Voting

The Note You're Voting On

15 years ago
A function among several others to parse a google results page, I wrote this some time ago - google has probably changed their site since then, but I thought this might be helpful to someone.

I'm moving servers, but I will probably throw this up on my blog when I get it back up.

<?php

function googleResult($listItem) {
    // given a LIST ITEM element, this will validate, and return an array for that LI entry as an inline result from google.
    /*
     * <li class='g w0'>
     *   <h3 class='r'>
     *     <a href='the URL' class='l'>
     *       Description <em>description</em>
     *     </a>
     *   </h3>
     * </li>
     *
       UPDATE:
       This function will now look for any subcontainer that has an href, it doesn't have to be an H3
       this will make it work with a few more formatted search results.
     */

    $listItem = $listItem->childNodes;
        // Yes I don't use instanceof - I guess you'll have to deal.
    foreach($listItem as $element) {
        if(is_object($element) && get_class($element) == 'DOMElement' && $element->hasChildNodes()) {
            $hrefContainer = $element->childNodes;
            foreach($hrefContainer as $element2) {
                if(is_object($element2) && get_class($element2) == 'DOMElement' && $element2->nodeName == 'a' && $element2->hasAttribute('href')) {
                    $anchor = $element2;
                    unset($h3);
                    unset($element2);
                    break;
                } else {
                  //print __LINE__ ." :: Breaking out of loop (normal result) element is not an annchor Element='".$element2->nodeName."'\n";
                }
            }
            unset($element);
            unset($listItem);
            break;
        }
    }
    if(empty($anchor) || !is_object($anchor) || get_class($anchor) != 'DOMElement') {
        //print __LINE__ ." :: Returning false, did not locate anchor through iteration...";
        return false;
    }
    $href = $anchor->getAttribute('href');
    if(empty($href)) {
        //print __LINE__ ." :: Found anchor object, could not read href attribute / href is empty? href='$href'\n";
        return false;
    }
    $description = $anchor->childNodes;
    $urlDescription = '';
    foreach($description as $words) {
        $name =  trim($words->nodeName);
        if($name == 'em' || $name == '#text' || $name == 'b') {
            if(!empty($words->nodeValue)) {
                $text = trim($words->nodeValue);
                $urlDescription = $urlDescription . $text . ' ';
            }
        }
    }
    $urlDescription = htmlspecialchars_decode($urlDescription, ENT_QUOTES);
    $urlDescription = trim($urlDescription);
    return array('description' => $urlDescription, 'href' => $href);
}
<< Back to user notes page