SQL笛卡儿类型联接

Question

我有两个表(这个例子简化了)，我想以一种特定的方式加入它们。

• 表1 ( ae )有学科编号，ae术语
• 表2 ( mh )有主题号，mh项

这里的构造有助于

INSERT INTO ae VALUES ('0001-0012','DRY EYE (12 OCT 2017)');
INSERT INTO ae VALUES  ('0001-0012', 'DEPRESSION (7 OCT 2017)');
INSERT INTO ae VALUES ('0001-0013','SKIN ATROPHY, LEFT UPPER ARM (4 OCT 2017 )');


INSERT INTO mh VALUES ('0001-0012', 'DIABETES MELLITUS (UN JUL 2007)');
INSERT INTO mh VALUES ('0001-0012', 'GASTRO-ESOPHAGEAL REFLUX INTERMITENT GR1 (18 AUG 2017)');
INSERT INTO mh VALUES ('0001-0012', 'ESOPHAGITIS GR 1 (18 AUG 2017)');
INSERT INTO mh VALUES ('0001-0012', 'DIARRHEA INITERMITTENT GR2 (5 JUL 2017 )');
INSERT INTO mh VALUES ('0001-0012', 'FATIGUE INTERMITTENT GR1 (18 AUG 2017)');
INSERT INTO mh VALUES ('0001-0013', 'VOMITING, INTERMITTENT GR1 (6 JUL 2017 )');

我希望我的输出是这样的：

​

有什么帮助吗？正如你所看到的，这是我所需要的一种极小的连接！

Answer 1

您可以使用row_number()和full join

select coalesce(ae.col1, mh.col1) as col1, ae.col2, mh.col2
from (select ae.*, row_number() over (partition by col1 order by col1) as seqnum
      from ae
     ) ae full join
     (select mh.*, row_number() over (partition by col1 order by col1) as seqnum
      from mh
     ) mh
     on mh.col1 = ae.col1 and mh.seqnum = ae.seqnum
order by coalesce(ae.col1, mh.col1), seqnum;

Answer 2

这是到目前为止我更新的代码，它现在给我正确的顺序编辑，只是在代码末尾添加了DESC，以给出结果。

select coalesce(ae.subject, mh.subject) as subject, ae.aeterms, mh.mhterms
from (select ae.*, row_number() over (partition by subject order by subject) 
as seqnum
from ae) ae full join
(select mh.*, row_number() over (partition by subject order by subject) as 
seqnum from mh) mh
on mh.subject = ae.subject and mh.seqnum = ae.seqnum
order by coalesce(ae.subject, mh.subject), ae.seqnum Desc;