You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

48 lines
1.9 KiB

  1. # coding: utf-8
  2. from __future__ import unicode_literals
  3. import re
  4. from .common import InfoExtractor
  5. class GazetaIE(InfoExtractor):
  6. _VALID_URL = r'(?P<url>https?://(?:www\.)?gazeta\.ru/(?:[^/]+/)?video/(?:main/)*(?:\d{4}/\d{2}/\d{2}/)?(?P<id>[A-Za-z0-9-_.]+)\.s?html)'
  7. _TESTS = [{
  8. 'url': 'http://www.gazeta.ru/video/main/zadaite_vopros_vladislavu_yurevichu.shtml',
  9. 'md5': 'd49c9bdc6e5a7888f27475dc215ee789',
  10. 'info_dict': {
  11. 'id': '205566',
  12. 'ext': 'mp4',
  13. 'title': '«70–80 процентов гражданских в Донецке на грани голода»',
  14. 'description': 'md5:38617526050bd17b234728e7f9620a71',
  15. 'thumbnail': r're:^https?://.*\.jpg',
  16. },
  17. 'skip': 'video not found',
  18. }, {
  19. 'url': 'http://www.gazeta.ru/lifestyle/video/2015/03/08/master-klass_krasivoi_byt._delaem_vesennii_makiyazh.shtml',
  20. 'only_matching': True,
  21. }, {
  22. 'url': 'http://www.gazeta.ru/video/main/main/2015/06/22/platit_ili_ne_platit_po_isku_yukosa.shtml',
  23. 'md5': '37f19f78355eb2f4256ee1688359f24c',
  24. 'info_dict': {
  25. 'id': '252048',
  26. 'ext': 'mp4',
  27. 'title': '"Если по иску ЮКОСа придется платить, это будет большой удар по бюджету"',
  28. },
  29. 'add_ie': ['EaglePlatform'],
  30. }]
  31. def _real_extract(self, url):
  32. mobj = re.match(self._VALID_URL, url)
  33. display_id = mobj.group('id')
  34. embed_url = '%s?p=embed' % mobj.group('url')
  35. embed_page = self._download_webpage(
  36. embed_url, display_id, 'Downloading embed page')
  37. video_id = self._search_regex(
  38. r'<div[^>]*?class="eagleplayer"[^>]*?data-id="([^"]+)"', embed_page, 'video id')
  39. return self.url_result(
  40. 'eagleplatform:gazeta.media.eagleplatform.com:%s' % video_id, 'EaglePlatform')