I was asked today how to host datafeed images on your local server. Starting with the script given in Step 2 of the datafeed series, we’re just going to add some code after the MySQL query has been executed without errors:
Read more

I’m so happy to hear from affiliates who have followed in the datafeed series and created websites of their own! The other day, Jen Goode showed me a site she was working on, and brought up a common problem with datafeeds: how to categorize the products.

I was able to do this using simple like queries:

select * from products where MerchantSubcategory like '%green tea%' or MerchantCategory like '%green tea%' order by Name;

But sometimes this isn’t the case. Sometimes you have to (gasp!) manually categorize the products. To do this, start with some changes to the products table:

ALTER TABLE  `products` ADD  `MyCategory` VARCHAR( 10 ) NOT NULL ;
ALTER TABLE  `products` ADD INDEX (  `MyCategory` ) ;

Then create an admin page for you to categorize the products. Use the admin-sas-datafeed.php as your template, but change the inside code to:

<?php

if (!empty($_POST['ProductID']) && !empty($_POST['MyCategory'])) {
	mysql_query("update products set MyCategory='" . myres($_POST['MyCategory']) . "' where ProductID=" . (int)$_POST['ProductID'] . " limit 1");
	echo("<p>The product has been updated.</p>");
} // ends

$cQuery = "select * from products where MyCategory='' limit 1";
$oResult = mysql_query($cQuery);
if ($rsData = mysql_fetch_array($oResult)) {
	?>
	<form name="productForm" method="post" action="<?= $_SERVER['PHP_SELF'] ?>">
		<input type="hidden" name="ProductID" value="<?= (int)$rsData['ProductID'] ?>" />
		<table width="500">
			<tr>
				<td><a href="/item.php?ProductID=<?= (int)$rsData['ProductID'] ?>" target="_blank"><?= $rsData['Name'] ?></a>: </td>
				<td><select name="MyCategory" onChange="document.forms['productForm'].submit();">
					<option value="">Choose</option>
					<option value="green-tea">Green Tea</option>
					<option value="white-tea">White Tea</option>
					<option value="black-tea">Black Tea</option>
					<option value="oolong-tea">Oolong Tea</option>
					<option value="tea-pot">Tea Pot</option>
					<option value="tea-cup">Tea Cup</option>
					<option value="junk">Do Not Use</option>
				</select></td>
			</tr>
			<tr>
				<td></td>
				<td><input type="submit" name="cAction" value="Submit" /></td>
			</tr>
		</table>
	</form>
	<?php
} // ends if ($rsData = mysql_fetch_array($oResult))
else {
	echo("<p><strong>Congratulations</strong>! All products have been categorized!</p>");
} // ends else from if ($rsData = mysql_fetch_array($oResult))
?>

What this does is update the “MyCategory” field for each product. Then, your SQL query on the browse page is simply:

select * from products where MyCategory='green-tea' order by Name;

Don’t worry – when you update the datafeed (by downloading a new zip file and uploading it in the admin tool you built before), the MyCategory field isn’t touched!

Now you can categorize products however you’d like – enjoy!

In step 3, we’re going to actually display the data saved in the database! So far I’ve held your hand along the way, but now I’m just going to point you in the right direction, because there isn’t one way to do this. Merchants handle categories in their datafeeds differently, so you’re going to have to look at what the merchants you’ve chosen to work with are doing, and adjust accordingly.

Looking at my tea example, I wanted pages for Green Tea, White Tea, Black Tea and Oolong Tea. So (in theory, I don’t do it this way) we could make the Green Tea page /browse.php?cCategory=green+tea, and then pull in all the green teas. How do we do that? This is where you have to think.

What I’m doing is the following:

select * from products where MerchantSubcategory like '%green tea%' or MerchantCategory like '%green tea%' order by Name;

That works nicely for my dataset. However, if you were using the Baghaus datafeed, your query may be

select * from products where Name like '%Tylie Malibu%' order by Price desc;

Once you have your data, loop through it & display the products on the page, with a link to the product details page. You can use a list, or a table… whatever you’d like. In the end, you’ll be linking to individual products, like: /item.php?ProductID=466820102

And on item.php, you’ll do a quick lookup and display the product:

$cQuery = "select * from products where ProductID=" . (int)$_GET['ProductID'] . " limit 1";
$oResult = mysql_query($cQuery);
$rsData = mysql_fetch_array($oResult);

<a href="<?= $rsData['Link'] ?>" title="<?= $rsData['Name'] ?>" rel="nofollow"><?= $rsData['Name'] ?></a>

If you look at my tea site, you’ll notice I’m not using links like those in the blog post. Instead, I’m using .htaccess to rewrite the URLs, to make them “pretty”. You can learn about this in my Follow-up on Datafeeds Podcast blog post.

Finally, one problem you’ll be sure to run into is that merchants often have broken images in their datafeeds. This sucks, but there isn’t much you can do about it.

At first, I had a PHP script that made sure the image existed, and if so displayed it and if not, displayed another one. But, that slowed down the page display considerably. So instead, I use a bit of Javascript to replace broken images with a pre-defined image:

<script language="JavaScript" type="text/javascript">
function ImgError(source){
	source.src = "/images/no-image.gif";
	source.onerror = "";
	return true;
}
</script>

Warren Buffett once commented, “I want to give my kids just enough so that they would feel that they could do anything, but not so much that they would feel like doing nothing”.

In this datafeed series, I want to think the same thing. If I give you a finished site, you haven’t learned anything. However, if I give you the framework, and set you in the right direction, the finished product will be something you developed, not something you copied.

If you’ve followed from Step 1, and I know some of you have, you should be well on your way to a finished site. But if you’re waiting for a template to work off of, here it is:

By no means is this a website that you can actually launch, but the framework is there for you to build off of.

As always, if you need help, leave a comment, email me, or find me on Twitter @esnagel

The next step in building a datafeed website is getting the data into the database, but to do this, we need to build an administrative section to your website, and add some helper functions & files.

vars.php was a file seen in Step 1, which I use to store the mysql connection and other variables. Here’s a simplified version of the one I’m using:

<?php
	// MySQL
	$DB_NAME = 'your_db_name';
	$DB_USER = 'your_db_username';
	$DB_PASSWORD = 'your_db_password';
	$DB_HOST = 'your_db_host';

	mysql_connect($DB_HOST, $DB_USER, $DB_PASSWORD);
	mysql_select_db($DB_NAME);

	// Your ShareASale ID
	$nSASID = 132296;

	// Shortcut / Alias
	function myres($cString) {
		return mysql_real_escape_string(stripslashes($cString));
	} // ends function mres($cString)
?>

Change lines 3-6 and 12.

Next, you’ll need to set-up an admin.php page, like so:

<?php
	include('./vars.php');
	include('./admin-password.php');
	$bNavHome = true;
	include('./header.php');
?>
			<div class="post">
				<h2 class="title">GreenWhiteAndBlackTea.com Admin</h2>
				<div class="entry">
					<ul>
						<li><a href="admin-sas-datafeed.php">Upload ShareASale Datafeed</a></li>
					</ul>
				</div>
			</div><!-- ends class="post" -->
<?php
	include('./footer.php');
?>

Change as necessary – this is just an index for your admin section.

and admin-password.php:

<?php
if (!$_SERVER['PHP_AUTH_USER']) {
        Header("WWW-authenticate: basic realm=\"" . $_SERVER['HTTP_HOST'] . " Admin\"");
        Header("HTTP/1.0 401 Unauthorized");
        exit;
}
else {
        if (($_SERVER['PHP_AUTH_USER'] == 'the-username-you-want')
        && ($_SERVER['PHP_AUTH_PW'] == 'ssssshhhhh-super-secret')) {
                // OK!!!!
				$_SESSION['bAdmin'] = true;
                ;
        }
        else {
                Header("WWW-authenticate: basic realm=\"" . $_SERVER['HTTP_HOST'] . " Admin\"");
                Header("HTTP/1.0 401 Unauthorized");
                echo("Login failed.\n");
                exit;
        }
}
?>

Set your username & password that you’d like on lines 8 & 9

Finally, in your MySQL database, create the products table:

CREATE TABLE `products` (
  `ProductID` int(10) unsigned NOT NULL,
  `Name` varchar(255) NOT NULL,
  `MerchantID` int(10) unsigned NOT NULL,
  `Merchant` varchar(255) NOT NULL,
  `Link` varchar(255) NOT NULL,
  `Thumbnail` varchar(255) NOT NULL,
  `BigImage` varchar(255) NOT NULL,
  `Price` float(8,2) unsigned NOT NULL,
  `RetailPrice` float(8,2) unsigned NOT NULL,
  `Category` varchar(50) NOT NULL,
  `Subcategory` varchar(50) NOT NULL,
  `Description` text NOT NULL,
  `Custom1` varchar(255) NOT NULL,
  `Custom2` varchar(255) NOT NULL,
  `Custom3` varchar(255) NOT NULL,
  `Custom4` varchar(255) NOT NULL,
  `Custom5` varchar(255) NOT NULL,
  `Lastupdated` datetime NOT NULL,
  `Status` varchar(50) NOT NULL,
  `Manufacturer` varchar(255) NOT NULL,
  `PartNumber` varchar(255) NOT NULL,
  `MerchantCategory` varchar(255) NOT NULL,
  `MerchantSubcategory` varchar(255) NOT NULL,
  `ShortDescription` varchar(255) NOT NULL,
  `ISBN` varchar(25) NOT NULL,
  `UPC` varchar(25) NOT NULL,
  `SKU` varchar(255) NOT NULL,
  `CrossSell` varchar(255) NOT NULL,
  `MerchantGroup` varchar(255) NOT NULL,
  `MerchantSubgroup` varchar(255) NOT NULL,
  `CompatibleWith` varchar(255) NOT NULL,
  `CompareTo` varchar(255) NOT NULL,
  `QuantityDiscount` varchar(255) NOT NULL,
  `Bestseller` tinyint(1) unsigned NOT NULL,
  `AddToCartURL` varchar(255) NOT NULL,
  `ReviewsRSSURL` varchar(255) NOT NULL,
  `Option1` varchar(255) NOT NULL,
  `Option2` varchar(255) NOT NULL,
  `Option3` varchar(255) NOT NULL,
  `Option4` varchar(255) NOT NULL,
  `Option5` varchar(255) NOT NULL,
  `bActiveProduct` tinyint(1) unsigned NOT NULL,
  UNIQUE KEY `ProductID` (`ProductID`)
) ENGINE=MyISAM DEFAULT CHARSET=utf8;

I’m wrapping up the script which loads the data, which will be posted tomorrow.