---

---

1.排行榜列表

1. 打开 kuwo.cn， 浏览器F12 监听网络请求，刷新一次页面。
2. 根据网页信息，定位接口地址。
3. F12界面全局搜索(ctrl+shift+F)排行榜中的向云端 (你看到什么就搜什么 不固定)

4. 点击搜索结果，跳转至关键词数据所在页面，在此页面搜索(ctrl+F)向云端，结论：数据来源于页面window._NUXT_.data.bang对象。如下图：
   
   
   页面用的是Nuxt服务器端渲染，看到a,b,c,d,e,f...了吧，这段js被混淆了

5. 数据提取。可能首先想到的是用正则匹配，那有没有办法办法直接获取到 __NUXT__ 对象呢？既然是js，那就想办法运行这段js。

6. 技术选型。那就用 nodejs 吧！ nodejs 爬虫实战，看代码：

    let url = 'http://www.kuwo.cn/';
    //模拟访问首页
    let result = await get(url);//get方法为request封装
    //正则匹配那段js
    var mc = result.match(/<script>.*?<\/script>/gi);
    //console.log(mc);
    var __NUXT__;//定义接收变量
    let sc = mc[0] ?? '';
    //一系列的替换
    if (sc) {
        sc = sc.replace('<script>', '');
        sc = sc.replace('</script>', '');
        sc = sc.replace('window.', '');//由于不是浏览器没有window对象，就替换为空
        sc = sc.replace(/\r/gi, '');
        sc = sc.replace(/\n/gi, '');
        //console.log(sc);
    }
    let js = eval(sc);//用nodejs的方便之处，eval搞定，直接把结果赋值给__NUXT__
    //console.log(js);
    let bangs = __NUXT__.data[0].bang;

结果：

[
    {
        "leader":"酷我热歌榜",
        "num":"300",
        "name":"酷我热歌榜",
        "pic":"https://img4.kuwo.cn/star/albumcover/120/1/29/1934979845.jpg",
        "id":"16",
        "pub":"2023-07-14",
        "musicList":[
            {
                "musicrid":"MUSIC_279292599",
                "barrage":"0",
                "ad_type":"",
                "artist":"小霞&海洋Bo",
                "mvpayinfo":{
                    "play":0,
                    "vid":0,
                    "down":0
                },
                "pic":"https://img4.kuwo.cn/star/albumcover/500/1/29/1934979845.jpg",
                "isstar":0,
                "rid":279292599,
                "duration":251,
                "score100":"93",
                "ad_subtype":"0",
                "content_type":"0",
                "track":1,
                "hasmv":0,
                "releaseDate":"2023-06-29",
                "album":"向云端",
                "albumid":38777035,
                "pay":"16515324",
                "artistid":8854278,
                "albumpic":"https://img4.kuwo.cn/star/albumcover/500/1/29/1934979845.jpg",
                "originalsongtype":0,
                "isListenFee":false,
                "pic120":"https://img4.kuwo.cn/star/albumcover/120/1/29/1934979845.jpg",
                "name":"向云端",
                "online":1,

2.排行榜榜单详情

1. 点击首页排行榜更多，监听网络请求
   

详情地址为：
GET => https://www.kuwo.cn/api/www/bang/bang/musicList?bangId=93&pn=1&rn=20&httpsStatus=1&reqId=16fffd81-21fa-11ee-9cdf-25abfdc1606c&plat=web_www&from=

bangId=93//上个步骤中的id
pn=1//页码
rn=20//每页多少条
reqId=16fffd81-21fa-11ee-9cdf-25abfdc1606//可有可无

同样用 nodejs 模拟这个请求，结果：

{
    "success":false,
    "message":"CSRF token Invalid!",
    "now":"2023-07-14T05:49:44.720Z"
}

2. 爬虫的关键部分来了，请求头模拟。简单来说，就是用程序把请求做的像浏览器发出的一样。通过分析请求头，以及反复测试，最终得出以下两个头信息为必备：(把以下所有头信息都模拟更好)
   
3. F12 界面全局搜Cross 定位出算法：
   
   由上面代码可以看出来，Cross 是通过以 Hm_token 为参数，调用A()方法生成的。打断点调试A():
   
4. 梳理这两个参数的关系：

Cross 是由 Hm_token 加密处理得到。

由经验得知：Cross = md5(sha1(Hm_token))

常用的加密算法：md5 aes des sha 不知道数据用什么加密方式就全局搜，或者一个一个加密方法试。最后才选择单步调试~

5. 加密算法有了，寻找 Hm_token :
   
   每次请求，都会往cookie里面写入Hm_token ，用程序模拟并获取cookie:

async function hmToken() {
    let url = `https://www.kuwo.cn/`;
    let result = await superagent.get(url);
    let cookie = result.header['set-cookie'][0] ?? '';
    //console.log(cookie);
    let mc = cookie.match(/Hm_token=(.*?);/);
    let hm = mc[1] ?? '';
    return hm;
}

6. 排行榜榜单详情最终代码：

async function getBangMusicList(id, pn, rn) {
    let hm = await hmToken();
    console.log(hm)
    
    var obj = crypto.createHash('sha1');
    obj.update(hm);
    let hmSha = obj.digest('hex');
    console.log(hmSha);

    obj = crypto.createHash('md5');
    obj.update(hmSha);
    var cross = obj.digest('hex');//hex是十六进制
    console.log(cross);

    let url = `http://www.kuwo.cn/api/www/bang/bang/musicList?bangId=${id}&pn=${pn}&rn=${rn}&httpsStatus=1&plat=web_www&reqId=`;
    var result = await get(url, {}, { Cross: cross, cookie: `Hm_token=${hm}` });

    var obj = JSON.parse(result);

    return obj;
}

3.根据rid获取歌曲mp3地址

1. 点首歌，监听网络请求：
   
2. 程序模拟：

async function getAudio(id) {
    let url = 'https://www.kuwo.cn/api/v1/www/music/playUrl?mid=' + id + '&type=music&httpsStatus=1&reqId=&plat=web_www&from=';
    let result = await get(url);
    //console.log(result);
    return result;
}
======================================
getAudio(279292599) result 结果：
{
    "code":200,
    "msg":"success",
    "reqId":"b9ca2ee89fec9ad131105ce494d05a2b",
    "data":{
        "url":"https://other-web-nf01-sycdn.kuwo.cn/b09d144c5526bbb95e8d1166f1c8db8a/64b0fa4a/resource/n2/60/65/902137489.mp3?from$unkown&plat$web_www"
    },
    "profileId":"site",
    "curTime":1689320311016,
    "success":true
}

4.2023.07.17更新，请求头由“Cross”改为“Secret”

F12定位Secret:

单步调试这个h()方法：

分析完毕，看具体实现：

1. 新增一个加密函数：(来源：某度)

function encrypt(str, pwd) {
    if (pwd == null || pwd.length <= 0) {
        alert("Please enter a password with which to encrypt the message.");
        return null;
    }
    var prand = "";
    for (var i = 0; i < pwd.length; i++) {
        prand += pwd.charCodeAt(i).toString();
    }
    var sPos = Math.floor(prand.length / 5);
    var mult = parseInt(prand.charAt(sPos) + prand.charAt(sPos * 2) + prand.charAt(sPos * 3) + prand.charAt(sPos * 4) + prand.charAt(sPos * 5));
    var incr = Math.ceil(pwd.length / 2);
    var modu = Math.pow(2, 31) - 1;
    if (mult < 2) {
        alert("Algorithm cannot find a suitable hash. Please choose a different password. \nPossible considerations are to choose a more complex or longer password.");
        return null;
    }
    var salt = Math.round(Math.random() * 1000000000) % 100000000;
    prand += salt;
    while (prand.length > 10) {
        prand = (parseInt(prand.substring(0, 10)) + parseInt(prand.substring(10, prand.length))).toString();
    }
    prand = (mult * prand + incr) % modu;
    var enc_chr = "";
    var enc_str = "";
    for (var i = 0; i < str.length; i++) {
        enc_chr = parseInt(str.charCodeAt(i) ^ Math.floor((prand / modu) * 255));
        if (enc_chr < 16) {
            enc_str += "0" + enc_chr.toString(16);
        } else enc_str += enc_chr.toString(16);
        prand = (mult * prand + incr) % modu;
    }
    salt = salt.toString(16);
    while (salt.length < 8) salt = "0" + salt;
    enc_str += salt;
    return enc_str;
}

2. 新增获取请求头方法：

async function getHmIuvt() {
    let url = `https://www.kuwo.cn/`;
    var result = await superagent.get(url);
    let cookie = result.header['set-cookie'][0] ?? '';
    //console.log(cookie);
    if (cookie) {
        cookie = cookie.replace(/;.*/ig, '');
    }
    let sp = cookie.split('=');
    let data = {
        key: sp[0] ?? '',
        token: sp[1] ?? ''
    };
    return data;
}

3. 完整代码示例（获取榜单列表）：

async function getBangMusicList(id, pn, rn) {
    let iuvt = await getHmIuvt();
    let secret = encrypt(iuvt.token, iuvt.key);
    let url = `http://www.kuwo.cn/api/www/bang/bang/musicList?bangId=${id}&pn=${pn}&rn=${rn}&httpsStatus=1&plat=web_www&reqId=`;
    var result = await get(url, {}, { Secret: secret, cookie: `${iuvt.key}=${iuvt.token}` });
    var obj = JSON.parse(result);
    return obj;
}

5.根据rid获取VIP/高音质音乐链接

rid:281472652 音乐Url，请打开以下链接↓↓↓↓

https://nmobi.kuwo.cn/mobi.s?f=kuwo&q=NI8S5evAnmGldi4g47EsqtfDbGsJckckbTQQd2LAgmDPITUWSd51OkjHRFj6xHPEQxNN6u+tD3K2e3HYhbE4U0pUYwqjd2kt

结果是有过期时间的哦~每打开一次会获取最新

format=flac
bitrate=2000
url=http://ll.sycdn.kuwo.cn/0f304a22b30673380bb9646d6737302d/64b512bc/resource/s1/77/85/2165046264.flac?bitrate$2000&format$flac&type$convert_url2
sig=9298802900736320721
rid=281472652
type=0

rid:228908 音乐Url，请打开以下链接↓↓↓↓

http://nmobi.kuwo.cn/mobi.s?f=kuwo&q=NI8S5evAnmGldi4g47EsqtfDbGsJckckbTQQd2LAgmDPITUWSd51OkjHRFj6xHPEQxNN6u+tD3KtKpZdbgOTWA==

结果是有过期时间的哦~每打开一次会获取最新

format=flac
bitrate=2000
url=http://sy.sycdn.kuwo.cn/188862a666b9fff315836ac0dc168441/64b51570/resource/s1/96/9/3483822377.flac?bitrate$2000&format$flac&type$convert_url2
sig=14962903177885333588
rid=260839262
type=1

注意：电脑端访问：F12 将浏览器切换为手机模式才可访问！！！！！；或者发送到手机端直接访问即可！

6.根据rid获取歌词

async function musicLrc(rid) {
    let url = `https://m.kuwo.cn/newh5/singles/songinfoandlrc?musicId=${rid}`;
    let result = await superagent.get(url);
    if (result?.text) {
        let data = JSON.parse(result.text);
        return data.data.lrclist;
    }
    return null;
}

rid:228908 歌词 musicLrc(228908)

[{"lineLyric":"晴天-周杰伦","time":"0.0"},{"lineLyric":"词：周杰伦","time":"1.88"},{"lineLyric":"曲：周杰伦","time":"3.76"},{"lineLyric":"编曲：周杰伦","time":"5.65"},{"lineLyric":"制作人：周杰伦","time":"7.53"},{"lineLyric":"合声：周杰伦","time":"9.42"},{"lineLyric":"合声编写：周杰伦","time":"11.3"},{"lineLyric":"吉他：蔡科俊Again","time":"13.18"},{"lineLyric":"贝斯：陈任佑","time":"15.07"},{"lineLyric":"鼓：陈柏州","time":"16.95"},{"lineLyric":"录音助理：刘勇志","time":"18.84"},{"lineLyric":"录音工程：杨瑞代Alfa Studio","time":"20.73"},{"lineLyric":"混音工程：杨大纬杨大纬录音工作室","time":"27.72"},{"lineLyric":"故事的小黄花","time":"29.22"},{"lineLyric":"从出生那年就飘着","time":"32.67"},{"lineLyric":"童年的荡秋千","time":"36.15"},{"lineLyric":"随记忆一直晃到现在","time":"39.6"},{"lineLyric":"Re So So Si Do Si La","time":"44.55"},{"lineLyric":"So La Si Si Si Si La Si La So","time":"45.72"},{"lineLyric":"吹着前奏望着天空","time":"49.74"},{"lineLyric":"我想起花瓣试着掉落","time":"53.1"},{"lineLyric":"为你翘课的那一天","time":"56.58"},{"lineLyric":"花落的那一天","time":"58.8"},{"lineLyric":"教室的那一间","time":"60.51"},{"lineLyric":"我怎么看不见","time":"62.16"},{"lineLyric":"消失的下雨天","time":"64.01"},{"lineLyric":"我好想再淋一遍","time":"65.76"},{"lineLyric":"没想到失去的勇气我还留着","time":"69.75"},{"lineLyric":"好想再问一遍","time":"75.81"},{"lineLyric":"你会等待还是离开","time":"77.88"},{"lineLyric":"刮风这天我试过握着你手","time":"84.75"},{"lineLyric":"但偏偏雨渐渐大到我看你不见","time":"90.51"},{"lineLyric":"还要多久我才能在你身边","time":"98.7"},{"lineLyric":"等到放晴的那天也许我会比较好一点","time":"105.39"},{"lineLyric":"从前从前有个人爱你很久","time":"112.740005"},{"lineLyric":"但偏偏风渐渐把距离吹得好远","time":"118.53"},{"lineLyric":"好不容易又能再多爱一天","time":"126.72"},{"lineLyric":"但故事的最后你好像还是说了拜拜","time":"133.38"},{"lineLyric":"为你翘课的那一天","time":"154.59"},{"lineLyric":"花落的那一天","time":"156.81"},{"lineLyric":"教室的那一间","time":"158.52"},{"lineLyric":"我怎么看不见","time":"160.32"},{"lineLyric":"消失的下雨天","time":"162.03"},{"lineLyric":"我好想再淋一遍","time":"163.83"},{"lineLyric":"没想到失去的勇气我还留着","time":"167.97"},{"lineLyric":"好想再问一遍","time":"172.14"},{"lineLyric":"你会等待还是离开","time":"175.98"},{"lineLyric":"刮风这天我试过握着你手","time":"182.82"},{"lineLyric":"但偏偏雨渐渐大到我看你不见","time":"188.37"},{"lineLyric":"还要多久我才能在你身边","time":"196.8"},{"lineLyric":"等到放晴的那天也许我会比较好一点","time":"203.45999"},{"lineLyric":"从前从前有个人爱你很久","time":"210.87"},{"lineLyric":"偏偏风渐渐把距离吹得好远","time":"216.99"},{"lineLyric":"好不容易又能再多爱一天","time":"224.79001"},{"lineLyric":"但故事的最后你好像还是说了拜拜","time":"231.45"},{"lineLyric":"刮风这天我试过握着你手","time":"238.95"},{"lineLyric":"但偏偏雨渐渐大到我看你不见","time":"241.77"},{"lineLyric":"还要多久我才能够在你身边","time":"245.43"},{"lineLyric":"等到放晴那天也许我会比较好一点","time":"248.94"},{"lineLyric":"从前从前有个人爱你很久","time":"252.69"},{"lineLyric":"但偏偏雨渐渐把距离吹得好远","time":"255.78"},{"lineLyric":"好不容易又能再多爱一天","time":"259.26"},{"lineLyric":"但故事的最后你好像还是说了拜","time":"262.74"}]

声明：仅用于学习交流，切勿用于其他用途~