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

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..utils import float_or_none
   7
   8
   9 class CanvasIE(InfoExtractor):
  10     IE_DESC = 'canvas.be and een.be'
  11     _VALID_URL = r'https?://(?:www\.)?(?P<site_id>canvas|een)\.be/(?:[^/]+/)*(?P<id>[^/?#&]+)'
  12     _TESTS = [{
  13         'url': 'http://www.canvas.be/video/de-afspraak/najaar-2015/de-afspraak-veilt-voor-de-warmste-week',
  14         'md5': 'ea838375a547ac787d4064d8c7860a6c',
  15         'info_dict': {
  16             'id': 'mz-ast-5e5f90b6-2d72-4c40-82c2-e134f884e93e',
  17             'display_id': 'de-afspraak-veilt-voor-de-warmste-week',
  18             'ext': 'mp4',
  19             'title': 'De afspraak veilt voor de Warmste Week',
  20             'description': 'md5:24cb860c320dc2be7358e0e5aa317ba6',
  21             'thumbnail': r're:^https?://.*\.jpg$',
  22             'duration': 49.02,
  23         }
  24     }, {
  25         # with subtitles
  26         'url': 'http://www.canvas.be/video/panorama/2016/pieter-0167',
  27         'info_dict': {
  28             'id': 'mz-ast-5240ff21-2d30-4101-bba6-92b5ec67c625',
  29             'display_id': 'pieter-0167',
  30             'ext': 'mp4',
  31             'title': 'Pieter 0167',
  32             'description': 'md5:943cd30f48a5d29ba02c3a104dc4ec4e',
  33             'thumbnail': r're:^https?://.*\.jpg$',
  34             'duration': 2553.08,
  35             'subtitles': {
  36                 'nl': [{
  37                     'ext': 'vtt',
  38                 }],
  39             },
  40         },
  41         'params': {
  42             'skip_download': True,
  43         }
  44     }, {
  45         'url': 'https://www.een.be/sorry-voor-alles/herbekijk-sorry-voor-alles',
  46         'info_dict': {
  47             'id': 'mz-ast-11a587f8-b921-4266-82e2-0bce3e80d07f',
  48             'display_id': 'herbekijk-sorry-voor-alles',
  49             'ext': 'mp4',
  50             'title': 'Herbekijk Sorry voor alles',
  51             'description': 'md5:8bb2805df8164e5eb95d6a7a29dc0dd3',
  52             'thumbnail': r're:^https?://.*\.jpg$',
  53             'duration': 3788.06,
  54         },
  55         'params': {
  56             'skip_download': True,
  57         }
  58     }, {
  59         'url': 'https://www.canvas.be/check-point/najaar-2016/de-politie-uw-vriend',
  60         'only_matching': True,
  61     }]
  62
  63     def _real_extract(self, url):
  64         mobj = re.match(self._VALID_URL, url)
  65         site_id, display_id = mobj.group('site_id'), mobj.group('id')
  66
  67         webpage = self._download_webpage(url, display_id)
  68
  69         title = (self._search_regex(
  70             r'<h1[^>]+class="video__body__header__title"[^>]*>(.+?)</h1>',
  71             webpage, 'title', default=None) or self._og_search_title(
  72             webpage)).strip()
  73
  74         video_id = self._html_search_regex(
  75             r'data-video=(["\'])(?P<id>(?:(?!\1).)+)\1', webpage, 'video id', group='id')
  76
  77         data = self._download_json(
  78             'https://mediazone.vrt.be/api/v1/%s/assets/%s'
  79             % (site_id, video_id), display_id)
  80
  81         formats = []
  82         for target in data['targetUrls']:
  83             format_url, format_type = target.get('url'), target.get('type')
  84             if not format_url or not format_type:
  85                 continue
  86             if format_type == 'HLS':
  87                 formats.extend(self._extract_m3u8_formats(
  88                     format_url, display_id, entry_protocol='m3u8_native',
  89                     ext='mp4', preference=0, fatal=False, m3u8_id=format_type))
  90             elif format_type == 'HDS':
  91                 formats.extend(self._extract_f4m_formats(
  92                     format_url, display_id, f4m_id=format_type, fatal=False))
  93             elif format_type == 'MPEG_DASH':
  94                 formats.extend(self._extract_mpd_formats(
  95                     format_url, display_id, mpd_id=format_type, fatal=False))
  96             else:
  97                 formats.append({
  98                     'format_id': format_type,
  99                     'url': format_url,
 100                 })
 101         self._sort_formats(formats)
 102
 103         subtitles = {}
 104         subtitle_urls = data.get('subtitleUrls')
 105         if isinstance(subtitle_urls, list):
 106             for subtitle in subtitle_urls:
 107                 subtitle_url = subtitle.get('url')
 108                 if subtitle_url and subtitle.get('type') == 'CLOSED':
 109                     subtitles.setdefault('nl', []).append({'url': subtitle_url})
 110
 111         return {
 112             'id': video_id,
 113             'display_id': display_id,
 114             'title': title,
 115             'description': self._og_search_description(webpage),
 116             'formats': formats,
 117             'duration': float_or_none(data.get('duration'), 1000),
 118             'thumbnail': data.get('posterImageUrl'),
 119             'subtitles': subtitles,
 120         }