_ 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 ..utils import (
   8     determine_ext,
   9     float_or_none,
  10     int_or_none,
  11 )
  12
  13
  14 class JWPlatformBaseIE(InfoExtractor):
  15     def _parse_jwplayer_data(self, jwplayer_data, video_id, require_title=True, m3u8_id=None, rtmp_params=None):
  16         video_data = jwplayer_data['playlist'][0]
  17
  18         formats = []
  19         for source in video_data['sources']:
  20             source_url = self._proto_relative_url(source['file'])
  21             source_type = source.get('type') or ''
  22             if source_type in ('application/vnd.apple.mpegurl', 'hls') or determine_ext(source_url) == 'm3u8':
  23                 formats.extend(self._extract_m3u8_formats(
  24                     source_url, video_id, 'mp4', 'm3u8_native', m3u8_id=m3u8_id, fatal=False))
  25             elif source_type.startswith('audio'):
  26                 formats.append({
  27                     'url': source_url,
  28                     'vcodec': 'none',
  29                 })
  30             else:
  31                 a_format = {
  32                     'url': source_url,
  33                     'width': int_or_none(source.get('width')),
  34                     'height': int_or_none(source.get('height')),
  35                 }
  36                 if source_url.startswith('rtmp'):
  37                     a_format['ext'] = 'flv',
  38
  39                     # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
  40                     # of jwplayer.flash.swf
  41                     rtmp_url_parts = re.split(
  42                         r'((?:mp4|mp3|flv):)', source_url, 1)
  43                     if len(rtmp_url_parts) == 3:
  44                         rtmp_url, prefix, play_path = rtmp_url_parts
  45                         a_format.update({
  46                             'url': rtmp_url,
  47                             'play_path': prefix + play_path,
  48                         })
  49                     if rtmp_params:
  50                         a_format.update(rtmp_params)
  51                 formats.append(a_format)
  52         self._sort_formats(formats)
  53
  54         subtitles = {}
  55         tracks = video_data.get('tracks')
  56         if tracks and isinstance(tracks, list):
  57             for track in tracks:
  58                 if track.get('file') and track.get('kind') == 'captions':
  59                     subtitles.setdefault(track.get('label') or 'en', []).append({
  60                         'url': self._proto_relative_url(track['file'])
  61                     })
  62
  63         return {
  64             'id': video_id,
  65             'title': video_data['title'] if require_title else video_data.get('title'),
  66             'description': video_data.get('description'),
  67             'thumbnail': self._proto_relative_url(video_data.get('image')),
  68             'timestamp': int_or_none(video_data.get('pubdate')),
  69             'duration': float_or_none(jwplayer_data.get('duration')),
  70             'subtitles': subtitles,
  71             'formats': formats,
  72         }
  73
  74
  75 class JWPlatformIE(JWPlatformBaseIE):
  76     _VALID_URL = r'(?:https?://content\.jwplatform\.com/(?:feeds|players|jw6)/|jwplatform:)(?P<id>[a-zA-Z0-9]{8})'
  77     _TEST = {
  78         'url': 'http://content.jwplatform.com/players/nPripu9l-ALJ3XQCI.js',
  79         'md5': 'fa8899fa601eb7c83a64e9d568bdf325',
  80         'info_dict': {
  81             'id': 'nPripu9l',
  82             'ext': 'mov',
  83             'title': 'Big Buck Bunny Trailer',
  84             'description': 'Big Buck Bunny is a short animated film by the Blender Institute. It is made using free and open source software.',
  85             'upload_date': '20081127',
  86             'timestamp': 1227796140,
  87         }
  88     }
  89
  90     @staticmethod
  91     def _extract_url(webpage):
  92         mobj = re.search(
  93             r'<script[^>]+?src=["\'](?P<url>(?:https?:)?//content.jwplatform.com/players/[a-zA-Z0-9]{8})',
  94             webpage)
  95         if mobj:
  96             return mobj.group('url')
  97
  98     def _real_extract(self, url):
  99         video_id = self._match_id(url)
 100         json_data = self._download_json('http://content.jwplatform.com/feeds/%s.json' % video_id, video_id)
 101         return self._parse_jwplayer_data(json_data, video_id)