Question

考虑这个小样本数据集：

| id | territory_id | signed_in
| 1  | 6            | 2010-12-22 01:00:00
| 2  | 6            | 2011-12-11 01:00:00
| 3  | 6            | 2013-03-13 01:00:00
| 4  | 6            | NULL
| 5  | 3            | 2013-03-06 01:00:00
| 6  | 3            | 2013-11-20 01:00:00

我希望将结果按territory_id分组，其中整个组不包含signed_in为NULL的行。或者基本上我想得到这些结果：

| 5  | 3            | 2013-03-06 01:00:00
| 6  | 3            | 2013-11-20 01:00:00

这是我当前的sql，它在加入signed_in表时查找每个组的最大territories值：

SELECT `territories`.`id`, `territories`.`label`, `territories`.`type_id`, `territories`.`area_type_id`, `territories`.`map_embed_id`, `tsio`.`signed_in`
FROM `territories` INNER JOIN (
    SELECT territory_id, MAX(signed_in) signed_in
    FROM `territories_sign_in_out`
    GROUP BY territory_id) tsio ON `territories`.`id` = `tsio`.`territory_id`
WHERE `territories`.`type_id` = ?
ORDER BY `tsio`.`signed_in` ASC 
LIMIT 15

Answer 1

试试这个：

select * from territories
where territory_id not in (
    select territory_id from territories where signed_in is null);

Answer 2

这是一个应该返回您正在寻找的结果的查询：

SELECT *
FROM territories T
LEFT OUTER JOIN (SELECT DISTINCT T2.territory_id
                 FROM territories T2
                 WHERE T2.signed_in IS NULL) TN ON TN.territory_id = T.territory_id
WHERE TN.territory_id IS NULL

希望这会对你有所帮助。

如果在组中找到NULL值，则Mysql排除组

2 个答案: