from .youtube import YoutubeIE
from ..compat import (
compat_etree_fromstring,
+ compat_str,
compat_urllib_parse_unquote,
compat_urlparse,
compat_xml_parse_error,
from .anvato import AnvatoIE
from .washingtonpost import WashingtonPostIE
from .wistia import WistiaIE
+from .mediaset import MediasetIE
class GenericIE(InfoExtractor):
},
'add_ie': [WashingtonPostIE.ie_key()],
},
+ {
+ # Mediaset embed
+ 'url': 'http://www.tgcom24.mediaset.it/politica/serracchiani-voglio-vivere-in-una-societa-aperta-reazioni-sproporzionate-_3071354-201702a.shtml',
+ 'info_dict': {
+ 'id': '720642',
+ 'ext': 'mp4',
+ 'title': 'Serracchiani: "Voglio vivere in una società aperta, con tutela del patto di fiducia"',
+ },
+ 'params': {
+ 'skip_download': True,
+ },
+ 'add_ie': [MediasetIE.ie_key()],
+ },
# {
# # TODO: find another test
# # http://schema.org/VideoObject
content_type = head_response.headers.get('Content-Type', '').lower()
m = re.match(r'^(?P<type>audio|video|application(?=/(?:ogg$|(?:vnd\.apple\.|x-)?mpegurl)))/(?P<format_id>[^;\s]+)', content_type)
if m:
- format_id = m.group('format_id')
+ format_id = compat_str(m.group('format_id'))
if format_id.endswith('mpegurl'):
formats = self._extract_m3u8_formats(url, video_id, 'mp4')
elif format_id == 'f4m':
formats = self._extract_f4m_formats(url, video_id)
else:
formats = [{
- 'format_id': m.group('format_id'),
+ 'format_id': format_id,
'url': url,
'vcodec': 'none' if m.group('type') == 'audio' else None
}]
return self.playlist_from_matches(
wapo_urls, video_id, video_title, ie=WashingtonPostIE.ie_key())
+ # Look for Mediaset embeds
+ mediaset_urls = MediasetIE._extract_urls(webpage)
+ if mediaset_urls:
+ return self.playlist_from_matches(
+ mediaset_urls, video_id, video_title, ie=MediasetIE.ie_key())
+
# Looking for http://schema.org/VideoObject
json_ld = self._search_json_ld(
webpage, video_id, default={}, expected_type='VideoObject')