download_music_from_163.py

#! /usr/bin/env python
# -*- coding: utf-8 -*-
# https://gist.github.com/tigerwang202/ba16e5a31e0da3c307cc#file-m163api-py

import re
import os
import sys
import md5
import json
import random
import requests

ENCODE = 'gbk' if sys.platform.startswith('win') else 'utf8'
CMD = 'curl -# -o "{name}" "{url}"'
SEARCH = 'http://music.163.com/api/search/get/web'
DETAIL = 'http://music.163.com/api/song/detail/?ids=[{}]'
LRC = 'http://music.163.com/api/song/media?id={}'
HEADERS = {'Referer': 'http://music.163.com'}
MP3 = 'http://m{}.music.126.net/{}/{}.mp3'


class Json:
    def __repr__(self):
        return json.dumps(self.json, indent=2)


def json2obj(json):
    if isinstance(json, dict):
        obj = Json()
        setattr(obj, 'json', json)
        for k, v in json.items():
            k = k.replace(' ', '_')
            setattr(obj, k, json2obj(v))
        return obj
    if isinstance(json, list):
        return map(json2obj, json)
    return json


def search(q, tp=1):
    '''
    >>> search('2375').result.songs[0].id
    365613
    '''
    return json2obj(requests.post(SEARCH, headers=HEADERS,
                                  data=dict(type=tp, s=q, offset=0,
                                            limit=30, total='true')
                                  ).json())


def searchAlbum(q):
    return search(q, tp=10)


def searchArtist(q):
    return search(q, tp=100)


def detail(sids):
    '''
    >>> detail([365613]).songs[0].name
    u'2375'
    '''
    return json2obj(requests.get(DETAIL.format(','.join(map(str, sids))),
                                 headers=HEADERS).json())


def url_2_sids(url):
    '''
    >>> url_2_sids('http://music.163.com/#/album?id=36197')
    ['365612', '365613', '365614', '365615', '365617']
    '''
    _id = url.split('=')[1]
    if 'song?id' in url:
        return [_id]
    else:
        r = requests.get(url.replace('/#/', '/'), headers=HEADERS)
        return re.findall(r'song\?id=(\d+)', r.content)


def encrypted_id(id):
    ''' from https://github.com/yanunon/NeteaseCloudMusic '''
    byte1 = bytearray('3go8&$8*3*3h0k(2)2')
    byte2 = bytearray(id)
    byte1_len = len(byte1)
    for i in xrange(len(byte2)):
        byte2[i] = byte2[i] ^ byte1[i % byte1_len]
    m = md5.new()
    m.update(byte2)
    result = m.digest().encode('base64')[:-1]
    result = result.replace('/', '_')
    result = result.replace('+', '-')
    return result


def download(sids, withlrc=False):
    '''
    >>> dfsId = str(detail([365613]).songs[0].bMusic.dfsId)
    >>> MP3.format(1, encrypted_id(dfsId), dfsId)
    'http://m1.music.126.net/n5usuzBfOxdzcGGnF_2hnQ==/2781764418296774.mp3'
    '''
    json = detail(sids)
    for i in json.songs:
        '''
        name = ''
        for c in i.name:
            try:
                name += c.encode(ENCODE)
            except:
                pass
        artist = ''
        for c in i.artists[0].name:
            try:
                artist += c.encode(ENCODE)
            except:
                pass
        '''
        name = i.name.encode(ENCODE)
        artist = i.artists[0].name.encode(ENCODE)
        # url = i.mp3Url

        mp3file = artist + ' - ' + name + '.mp3'
        try:
            dfsId = str(i.bMusic.dfsId)
        except:
            # 对于已下架歌曲只下载m4a格式，质量较差
            dfsId = str(i.audition.dfsId)
        url = MP3.format(random.randrange(1, 3), encrypted_id(dfsId), dfsId)
        print artist, name, url

        cc = os.system(CMD.format(name=mp3file, url=url))
        # assert not cc, 'Interrupted'
        if cc:
            print 'download error, skip...'
            if os.path.exists(mp3file):
                os.remove(mp3file)
            continue

        if withlrc:
            lrc = requests.get(LRC.format(i.id), headers=HEADERS).json()
            lrc = json2obj(lrc)
            if hasattr(lrc, 'lyric'):
                lrcfile = artist + ' - ' + name + '.lrc'
                open(lrcfile, 'w').write(lrc.lyric.encode('utf-8'))


def main():
    # print(encrypted_id('5794426278434182'))

    if len(sys.argv) != 2:
        print 'Usage: download_music_from_163 url'
        print 'url: song or album \'s page url.'
        exit(1)
    else:
        param = sys.argv[1]
        if param.startswith('http'):
            songids = url_2_sids(param)
            download(songids)
        elif param.endswith('.txt'):
            ''' batch download urls in txt file '''
            with open(param, 'r') as f:
                lines = f.readlines()
                for url in lines:
                    if url.startswith('http'):
                        songids = url_2_sids(url)
                        download(songids)
            f.close()
        else:
            print 'unknown param, quit!'


if __name__ == '__main__':
    main()