_date` desc
    )
) as `rTmp` order by `privmsgs_date` desc;

有两个表tmp和tmp1被union合并，但是字段contact翻了一番：

privmsgs_id contact privmsgs_date   contentType
21490780    7070    1315207813      message
21556868    7070    1315215266      postcard
21226460    7754    1312025735      message
21539085    15588   1314615528      postcard
21489812    15588   1315208838      message

所以，我只需要最后一条记录（消息或明信片 - 无关紧要）和最后一条记录的 id（有问题 - 我可以在消息和明信片中分别获取 max(id)，但不能这样做在合并表中）：

privmsgs_id contact privmsgs_date   contentType
21556868    7070    1315215266      postcard
21226460    7754    1312025735      message
21489812    15588   1315208838      message

原因，我为什么不通过简化查询来做是因为我需要特定数量的结果，所以我只能通过一个查询来做到这一点。

Answer 1

您必须将结果按contact 分组并选择最大privmsgd_id，然后仅从privmsgd_id 中选择信息。

查看GROUP BY了解更多信息。