获取关键词及所属来源搜索引擎名称 

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/html4/strict.dtd">
<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=gb2312" />
    <title>greengnn codes</title>
  </head>
  <body>
<form action="index.php" method="get">
    <input name="url" size="80"><button type="submit">GO</button>
</form>

<?php
//获取关键词及所属来源搜索引擎名称

$search_url = isset($_GET['url'])?$_GET['url']:''; //代表传入的地址
/*$search_url = urldecode($search_url);
print_r($search_url);
*/

$config = array(
    "s1"=>array(
        "domain" => "google.com",
        "kw" => "q",
        "charset" => "utf-8"
    ),
    "s3"=>array(
        "domain" => "google.cn",
        "kw" => "q",
        "charset" => "utf-8"
    ),
    "s4"=>array(
        "domain" => "baidu.com",
        "kw" => "wd",
        "charset" => "gbk"
    ),
    "s5"=>array(
        "domain" => "soso.com",
        "kw" => "q",
        "charset" => "utf-8"
    ),
    "s6"=>array(
        "domain" => "yahoo.com",
        "kw" => "q",
        "charset" => "utf-8"
    ),
    "s7"=>array(
        "domain" => "bing.com",
        "kw" => "q",
        "charset" => "utf-8"
    ),
    "s8"=>array(
        "domain" => "sogou.com",
        "kw" => "query",
        "charset" => "gbk"
    ),
    "s9"=>array(
        "domain" => "youdao.com",
        "kw" => "q",
        "charset" => "utf-8"
    ),
);

//函数作用：从url中提取关键词。参数说明：url及关键词前的字符。
function get_keyword($url,$kw_start)

{
    $start = stripos($url,$kw_start);
    $url = substr($url,$start+strlen($kw_start));
    $start = stripos($url,'&');
    if ($start>0)
    {
        $start=stripos($url,'&');
        $s_s_keyword=substr($url,0,$start);
    }
    else 
    {
        $s_s_keyword=substr($url,0);
    }
    return $s_s_keyword;
}

$arr_key = array();
foreach($config as $item){
    $sh = preg_match("/\b{$item['domain']}\b/",$search_url);
    if($sh){
        $query = $item['kw']."=";

        $s_s_keyword = get_keyword($search_url,$query);
        $F_Skey=urldecode($s_s_keyword);
        if($item['charset']=="utf-8"){
            $F_Skey=iconv( "UTF-8","gb2312//IGNORE",$F_Skey); //最终提取的关键词
        }
        $keys = explode(" ",$F_Skey);
        $arr_key[$item['domain']] = $keys;
    }
}
echo "<pre>";
print_r($arr_key);
posted @ 2011-02-14 18:30  greengnn  阅读(699)  评论(0)    收藏  举报