Question

我试图弄清楚为什么Chrome会发生这种情况，而不是FF或IE。在以下脚本中，$ this-＆gt; referrer = $ _SESSION ['URL'];解析为/第一次加载页面时，如果我刷新页面，它会解析为/404.php

function startSession(){
  global $database;  //The database connection
  session_name("Training");
  session_start();   //Tell PHP to start the session

  /* Determine if user is logged in */
  $this->logged_in = $this->checkLogin();

  /**
   * Set guest value to users not logged in, and update
   * active guests table accordingly.
   */
  if(!$this->logged_in){
     $this->username = $_SESSION['username'] = GUEST_NAME;
     $this->userlevel = GUEST_LEVEL;
     $database->addActiveGuest($_SERVER['REMOTE_ADDR'], $this->time);
  }
  /* Update users last active timestamp */
  else{
     $this->addLogCount($this->id);
     $database->addActiveUser($this->username, $this->time);
  }

  /* Remove inactive visitors from database */
  $database->removeInactiveUsers();
  $database->removeInactiveGuests();

  /* Set referrer page */
  if(isset($_SESSION['URL'])){
     $this->referrer = $_SESSION['URL'];
  }else{
     $this->referrer = "/";
  }

  /* Set current url */
  $this->url = $_SESSION['URL'] = $_SERVER['PHP_SELF'];
}

对于FF和IE，它首先解析为/，然后解析为刷新时调用startSession（）的页面的地址。有什么方法可以解决Chrome需要考虑的问题吗？

Chrome中的

$ session：

Session Object
(
    [id] => 
    [tracksid] => 
    [username] => Guest
    [userid] => 
    [userlevel] => 0
    [time] => 1297103371
    [logged_in] => 
    [userinfo] => Array
        (
        )

    [url] => /Community/login.php
    [referrer] => /404.php
    [type] => 
    [company] => 
    [ip] => 
    [badip] => 
)

FireFox中的

$ session：

Session Object
(
    [id] => 
    [tracksid] => 
    [username] => Guest
    [userid] => 
    [userlevel] => 0
    [time] => 1297096106
    [logged_in] => 
    [userinfo] => Array
        (
        )

    [url] => /Community/login.php
    [referrer] => /Community/login.php
    [type] => 
    [company] => 
    [ip] => 
    [badip] => 
)

Answer 1

记录对404.php的所有访问权限。另请查看访问日志和错误日志。通过这种方式，您知道谁在哪个时间访问了哪个网址（我猜无效网址由404.php处理）。我怀疑某种类型的Web加速器试图下载不存在的文件，或者HTML中没有指定favicon，因此浏览器会尝试猜测并下载标准文件。

Chrome使用$ _SESSION ['URL']执行奇怪的操作，返回404页面

1 个答案: