摘自织梦CMS的HTTP文件下载类

<?php if(!defined('dedeinc')) exit("request error!");
/**
 * 织梦http下载类
 *
 * @version  $id: dedehttpdown.class.php 1 11:42 2010年7月6日z tianya $
 * @package  dedecms.libraries
 * @copyright  copyright (c) 2007 - 2010, desdev, inc.
 * @license  http://help.dedecms.com/usersguide/license.html
 * @link   http://www.dedecms.com
 */
@set_time_limit(0);
class dedehttpdown
{
 var $m_url = '';
 var $m_urlpath = '';
 var $m_scheme = 'http';
 var $m_host = '';
 var $m_port = '80';
 var $m_user = '';
 var $m_pass = '';
 var $m_path = '/';
 var $m_query = '';
 var $m_fp = '';
 var $m_error = '';
 var $m_httphead = '';
 var $m_html = '';
 var $m_puthead = '';
 var $baseurlpath = '';
 var $homeurl = '';
 var $retry = 0;
 var $jumpcount = 0;
 /**
  * 初始化系统
  *
  * @access public
  * @param  string $url 需要下载的地址
  * @return string
  */
 function privateinit($url)
 {
  if($url=='') {
   return ;
  }
  $urls = '';
  $urls = @parse_url($url);
  $this->m_url = $url;
  if(is_array($urls))
  {
   $this->m_host = $urls["host"];
   if(!empty($urls["scheme"]))
   {
    $this->m_scheme = $urls["scheme"];
   }
   if(!empty($urls["user"]))
   {
    $this->m_user = $urls["user"];
   }
   if(!empty($urls["pass"]))
   {
    $this->m_pass = $urls["pass"];
   }
   if(!empty($urls["port"]))
   {
    $this->m_port = $urls["port"];
   }
   if(!empty($urls["path"]))
   {
    $this->m_path = $urls["path"];
   }
   $this->m_urlpath = $this->m_path;
   if(!empty($urls["query"]))
   {
    $this->m_query = $urls["query"];
    $this->m_urlpath .= "?".$this->m_query;
   }
   $this->homeurl = $urls["host"];
   $this->baseurlpath = $this->homeurl.$urls["path"];
   $this->baseurlpath = preg_replace("/\/([^\/]*)\.(.*)$/","/",$this->baseurlpath);
   $this->baseurlpath = preg_replace("/\/$/","",$this->baseurlpath);
  }
 }
 /**
  * 重设各参数
  *
  * @access public
  * @return void
  */
 function resetany()
 {
  $this->m_url = "";
  $this->m_urlpath = "";
  $this->m_scheme = "http";
  $this->m_host = "";
  $this->m_port = "80";
  $this->m_user = "";
  $this->m_pass = "";
  $this->m_path = "/";
  $this->m_query = "";
  $this->m_error = "";
 }
 /**
  * 打开指定网址
  *
  * @access public
  * @param  string $url 地址
  * @param  string $requesttype 请求类型
  * @return string
  */
 function openurl($url,$requesttype="get")
 {
  $this->resetany();
  $this->jumpcount = 0;
  $this->m_httphead = array() ;
  $this->m_html = '';
  $this->retry = 0;
  $this->close();
  //初始化系统
  $this->privateinit($url);
  $this->privatestartsession($requesttype);
 }
 /**
  * 转到303重定向网址
  *
  * @access public
  * @param  string $url 地址
  * @return string
  */
 function jumpopenurl($url)
 {
  $this->resetany();
  $this->jumpcount++;
  $this->m_httphead = array() ;
  $this->m_html = "";
  $this->close();
  //初始化系统
  $this->privateinit($url);
  $this->privatestartsession('get');
 }
 /**
  * 获得某操作错误的原因
  *
  * @access public
  * @return void
  */
 function printerror()
 {
  echo "错误信息：".$this->m_error;
  echo "<br/>具体返回头：<br/>";
  foreach($this->m_httphead as $k=>$v){ echo "$k => $v <br/>\r\n"; }
 }
 /**
  * 判别用get方法发送的头的应答结果是否正确
  *
  * @access public
  * @return bool
  */
 function isgetok()
 {
  if( preg_match("/^2/",$this->gethead("http-state")) )
  {
   return true;
  }
  else
  {
   $this->m_error .= $this->gethead("http-state")." - ".$this->gethead("http-describe")."<br/>";
   return false;
  }
 }
 /**
  * 看看返回的网页是否是text类型
  *
  * @access public
  * @return bool
  */
 function istext()
 {
  if( preg_match("/^2/",$this->gethead("http-state")) && preg_match("/text|xml/i",$this->gethead("content-type")) )
  {
   return true;
  }
  else
  {
   $this->m_error .= "内容为非文本类型或网址重定向<br/>";
   return false;
  }
 }
 /**
  * 判断返回的网页是否是特定的类型
  *
  * @access public
  * @param  string $ctype 内容类型
  * @return string
  */
 function iscontenttype($ctype)
 {
  if(preg_match("/^2/",$this->gethead("http-state"))
  && $this->gethead("content-type")==strtolower($ctype))
  { return true; }
  else
  {
   $this->m_error .= "类型不对 ".$this->gethead("content-type")."<br/>";
   return false;
  }
 }
 /**
  * 用http协议下载文件
  *
  * @access public
  * @param  string $savefilename 保存文件名称
  * @return string
  */
 function savetobin($savefilename)
 {
  if(!$this->isgetok())
  {
   return false;
  }
  if(@feof($this->m_fp))
  {
   $this->m_error = "连接已经关闭！"; return false;
  }
  $fp = fopen($savefilename,"w");
  while(!feof($this->m_fp))
  {
   fwrite($fp, fread($this->m_fp, 1024));
  }
  fclose($this->m_fp);
  fclose($fp);
  return true;
 }
 /**
  * 保存网页内容为text文件
  *
  * @access public
  * @param  string $savefilename 保存文件名称
  * @return string
  */
 function savetotext($savefilename)
 {
  if($this->istext())
  {
   $this->savebinfile($savefilename);
  }
  else
  {
   return "";
  }
 }
 /**
  * 用http协议获得一个网页的内容
  *
  * @access public
  * @return string
  */
 function gethtml()
 {
  if(!$this->istext())
  {
   return '';
  }
  if($this->m_html!='')
  {
   return $this->m_html;
  }
  if(!$this->m_fp||@feof($this->m_fp))
  {
   return '';
  }
  while(!feof($this->m_fp))
  {
   $this->m_html .= fgets($this->m_fp,256);
  }
  @fclose($this->m_fp);
  return $this->m_html;
 }
 /**
  * 开始http会话
  *
  * @access public
  * @param  string $requesttype 请求类型
  * @return string
  */
 function privatestartsession($requesttype="get")
 {
  if(!$this->privateopenhost())
  {
   $this->m_error .= "打开远程主机出错!";
   return false;
  }
  $this->retry++;
  if($this->gethead("http-edition")=="http/1.1")
  {
   $httpv = "http/1.1";
  }
  else
  {
   $httpv = "http/1.0";
  }
  $ps = explode('?',$this->m_urlpath);
  $headstring = '';
  //发送固定的起始请求头get、host信息
  if($requesttype=="get")
  {
   $headstring .= "get ".$this->m_urlpath." $httpv\r\n";
  }
  else
  {
   $headstring .= "post ".$ps[0]." $httpv\r\n";
  }
  $this->m_puthead["host"] = $this->m_host;
  //发送用户自定义的请求头
  if(!isset($this->m_puthead["accept"]))
  {
   $this->m_puthead["accept"] = "*/*";
  }
  if(!isset($this->m_puthead["user-agent"]))
  {
   $this->m_puthead["user-agent"] = "mozilla/4.0 (compatible; msie 6.0; windows nt 5.2)";
  }
  if(!isset($this->m_puthead["refer"]))
  {
   $this->m_puthead["refer"] = "http://".$this->m_puthead["host"];
  }
  foreach($this->m_puthead as $k=>$v)
  {
   $k = trim($k);
   $v = trim($v);
   if($k!=""&&$v!="")
   {
    $headstring .= "$k: $v\r\n";
   }
  }
  fputs($this->m_fp, $headstring);
  if($requesttype=="post")
  {
   $postdata = "";
   if(count($ps)>1)
   {
    for($i=1;$i<count($ps);$i++)
    {
     $postdata .= $ps[$i];
    }
   }
   else
   {
    $postdata = "ok";
   }
   $plen = strlen($postdata);
   fputs($this->m_fp,"content-type: application/x-www-form-urlencoded\r\n");
   fputs($this->m_fp,"content-length: $plen\r\n");
  }
  //发送固定的结束请求头
  //http1.1协议必须指定文档结束后关闭链接,否则读取文档时无法使用feof判断结束
  if($httpv=="http/1.1")
  {
   fputs($this->m_fp,"connection: close\r\n\r\n");
  }
  else
  {
   fputs($this->m_fp,"\r\n");
  }
  if($requesttype=="post")
  {
   fputs($this->m_fp,$postdata);
  }
  //获取应答头状态信息
  $httpstas = explode(" ",fgets($this->m_fp,256));
  $this->m_httphead["http-edition"] = trim($httpstas[0]);
  $this->m_httphead["http-state"] = trim($httpstas[1]);
  $this->m_httphead["http-describe"] = "";
  for($i=2;$i<count($httpstas);$i++)
  {
   $this->m_httphead["http-describe"] .= " ".trim($httpstas[$i]);
  }
  //获取详细应答头
  while(!feof($this->m_fp))
  {
   $line = trim(fgets($this->m_fp,256));
   if($line == "")
   {
    break;
   }
   $hkey = "";
   $hvalue = "";
   $v = 0;
   for($i=0;$i<strlen($line);$i++)
   {
    if($v==1)
    {
     $hvalue .= $line[$i];
    }
    if($line[$i]==":")
    {
     $v = 1;
    }
    if($v==0)
    {
     $hkey .= $line[$i];
    }
   }
   $hkey = trim($hkey);
   if($hkey!="")
   {
    $this->m_httphead[strtolower($hkey)] = trim($hvalue);
   }
  }
  //如果连接被不正常关闭，重试
  if(feof($this->m_fp))
  {
   if($this->retry > 10)
   {
    return false;
   }
   $this->privatestartsession($requesttype);
  }
  //判断是否是3xx开头的应答
  if(preg_match("/^3/",$this->m_httphead["http-state"]))
  {
   if($this->jumpcount > 3)
   {
    return;
   }
   if(isset($this->m_httphead["location"]))
   {
    $newurl = $this->m_httphead["location"];
    if(preg_match("/^http/i",$newurl))
    {
     $this->jumpopenurl($newurl);
    }
    else
    {
     $newurl = $this->fillurl($newurl);
     $this->jumpopenurl($newurl);
    }
   }
   else
   {
    $this->m_error = "无法识别的答复！";
   }
  }
 }
 /**
  * 获得一个http头的值
  *
  * @access public
  * @param  string $headname 头文件名称
  * @return string
  */
 function gethead($headname)
 {
  $headname = strtolower($headname);
  return isset($this->m_httphead[$headname]) ? $this->m_httphead[$headname] : '';
 }
 /**
  * 设置http头的值
  *
  * @access public
  * @param  string $skey 键
  * @param  string $svalue 值
  * @return string
  */
 function sethead($skey,$svalue)
 {
  $this->m_puthead[$skey] = $svalue;
 }
 /**
  * 打开连接
  *
  * @access public
  * @return bool
  */
 function privateopenhost()
 {
  if($this->m_host=="")
  {
   return false;
  }
  $errno = "";
  $errstr = "";
  $this->m_fp = @fsockopen($this->m_host, $this->m_port, $errno, $errstr,10);
  if(!$this->m_fp)
  {
   $this->m_error = $errstr;
   return false;
  }
  else
  {
   return true;
  }
 }
 /**
  * 关闭连接
  *
  * @access public
  * @return void
  */
 function close()
 {
  @fclose($this->m_fp);
 }
 /**
  * 补全相对网址
  *
  * @access public
  * @param  string $surl 需要不全的地址
  * @return string
  */
 function fillurl($surl)
 {
  $i = 0;
  $dstr = "";
  $pstr = "";
  $okurl = "";
  $pathstep = 0;
  $surl = trim($surl);
  if($surl=="")
  {
   return "";
  }
  $pos = strpos($surl,"#");
  if($pos>0)
  {
   $surl = substr($surl,0,$pos);
  }
  if($surl[0]=="/")
  {
   $okurl = "http://".$this->homeurl.$surl;
  }
  else if($surl[0]==".")
  {
   if(strlen($surl)<=1)
   {
    return "";
   }
   else if($surl[1]=="/")
   {
    $okurl = "http://".$this->baseurlpath."/".substr($surl,2,strlen($surl)-2);
   }
   else
   {
    $urls = explode("/",$surl);
    foreach($urls as $u)
    {
     if($u=="..")
     {
      $pathstep++;
     }
     else if($i<count($urls)-1)
     {
      $dstr .= $urls[$i]."/";
     }
     else
     {
      $dstr .= $urls[$i];
     }
     $i++;
    }
    $urls = explode("/",$this->baseurlpath);
    if(count($urls) <= $pathstep)
    {
     return "";
    }
    else
    {
     $pstr = "http://";
     for($i=0;$i<count($urls)-$pathstep;$i++)
     {
      $pstr .= $urls[$i]."/";
     }
     $okurl = $pstr.$dstr;
    }
   }
  }
  else
  {
   if(strlen($surl)<7)
   {
    $okurl = "http://".$this->baseurlpath."/".$surl;
   }
   else if(strtolower(substr($surl,0,7))=="http://")
   {
    $okurl = $surl;
   }
   else
   {
    $okurl = "http://".$this->baseurlpath."/".$surl;
   }
  }
  $okurl = preg_replace("/^(http:\/\/)/i","",$okurl);
  $okurl = preg_replace("/\/{1,}/", "/", $okurl);
  return "http://".$okurl;
 }
}//end class