youtube_dlc/extractor/streetvoice.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     int_or_none,
   7     parse_iso8601,
   8     str_or_none,
   9     strip_or_none,
  10     try_get,
  11     urljoin,
  12 )
  13
  14
  15 class StreetVoiceIE(InfoExtractor):
  16     _VALID_URL = r'https?://(?:.+?\.)?streetvoice\.com/[^/]+/songs/(?P<id>[0-9]+)'
  17     _TESTS = [{
  18         'url': 'https://streetvoice.com/skippylu/songs/123688/',
  19         'md5': '0eb535970629a5195685355f3ed60bfd',
  20         'info_dict': {
  21             'id': '123688',
  22             'ext': 'mp3',
  23             'title': '流浪',
  24             'description': 'md5:8eb0bfcc9dcd8aa82bd6efca66e3fea6',
  25             'thumbnail': r're:^https?://.*\.jpg',
  26             'duration': 270,
  27             'upload_date': '20100923',
  28             'uploader': 'Crispy脆樂團',
  29             'uploader_id': '627810',
  30             'uploader_url': 're:^https?://streetvoice.com/skippylu/',
  31             'timestamp': 1285261661,
  32             'view_count': int,
  33             'like_count': int,
  34             'comment_count': int,
  35             'repost_count': int,
  36             'track': '流浪',
  37             'track_id': '123688',
  38             'album': '2010',
  39         }
  40     }, {
  41         'url': 'http://tw.streetvoice.com/skippylu/songs/94440/',
  42         'only_matching': True,
  43     }]
  44
  45     def _real_extract(self, url):
  46         song_id = self._match_id(url)
  47         base_url = 'https://streetvoice.com/api/v4/song/%s/' % song_id
  48         song = self._download_json(base_url, song_id, query={
  49             'fields': 'album,comments_count,created_at,id,image,length,likes_count,name,nickname,plays_count,profile,share_count,synopsis,user,username',
  50         })
  51         title = song['name']
  52
  53         formats = []
  54         for suffix, format_id in [('hls/file', 'hls'), ('file', 'http'), ('file/original', 'original')]:
  55             f_url = (self._download_json(
  56                 base_url + suffix + '/', song_id,
  57                 'Downloading %s format URL' % format_id,
  58                 data=b'', fatal=False) or {}).get('file')
  59             if not f_url:
  60                 continue
  61             f = {
  62                 'ext': 'mp3',
  63                 'format_id': format_id,
  64                 'url': f_url,
  65                 'vcodec': 'none',
  66             }
  67             if format_id == 'hls':
  68                 f['protocol'] = 'm3u8_native'
  69             abr = self._search_regex(r'\.mp3\.(\d+)k', f_url, 'bitrate', default=None)
  70             if abr:
  71                 abr = int(abr)
  72                 f.update({
  73                     'abr': abr,
  74                     'tbr': abr,
  75                 })
  76             formats.append(f)
  77
  78         user = song.get('user') or {}
  79         username = user.get('username')
  80         get_count = lambda x: int_or_none(song.get(x + '_count'))
  81
  82         return {
  83             'id': song_id,
  84             'formats': formats,
  85             'title': title,
  86             'description': strip_or_none(song.get('synopsis')),
  87             'thumbnail': song.get('image'),
  88             'duration': int_or_none(song.get('length')),
  89             'timestamp': parse_iso8601(song.get('created_at')),
  90             'uploader': try_get(user, lambda x: x['profile']['nickname']),
  91             'uploader_id': str_or_none(user.get('id')),
  92             'uploader_url': urljoin(url, '/%s/' % username) if username else None,
  93             'view_count': get_count('plays'),
  94             'like_count': get_count('likes'),
  95             'comment_count': get_count('comments'),
  96             'repost_count': get_count('share'),
  97             'track': title,
  98             'track_id': song_id,
  99             'album': try_get(song, lambda x: x['album']['name']),
 100         }