Help with code

[Pro.Luv]

Hi,

 

I have this code that gets links from web pages but it also gets javascript links like if a link is:

 

 document.form1.brand.options[document.form1.brand.selectedIndex].value 

 

I don't want it to fetch those links how can I stop it ?

 

<?

    function getLinks($link)
    {
        /*** return array ***/
        $ret = array();

        /*** a new dom object ***/
        $dom = new domDocument;

        /*** get the HTML (suppress errors) ***/
        @$dom->loadHTML(file_get_contents($link));

        /*** remove silly white space ***/
        $dom->preserveWhiteSpace = false;

        /*** get the links from the HTML ***/
        $links = $dom->getElementsByTagName('a');
    
        /*** loop over the links ***/
        foreach ($links as $tag)
        {
            $ret[$tag->getAttribute('href')] = $tag->childNodes->item(0)->nodeValue;
        }

        return $ret;
    }


/*** a link to search ***/
    $link = "http://www.website.com";

    /*** get the links ***/
    $urls = getLinks($link);

    /*** check for results ***/
    if(sizeof($urls) > 0)
    {
        foreach($urls as $key => $value)
        {
            echo $key.'<br >';
        }
    }
    else
    {
        echo "No links found at $link";
    }
?>

 

thanks

 

 

[jonsjava]

My IDE is broke (switched to Gnome from KDE, and Zend died), so I couldn't test this, but it should do the trick.

<?php

    function getLinks($link)
    {
        /*** return array ***/
        $ret = array();

        /*** a new dom object ***/
        $dom = new domDocument;

        /*** get the HTML (suppress errors) ***/
        @$dom->loadHTML(file_get_contents($link));

        /*** remove silly white space ***/
        $dom->preserveWhiteSpace = false;

        /*** get the links from the HTML ***/
        $links = $dom->getElementsByTagName('a');
   
        /*** loop over the links ***/
        foreach ($links as $tag)
        {
    if (!stristr($tag, "document."){
      $ret[$tag->getAttribute('href')] = $tag->childNodes->item(0)->nodeValue;
    }
        }

        return $ret;
    }


/*** a link to search ***/
    $link = "http://www.website.com";

    /*** get the links ***/
    $urls = getLinks($link);

    /*** check for results ***/
    if(sizeof($urls) > 0)
    {
        foreach($urls as $key => $value)
        {
            echo $key.'<br >';
        }
    }
    else
    {
        echo "No links found at $link";
    }
?>