_ 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                     # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
  38                     # of jwplayer.flash.swf
  39                     rtmp_url, prefix, play_path = re.split(
  40                         r'((?:mp4|mp3|flv):)', source_url, 1)
  41                     a_format.update({
  42                         'url': rtmp_url,
  43                         'ext': 'flv',
  44                         'play_path': prefix + play_path,
  45                     })
  46                     if rtmp_params:
  47                         a_format.update(rtmp_params)
  48                 formats.append(a_format)
  49         self._sort_formats(formats)
  50
  51         subtitles = {}
  52         tracks = video_data.get('tracks')
  53         if tracks and isinstance(tracks, list):
  54             for track in tracks:
  55                 if track.get('file') and track.get('kind') == 'captions':
  56                     subtitles.setdefault(track.get('label') or 'en', []).append({
  57                         'url': self._proto_relative_url(track['file'])
  58                     })
  59
  60         return {
  61             'id': video_id,
  62             'title': video_data['title'] if require_title else video_data.get('title'),
  63             'description': video_data.get('description'),
  64             'thumbnail': self._proto_relative_url(video_data.get('image')),
  65             'timestamp': int_or_none(video_data.get('pubdate')),
  66             'duration': float_or_none(jwplayer_data.get('duration')),
  67             'subtitles': subtitles,
  68             'formats': formats,
  69         }
  70
  71
  72 class JWPlatformIE(JWPlatformBaseIE):
  73     _VALID_URL = r'(?:https?://content\.jwplatform\.com/(?:feeds|players|jw6)/|jwplatform:)(?P<id>[a-zA-Z0-9]{8})'
  74     _TEST = {
  75         'url': 'http://content.jwplatform.com/players/nPripu9l-ALJ3XQCI.js',
  76         'md5': 'fa8899fa601eb7c83a64e9d568bdf325',
  77         'info_dict': {
  78             'id': 'nPripu9l',
  79             'ext': 'mov',
  80             'title': 'Big Buck Bunny Trailer',
  81             'description': 'Big Buck Bunny is a short animated film by the Blender Institute. It is made using free and open source software.',
  82             'upload_date': '20081127',
  83             'timestamp': 1227796140,
  84         }
  85     }
  86
  87     @staticmethod
  88     def _extract_url(webpage):
  89         mobj = re.search(
  90             r'<script[^>]+?src=["\'](?P<url>(?:https?:)?//content.jwplatform.com/players/[a-zA-Z0-9]{8})',
  91             webpage)
  92         if mobj:
  93             return mobj.group('url')
  94
  95     def _real_extract(self, url):
  96         video_id = self._match_id(url)
  97         json_data = self._download_json('http://content.jwplatform.com/feeds/%s.json' % video_id, video_id)
  98         return self._parse_jwplayer_data(json_data, video_id)