crawler.php

<?php
/**
      Copyright 2011 Audun Larsen. All rights reserved.
      [email protected]

   Redistribution and use, with or without modification,
   are permitted provided that the following condition is met:

   * Redistribution and use of source code must retain the above copyright notice,
     this list of conditions and the following disclaimer.

   THIS SOFTWARE IS PROVIDED BY ``AS IS''
   IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY TYPE OF
   DAMAGE ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE.

 */

require 'Snoopy.class.php';
 
/* This is gonna be a really simple crawler. We are going to store all URIs in this array */
$quee = array(
  'http://www.joomlainorge.no',
  'http://www.google.no/search?hl=no&client=opera&hs=4k2&rls=en&channel=suggest&q=inurl%3Akommune.no&aq=f&aqi=&aql=&oq=',
);
 
 /* All the URIs we have crawled will be in this */
$crawled = array();

while(list($key, $uri) = each($quee)) {
  $snoopy = new Snoopy;
  $crawled[] = $uri;
  unset($quee[$key]);  
  
  echo "Crawling $uri. (Crawled $key pages. ".sizeof($quee)." in quee.)\n";
  
  $snoopy->fetchlinks($uri);
  $links = $snoopy->results;
  foreach($links as $link) {
    if(strpos($link, 'kommune.no') !== false) {
      addQuee(expandUri($link, $uri));
    }
  }
  
  /* Now, check if site uses Joomla */
  /* First, get the base URL */
  $url = parse_url($uri);
  $testUrl = $url['scheme'].'://'.$url['host'].'/index.php?tp=1';
  $snoopy = new Snoopy;
  $snoopy->fetch($testUrl);
	$text = $snoopy->results;
  if(strpos($text, '<div class="mod-preview-info">') !== false) {
    echo "Joomla!!!\n\n";
  } else {
    echo "Not joomla.\n\n";
  }
  sleep(3);
}

function addQuee($uri) {
  global $crawled, $quee;
  if(!in_array($uri, $crawled) && !in_array($uri, $quee)) {
    $quee[] = $uri;
  }
}


function expandUri($uri, $base) {  
  return $uri;
}