Menu
Community /Pin to ProfileBookmark
@NogDogJul 09.2010 — #You could do it pretty easily via the [url=http://php.net/dom]DOM extension[/url] , though I'll note the "type" attribute for an <a> element is optional, and there's no guarantee that even if it is set it's correct.@NogDogJul 09.2010 — #PS: I just threw this together, seems to work with my simple test data:
rss stripper
Anyone have a function that will search an html page for all rss links?
Steps:
1) Load html page
2) look for type=”application/rss+xml”
3) pull href link
Thanks
Sign in
to post a comment3 Comments(s) ↴
0
@NogDogJul 09.2010 — #You could do it pretty easily via the 0
@NogDogJul 09.2010 — #PS: I just threw this together, seems to work with my simple test data:[code=php]
<?php
class DOMScraper
{
/**
* @var object $dom DOMDocument
*/
private $dom;
/**
* Constructor
* @return void
* @param object $dom DOMDocument
*/
public function __construct(DOMDocument $dom)
{
$this->dom = $dom;
}
/**
* Get element by Tag name
* @return array
* @param array $attrs Attribute_name => Attribute_value
*/
public function byTag($tag, $attrCriteria = null)
{
$elements = $this->dom->getElementsByTagName($tag);
$data = array();
foreach($elements as $elem)
{
$attrs = array();
foreach($elem->attributes as $name => $value)
{
$attrs[$name] = $value->textContent;
}
$attrs['textContent'] = $elem->textContent;
if(!empty($attrCriteria))
{
if(!is_array($attrCriteria))
{
user_error('attrCriteria must be an array');
return false;
}
if(!$this->criteraMatch($attrs, $attrCriteria))
{
continue;
}
}
$data[] = $attrs;
}
return $data;
}
/**
* Test if node has desired attributes
* @return bool
* @param array $attrs
* @param array $attrCritera
*/
private function criteraMatch($attrs, $attrCritera)
{
ksort($attrs);
ksort($attrCritera);
return count(array_intersect_assoc($attrCritera, $attrs)) ==
count($attrCritera);
}
}
// USAGE:
$dom = new DOMDocument();
$dom->loadHTMLFile('test.html');
$test = new DOMScraper($dom);
$result = $test->byTag('a', array('type' => 'application/rss+xml'));
echo "<pre>".print_r($result,1)."</pre>";
[/code]