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