git.bitcoin.ninja Git - youtube-dl/blob - youtube_dl/extractor/jwplatform.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import compat_urlparse
   8 from ..utils import (
   9     determine_ext,
  10     float_or_none,
  11     int_or_none,
  12     js_to_json,
  13     mimetype2ext,
  14 )
  15
  16
  17 class JWPlatformBaseIE(InfoExtractor):
  18     @staticmethod
  19     def _find_jwplayer_data(webpage):
  20         # TODO: Merge this with JWPlayer-related codes in generic.py
  21
  22         mobj = re.search(
  23             r'jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)\.setup\s*\((?P<options>[^)]+)\)',
  24             webpage)
  25         if mobj:
  26             return mobj.group('options')
  27
  28     def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
  29         jwplayer_data = self._parse_json(
  30             self._find_jwplayer_data(webpage), video_id,
  31             transform_source=js_to_json)
  32         return self._parse_jwplayer_data(
  33             jwplayer_data, video_id, *args, **kwargs)
  34
  35     def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True, m3u8_id=None, rtmp_params=None, base_url=None):
  36         # JWPlayer backward compatibility: flattened playlists
  37         # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
  38         if 'playlist' not in jwplayer_data:
  39             jwplayer_data = {'playlist': [jwplayer_data]}
  40
  41         entries = []
  42         for video_data in jwplayer_data['playlist']:
  43             # JWPlayer backward compatibility: flattened sources
  44             # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
  45             if 'sources' not in video_data:
  46                 video_data['sources'] = [video_data]
  47
  48             this_video_id = video_id or video_data['mediaid']
  49
  50             formats = []
  51             for source in video_data['sources']:
  52                 source_url = self._proto_relative_url(source['file'])
  53                 if base_url:
  54                     source_url = compat_urlparse.urljoin(base_url, source_url)
  55                 source_type = source.get('type') or ''
  56                 ext = mimetype2ext(source_type) or determine_ext(source_url)
  57                 if source_type == 'hls' or ext == 'm3u8':
  58                     formats.extend(self._extract_m3u8_formats(
  59                         source_url, this_video_id, 'mp4', 'm3u8_native', m3u8_id=m3u8_id, fatal=False))
  60                 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
  61                 elif source_type.startswith('audio') or ext in ('oga', 'aac', 'mp3', 'mpeg', 'vorbis'):
  62                     formats.append({
  63                         'url': source_url,
  64                         'vcodec': 'none',
  65                         'ext': ext,
  66                     })
  67                 else:
  68                     height = int_or_none(source.get('height'))
  69                     if height is None:
  70                         # Often no height is provided but there is a label in
  71                         # format like 1080p.
  72                         height = int_or_none(self._search_regex(
  73                             r'^(\d{3,})[pP]$', source.get('label') or '',
  74                             'height', default=None))
  75                     a_format = {
  76                         'url': source_url,
  77                         'width': int_or_none(source.get('width')),
  78                         'height': height,
  79                         'ext': ext,
  80                     }
  81                     if source_url.startswith('rtmp'):
  82                         a_format['ext'] = 'flv'
  83
  84                         # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
  85                         # of jwplayer.flash.swf
  86                         rtmp_url_parts = re.split(
  87                             r'((?:mp4|mp3|flv):)', source_url, 1)
  88                         if len(rtmp_url_parts) == 3:
  89                             rtmp_url, prefix, play_path = rtmp_url_parts
  90                             a_format.update({
  91                                 'url': rtmp_url,
  92                                 'play_path': prefix + play_path,
  93                             })
  94                         if rtmp_params:
  95                             a_format.update(rtmp_params)
  96                     formats.append(a_format)
  97             self._sort_formats(formats)
  98
  99             subtitles = {}
 100             tracks = video_data.get('tracks')
 101             if tracks and isinstance(tracks, list):
 102                 for track in tracks:
 103                     if track.get('file') and track.get('kind') == 'captions':
 104                         subtitles.setdefault(track.get('label') or 'en', []).append({
 105                             'url': self._proto_relative_url(track['file'])
 106                         })
 107
 108             entries.append({
 109                 'id': this_video_id,
 110                 'title': video_data['title'] if require_title else video_data.get('title'),
 111                 'description': video_data.get('description'),
 112                 'thumbnail': self._proto_relative_url(video_data.get('image')),
 113                 'timestamp': int_or_none(video_data.get('pubdate')),
 114                 'duration': float_or_none(jwplayer_data.get('duration')),
 115                 'subtitles': subtitles,
 116                 'formats': formats,
 117             })
 118         if len(entries) == 1:
 119             return entries[0]
 120         else:
 121             return self.playlist_result(entries)
 122
 123
 124 class JWPlatformIE(JWPlatformBaseIE):
 125     _VALID_URL = r'(?:https?://content\.jwplatform\.com/(?:feeds|players|jw6)/|jwplatform:)(?P<id>[a-zA-Z0-9]{8})'
 126     _TEST = {
 127         'url': 'http://content.jwplatform.com/players/nPripu9l-ALJ3XQCI.js',
 128         'md5': 'fa8899fa601eb7c83a64e9d568bdf325',
 129         'info_dict': {
 130             'id': 'nPripu9l',
 131             'ext': 'mov',
 132             'title': 'Big Buck Bunny Trailer',
 133             'description': 'Big Buck Bunny is a short animated film by the Blender Institute. It is made using free and open source software.',
 134             'upload_date': '20081127',
 135             'timestamp': 1227796140,
 136         }
 137     }
 138
 139     @staticmethod
 140     def _extract_url(webpage):
 141         mobj = re.search(
 142             r'<script[^>]+?src=["\'](?P<url>(?:https?:)?//content.jwplatform.com/players/[a-zA-Z0-9]{8})',
 143             webpage)
 144         if mobj:
 145             return mobj.group('url')
 146
 147     def _real_extract(self, url):
 148         video_id = self._match_id(url)
 149         json_data = self._download_json('http://content.jwplatform.com/feeds/%s.json' % video_id, video_id)
 150         return self._parse_jwplayer_data(json_data, video_id)