子查询中的`SELECT`和`FROM`之间是否存在性能差异？

Question

这些查询返回相同的结果，这是更好的性能吗？

SELECT

中的子查询

SELECT
    books.id,
    books.title,
    (SELECT COUNT(*) FROM subscriptions WHERE books.id = subscriptions.book_id) AS subscription_count,
    (SELECT COUNT(*) FROM books) AS book_count
FROM books
ORDER BY id;

FROM

中的子查询

SELECT
    *
FROM
    (SELECT
        books.id,
        books.title,
        (SELECT COUNT(*) FROM subscriptions WHERE books.id = subscriptions.book_id) AS subscription_count,
        (SELECT COUNT(*) FROM books) AS book_count
    FROM books) AS tmp
ORDER BY id;

在这种情况下，我应该从其他sql获得book_count吗？

Answer 1

您正在使用的是一个派生表，对于现代DBMS，没有任何性能差异。所以

select * 
from some_table
order by some_column;

与完全的性能相同：

select *
from (
  select *
  from some_table
) t
order by some_column;