php实现递归抓取

php实现递归抓取

<?php

class crawler{

 private $_depth=5;

 private $_urls=array();

 function extract_links($url)

 {

  if(!$this->_started){

   $this->_started=1;

   $curr_depth=0;

  }else{

   $curr_depth++;

  }

  if($curr_depth_depth)

  {

   $data=file_get_contents($url);

   if(preg_match_all('/((?:http|https)://(?:www.)*(?:[a-zA-Z0-9_-]{1,15}.+[a-zA-Z0-9_]{1,}){1,}(?:[a-zA-Z0-9_/.-?&:%,!;]*))/',$data,$urls12))

   {

    foreach($urls12[0] as $k=>$v){

     $check=get_headers($v,1);

     if(strstr($v,$url) && $check[0]=='HTTP/1.1 200 OK' && !array_search($v,$this->_urls) && $curr_depth_depth){

      $this->_urls[]=$v;

      $this->extract_links($v);

     }//codego.net

    }

   }

  }

  return $this->_urls;

 }

} 

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/29973590/viewspace-1485210/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/29973590/viewspace-1485210/