最近在开发团购网站的过程中,有个把原文长链接转换成短链接的需求。研究了下,大概有下面几种思路。
一、随机生成
function random($length, $pool = '') { $random = ''; $pool = 'abcdefghkmnpqrstuvwxyz123456789'; for($i = 0; $i < $length; $i++) { $j = rand(0,strlen($pool)); $random .= $pool[$j]; } return $random; } $a=random(6); print_r($a);
二、 枚举生成方法
将长网址md5生成32位签名串,分为4段,每段8个字节;
对这四段循环处理,取8个字节,将他看成16进制串与0x3fffffff(30位1)与操作,即超过30位的忽略处理;
这30位分成6段,每5位的数字作为字母表的索引取得特定字符,依次进行获得6位字符串;
总的md5串可以获得4个6位串;取里面的任意一个就可作为这个长url的短url地址;
function shorturl($url) { $base32 = array ( "a","b","c","d","e","f","g","h", "i","j","k","l","m","n","o","p", "q","r","s","t","u","v","w","x", "y","z","0","1","2","3","4","5", "6","7","8","9","A","B","C","D", "E","F","G","H","I","J","K","L", "M","N","O","P","Q","R","S","T", "U","V","W","X","Y","Z" ); $hex = md5($url); $output = array(); for ($i = 0; $i < 4; $i++) { $subHex = substr ($hex, $i * 8, 8); echo $subHex; //把加密字符按照8位一组16进制与0x3FFFFFFF进行位与运算 $int = 0x3FFFFFFF & (1 * ('0x'.$subHex)); echo $int; $out = ''; for ($j = 0; $j < 6; $j++) { //把得到的值与0x0000003D进行位与运算,取得字符数组chars索引 $val = 0x0000001f & $int; echo $val; //把取得的字符相加 $out .= $base32[$val]; //每次循环按位右移5位 $int = $int >> 5; } $output[] = $out; } return $output; } $a=shorturl("http://www.baidu.com"); print_r($a);
三、62 位生成方法
function base62($x) { $show= ''; while($x> 0) { $s= $x% 62; if($s> 35) { $s= chr($s+61); } elseif($s> 9 && $s<=35) { $s= chr($s+ 55); } $show.= $s; $x= floor($x/62); } return $show; } function urlShort($url) { $url= crc32($url); $result= sprintf("%u", $url); return base62($result); } echo urlShort("http://www.baidu.com/");
四、总结
长网址转短网址,其实就是个62位的转换存储问题,下面来说一下PHP的62进制与10进制互相转化的实现。
/** * 十进制数转换成62进制 * * @param integer $num * @return string */ function from10_to62($num) { $to = 62; $dict = '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'; $ret = ''; do { $ret = $dict[bcmod($num, $to)] . $ret; $num = bcdiv($num, $to); } while ($num > 0); return $ret; } /** * 62进制数转换成十进制数 * * @param string $num * @return string */ function from62_to10($num) { $from = 62; $num = strval($num); $dict = '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'; $len = strlen($num); $dec = 0; for($i = 0; $i < $len; $i++) { $pos = strpos($dict, $num[$i]); $dec = bcadd(bcmul(bcpow($from, $len - $i - 1), $pos), $dec); } return $dec; }