网易云音乐是广大网友喜闻乐见的音乐平台，区别于别的音乐平台的最大特点，除了“它比我还懂我的音乐喜好”、“小清新的界面设计”就是它独有的评论区了——————各种故事汇，各种金句频出。我们可以透过歌曲的评论数来判断一个歌者的市场分量和歌曲的流行度。言归正传，如果我们想要简单爬取指定歌曲的评论内容来做词云或者其他相关数据分析，有没有容易上手的好方法呢？

> function a(a){var d,e,b="abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789",c="";for(d=0;a>d;d+=1)e=Math.random()*b.length,e=Math.floor(e),c+=b.charAt(e);return c}
< undefined
> a(16)
< "WugFoWvqBro7YomO"
> a(16)
< "TNIMHadoojG3O1yJ"

function b(a,b){var c=CryptoJS.enc.Utf8.parse(b),d=CryptoJS.enc.Utf8.parse("0102030405060708"),e=CryptoJS.enc.Utf8.parse(a),f=CryptoJS.AES.encrypt(e,c,{iv:d,mode:CryptoJS.mode.CBC});return f.toString()}

function e(a,b,d,e){var f={};return f.encText=c(a+e,b,d),f}

function c(a,b,c){var d,e;return setMaxDigits(131),d=new RSAKeyPair(b,"",c),e=encryptedString(d,a)}

function d(d,e,f,g){var h={},i=a(16);return h.encText=b(d,g),h.encText=b(h.encText,i),h.encSecKey=c(i,e,f),h}

# 参考地址：https://www.zhihu.com/question/36081767
# 完美兼容win10、python3.6，由于python3.6下pycrypto库已经停止维护，可以安装pyCryptodome库代替，pyCyrpto库的后续分支，有一个叫pyCryptodome的库，是前代的延伸版。
import sys
import codecs
import requests,json,os
import base64
import Crypto
from Crypto.Cipher import AESclass Spider():def __init__(self,idNum):#user-Agent字段直接从浏览器中复制过来即可，请求头中其他字段非必须项，也可以从浏览器中找到所有字段都放到Request Headersself.header = {'User-Agent':'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/49.0.2623.221 Safari/537.36 SE 2.X MetaSr 1.0','Referer': 'http://music.163.com/'}self.url = 'https://music.163.com/weapi/v1/resource/comments/R_SO_4_'+idNum+'?csrf_token='   #每一次的base_url只有歌曲id不同，构造url即可。def __get_jsons(self,url,page):# 获取两个参数music = WangYiYun()text = music.create_random_16()params = music.get_params(text,page)encSecKey = music.get_encSEcKey(text)fromdata = {'params' : params,'encSecKey' : encSecKey}jsons = requests.post(url, data=fromdata, headers=self.header)#print(jsons.raise_for_status())# 打印返回来的内容，是个json格式的#print(jsons.content)return jsons.textdef json2list(self,jsons):'''把json转成字典，并把他重要的信息获取出来存入列表'''# 可以用json.loads()把它转成字典#print(json.loads(jsons.text))users = json.loads(jsons)comments = []for user in users['comments']:# print(user['user']['nickname']+' : '+user['content']+'   点赞数：'+str(user['likedCount']))name = user['user']['nickname']content = user['content']# 点赞数likedCount = user['likedCount']#提取所需json中所需的字段构造字典user_dict = {'name': name, 'content': content, 'likedCount': likedCount}#将提取的字典信息追加到列表中
            comments.append(user_dict)return commentsdef run(self,idNum):self.page = 1while True:jsons = self.__get_jsons(self.url,self.page)comments = self.json2list(jsons)non_bmp_map = dict.fromkeys(range(0x10000, sys.maxunicode + 1), 0xfffd)##            print(str(comments[0]).translate(non_bmp_map))print('self.page = '+str(self.page)) #控制台打印正在爬取的页码数print(idNum) #打印正在爬取的歌曲id#在该脚本同级目录下生成“comments”文件夹dirName = u'{}'.format('comments')if not os.path.exists(dirName):os.makedirs(dirName)with open(".\comments\\"+idNum+".txt","a",encoding='utf-8') as f:  #结果写入txt文件
##                print(len(comments))for ii in range(len(comments)):f.write(str(comments[ii]).translate(non_bmp_map))f.write('\n')
##                    print(ii)
                f.close()# 当这一页的评论数少于20条时，证明已经获取完
##            self.write2sql(comments)if len(comments) < 100 :   #当limits设置为100时，默认每次服务器请求结果100条comments，当小于此数，意味爬到最后一页。print('评论已经获取完')breakself.page +=1# 找出post的两个参数params和encSecKey
class WangYiYun():def __init__(self):# 在网易云获取的三个参数
self.second_param = '010001'self.third_param = '00e0b509f6259df8642dbc35662901477df22677ec152b5ff68ace615bb7b725152b3ab17a876aea8a5aa76d2e417629ec4ee341f56135fccf695280104e0312ecbda92557c93870114af6c9d05c4f7f0c3685b7a46bee255932575cce10b424d813cfe4875d3e82047b97ddef52741d546b8e289dc6935b3ece0462db0a22b8e7'self.fourth_param = '0CoJUm6Qyw8W8jud'def create_random_16(self):'''获取随机十六个字母拼接成的字符串'''return (''.join(map(lambda xx: (hex(ord(xx))[2:]), str(os.urandom(16)))))[0:16]def aesEncrypt(self, text, key):# 偏移量iv = '0102030405060708'# 文本
pad = 16 - len(text) % 16text = text + pad * chr(pad)  #补齐文本长度
encryptor = AES.new(bytearray(key,'utf-8'), AES.MODE_CBC, bytearray(iv,'utf-8'))# encryptor = AES.new(key, 2, iv)
ciphertext = encryptor.encrypt(bytearray(text,'utf-8'))
##        print(bytearray(key,'utf-8'))ciphertext = base64.b64encode(ciphertext)return ciphertextdef get_params(self,text,page):'''获取网易云第一个参数'''# 第一个参数if page == 1:self.first_param = '{rid: "", offset: "0", total: "true", limit: "100", csrf_token: ""}'#rid: "R_SO_4_557581284"，经测试该值可以置空，不影响结果的执行。else:self.first_param = '{rid: "", offset:%s, total: "false", limit: "100", csrf_token: ""}'%str((page-1)*20)  #limit参数可以灵活设置，默认为20，设置为100，爬取效率可以提高
params = self.aesEncrypt(self.first_param, self.fourth_param).decode('utf-8')params = self.aesEncrypt(params, text)return paramsdef rsaEncrypt(self, pubKey, text, modulus):'''进行rsa加密'''text = text[::-1]rs = int(codecs.encode(text.encode('utf-8'), 'hex_codec'), 16) ** int(pubKey, 16) % int(modulus, 16)return format(rs, 'x').zfill(256)def get_encSEcKey(self,text):'''获取第二个参数'''pubKey = self.second_parammoudulus = self.third_paramencSecKey = self.rsaEncrypt(pubKey, text, moudulus)return encSecKeydef main():idPs = ['557581284','32019002']   #花粥《纸短情长》以及Zedd / Jon Bellion的《beautiful now》，可根据需要在网易云音乐查找歌曲ID后替换，列表元素越多，爬取的循环次数越多for jj in range(len(idPs)):idNum = idPs[jj]spider = Spider(idNum)  #根据Spider类实例化spider对象spider.run(idNum) #调用spider对象的run方法if __name__ == '__main__':main()