Raphaël G. Git Repositories - youtubedl/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     urljoin,
  15 )
  16
  17
  18 class JWPlatformBaseIE(InfoExtractor):
  19     @staticmethod
  20     def _find_jwplayer_data(webpage):
  21         # TODO: Merge this with JWPlayer-related codes in generic.py
  22
  23         mobj = re.search(
  24             r'jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)\.setup\s*\((?P<options>[^)]+)\)',
  25             webpage)
  26         if mobj:
  27             return mobj.group('options')
  28
  29     def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
  30         jwplayer_data = self._parse_json(
  31             self._find_jwplayer_data(webpage), video_id,
  32             transform_source=js_to_json)
  33         return self._parse_jwplayer_data(
  34             jwplayer_data, video_id, *args, **kwargs)
  35
  36     def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
  37                              m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
  38         # JWPlayer backward compatibility: flattened playlists
  39         # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
  40         if 'playlist' not in jwplayer_data:
  41             jwplayer_data = {'playlist': [jwplayer_data]}
  42
  43         entries = []
  44
  45         # JWPlayer backward compatibility: single playlist item
  46         # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
  47         if not isinstance(jwplayer_data['playlist'], list):
  48             jwplayer_data['playlist'] = [jwplayer_data['playlist']]
  49
  50         for video_data in jwplayer_data['playlist']:
  51             # JWPlayer backward compatibility: flattened sources
  52             # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
  53             if 'sources' not in video_data:
  54                 video_data['sources'] = [video_data]
  55
  56             this_video_id = video_id or video_data['mediaid']
  57
  58             formats = []
  59             for source in video_data['sources']:
  60                 source_url = self._proto_relative_url(source['file'])
  61                 if base_url:
  62                     source_url = compat_urlparse.urljoin(base_url, source_url)
  63                 source_type = source.get('type') or ''
  64                 ext = mimetype2ext(source_type) or determine_ext(source_url)
  65                 if source_type == 'hls' or ext == 'm3u8':
  66                     formats.extend(self._extract_m3u8_formats(
  67                         source_url, this_video_id, 'mp4', 'm3u8_native', m3u8_id=m3u8_id, fatal=False))
  68                 elif ext == 'mpd':
  69                     formats.extend(self._extract_mpd_formats(
  70                         source_url, this_video_id, mpd_id=mpd_id, fatal=False))
  71                 # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
  72                 elif source_type.startswith('audio') or ext in ('oga', 'aac', 'mp3', 'mpeg', 'vorbis'):
  73                     formats.append({
  74                         'url': source_url,
  75                         'vcodec': 'none',
  76                         'ext': ext,
  77                     })
  78                 else:
  79                     height = int_or_none(source.get('height'))
  80                     if height is None:
  81                         # Often no height is provided but there is a label in
  82                         # format like 1080p.
  83                         height = int_or_none(self._search_regex(
  84                             r'^(\d{3,})[pP]$', source.get('label') or '',
  85                             'height', default=None))
  86                     a_format = {
  87                         'url': source_url,
  88                         'width': int_or_none(source.get('width')),
  89                         'height': height,
  90                         'ext': ext,
  91                     }
  92                     if source_url.startswith('rtmp'):
  93                         a_format['ext'] = 'flv'
  94
  95                         # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
  96                         # of jwplayer.flash.swf
  97                         rtmp_url_parts = re.split(
  98                             r'((?:mp4|mp3|flv):)', source_url, 1)
  99                         if len(rtmp_url_parts) == 3:
 100                             rtmp_url, prefix, play_path = rtmp_url_parts
 101                             a_format.update({
 102                                 'url': rtmp_url,
 103                                 'play_path': prefix + play_path,
 104                             })
 105                         if rtmp_params:
 106                             a_format.update(rtmp_params)
 107                     formats.append(a_format)
 108             self._sort_formats(formats)
 109
 110             subtitles = {}
 111             tracks = video_data.get('tracks')
 112             if tracks and isinstance(tracks, list):
 113                 for track in tracks:
 114                     if track.get('kind') != 'captions':
 115                         continue
 116                     track_url = urljoin(base_url, track.get('file'))
 117                     if not track_url:
 118                         continue
 119                     subtitles.setdefault(track.get('label') or 'en', []).append({
 120                         'url': self._proto_relative_url(track_url)
 121                     })
 122
 123             entries.append({
 124                 'id': this_video_id,
 125                 'title': video_data['title'] if require_title else video_data.get('title'),
 126                 'description': video_data.get('description'),
 127                 'thumbnail': self._proto_relative_url(video_data.get('image')),
 128                 'timestamp': int_or_none(video_data.get('pubdate')),
 129                 'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')),
 130                 'subtitles': subtitles,
 131                 'formats': formats,
 132             })
 133         if len(entries) == 1:
 134             return entries[0]
 135         else:
 136             return self.playlist_result(entries)
 137
 138
 139 class JWPlatformIE(JWPlatformBaseIE):
 140     _VALID_URL = r'(?:https?://content\.jwplatform\.com/(?:feeds|players|jw6)/|jwplatform:)(?P<id>[a-zA-Z0-9]{8})'
 141     _TEST = {
 142         'url': 'http://content.jwplatform.com/players/nPripu9l-ALJ3XQCI.js',
 143         'md5': 'fa8899fa601eb7c83a64e9d568bdf325',
 144         'info_dict': {
 145             'id': 'nPripu9l',
 146             'ext': 'mov',
 147             'title': 'Big Buck Bunny Trailer',
 148             'description': 'Big Buck Bunny is a short animated film by the Blender Institute. It is made using free and open source software.',
 149             'upload_date': '20081127',
 150             'timestamp': 1227796140,
 151         }
 152     }
 153
 154     @staticmethod
 155     def _extract_url(webpage):
 156         mobj = re.search(
 157             r'<script[^>]+?src=["\'](?P<url>(?:https?:)?//content.jwplatform.com/players/[a-zA-Z0-9]{8})',
 158             webpage)
 159         if mobj:
 160             return mobj.group('url')
 161
 162     def _real_extract(self, url):
 163         video_id = self._match_id(url)
 164         json_data = self._download_json('http://content.jwplatform.com/feeds/%s.json' % video_id, video_id)
 165         return self._parse_jwplayer_data(json_data, video_id)