Question

I have a massive table with thousands of rows. Each row has a unique path. For example:

electronics/samsung/tv/lcd-23384
electronics/philips/tv/lcd-12ger
etc...

The problem is, a recent query was executed incorrectly, resulting in some of the fields having duplicate paths. So, what i want to know is, is there a query I can execute to show ALL the duplicates? In other words, if I have this:

ID     | PATH 
1      | path_1
2      | path_2
3      | path_3
4      | path_3
5      | path_3
6      | path_4
7      | path_4
8      | path_5
9      | path_6

I want this result:

ID     | PATH 
3      | path_3
4      | path_3
5      | path_3
6      | path_4
7      | path_4

Notice that all NON-duplicates have been removed. Ie…. these ones:

ID     | PATH 
1      | path_1
2      | path_2
8      | path_5
9      | path_6

What SQL query can accomplish this?

Answer 1

SELECT * FROM mytable NATURAL JOIN (
  SELECT PATH FROM mytable GROUP BY PATH HAVING COUNT(*) > 1
) dupes

See it on sqlfiddle.

---

To perform the update you request in the comments below:

UPDATE mytable NATURAL JOIN (
  SELECT PATH FROM mytable GROUP BY PATH HAVING COUNT(*) > 1
) dupes, (SELECT @r:=0) init
SET mytable.PATH = CONCAT(PATH, '-', @r:=@r+1);

See it on sqlfiddle.