sql - 一个简单的选择 SQL 语句

Question

我正在使用 sql server 2008 R2 并且我有表：

Book: id, name, ...  >> 3500 rows
BookChapter: id, bookid, content(nvarchar(max)... >> it has about 300000 rows

2之间没有关系。

现在我想计算所有目前没有章节的书。

SELECT     COUNT(Id) AS Expr1
FROM         dbo.Book
WHERE     (Id NOT IN  (SELECT BookId FROM dbo.BookChapter))

总是给超时过期。我怎样才能实现它？

score 4 · Accepted Answer

尝试使用与不存在类似但有时性能更好的左连接。

select count(1) as Counts
from 
    dbo.Book B
    left join dbo.BookChapter BC on BC.BookID = B.ID
where
    BC.BookID is null

score 0 · Accepted Answer

您的子查询变得非常大。您可以尝试按 BookId 对其进行分组。

SELECT COUNT(Id) AS Expr1
FROM   dbo.Book
WHERE  (Id NOT IN  (SELECT BookId FROM dbo.BookChapter GROUP BY BookId))

另一个选项使用NOT EXISTS;

SELECT COUNT(B.Id) AS count
FROM   dbo.Book B
WHERE  NOT EXISTS (SELECT * FROM dbo.BookChapter WHERE BookId=B.BookId))

score 0 · Accepted Answer

不要使用“不在”。它太慢了。您可以按照 Apartidge 的第二个建议使用“不存在”，或者您可以做一些完全不直观的事情，如下所示：

select count(*) BooksWithoutChapters
from book
where id in 
(select id
from book
minus 
select bookid
from BookChapter
)

3 回答 3