mysql唯一最新条目

Question

I have a table of urls. There is an extra column with the parsed domain name (foreign key to other table). What i want are the freshest urls, but only one url of each domain name. I have following schema:

id|url|id_domain

I tried severeal versions of my query but none of them seemed to work performant, such as:

SELECT * from urls GROUP BY id_domain ORDER BY id DESC

Explain tells me that it uses a temporary table.

There are indices on id, id_domain, (id, id_domain) and (id_domain, id)

Does anybody know a more efficient way to query such data?

Edit Here is some sample data with expected output: Sample data:

id | url | id_domain
1 | www.example1.com/link1/index.html | 1
2 | www.example1.com/link2/index.html | 1
3 | www.example1.com/link3/index.html | 1
4 | www.example2.com/link1/index.html | 2  

i would expect the newest urls (id is auto increment, therefore higher id = newer.
In this case i would expect to have urls with id 3 and id 4 returned

Thank you very much

Answer 1

您可以在查询中使用DISTINCT。

SELECT DISTINCT id,url,id_domain from urls GROUP BY id_domain ORDER BY id DESC