Question

I am trying to create a simple player profile website and am running into a problem with my SQL.

I have a player table and a previous clubs table. As this is a many-to-many relationship, I am using a look up table (player previous clubs) between the player and previous clubs table.

This all seemed to be working fine until I ran my sql which looks like this:

SELECT player.id, player.name AS name, age, position, height, weight, previousclubs.name AS previousclubs, satscore, gpa
FROM player 
INNER JOIN playerpreviousclubs
ON player.id = playerid
INNER JOIN previousclubs
ON previousclubid = previousclubs.id

The problem is that if a player has more then one previous club, then all of the information (apart from the previous club) is duplicated so I end up with something like this.

RESULT SET 1

Name: John Smith
Age: 23
Previous Clubs: Arsenal

RESULT SET 2

Name: John Smith
Age: 23
Previous Clubs: Reading

What I would like it to do is the following:

RESULT SET 1

Name: John Smith
Age: 23
Previous Clubs: Arsenal, Reading

Can anyone explain to me why this is currently happening please and how I go about rectifying it.

Thanks

Answer 1

You have to use GROUP BY combined with aggregate function in order to achieve wished behaviour (i.e. you need to “aggregate” n-tuples retrieved from DB, that’s why the functions are called aggregate :-)). Choice of aggregate function depends on RDBMS you are using, for example:

MySQL:

SELECT player.id, player.name AS name, GROUP_CONCAT(previousclubs.name)
FROM player
INNER JOIN playerpreviousclubs ON player.id = playerid
INNER JOIN previousclubs ON previousclubid = previousclubs.id
GROUP BY player.id, player.name

PostgreSQL (v8.4 and higher):

SELECT player.id, player.name AS name, array_agg(previousclubs.name)
FROM player
INNER JOIN playerpreviousclubs ON player.id = playerid
INNER JOIN previousclubs ON previousclubid = previousclubs.id
GROUP BY player.id, player.name