youtube_dl/extractor/vlive.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     dict_get,
   9     ExtractorError,
  10     float_or_none,
  11     int_or_none,
  12     remove_start,
  13 )
  14 from ..compat import compat_urllib_parse_urlencode
  15
  16
  17 class VLiveIE(InfoExtractor):
  18     IE_NAME = 'vlive'
  19     _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/video/(?P<id>[0-9]+)'
  20     _TEST = {
  21         'url': 'http://www.vlive.tv/video/1326',
  22         'md5': 'cc7314812855ce56de70a06a27314983',
  23         'info_dict': {
  24             'id': '1326',
  25             'ext': 'mp4',
  26             'title': "[V LIVE] Girl's Day's Broadcast",
  27             'creator': "Girl's Day",
  28             'view_count': int,
  29         },
  30     }
  31
  32     def _real_extract(self, url):
  33         video_id = self._match_id(url)
  34
  35         webpage = self._download_webpage(
  36             'http://www.vlive.tv/video/%s' % video_id, video_id)
  37
  38         video_params = self._search_regex(
  39             r'\bvlive\.video\.init\(([^)]+)\)',
  40             webpage, 'video params')
  41         status, _, _, live_params, long_video_id, key = re.split(
  42             r'"\s*,\s*"', video_params)[2:8]
  43         status = remove_start(status, 'PRODUCT_')
  44
  45         if status == 'LIVE_ON_AIR' or status == 'BIG_EVENT_ON_AIR':
  46             live_params = self._parse_json('"%s"' % live_params, video_id)
  47             live_params = self._parse_json(live_params, video_id)
  48             return self._live(video_id, webpage, live_params)
  49         elif status == 'VOD_ON_AIR' or status == 'BIG_EVENT_INTRO':
  50             if long_video_id and key:
  51                 return self._replay(video_id, webpage, long_video_id, key)
  52             else:
  53                 status = 'COMING_SOON'
  54
  55         if status == 'LIVE_END':
  56             raise ExtractorError('Uploading for replay. Please wait...',
  57                                  expected=True)
  58         elif status == 'COMING_SOON':
  59             raise ExtractorError('Coming soon!', expected=True)
  60         elif status == 'CANCELED':
  61             raise ExtractorError('We are sorry, '
  62                                  'but the live broadcast has been canceled.',
  63                                  expected=True)
  64         else:
  65             raise ExtractorError('Unknown status %s' % status)
  66
  67     def _get_common_fields(self, webpage):
  68         title = self._og_search_title(webpage)
  69         creator = self._html_search_regex(
  70             r'<div[^>]+class="info_area"[^>]*>\s*<a\s+[^>]*>([^<]+)',
  71             webpage, 'creator', fatal=False)
  72         thumbnail = self._og_search_thumbnail(webpage)
  73         return {
  74             'title': title,
  75             'creator': creator,
  76             'thumbnail': thumbnail,
  77         }
  78
  79     def _live(self, video_id, webpage, live_params):
  80         formats = []
  81         for vid in live_params.get('resolutions', []):
  82             formats.extend(self._extract_m3u8_formats(
  83                 vid['cdnUrl'], video_id, 'mp4',
  84                 m3u8_id=vid.get('name'),
  85                 fatal=False, live=True))
  86         self._sort_formats(formats)
  87
  88         return dict(self._get_common_fields(webpage),
  89                     id=video_id,
  90                     formats=formats,
  91                     is_live=True)
  92
  93     def _replay(self, video_id, webpage, long_video_id, key):
  94         playinfo = self._download_json(
  95             'http://global.apis.naver.com/rmcnmv/rmcnmv/vod_play_videoInfo.json?%s'
  96             % compat_urllib_parse_urlencode({
  97                 'videoId': long_video_id,
  98                 'key': key,
  99                 'ptc': 'http',
 100                 'doct': 'json',  # document type (xml or json)
 101                 'cpt': 'vtt',  # captions type (vtt or ttml)
 102             }), video_id)
 103
 104         formats = [{
 105             'url': vid['source'],
 106             'format_id': vid.get('encodingOption', {}).get('name'),
 107             'abr': float_or_none(vid.get('bitrate', {}).get('audio')),
 108             'vbr': float_or_none(vid.get('bitrate', {}).get('video')),
 109             'width': int_or_none(vid.get('encodingOption', {}).get('width')),
 110             'height': int_or_none(vid.get('encodingOption', {}).get('height')),
 111             'filesize': int_or_none(vid.get('size')),
 112         } for vid in playinfo.get('videos', {}).get('list', []) if vid.get('source')]
 113         self._sort_formats(formats)
 114
 115         view_count = int_or_none(playinfo.get('meta', {}).get('count'))
 116
 117         subtitles = {}
 118         for caption in playinfo.get('captions', {}).get('list', []):
 119             lang = dict_get(caption, ('language', 'locale', 'country', 'label'))
 120             if lang and caption.get('source'):
 121                 subtitles[lang] = [{
 122                     'ext': 'vtt',
 123                     'url': caption['source']}]
 124
 125         return dict(self._get_common_fields(webpage),
 126                     id=video_id,
 127                     formats=formats,
 128                     view_count=view_count,
 129                     subtitles=subtitles)