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

   1 import re
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     ExtractorError,
   6 )
   7
   8
   9 class MDRIE(InfoExtractor):
  10     _VALID_URL = r'^(?P<domain>(?:https?://)?(?:www\.)?mdr\.de)/mediathek/(?:.*)/(?P<type>video|audio)(?P<video_id>[^/_]+)_.*'
  11
  12     # No tests, MDR regularily deletes its videos
  13
  14     def _real_extract(self, url):
  15         m = re.match(self._VALID_URL, url)
  16         video_id = m.group('video_id')
  17         domain = m.group('domain')
  18
  19         # determine title and media streams from webpage
  20         html = self._download_webpage(url, video_id)
  21
  22         title = self._html_search_regex(r'<h2>(.*?)</h2>', html, u'title')
  23         xmlurl = self._search_regex(
  24             r'(/mediathek/(?:.+)/(?:video|audio)[0-9]+-avCustom.xml)', html, u'XML URL')
  25
  26         doc = self._download_xml(domain + xmlurl, video_id)
  27         formats = []
  28         for a in doc.findall('./assets/asset'):
  29             url_el = a.find('.//progressiveDownloadUrl')
  30             if url_el is None:
  31                 continue
  32             abr = int(a.find('bitrateAudio').text) // 1000
  33             media_type = a.find('mediaType').text
  34             format = {
  35                 'abr': abr,
  36                 'filesize': int(a.find('fileSize').text),
  37                 'url': url_el.text,
  38             }
  39
  40             vbr_el = a.find('bitrateVideo')
  41             if vbr_el is None:
  42                 format.update({
  43                     'vcodec': 'none',
  44                     'format_id': u'%s-%d' % (media_type, abr),
  45                 })
  46             else:
  47                 vbr = int(vbr_el.text) // 1000
  48                 format.update({
  49                     'vbr': vbr,
  50                     'width': int(a.find('frameWidth').text),
  51                     'height': int(a.find('frameHeight').text),
  52                     'format_id': u'%s-%d' % (media_type, vbr),
  53                 })
  54             formats.append(format)
  55         formats.sort(key=lambda f: (f.get('vbr'), f['abr']))
  56         if not formats:
  57             raise ExtractorError(u'Could not find any valid formats')
  58
  59         return {
  60             'id': video_id,
  61             'title': title,
  62             'formats': formats,
  63         }