_ 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     @staticmethod
  16     def _find_jwplayer_data(webpage):
  17         # TODO: Merge this with JWPlayer-related codes in generic.py
  18
  19         mobj = re.search(
  20             'jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)\.setup\((?P<options>[^)]+)\)',
  21             webpage)
  22         if mobj:
  23             return mobj.group('options')
  24
  25     def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
  26         jwplayer_data = self._parse_json(
  27             self._find_jwplayer_data(webpage), video_id)
  28         return self._parse_jwplayer_data(
  29             jwplayer_data, video_id, *args, **kwargs)
  30
  31     def _parse_jwplayer_data(self, jwplayer_data, video_id, require_title=True, m3u8_id=None, rtmp_params=None):
  32         # JWPlayer backward compatibility: flattened playlists
  33         # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
  34         if 'playlist' not in jwplayer_data:
  35             jwplayer_data = {'playlist': [jwplayer_data]}
  36
  37         video_data = jwplayer_data['playlist'][0]
  38
  39         # JWPlayer backward compatibility: flattened sources
  40         # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
  41         if 'sources' not in video_data:
  42             video_data['sources'] = [video_data]
  43
  44         formats = []
  45         for source in video_data['sources']:
  46             source_url = self._proto_relative_url(source['file'])
  47             source_type = source.get('type') or ''
  48             if source_type in ('application/vnd.apple.mpegurl', 'hls') or determine_ext(source_url) == 'm3u8':
  49                 formats.extend(self._extract_m3u8_formats(
  50                     source_url, video_id, 'mp4', 'm3u8_native', m3u8_id=m3u8_id, fatal=False))
  51             elif source_type.startswith('audio'):
  52                 formats.append({
  53                     'url': source_url,
  54                     'vcodec': 'none',
  55                 })
  56             else:
  57                 a_format = {
  58                     'url': source_url,
  59                     'width': int_or_none(source.get('width')),
  60                     'height': int_or_none(source.get('height')),
  61                 }
  62                 if source_url.startswith('rtmp'):
  63                     a_format['ext'] = 'flv',
  64
  65                     # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
  66                     # of jwplayer.flash.swf
  67                     rtmp_url_parts = re.split(
  68                         r'((?:mp4|mp3|flv):)', source_url, 1)
  69                     if len(rtmp_url_parts) == 3:
  70                         rtmp_url, prefix, play_path = rtmp_url_parts
  71                         a_format.update({
  72                             'url': rtmp_url,
  73                             'play_path': prefix + play_path,
  74                         })
  75                     if rtmp_params:
  76                         a_format.update(rtmp_params)
  77                 formats.append(a_format)
  78         self._sort_formats(formats)
  79
  80         subtitles = {}
  81         tracks = video_data.get('tracks')
  82         if tracks and isinstance(tracks, list):
  83             for track in tracks:
  84                 if track.get('file') and track.get('kind') == 'captions':
  85                     subtitles.setdefault(track.get('label') or 'en', []).append({
  86                         'url': self._proto_relative_url(track['file'])
  87                     })
  88
  89         return {
  90             'id': video_id,
  91             'title': video_data['title'] if require_title else video_data.get('title'),
  92             'description': video_data.get('description'),
  93             'thumbnail': self._proto_relative_url(video_data.get('image')),
  94             'timestamp': int_or_none(video_data.get('pubdate')),
  95             'duration': float_or_none(jwplayer_data.get('duration')),
  96             'subtitles': subtitles,
  97             'formats': formats,
  98         }
  99
 100
 101 class JWPlatformIE(JWPlatformBaseIE):
 102     _VALID_URL = r'(?:https?://content\.jwplatform\.com/(?:feeds|players|jw6)/|jwplatform:)(?P<id>[a-zA-Z0-9]{8})'
 103     _TEST = {
 104         'url': 'http://content.jwplatform.com/players/nPripu9l-ALJ3XQCI.js',
 105         'md5': 'fa8899fa601eb7c83a64e9d568bdf325',
 106         'info_dict': {
 107             'id': 'nPripu9l',
 108             'ext': 'mov',
 109             'title': 'Big Buck Bunny Trailer',
 110             'description': 'Big Buck Bunny is a short animated film by the Blender Institute. It is made using free and open source software.',
 111             'upload_date': '20081127',
 112             'timestamp': 1227796140,
 113         }
 114     }
 115
 116     @staticmethod
 117     def _extract_url(webpage):
 118         mobj = re.search(
 119             r'<script[^>]+?src=["\'](?P<url>(?:https?:)?//content.jwplatform.com/players/[a-zA-Z0-9]{8})',
 120             webpage)
 121         if mobj:
 122             return mobj.group('url')
 123
 124     def _real_extract(self, url):
 125         video_id = self._match_id(url)
 126         json_data = self._download_json('http://content.jwplatform.com/feeds/%s.json' % video_id, video_id)
 127         return self._parse_jwplayer_data(json_data, video_id)