First step is to get the product collection,

$collection = Mage::getModel('catalog/product')->getCollection();

Now we need to modify the collection to add a new select column. The mysql is also important here. You’ll see that I’m using GROUP_CONCAT and DISTINCT, and towards the bottom specifying to group by ‘e.entity_id’. I’m not going to dive into why this is necessary or how it all works. You’ll be better off visiting the mysql documentation on those.

$collection = Mage::getModel('catalog/product')->getCollection();
 
$collection->getSelect()
    ->columns('GROUP_CONCAT(DISTINCT (SELECT sku FROM catalog_product_entity WHERE entity_id = r.linked_product_id) SEPARATOR \', \') AS related_skus')
    ->joinLeft(array('r' => 'catalog_product_link'), 'r.product_id = e.entity_id AND r.link_type_id = 1')
    ->group('e.entity_id');

The result of this is an additional field in the collection called ‘related_skus’ that is a comma-space separated list of skus that are related products for each product in the collection.

If you want to log the actual query that is generated by this, simply add this line below the code above:

Mage::log($collection->getSelect()->__toString());

Hopefully this helps you in some way!

mysqldump -u [user] --password=[password] databasetodupe | mysql -u [user] --password=[password] newdatabase

I did some Googling around and came across someone that seemed to have a similar issue, though not with Magento. So I figured I’d give it a shot. Well, it worked. The unfortunate part is that, being not all too familiar with how all this configuration stuff works, I don’t know why it works (if you know why, please let me know).

The Solution

Open up your php.ini file (mine was in /etc/). Look for the following line:

pdo_mysql.default_socket=/var/mysql/mysql.sock

If that line exists for you, try changing it to:

pdo_mysql.default_socket=/tmp/mysql.sock

Restart apache after saving, and that’s all I had to do. One thing to keep in mind though is that I am using the stock php (with 10.6), and I’m using mysql installed with homebrew.

I just imported a ton of products for new website that I’m working on, and it turns out that the client misnamed a bunch of images, causing Magento to ignore the image import for those ones. I don’t want to scour through the website and the CSV (and the client wouldn’t want to pay us to do that either), so I figured if I could just come up with a list of SKUs for products missing an image, I could hand it off to them so they could correct everything.

This isn’t foolproof, but it does the job. This query will give you a list of SKUs where there is no record in the database for that product’s ‘image’ attribute (the main large product image).

SELECT `sku` FROM `catalog_product_entity` AS a LEFT JOIN `catalog_product_entity_varchar` AS b ON a.entity_id = b.entity_id AND b.attribute_id = 74 WHERE b.entity_id IS NULL

I was dumping a database today to put onto a different server. I am currently running 5.1.45 locally, and the server I am moving to has 5.0.58. When trying to import the database to the new server, I was getting the following error:

ERROR 1064 (42000) at line 576: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'USING BTREE,
KEY `FK_ATTRIBUTE_VARCHAR_ENTITY` (`entity_id`),
KEY `FK_CATALO' at line 9

After doing some checking around, I found that the problem was the ‘USING BTREE’. Mysql 5.0 and lower doesn’t seem to understand what that is about and will fail. The solution: Just remove ‘USING BTREE’, and you’ll be good to go.

It sounds like versions after mine (5.1.45) have fixed this and omit the ‘USING BTREE’ from the dumps.

I’m not sure if it is possible to make that title much shorter than that… Anyway, I had some old data that I imported to an installation of Magento. All of the product titles were all uppercase, which was quite ugly. I needed a way to convert all of those values to only have uppercase first letter of each word.

Programming that was quite easy, but I noticed that there were also some parts of the string which were not converted like “On/off”, “D.e.”, “Quick-connect”, etc. The ucwords() function doesn’t recognize that kind of stuff. So, I had to find a way around those. I created an array that houses whatever characters you’d like to capitalize after (as long as there is a letter immediately after it). So, those three examples become “On/Off”, “D.E.”, and “Quick-Connect”.

Here’s what I came up with. Feel free to use it if you wish – you’ll just have to change the database credentials, change the SELECT and UPDATE queries to suit your needs, and edit the $also_after array to uppercase after any characters you want:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

<?php
// Database Connection
$db = new mysqli('localhost','username','password','database');
 
// What Ever Query You Want
$query = $db->query("SELECT value_id, value FROM catalog_product_entity_varchar WHERE attribute_id = 45");
 
// Run For Each Result
while ($line = $query->fetch_object()) {
	$value_id 	= $line->value_id;
	$value		= $line->value;
 
	// Populate this array with characters that you want this script to capitalize after.
	// If in the string there is "onn/off", put the "/" character in the array to make "On/Off"
	// otherwise, it would only do "On/off".
	$also_after	= array('-', '/', '(', '.');
 
	// Run this function for each value. Returns new string.
	$string = convert_to_uc($value, $also_after);
 
	// Escape characters for database entry
	$string = $db->real_escape_string($new_string);
 
	// Update the database with your new string
	$db->query("UPDATE catalog_product_entity_varchar SET value = '$string' WHERE value_id = '$value_id'");
}
 
// Convert To Uppercase function
function convert_to_uc($value, $also_after)
{
	$string 	= '';
 
	// Convert thet string to all lowercase, then uppercase the first letter of every word.
	$string 	= ucwords(strtolower($value));
 
	// For each $also_after character, run the uc_match function
	foreach ($also_after as $char) {
		$string = preg_replace_callback('/(\\'.$char.')(\w)/',"uc_match",$string);
	}
 
	// Returns new formatted string
	return $string;
}
 
// Uppercase Match function - used by preg_replace_callback above
function uc_match($match)
{
	return $match[1].strtoupper($match[2]);
}

If you have ever tried to delete a large amount of products out of Magento, you’ll realize that it takes forever, and in many cases, you have to do it in small quantities at a time otherwise it will freeze up. I needed to delete all the products out today but found that I could only delete about 30 at a time. This just wasn’t going to work since there are about 2500 products in the DB. I found something out today by accident that turned out to be extremely helpful.

Due to the nature of the InnoDB mysql engine that Magento uses, and the way they setup their foreign keys, all it takes is one simple mysql command, and mysql will automatically clear out all of the product data from all of the different tables it uses. On top of that, it resets the auto-increment values back to 1 so that you can start off with entity_id 1 again.

I haven’t given it extensive testing, but I checked every table that I knew that stored product information and they were all clear. So, if you need to clear out all your products, open up mysql and just enter this one line in:

TRUNCATE TABLE `catalog_product_entity`;

For more information on how it works, you can visit the Truncate Syntax page on the Mysql website.

Mysqli stands for Mysql Improved. It came out a while ago, I believe with Mysql 4.1.3 and up. There are a lot of great benefits to using mysqli, which I won’t get into now. Maybe later. Here’s just a little to get you started using it:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

<?php
 
// Connect to Database
$db = new mysqli('host', 'username', 'password', 'database');
 
// Perform a query
$items = $db->query("SELECT * FROM cp_items WHERE id > 100");
 
// Run a loop through the results
while($item = $items->fetch_object()) {
    echo '<ul>';
    echo '<li>' . $item->id . '</li>';
    echo '<li>' . $item->name . '</li>';
    echo '</ul>';
}

Connecting to multiple databases at the same time is simple. Just create a new database object, and call the mysqli functions from it. You can read the full documentation of everything available with mysqli to learn more about it.

As I said in the post prior to this, I have really been enjoying using Terminal for accessing MySQL databases. I found a problem with it though. If you have a table with a lot of columns, or a column with a lot of data in it, Terminal will wrap the results of your query making it almost completely unreadable (or a pain to read at best).

I found a work-around. Though it isn’t as good as it could be, it does work. It prevents the wrap, but, everything beyond the right-side of your terminal window is completely removed. It doesn’t give you a scroll bar to see beyond, it is literally gone. If you try to resize your window, you’ll just get blank space.

Why Terminal doesn’t have the ability to turn it on or off is a mystery to me. But, that said, this fix does have it’s place, and it is fairly easy to turn on and off. Here’s what you do:

Open this file: $YOURHOME/.bashrc and place the following in there and save it:

# Wordwrap off
echo -e "\e[?7l\c"

After it is saved, you can close Terminal and re-open it, or restart bash with this command:

. ~/.bashrc

To turn it off again, just remove or comment out the echo statement you added to .bashrc and restart bash again.

If you have any better suggestions for this, please let me know! I’d love to have a better solution.