X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/684fbbb940adb6ea4043bc437e527888687a53da..39393b81acfaf4045fb7f20454a0226f0dc9142e:/youtube_dl/extractor/go.py diff --git a/youtube_dl/extractor/go.py b/youtube_dl/extractor/go.py index 206d89e..03cfba9 100644 --- a/youtube_dl/extractor/go.py +++ b/youtube_dl/extractor/go.py @@ -34,10 +34,23 @@ class GoIE(AdobePassIE): 'watchdisneyxd': { 'brand': '009', 'resource_id': 'DisneyXD', + }, + 'disneynow': { + 'brand': '011', + 'resource_id': 'Disney', } } - _VALID_URL = r'https?://(?:(?P%s)\.)?go\.com/(?:(?:[^/]+/)*(?Pvdka\w+)|(?:[^/]+/)*(?P[^/?#]+))'\ - % '|'.join(list(_SITE_INFO.keys()) + ['disneynow']) + _VALID_URL = r'''(?x) + https?:// + (?: + (?:(?P%s)\.)?go| + (?Pabc|freeform|disneynow) + )\.com/ + (?: + (?:[^/]+/)*(?P[Vv][Dd][Kk][Aa]\w+)| + (?:[^/]+/)*(?P[^/?\#]+) + ) + ''' % '|'.join(list(_SITE_INFO.keys())) _TESTS = [{ 'url': 'http://abc.go.com/shows/designated-survivor/video/most-recent/VDKA3807643', 'info_dict': { @@ -50,6 +63,7 @@ class GoIE(AdobePassIE): # m3u8 download 'skip_download': True, }, + 'skip': 'This content is no longer available.', }, { 'url': 'http://watchdisneyxd.go.com/doraemon', 'info_dict': { @@ -57,6 +71,34 @@ class GoIE(AdobePassIE): 'id': 'SH55574025', }, 'playlist_mincount': 51, + }, { + 'url': 'http://freeform.go.com/shows/shadowhunters/episodes/season-2/1-this-guilty-blood', + 'info_dict': { + 'id': 'VDKA3609139', + 'ext': 'mp4', + 'title': 'This Guilty Blood', + 'description': 'md5:f18e79ad1c613798d95fdabfe96cd292', + 'age_limit': 14, + }, + 'params': { + 'geo_bypass_ip_block': '3.244.239.0/24', + # m3u8 download + 'skip_download': True, + }, + }, { + 'url': 'https://abc.com/shows/the-rookie/episode-guide/season-02/03-the-bet', + 'info_dict': { + 'id': 'VDKA13435179', + 'ext': 'mp4', + 'title': 'The Bet', + 'description': 'md5:c66de8ba2e92c6c5c113c3ade84ab404', + 'age_limit': 14, + }, + 'params': { + 'geo_bypass_ip_block': '3.244.239.0/24', + # m3u8 download + 'skip_download': True, + }, }, { 'url': 'http://abc.go.com/shows/the-catch/episode-guide/season-01/10-the-wedding', 'only_matching': True, @@ -71,6 +113,9 @@ class GoIE(AdobePassIE): # brand 008 'url': 'http://disneynow.go.com/shows/minnies-bow-toons/video/happy-campers/vdka4872013', 'only_matching': True, + }, { + 'url': 'https://disneynow.com/shows/minnies-bow-toons/video/happy-campers/vdka4872013', + 'only_matching': True, }] def _extract_videos(self, brand, video_id='-1', show_id='-1'): @@ -80,16 +125,21 @@ class GoIE(AdobePassIE): display_id)['video'] def _real_extract(self, url): - sub_domain, video_id, display_id = re.match(self._VALID_URL, url).groups() + mobj = re.match(self._VALID_URL, url) + sub_domain = mobj.group('sub_domain') or mobj.group('sub_domain_2') + video_id, display_id = mobj.group('id', 'display_id') site_info = self._SITE_INFO.get(sub_domain, {}) brand = site_info.get('brand') if not video_id or not site_info: webpage = self._download_webpage(url, display_id or video_id) video_id = self._search_regex( - # There may be inner quotes, e.g. data-video-id="'VDKA3609139'" - # from http://freeform.go.com/shows/shadowhunters/episodes/season-2/1-this-guilty-blood - r'data-video-id=["\']*(VDKA\w+)', webpage, 'video id', - default=None) + ( + # There may be inner quotes, e.g. data-video-id="'VDKA3609139'" + # from http://freeform.go.com/shows/shadowhunters/episodes/season-2/1-this-guilty-blood + r'data-video-id=["\']*(VDKA\w+)', + # https://abc.com/shows/the-rookie/episode-guide/season-02/03-the-bet + r'\b(?:video)?id["\']\s*:\s*["\'](VDKA\w+)' + ), webpage, 'video id', default=video_id) if not site_info: brand = self._search_regex( (r'data-brand=\s*["\']\s*(\d+)',