Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/rtlnl.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     int_or_none,
   7     parse_duration,
   8 )
   9
  10
  11 class RtlNlIE(InfoExtractor):
  12     IE_NAME = 'rtl.nl'
  13     IE_DESC = 'rtl.nl and rtlxl.nl'
  14     _VALID_URL = r'''(?x)
  15         https?://(?:www\.)?
  16         (?:
  17             rtlxl\.nl/\#!/[^/]+/|
  18             rtl\.nl/system/videoplayer/(?:[^/]+/)+(?:video_)?embed\.html\b.+?\buuid=
  19         )
  20         (?P<id>[0-9a-f-]+)'''
  21
  22     _TESTS = [{
  23         'url': 'http://www.rtlxl.nl/#!/rtl-nieuws-132237/6e4203a6-0a5e-3596-8424-c599a59e0677',
  24         'md5': 'cc16baa36a6c169391f0764fa6b16654',
  25         'info_dict': {
  26             'id': '6e4203a6-0a5e-3596-8424-c599a59e0677',
  27             'ext': 'mp4',
  28             'title': 'RTL Nieuws - Laat',
  29             'description': 'md5:6b61f66510c8889923b11f2778c72dc5',
  30             'timestamp': 1408051800,
  31             'upload_date': '20140814',
  32             'duration': 576.880,
  33         },
  34     }, {
  35         'url': 'http://www.rtl.nl/system/videoplayer/derden/rtlnieuws/video_embed.html#uuid=84ae5571-ac25-4225-ae0c-ef8d9efb2aed/autoplay=false',
  36         'md5': 'dea7474214af1271d91ef332fb8be7ea',
  37         'info_dict': {
  38             'id': '84ae5571-ac25-4225-ae0c-ef8d9efb2aed',
  39             'ext': 'mp4',
  40             'timestamp': 1424039400,
  41             'title': 'RTL Nieuws - Nieuwe beelden Kopenhagen: chaos direct na aanslag',
  42             'thumbnail': 're:^https?://screenshots\.rtl\.nl/system/thumb/sz=[0-9]+x[0-9]+/uuid=84ae5571-ac25-4225-ae0c-ef8d9efb2aed$',
  43             'upload_date': '20150215',
  44             'description': 'Er zijn nieuwe beelden vrijgegeven die vlak na de aanslag in Kopenhagen zijn gemaakt. Op de video is goed te zien hoe omstanders zich bekommeren om één van de slachtoffers, terwijl de eerste agenten ter plaatse komen.',
  45         }
  46     }, {
  47         # empty synopsis and missing episodes (see https://github.com/rg3/youtube-dl/issues/6275)
  48         'url': 'http://www.rtl.nl/system/videoplayer/derden/rtlnieuws/video_embed.html#uuid=f536aac0-1dc3-4314-920e-3bd1c5b3811a/autoplay=false',
  49         'info_dict': {
  50             'id': 'f536aac0-1dc3-4314-920e-3bd1c5b3811a',
  51             'ext': 'mp4',
  52             'title': 'RTL Nieuws - Meer beelden van overval juwelier',
  53             'thumbnail': 're:^https?://screenshots\.rtl\.nl/system/thumb/sz=[0-9]+x[0-9]+/uuid=f536aac0-1dc3-4314-920e-3bd1c5b3811a$',
  54             'timestamp': 1437233400,
  55             'upload_date': '20150718',
  56             'duration': 30.474,
  57         },
  58         'params': {
  59             'skip_download': True,
  60         },
  61     }, {
  62         # encrypted m3u8 streams, georestricted
  63         'url': 'http://www.rtlxl.nl/#!/afl-2-257632/52a74543-c504-4cde-8aa8-ec66fe8d68a7',
  64         'only_matching': True,
  65     }, {
  66         'url': 'http://www.rtl.nl/system/videoplayer/derden/embed.html#!/uuid=bb0353b0-d6a4-1dad-90e9-18fe75b8d1f0',
  67         'only_matching': True,
  68     }]
  69
  70     def _real_extract(self, url):
  71         uuid = self._match_id(url)
  72         info = self._download_json(
  73             'http://www.rtl.nl/system/s4m/vfd/version=2/uuid=%s/fmt=adaptive/' % uuid,
  74             uuid)
  75
  76         material = info['material'][0]
  77         title = info['abstracts'][0]['name']
  78         subtitle = material.get('title')
  79         if subtitle:
  80             title += ' - %s' % subtitle
  81         description = material.get('synopsis')
  82
  83         meta = info.get('meta', {})
  84
  85         # m3u8 streams are encrypted and may not be handled properly by older ffmpeg/avconv.
  86         # To workaround this previously adaptive -> flash trick was used to obtain
  87         # unencrypted m3u8 streams (see https://github.com/rg3/youtube-dl/issues/4118)
  88         # and bypass georestrictions as well.
  89         # Currently, unencrypted m3u8 playlists are (intentionally?) invalid and therefore
  90         # unusable albeit can be fixed by simple string replacement (see
  91         # https://github.com/rg3/youtube-dl/pull/6337)
  92         # Since recent ffmpeg and avconv handle encrypted streams just fine encrypted
  93         # streams are used now.
  94         videopath = material['videopath']
  95         m3u8_url = meta.get('videohost', 'http://manifest.us.rtl.nl') + videopath
  96
  97         formats = self._extract_m3u8_formats(m3u8_url, uuid, ext='mp4')
  98
  99         video_urlpart = videopath.split('/adaptive/')[1][:-5]
 100         PG_URL_TEMPLATE = 'http://pg.us.rtl.nl/rtlxl/network/%s/progressive/%s.mp4'
 101
 102         formats.extend([
 103             {
 104                 'url': PG_URL_TEMPLATE % ('a2m', video_urlpart),
 105                 'format_id': 'pg-sd',
 106             },
 107             {
 108                 'url': PG_URL_TEMPLATE % ('a3m', video_urlpart),
 109                 'format_id': 'pg-hd',
 110                 'quality': 0,
 111             }
 112         ])
 113         self._sort_formats(formats)
 114
 115         thumbnails = []
 116
 117         for p in ('poster_base_url', '"thumb_base_url"'):
 118             if not meta.get(p):
 119                 continue
 120
 121             thumbnails.append({
 122                 'url': self._proto_relative_url(meta[p] + uuid),
 123                 'width': int_or_none(self._search_regex(
 124                     r'/sz=([0-9]+)', meta[p], 'thumbnail width', fatal=False)),
 125                 'height': int_or_none(self._search_regex(
 126                     r'/sz=[0-9]+x([0-9]+)',
 127                     meta[p], 'thumbnail height', fatal=False))
 128             })
 129
 130         return {
 131             'id': uuid,
 132             'title': title,
 133             'formats': formats,
 134             'timestamp': material['original_date'],
 135             'description': description,
 136             'duration': parse_duration(material.get('duration')),
 137             'thumbnails': thumbnails,
 138         }