I was tired of living in the early 2000’s so I changed my kpages rss feed service to AJAX….
But lo and behold, I like it a lot better the way it was before. I used to load the entire page, and every feed all at once. When a user wanted to read a feed it would appear. Now it loads the first page REALLY fast, but loads the feeds a little too slow for my liking.

Compare old vs. new.

But you probably want to know how I did it.

1. Write an RSS Parser. This was probably the biggest part, but also the most interesting. RSS feeds are very complex, so I had to interpret them in many different ways. When it was of type RDF I was able to strictly use the DOM functions that are built into PHP 5 to read them.

   case 'rdf:RDF': #Stupid RSS
     $title = $dom->getElementsByTagName('title')->item(0)->nodeValue;
     $link = $dom->getElementsByTagName('link')->item(0)->nodeValue;
     $desc = $dom->getElementsByTagName('description')->item(0)->nodeValue;
     break;
   

   When it was an ATOM feed it was a little more tricky because so many sites’ feeds are uncompliant. I had to triple check to get the link back to the homepage. A like so…
   

   case 'feed': #ATOM
     $title = $dom->getElementsByTagName('title')->item(0)->nodeValue;
     $link = $dom->getElementsByTagName('link')->item(0)->nodeValue;
     if($link == '')
     {
       $link_list = $dom->getElementsByTagName('link');
       foreach($link_list as $link_node)
       {
       	$link_type = $link_node->getAttribute('type');
       	$link_href = $link_node->getAttribute('href');
       	if($link_type == 'text/html')
       	{
       		$link = $link_href;
       		break;
       	}
       }
      }
      $desc = $dom->getElementsByTagName('tagline')->item(0)->nodeValue;
      break;
   

   Regular RSS feeds were the easiest to read. I even got to use the xpath functions of the DOM XML module. Piece of cake, really.
   

   default: #Nice RSS
     $tmp_title = $xpath->query("/$type/channel/title");
     $title = $tmp_title->item(0)->nodeValue;
     $tmp_link = $xpath->query("/$type/channel/link");
     $link = $tmp_link->item(0)->nodeValue;
     $tmp_desc = $xpath->query("/$type/channel/description");
     $desc = $tmp_desc->item(0)->nodeValue;
     break;
   

   
   The next part of the parser is arguably the most important step, actually reading the feeds. I accomplished this by getting all of the “stories” from the rss
   

   function get_items($dom)
   {
       switch($this->feed_type)
       {
         case 'feed':
           $item_name = 'entry';
           $desc_name = 'content';
           break;
         default:
           $item_name = 'item';
           $desc_name = 'description';
       }
       $items = $dom->getElementsByTagName("$item_name");
       $item_list = array();
       foreach ($items as $item)
       {
       	$title = $item->getElementsByTagName('title')->item(0)->nodeValue;
       	//First Try to get Link
       	$link = $item->getElementsByTagName('link')->item(0)->nodeValue;
       	//Try again
      	    if($link == '')
      	    {
       	   	$link_list = $item->getElementsByTagName('link');
       	   	foreach($link_list as $link_node)
       	   	{
                      $link_type = $link_node->getAttribute('type');
       	   	    $link_href = $link_node->getAttribute('href');
       	   	    if($link_type == 'text/html')
       	   	    {
       	   		$link = $link_href;
       	   		break;
       	   	    }
       	   	}
      	    }
      	    //Last try
      	    if($link == '')
      	    {
      	    	$link_node = $item->getElementsByTagName('link')->item(0);
      	    	$link = $link_node->getAttribute('href');
      	    }
   
       	$desc = $item->getElementsByTagName("$desc_name")->item(0)->nodeValue;
       	$hash = array(
       	       'title' => $title,
       	       'link'  => $link,
       	       'desc'  => $desc
               );
            array_push($item_list, $hash);
       }
       return $item_list;
   
   }

This entry was posted on Thursday, April 20th, 2006 at 10:42 pm and is filed under Ideas, Web Development, Web Projects. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.