Doing INTERSECT and MINUS in MySQL

linxh

本文转自：
http://www.bitbybit.dk/carsten/blog/?p=71

Doing INTERSECT and MINUS in MySQL
By Carsten | October 3, 2005
Doing an INTERSECT
An INTERSECT is simply an inner join where we compare the tuples of one table with those of the other, and select those that appear in both while weeding out duplicates. So
SELECT member_id, name FROM a
INTERSECT
SELECT member_id, name FROM b
can simply be rewritten to
SELECT a.member_id, a.name
FROM a INNER JOIN b
USING (member_id, name)
Performing a MINUS
To transform the statement
SELECT member_id, name FROM a
MINUS
SELECT member_id, name FROM b
into something that MySQL can process, we can utilize subqueries (available from MySQL 4.1 onward). The easy-to-understand transformation is:
SELECT DISTINCT member_id, name
FROM a
WHERE (member_id, name) NOT IN
(SELECT member_id, name FROM table2);
Of course, to any long-time MySQL user, this is immediately obvious as the classical use-left-join-to-find-what-isn’t-in-the-other-table:
SELECT DISTINCT a.member_id, a.name
FROM a LEFT JOIN b USING (member_id, name)
WHERE b.member_id IS NULL
which tends to be a lot more efficient.

Robin
Says:
October 16th, 2005 at 9:31 pm

These examples are very usefull, however, I think it is important to comment that internally something totally different happens. When you would do and INTERSECT query for example, it would query the results from table A and table B and then compare both sets. With the MySQL solution you describe it would query all results from the first table and then do a lookup for each result in the second table. This results in worse performance when your query returns a huge number of results.


本文来自ChinaUnix博客，如果查看原文请点：http://blog.chinaunix.net/u/8780/showart_345985.html