import re from .common import InfoExtractor class OnceIE(InfoExtractor): # XXX: Conventionally, base classes should end with BaseIE/InfoExtractor _VALID_URL = r'https?://.+?\.unicornmedia\.com/now/(?:ads/vmap/)?[^/]+/[^/]+/(?P<domain_id>[^/]+)/(?P<application_id>[^/]+)/(?:[^/]+/)?(?P<media_item_id>[^/]+)/content\.(?:once|m3u8|mp4)' ADAPTIVE_URL_TEMPLATE = 'http://once.unicornmedia.com/now/master/playlist/%s/%s/%s/content.m3u8' PROGRESSIVE_URL_TEMPLATE = 'http://once.unicornmedia.com/now/media/progressive/%s/%s/%s/%s/content.mp4' def _extract_once_formats(self, url, http_formats_preference=None): domain_id, application_id, media_item_id = re.match( OnceIE._VALID_URL, url).groups() formats = self._extract_m3u8_formats( self.ADAPTIVE_URL_TEMPLATE % ( domain_id, application_id, media_item_id), media_item_id, 'mp4', m3u8_id='hls', fatal=False) progressive_formats = [] for adaptive_format in formats: # Prevent advertisement from embedding into m3u8 playlist (see # https://github.com/ytdl-org/youtube-dl/issues/8893#issuecomment-199912684) adaptive_format['url'] = re.sub( r'\badsegmentlength=\d+', r'adsegmentlength=0', adaptive_format['url']) rendition_id = self._search_regex( r'/now/media/playlist/[^/]+/[^/]+/([^/]+)', adaptive_format['url'], 'redition id', default=None) if rendition_id: progressive_format = adaptive_format.copy() progressive_format.update({ 'url': self.PROGRESSIVE_URL_TEMPLATE % ( domain_id, application_id, rendition_id, media_item_id), 'format_id': adaptive_format['format_id'].replace( 'hls', 'http'), 'protocol': 'http', 'preference': http_formats_preference, }) progressive_formats.append(progressive_format) self._check_formats(progressive_formats, media_item_id) formats.extend(progressive_formats) return formats