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;
}