连接具有不同列值的行

Question

我正在使搜索引擎可以使用过滤器。

我的SQL看起来像这样；

SELECT l.location_id, og.*,f.*
        FROM object_types ot
        LEFT JOIN object_group_types ogt ON ogt.object_type_id = ot.object_type_id
        LEFT JOIN object_groups og ON og.object_group_type_id = ogt.object_group_type_id
        LEFT JOIN object_groups2filters og2f ON og2f.object_group_id = og.object_group_id
        LEFT JOIN filters f ON f.filter_id = og2f.filter_id
        LEFT JOIN locations l ON l.location_id = og.location_id
        WHERE ot.object_type_key = 'TYPE_TENNIS';

现在基于用户过滤器输入，我想为其选择正确的位置。 但是因为我将所有内容都用LEFT JOIN连接，所以所有过滤器项都位于不同的行中，请参见图片。

所以我想选择一个location_id同时具有filter_key FILTER_GRASS和FILTER_PARKING的位置。

如果我使用“ AND f.filter_key ='FILTER_PARKING'AND f.filter_key ='FILTER_GRASS'”，则将无法正常工作，因为过滤器值位于单独的行中。

任何人都有线索来选择location_id具有两个filter_keys的位置吗？

Answer 1

您需要将Group By与Having结合使用，以过滤出位置。

尝试：

SELECT l.location_id 
FROM object_types ot
LEFT JOIN object_group_types ogt ON ogt.object_type_id = ot.object_type_id
LEFT JOIN object_groups og ON og.object_group_type_id = ogt.object_group_type_id
LEFT JOIN object_groups2filters og2f ON og2f.object_group_id = og.object_group_id
LEFT JOIN filters f ON f.filter_id = og2f.filter_id
LEFT JOIN locations l ON l.location_id = og.location_id
WHERE ot.object_type_key = 'TYPE_TENNIS'
GROUP BY l.location_id 
HAVING SUM(f.filter_key = 'FILTER_PARKING') AND 
       SUM(f.filter_key = 'FILTER_GRASS')

Answer 2

您可以使用in和条件聚合

    SELECT l.location_id 
    FROM object_types ot
    LEFT JOIN object_group_types ogt ON ogt.object_type_id = ot.object_type_id
    LEFT JOIN object_groups og ON og.object_group_type_id = ogt.object_group_type_id
    LEFT JOIN object_groups2filters og2f ON og2f.object_group_id = og.object_group_id
    LEFT JOIN filters f ON f.filter_id = og2f.filter_id
    LEFT JOIN locations l ON l.location_id = og.location_id
    WHERE ot.object_type_key = 'TYPE_TENNIS'
   and f.filter_key in ('FILTER_PARKING','FILTER_GRAS')
    GROUP BY l.location_id 
    HAVING SUM(case when f.filter_key = 'FILTER_PARKING' then 1 else 0 end)>=1 AND 
           SUM(case whenf.filter_key = 'FILTER_GRAS' then 1 else 0 end)>=1