查找所有共同作者 - Faceting/Grouping 用于多对多映射 table

Find all co authors - Faceting/Grouping for many to many mapping table

Table : Books 
-------------
ID    Name
1     Book1
2     Book2
3     Book3
4     Book4

Table : Authors 
-------------
ID    Name
1     Author1
2     Author2
3     Author3
4     Author4


Table :  BookAuthorMapping
---------------------------
ID    BookId    AuthorId
1     1         1
2     1         2
3     1         3
4     2         2
5     2         3
6     3         3
7     4         4  

所以,

问题是: 给定一个 AuthorId,我需要找出其他作者与他合着了多少本书:

示例:

Given AuthorId: 1
-------------------
AuthorId   Count 
2          1           // 2 has co-authored only book1
3          1           // 3 has co-authored only book1
1          1           // Its okay, if i get author1 in the result

Given AuthorId: 2
-------------------
AuthorId   Count 
1          1           // 1 has co-authored only book1
3          2           // 3 has co-authored book1 and book2
2          2           // Its okay if i get author 2 in the result

Given AuthorId: 3
-------------------
AuthorId   Count 
1          1          // 1 has co-authored only book1
2          2          // 2 has co-authored book1 and 2
3          3          // Its okay if i get author 3 in the result.

Given AuthorId: 4
-------------------
AuthorId   Count 
4          1     // I should not get author1 0 author2 0 , author3 0 for this

-- 编辑 -- 我有一个像这样的解决方案:select aId, count(mapping.bId) from mapping join (select bId from mapping where aId = ?) as tmp on mapping.bId = tmp.bId group by aId;

此外,@Giorgos Betsos 在回复中也提到了同样的问题。

我很好奇如果没有内部查询是否可行。

试试这个:

SELECT "AuthorId", COUNT(*)
FROM BookAuthorMapping
WHERE "BookId" IN (SELECT "BookId" FROM BookAuthorMapping WHERE "AuthorId" = 1)
GROUP BY "AuthorId"

Demo here

您也可以使用 INNER JOIN:

SELECT t1."AuthorId", COUNT(*)
FROM BookAuthorMapping AS t1
INNER JOIN BookAuthorMapping AS t2 ON t1."BookId" = t2."BookId" AND t2."AuthorId" = 1
GROUP BY t1."AuthorId"

Demo here

您也可以使用 window 函数执行此操作:

select authorid, count(*)
from (select bam.*,
             max(case when AuthorId = 1 then 1 else 0 end) over (partition by bookid) as HasAuthor1
      from BookAuthorMapping bam
     ) bam
where hasAuthor1 = 1
group by authorid;

在某些情况下,window 函数比使用 joinin 的方法更快。