MichaelEdenfield'sWeblog

One of the sites I'm working on needs a photo gallery consisting of about 300 images. However, the images they need in the gallery are stored on their old server which we don't readily have access to. So rather than copying each photo by hand, we can use PHP to copy them directly to our server without having to move each one manually. Here is a script I wrote yesterday that saved me a ton of time.

Once again, I opted to use the simple_html_dom parser, which uses selectors like jQuery to match elements on a page and allows one to do fun stuff to manipulate the data.

On the site I need to take the images from, they are listed like this:

<a href='full/photo055.jpg' rel='lightbox' style='color:#ffffff;'><img src='thumbs/photo055.jpg' border='0' width='100' height='100'>	055	</a>
<a href='full/photo294.jpg' rel='lightbox' style='color:#ffffff;'><img src='thumbs/photo294.jpg' border='0' width='100' height='100'>	294	</a>
...

And so on about 300 more times. It's a thumbnail image linked to a larger image that appears in a lightbox when the user clicks. All we're interested in, really, is grabbing the large image and copying to our server.

So here's a script that parses this page and matches each <a> tag. It examines each match again, this time only using matches that end with 'jpg'. Then each of those matches are striped of extraneous code, starting with <a href='full/photo055.jpg' rel='lightbox' style='color:#ffffff;'><img src='thumbs/photo055.jpg' border='0' width='100' height='100'> 055 </a> and whittling down to just 'full/photo055.jpg'. Once that's done, we can add on a new path of where we want the copied photo to live on our server.

<?php 
include ('includes/simple_html_dom.php');
$html = file_get_html('http://www.somedomain.com/photos/index.html');
$frontremove = "<a href='";
$backremove = "' rel='lightbox' style='color:#ffffff;'>";
$x=0;
foreach($html->find('a') as $element) {
	if(preg_match('/jpg/',$element)) {
		$image[$x] = str_replace($frontremove,"",$element);
		$image[$x] = str_replace($backremove,"",$image[$x]);
		$image[$x] = substr($image[$x], 0, 17);
		$x++;
	}
}
for($n=0;$n<=$x;$n++) {
	$file = 'http://www.somedomain.com/photos/'.$image[$n];
	$newfile = $_SERVER['DOCUMENT_ROOT'] . '/photos/'.$image[$n];
	echo $newfile;
	fopen($newfile, "w");
	if (!copy($file, $newfile)) {
		echo "failed to copy $file...\n";
	}
	fclose($newfile);
}
?>

Now that I'm looking at this script again under a well-rested mindset, I think I could simplify this and avoid using DOM parsing altogether since we know the remote directory. This simpler method would save memory and probably work more efficiently. I probably overcomplicated this script.

However, an advantage to using DOM parsing is that you would only copy over the files that you want from the page, rather than running the risk of copying any other files that may be living in that directory that you may not want.

Oh well, it's all a learning experience. I'll play with a simplified version later.

These are the tags assigned to a sample post. The tag "sonoma" appears slightly larger than the other tags assigned to this post because it appears more frequently throughout all other posts.

A recent development project called for a tag cloud on the right rail. Easy to do generally; drupal has plenty of taxonomy modules available to take care of this for the most part.

What I needed that was different than any existing functionality I found was a tag cloud that showed:

1) A list of the node's terms in a block. These would be terms assigned to this post rather than an entire vocabulary.
2) The relative weights of those terms (and displayed visually using font size) as in a typical tag cloud.

The challenge was combining parts one and two. If you needed to apply this, the only thing in theory you'd need to change would be the vid = 5 in the database query to what ever vocabulary ID in which you want to run your query. So here is some code that does the trick (it could be simplified I think):

<?php
class tag
{
    var $name = "";
    var $count = 0;
    var $tid = 0;

    function getCount()
    {
        return $this->count;
    }
    function getName()
    {
        return $this->name;
    }
    function setCount($var)
    {
        $this->count = $var;
    }
    function setName($var)
    {
        $this->name = $var;
    }
    function setTID($var)
    {
        $this->tid = $var;
    }
    function getTID()
    {
        return $this->tid;
    }
}
$count = 0; // Overall count of used tags
$threshold = 0; // How many tags are needed to get displayed
$font_size = 0.8;
$node = node_load(arg(1));
$nid = $node->nid;
$query = "SELECT d.name,d.tid, dn.nid FROM {term_data} d, {term_node} dn WHERE dn.tid=d.tid and d.vid='5';";
$result = db_query($query);
$tags['Test'] = 0;
while($node = db_fetch_array($result))
{
    
    if ($tags[$node['name']] == NULL)
    {
        $tags[$node['name']] = new tag();
        $tags[$node['name']]->setName($node['name']);
        $tags[$node['name']]->setCount(1);
        $tags[$node['name']]->setTID($node['tid']);
        $count = $count + 1;
    }
    else
    {
        $tags[$node['name']]->setCount(
            $tags[$node['name']]->getCount() + 1
        );
        $count = $count + 1;
    }
}

$query2 = "SELECT d.name,d.tid, dn.nid FROM {term_data} d, {term_node} dn WHERE dn.tid=d.tid and dn.nid = '$nid' and d.vid='5';";
$result2 = db_query($query2);
$match1 =  $node['name'];
$match2 = $tag->name;
	while($node = db_fetch_array($result2)) {
			$termnames[] = $node['name'];
	}

foreach($tags as $tag)
{
    foreach($termnames as $termname){              
                 if($tag->name == $termname){
                 $matches[] = $tag; 
				}	
		}
}

foreach($matches as $match)
 {
    $mycount = $match->count;

    if($mycount > $threshold) 

    {
        $fraction = ((int)(($mycount / $count) * 60)) / 10;
        if($fraction < $font_size)
        {
            $fraction = $font_size;
        }
        echo '<span style="font-size: ' . $fraction . 'em;">' . l($match->name,'taxonomy/term/' . $match->tid,array('title="' . $match->count . ' Nodes"')).'</span> ';
    }
}

?>

The first half of the code I lifted from a snippet or forum post on Drupal.org many months ago, and the second half I managed to cobble together somehow on a cloudy day when the code gods were looking the other way.