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