Question

I have the following code:

$sth = $dbh->prepare("SELECT * FROM stats WHERE player_id = :player_id AND data_type = :data_type");

$sth->bindParam(':player_id', $player_id);
$sth->bindParam(':data_type', $total_time_data_type_id);
$sth->execute();

$result = $sth->fetch();            

if(!$result){
    $sth = $dbh->prepare("INSERT INTO stats (player_id, offset, created, modified, last_check, data_type, data) VALUES (:player_id, :offset, UNIX_TIMESTAMP(), UNIX_TIMESTAMP(), '1', :total_time_data_type_id, '0')");

    $sth->bindParam(':player_id', $player_id);
    $sth->bindParam(':offset', $offset);
    $sth->bindParam(':total_time_data_type_id', $total_time_data_type_id);
    $sth->execute();

    if(!$sth){
        return false;
    }
    $sth = $dbh->prepare("SELECT * FROM stats WHERE player_id = :player_id AND data_type = :data_type");

    $sth->bindParam(':player_id', $player_id);
    $sth->bindParam(':data_type', $total_time_data_type_id);
    $sth->execute();

    $result = $sth->fetch();
    if(!$result){
        return false;
    }
}else{
    $sth = $dbh->prepare("UPDATE stats SET .....");
    //Do more stuff
}

Now, occasionally this is creating duplicate rows (out of ~600 rows, there are 23 duplicates). This confuses me since before inserting the row I explicitly check for a row with the same player_id and data_type.

Multiple rows can exist for either the same player_id or data_type, but not the same of each.

I.E. This would be valid:

ID | PLAYER_ID | DATA_TYPE
---|-----------|----------
1  | 15        | 7
2  | 15        | 18
3  | 92        | 7
4  | 115       | 23

While this would not:

ID | PLAYER_ID | DATA_TYPE
---|-----------|----------
1  | 15        | 7        
2  | 32        | 18       
3  | 15        | 7        
4  | 115       | 23

Because of this I can’t simply declare the player_id field as unique.

The only thing I can think of that may be causing this problem is the fact that the code snippet above is inside of a foreach loop averaging around 115 iterations, and that this code is called again within seconds. Is there a way to prevent this programmatically?

Answer 1

Thanks to everyone for your help, especially @nikita2206. This is how I solved my problem:

1. Dropped column ID which was Primary Key.
2. Created new primary key on (Player_id, data_type).
3. Changed INSERT Statement:

From

$sth = $dbh->prepare("INSERT INTO stats (player_id, offset, created, modified, last_check, data_type, data) VALUES (:player_id, :offset, UNIX_TIMESTAMP(), UNIX_TIMESTAMP(), '1', :total_time_data_type_id, '0')");

to

$sth = $dbh->prepare("INSERT INTO stats (player_id, offset, created, modified, last_check, data_type, data) VALUES (:player_id, :offset, UNIX_TIMESTAMP(), UNIX_TIMESTAMP(), '1', :total_time_data_type_id, '0') ON DUPLICATE KEY UPDATE player_id = player_id");

---

Tested by calling the code several times very quickly and no duplicates were created.