优酷的视频下载有以下几个特点：
- 地址动态生成，每次请求返回的地址都不一样。
- 有效时间短，得到的下载地址大约只有1小时的有效时间。
- 视频地址经过加密，需在客户（用户）端进行解密。
- 长视频会被分割成多段短视频。
- 对视频下载没有限制，即用户A得到的下载地址，用户B也可以下载。
先来看一下解析后的视频地址：
http://f.youku.com/player/getFlvPath/sid/130086939328910582812_00/st/flv/fileid/03000201004D8858360BD1047C4F5FF471CDD7-C742-8D74-3EED-90A9EF54EEC1?K=de1515a31372faac182698bc
以上三段红色部分分别代表sid、fileid和key。
我们来分析一下这个地址，除了固定的部分以外，整个地址由sid、fileid和key三部分组成，下面我们逐一来分析如何解析这三个值。
以普通的优酷视频播放地址为例，
http://v.youku.com/v_show/id_XMjUyODAzNDg0.html
把其中的红色部分复制出来，拼在
http://v.youku.com/player/getPlayList/VideoIDS/
后面，得到
http://v.youku.com/player/getPlayList/VideoIDS/XMjUyODAzNDg0
访问该地址得到json格式的字符串，其中我们感兴趣的内容是：
"seed":6302, "key1":"bd7c3d19", "key2":"de1515a31372faac", "fileid":"13*18*13*13*13*11*13*42*13*13*39*44*41*41*47*41*18*29*13*60*44*42*13*39*17*33*39*56*47*56*56*39*17*42*33*44*44*17*54*33*17*39*11*54*41*44*17*39*54*18*55*55*44*54*38*13*57*38*55*56*47*39*55*55*33*42*",
我们的解析工作需要用到上面的这些内容。这里要说明一下，因为地址是动态生成的，每次请求返回的结果都不一样，所以你看到的和上面是不一样的，但是不影响解析的过程。
生成sid
sid是一个随机数，我们可以这样获得
private String genSid() {
int i1 = (int)(1000+Math.floor(Math.random()*999));
int i2 = (int)(1000+Math.floor(Math.random()*9000));
return System.currentTimeMillis()+"" + i1+"" + i2;
}生成fileid
优酷返回的fileid已经做了加密工作，好在并不难解，利用上面得到的fileid和seed
private String getFileID(String fileid,double seed){
String mixed = getFileIDMixString(seed);
String[] ids= fileid.split("\\*");
StringBuilder realId = new StringBuilder();
int idx;
for (int i=0; i< ids.length; i++){
idx = Integer.parseInt(ids[i]);
realId.append(mixed.charAt(idx));
}
return realId.toString();
}
private String getFileIDMixString(double seed){
StringBuilder mixed = new StringBuilder();
StringBuilder source = new StringBuilder(
"abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\\:._-1234567890");
int index, len = source.length();
for (int i=0; i< len;++i){
seed = (seed * 211 + 30031) % 65536;
index = (int)Math.floor(seed/65536 * source.length());
mixed.append(source.charAt(index));
source.deleteCharAt(index);
}
return mixed.toString();
}生成key
利用上面得到的key1和key2
private String genKey(String key1,String key2){
int key = Long.valueOf(key1,16).intValue();
key ^= 0xA55AA5A5;
return key2 + Long.toHexString(key);
}好了，把sid，fileid，key合并起来，就可以得到一开始的下载地址了。
接下来就是分段视频的问题了，如果一个视频分成几段，在返回的json对象中可以找到类似的内容：
"segs":{
"mp4":[{"no":"0","size":"39095085","seconds":"426"},{"no":"1","size":"22114342","seconds":"426"},{"no":"2","size":"23296715","seconds":"424"},{"no":"3","size":"18003234","seconds":"426"},{"no":"4","size":"31867294","seconds":"423"},{"no":"5","size":"14818514","seconds":"248"}],
"flv":[{"no":"0","size":"19739080","seconds":"425"},{"no":"1","size":"11506385","seconds":"426"},{"no":"2","size":"11821267","seconds":"426"},{"no":"3","size":"8988612","seconds":"426"},{"no":"4","size":"16078739","seconds":"425"},{"no":"5","size":"7634043","seconds":"245"}]}如果想下载mp4格式的，只要把下载地址中的/flv/换成/mp4/，当然你要确定该视频有mp4格式。
update:
php版代码如下：
function getSid(){
$sid = time().(rand(0,9000)+10000);
return $sid;
}
function getkey($key1,$key2){
$a = hexdec($key1);
$b = $a ^0xA55AA5A5;
$b = dechex($b);
return $key2.$b;
}
function getfileid($fileId,$seed){
$mixed = getMixString($seed);
$ids = explode("*",$fileId);
unset($ids[count($ids)-1]);
$realId = "";
for ($i=0;$i<count($ids);++$i){
$idx = $ids[$i];
$realId .= substr($mixed,$idx,1);
}
return $realId;
}
function getMixString($seed){
$mixed = "";
$source = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\\:._-1234567890";
$len = strlen($source);
for($i=0;$i<$len;++$i){
$seed = ($seed * 211 + 30031)%65536;
$index = ($seed / 65536 * strlen($source));
$c = substr($source,$index,1);
$mixed .= $c;
$source = str_replace($c,"",$source);
}
return $mixed;
}update2:
c#版代码如下：
private String genSid()
{
int i1 = (int)(1000+ Math.Floor((double)(new Random().Next(999))));
int i2 = (int)(1000+ Math.Floor((double)(new Random().Next(9000))));
TimeSpan ts = new TimeSpan(System.DateTime.UtcNow.Ticks-new DateTime(1970,1,1, 0,0,0).Ticks);
return Convert.ToInt64(ts.TotalMilliseconds).ToString()+"" + i1+"" + i2;
}
private String getFileID(String fileid,double seed)
{
String mixed = getFileIDMixString(seed);
String[] ids= fileid.Split('*');
StringBuilder realId = new StringBuilder();
int idx;
for (int i=0; i< ids.Length-1; i++)
{
idx = int.Parse(ids[i]);
realId.Append(mixed[idx]);
}
return realId.ToString();
}
private String getFileIDMixString(double seed)
{
StringBuilder mixed = new StringBuilder();
StringBuilder source = new StringBuilder("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\\:._-1234567890");
int index, len = source.Length;
for (int i=0; i< len;++i)
{
seed = (seed * 211 + 30031) % 65536;
index = (int)Math.Floor(seed/65536 * source.Length);
mixed.Append(source[index]);
source.Remove(index,1);
}
return mixed.ToString();
}
private String genKey(String key1,String key2)
{
int key = Convert.ToInt32(key1,16);
var tempkey = key ^ 0xA55AA5A5;
return key2 + Convert.ToString(tempkey,16).Substring(8);
}python版代码如下：
import time
import random
import math
def createSid():
nowTime = int(time.time() *1000)
random1 = random.randint(1000,1998)
random2 = random.randint(1000,9999)
return "%d%d%d" %(nowTime,random1,random2)
def getFileIDMixString(seed):
mixed=[]
source=list("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\:._-1234567890")
seed=float(seed)
for i in range(len(source)):
seed = (seed * 211 + 30031 ) % 65536
index = math.floor(seed /65536 *len(source))
mixed.append(source[int(index)])
source.remove(source[int(index)])
#return ''.join(mixed)
return mixed
def getFileId(fileId,seed):
mixed=getFileIDMixString(seed)
ids=fileId.split('*')
realId=[]
for ch in ids:
realId.append(mixed[int(ch)])
return ''.join(realId)
if __name__ == '__main__':
#print createSid()
#print getFileIDMixString(4528)
fileId='3*31*3*3*3*61*3*13*3*3*36*17*48*21*17*55*31*17*61*31*14*14*3*3*36*13*67*31*31*10*21*32*58*31*13*14*3*48*15*13*10*48*55*15*55*10*36*31*15*31*61*10*67*15*3*61*17*13*13*14*11*36*48*21*36*10'
seed=4528
print getFileId(fileId,seed)
