MySQL copy multiple rows values with same ID into new columns?

ItemID    |    File                    

---

1           /storage/somefile1.jpg
1           /storage/somefile2.jpg
1           /storage/somefile3.jpg
1           /storage/somefile5.jpg  

2           /storage/somerandomfile.jpg
2           /storage/anotherrandomfile.jpg
2           /storage/yetanotherrandomfile.jpg开发者_Go百科
2           /storage/somefile.jpg  

I am wanting to create a new column for each file rather than having 1 row per file. Such as:

ItemID    |    File                     |    File2                        | File3 etc...

---

1           /storage/somefile1.jpg      |   /storage/somefile2.jpg        |   /storage/..


2           /storage/somerandomfile.jpg |  /storage/anotherrandomfile.jpg | /storage/..

Is there any way to automate this with a query?

---

After seeing your comment and that you're just trying to make a CSV out of it, you can do something like this:

<?php
$query = $db->query('SELECT ItemID, GROUP_CONCAT(File SEPARATOR \'|$|\') AS Files FROM Table GROUP BY ItemID');
// Use a string that cant appear as part of the filename as the separator
$fh = fopen('items.csv', 'w');
foreach ($query->fetchAll(PDO::FETCH_ASSOC) as $row) {
    $files = explode('|$|', $row['Files']);
    fputcsv($fh, array_merge(array($row['ItemID']), $files));
}
fclose($fh);

---

No. You can't do it(change mysql scheme) on fly.

---

Following my comment on the question, you can create a new table with your desired schema, and run something like that (assuming PDO):

<?php
$db->beginTransaction();
$query = $db->query('SELECT ItemID, GROUP_CONCAT(File SEPARATOR \'|$|\') AS Files FROM Table GROUP BY ItemID');
// Use a string that cant appear as part of the filename as the separator
$sth = $db->prepare('INSERT INTO NewTable (ItemID, File1, File2, File3) VALUES (:ItemID, :File1, :File2, :File3)');
foreach ($query->fetchAll(PDO::FETCH_ASSOC) as $row) {
    $files = explode('|$|', $row['Files']);
    $sth->execute(array(
        'ItemID' => $row['ItemID'],
        'File1' => isset($files[0]) ? $files[0] : NULL,
        'File2' => isset($files[1]) ? $files[1] : NULL,
        'File3' => isset($files[2]) ? $files[2] : NULL,
    ));
}
$db->commit();

However, I do think your current schema design is better. Why are you trying to change it?

---

If you want to generate your CSV right in SQL, you can use the following query as starting point:

SELECT CONCAT_WS(',', ItemID, GROUP_CONCAT(File ORDER BY File SEPARATOR ',')) AS FullCSVRow
FROM ItemToFile
GROUP BY ItemID

... which returns:

FullCSVRow
1,/storage/somefile1.jpg,/storage/somefile2.jpg,/storage/somefile3.jpg,/storage/somefile5.jpg
2,/storage/somerandomfile.jpg,/storage/somerandomfile.jpg,/storage/somerandomfile.jpg,/storage/somerandomfile.jpg
3,/storage/file-1-of-2.jpg,/storage/file-2-of-2.jpg

And if you are using PHP you have full control on what queries you do and how you build them. You don't need to change the DB design.

---

SELECT
   ItemID,

   ( SELECT t2.File FROM ( SELECT @row := @row + 1 AS r, File FROM my_table AS t3
     JOIN ( SELECT @row := 0 ) AS init
     WHERE t1.ItemID = t3.ItemID ORDER BY t3.File ) AS t2 WHERE t2.r = 1 ),

   ( SELECT t2.File FROM ( SELECT @row := @row + 1 AS r, File FROM my_table AS t3
     JOIN ( SELECT @row := 0 ) AS init
     WHERE t1.ItemID = t3.ItemID ORDER BY t3.File ) AS t2 WHERE t2.r = 2 ),

   ( SELECT t2.File FROM ( SELECT @row := @row + 1 AS r, File FROM my_table AS t3
     JOIN ( SELECT @row := 0 ) AS init
     WHERE t1.ItemID = t3.ItemID ORDER BY t3.File ) AS t2 WHERE t2.r = 3 ),

   ( SELECT t2.File FROM ( SELECT @row := @row + 1 AS r, File FROM my_table AS t3
     JOIN ( SELECT @row := 0 ) AS init
     WHERE t1.ItemID = t3.ItemID ORDER BY t3.File ) AS t2 WHERE t2.r = 4 )

   FROM my_table AS t1