X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/fe979149c83b5a935f7d28baf75848a9137316fd..ab926e5dbeca04924d1628992210084e67368d25:/youtube_dl/extractor/twitter.py diff --git a/youtube_dl/extractor/twitter.py b/youtube_dl/extractor/twitter.py index b738429..41d0b6b 100644 --- a/youtube_dl/extractor/twitter.py +++ b/youtube_dl/extractor/twitter.py @@ -4,25 +4,46 @@ from __future__ import unicode_literals import re from .common import InfoExtractor +from ..compat import compat_urlparse from ..utils import ( determine_ext, + dict_get, + ExtractorError, float_or_none, - xpath_text, - remove_end, int_or_none, - ExtractorError, + remove_end, + try_get, + xpath_text, ) +from .periscope import PeriscopeIE + class TwitterBaseIE(InfoExtractor): - def _get_vmap_video_url(self, vmap_url, video_id): + def _extract_formats_from_vmap_url(self, vmap_url, video_id): vmap_data = self._download_xml(vmap_url, video_id) - return xpath_text(vmap_data, './/MediaFile').strip() + video_url = xpath_text(vmap_data, './/MediaFile').strip() + if determine_ext(video_url) == 'm3u8': + return self._extract_m3u8_formats( + video_url, video_id, ext='mp4', m3u8_id='hls', + entry_protocol='m3u8_native') + return [{ + 'url': video_url, + }] + + @staticmethod + def _search_dimensions_in_video_url(a_format, video_url): + m = re.search(r'/(?P\d+)x(?P\d+)/', video_url) + if m: + a_format.update({ + 'width': int(m.group('width')), + 'height': int(m.group('height')), + }) class TwitterCardIE(TwitterBaseIE): IE_NAME = 'twitter:card' - _VALID_URL = r'https?://(?:www\.)?twitter\.com/i/(?:cards/tfw/v1|videos/tweet)/(?P\d+)' + _VALID_URL = r'https?://(?:www\.)?twitter\.com/i/(?Pcards/tfw/v1|videos(?:/tweet)?)/(?P\d+)' _TESTS = [ { 'url': 'https://twitter.com/i/cards/tfw/v1/560070183650213889', @@ -30,10 +51,10 @@ class TwitterCardIE(TwitterBaseIE): 'info_dict': { 'id': '560070183650213889', 'ext': 'mp4', - 'title': 'Twitter Card', - 'thumbnail': 're:^https?://.*\.jpg$', + 'title': 'Twitter web player', + 'thumbnail': r're:^https?://.*\.jpg$', 'duration': 30.033, - } + }, }, { 'url': 'https://twitter.com/i/cards/tfw/v1/623160978427936768', @@ -41,19 +62,18 @@ class TwitterCardIE(TwitterBaseIE): 'info_dict': { 'id': '623160978427936768', 'ext': 'mp4', - 'title': 'Twitter Card', - 'thumbnail': 're:^https?://.*\.jpg', - 'duration': 80.155, + 'title': 'Twitter web player', + 'thumbnail': r're:^https?://.*$', }, }, { 'url': 'https://twitter.com/i/cards/tfw/v1/654001591733886977', - 'md5': 'd4724ffe6d2437886d004fa5de1043b3', + 'md5': 'b6d9683dd3f48e340ded81c0e917ad46', 'info_dict': { 'id': 'dq4Oj5quskI', 'ext': 'mp4', 'title': 'Ubuntu 11.10 Overview', - 'description': 'Take a quick peek at what\'s new and improved in Ubuntu 11.10.\n\nOnce installed take a look at 10 Things to Do After Installing: http://www.omgubuntu.co.uk/2011/10/10...', + 'description': 'md5:a831e97fa384863d6e26ce48d1c43376', 'upload_date': '20111013', 'uploader': 'OMG! Ubuntu!', 'uploader_id': 'omgubuntu', @@ -62,7 +82,7 @@ class TwitterCardIE(TwitterBaseIE): }, { 'url': 'https://twitter.com/i/cards/tfw/v1/665289828897005568', - 'md5': 'ab2745d0b0ce53319a534fccaa986439', + 'md5': '6dabeaca9e68cbb71c99c322a4b42a11', 'info_dict': { 'id': 'iBb2x00UVlv', 'ext': 'mp4', @@ -70,106 +90,185 @@ class TwitterCardIE(TwitterBaseIE): 'uploader_id': '1189339351084113920', 'uploader': 'ArsenalTerje', 'title': 'Vine by ArsenalTerje', + 'timestamp': 1447451307, }, 'add_ie': ['Vine'], }, { 'url': 'https://twitter.com/i/videos/tweet/705235433198714880', - 'md5': '3846d0a07109b5ab622425449b59049d', + 'md5': '884812a2adc8aaf6fe52b15ccbfa3b88', 'info_dict': { 'id': '705235433198714880', 'ext': 'mp4', 'title': 'Twitter web player', - 'thumbnail': 're:^https?://.*\.jpg', + 'thumbnail': r're:^https?://.*', }, + }, { + 'url': 'https://twitter.com/i/videos/752274308186120192', + 'only_matching': True, }, ] + _API_BASE = 'https://api.twitter.com/1.1' + + def _parse_media_info(self, media_info, video_id): + formats = [] + for media_variant in media_info.get('variants', []): + media_url = media_variant['url'] + if media_url.endswith('.m3u8'): + formats.extend(self._extract_m3u8_formats(media_url, video_id, ext='mp4', m3u8_id='hls')) + elif media_url.endswith('.mpd'): + formats.extend(self._extract_mpd_formats(media_url, video_id, mpd_id='dash')) + else: + tbr = int_or_none(dict_get(media_variant, ('bitRate', 'bitrate')), scale=1000) + a_format = { + 'url': media_url, + 'format_id': 'http-%d' % tbr if tbr else 'http', + 'tbr': tbr, + } + # Reported bitRate may be zero + if not a_format['tbr']: + del a_format['tbr'] + + self._search_dimensions_in_video_url(a_format, media_url) + + formats.append(a_format) + return formats + + def _extract_mobile_formats(self, username, video_id): + webpage = self._download_webpage( + 'https://mobile.twitter.com/%s/status/%s' % (username, video_id), + video_id, 'Downloading mobile webpage', + headers={ + # A recent mobile UA is necessary for `gt` cookie + 'User-Agent': 'Mozilla/5.0 (Android 6.0.1; Mobile; rv:54.0) Gecko/54.0 Firefox/54.0', + }) + main_script_url = self._html_search_regex( + r']+src="([^"]+main\.[^"]+)"', webpage, 'main script URL') + main_script = self._download_webpage( + main_script_url, video_id, 'Downloading main script') + bearer_token = self._search_regex( + r'BEARER_TOKEN\s*:\s*"([^"]+)"', + main_script, 'bearer token') + # https://developer.twitter.com/en/docs/tweets/post-and-engage/api-reference/get-statuses-show-id + api_data = self._download_json( + '%s/statuses/show/%s.json' % (self._API_BASE, video_id), + video_id, 'Downloading API data', + headers={ + 'Authorization': 'Bearer ' + bearer_token, + }) + media_info = try_get(api_data, lambda o: o['extended_entities']['media'][0]['video_info']) or {} + return self._parse_media_info(media_info, video_id) + def _real_extract(self, url): - video_id = self._match_id(url) + path, video_id = re.search(self._VALID_URL, url).groups() config = None formats = [] duration = None - webpage = self._download_webpage(url, video_id) + urls = [url] + if path.startswith('cards/'): + urls.append('https://twitter.com/i/videos/' + video_id) - iframe_url = self._html_search_regex( - r']+src="((?:https?:)?//(?:www.youtube.com/embed/[^"]+|(?:www\.)?vine\.co/v/\w+/card))"', - webpage, 'video iframe', default=None) - if iframe_url: - return self.url_result(iframe_url) + for u in urls: + webpage = self._download_webpage( + u, video_id, headers={'Referer': 'https://twitter.com/'}) - config = self._parse_json(self._html_search_regex( - r'data-(?:player-)?config="([^"]+)"', webpage, 'data player config'), - video_id) + iframe_url = self._html_search_regex( + r']+src="((?:https?:)?//(?:www\.youtube\.com/embed/[^"]+|(?:www\.)?vine\.co/v/\w+/card))"', + webpage, 'video iframe', default=None) + if iframe_url: + return self.url_result(iframe_url) - if config.get('source_type') == 'vine': - return self.url_result(config['player_url'], 'Vine') + config = self._parse_json(self._html_search_regex( + r'data-(?:player-)?config="([^"]+)"', webpage, + 'data player config', default='{}'), + video_id) - def _search_dimensions_in_video_url(a_format, video_url): - m = re.search(r'/(?P\d+)x(?P\d+)/', video_url) - if m: - a_format.update({ - 'width': int(m.group('width')), - 'height': int(m.group('height')), - }) + if config.get('source_type') == 'vine': + return self.url_result(config['player_url'], 'Vine') - video_url = config.get('video_url') or config.get('playlist', [{}])[0].get('source') + periscope_url = PeriscopeIE._extract_url(webpage) + if periscope_url: + return self.url_result(periscope_url, PeriscopeIE.ie_key()) - if video_url: - if determine_ext(video_url) == 'm3u8': - formats.extend(self._extract_m3u8_formats(video_url, video_id, ext='mp4', m3u8_id='hls')) - else: - f = { - 'url': video_url, - } + video_url = config.get('video_url') or config.get('playlist', [{}])[0].get('source') - _search_dimensions_in_video_url(f, video_url) + if video_url: + if determine_ext(video_url) == 'm3u8': + formats.extend(self._extract_m3u8_formats(video_url, video_id, ext='mp4', m3u8_id='hls')) + else: + f = { + 'url': video_url, + } - formats.append(f) + self._search_dimensions_in_video_url(f, video_url) - vmap_url = config.get('vmapUrl') or config.get('vmap_url') - if vmap_url: - formats.append({ - 'url': self._get_vmap_video_url(vmap_url, video_id), - }) + formats.append(f) - media_info = None + vmap_url = config.get('vmapUrl') or config.get('vmap_url') + if vmap_url: + formats.extend( + self._extract_formats_from_vmap_url(vmap_url, video_id)) - for entity in config.get('status', {}).get('entities', []): - if 'mediaInfo' in entity: - media_info = entity['mediaInfo'] + media_info = None - if media_info: - for media_variant in media_info['variants']: - media_url = media_variant['url'] - if media_url.endswith('.m3u8'): - formats.extend(self._extract_m3u8_formats(media_url, video_id, ext='mp4', m3u8_id='hls')) - elif media_url.endswith('.mpd'): - formats.extend(self._extract_mpd_formats(media_url, video_id, mpd_id='dash')) - else: - vbr = int_or_none(media_variant.get('bitRate'), scale=1000) - a_format = { - 'url': media_url, - 'format_id': 'http-%d' % vbr if vbr else 'http', - 'vbr': vbr, - } - # Reported bitRate may be zero - if not a_format['vbr']: - del a_format['vbr'] + for entity in config.get('status', {}).get('entities', []): + if 'mediaInfo' in entity: + media_info = entity['mediaInfo'] - _search_dimensions_in_video_url(a_format, media_url) + if media_info: + formats.extend(self._parse_media_info(media_info, video_id)) + duration = float_or_none(media_info.get('duration', {}).get('nanos'), scale=1e9) - formats.append(a_format) + username = config.get('user', {}).get('screen_name') + if username: + formats.extend(self._extract_mobile_formats(username, video_id)) - duration = float_or_none(media_info.get('duration', {}).get('nanos'), scale=1e9) + if formats: + title = self._search_regex(r'([^<]+)', webpage, 'title') + thumbnail = config.get('posterImageUrl') or config.get('image_src') + duration = float_or_none(config.get('duration'), scale=1000) or duration + break + if not formats: + headers = { + 'Authorization': 'Bearer AAAAAAAAAAAAAAAAAAAAAPYXBAAAAAAACLXUNDekMxqa8h%2F40K4moUkGsoc%3DTYfbDKbT3jJPCEVnMYqilB28NHfOPqkca3qaAxGfsyKCs0wRbw', + 'Referer': url, + } + ct0 = self._get_cookies(url).get('ct0') + if ct0: + headers['csrf_token'] = ct0.value + guest_token = self._download_json( + '%s/guest/activate.json' % self._API_BASE, video_id, + 'Downloading guest token', data=b'', + headers=headers)['guest_token'] + headers['x-guest-token'] = guest_token + self._set_cookie('api.twitter.com', 'gt', guest_token) + config = self._download_json( + '%s/videos/tweet/config/%s.json' % (self._API_BASE, video_id), + video_id, headers=headers) + track = config['track'] + vmap_url = track.get('vmapUrl') + if vmap_url: + formats = self._extract_formats_from_vmap_url(vmap_url, video_id) + else: + playback_url = track['playbackUrl'] + if determine_ext(playback_url) == 'm3u8': + formats = self._extract_m3u8_formats( + playback_url, video_id, 'mp4', + entry_protocol='m3u8_native', m3u8_id='hls') + else: + formats = [{ + 'url': playback_url, + }] + title = 'Twitter web player' + thumbnail = config.get('posterImage') + duration = float_or_none(track.get('durationMs'), scale=1000) + + self._remove_duplicate_formats(formats) self._sort_formats(formats) - title = self._search_regex(r'([^<]+)', webpage, 'title') - thumbnail = config.get('posterImageUrl') or config.get('image_src') - duration = float_or_none(config.get('duration')) or duration - return { 'id': video_id, 'title': title, @@ -181,8 +280,9 @@ class TwitterCardIE(TwitterBaseIE): class TwitterIE(InfoExtractor): IE_NAME = 'twitter' - _VALID_URL = r'https?://(?:www\.|m\.|mobile\.)?twitter\.com/(?P[^/]+)/status/(?P\d+)' + _VALID_URL = r'https?://(?:www\.|m\.|mobile\.)?twitter\.com/(?:i/web|(?P[^/]+))/status/(?P\d+)' _TEMPLATE_URL = 'https://twitter.com/%s/status/%s' + _TEMPLATE_STATUSES_URL = 'https://twitter.com/statuses/%s' _TESTS = [{ 'url': 'https://twitter.com/freethenipple/status/643211948184596480', @@ -190,13 +290,11 @@ class TwitterIE(InfoExtractor): 'id': '643211948184596480', 'ext': 'mp4', 'title': 'FREE THE NIPPLE - FTN supporters on Hollywood Blvd today!', - 'thumbnail': 're:^https?://.*\.jpg', + 'thumbnail': r're:^https?://.*\.jpg', 'description': 'FREE THE NIPPLE on Twitter: "FTN supporters on Hollywood Blvd today! http://t.co/c7jHH749xJ"', 'uploader': 'FREE THE NIPPLE', 'uploader_id': 'freethenipple', - }, - 'params': { - 'skip_download': True, # requires ffmpeg + 'duration': 12.922, }, }, { 'url': 'https://twitter.com/giphz/status/657991469417025536/photo/1', @@ -206,7 +304,7 @@ class TwitterIE(InfoExtractor): 'ext': 'mp4', 'title': 'Gifs - tu vai cai tu vai cai tu nao eh capaz disso tu vai cai', 'description': 'Gifs on Twitter: "tu vai cai tu vai cai tu nao eh capaz disso tu vai cai https://t.co/tM46VHFlO5"', - 'thumbnail': 're:^https?://.*\.png', + 'thumbnail': r're:^https?://.*\.png', 'uploader': 'Gifs', 'uploader_id': 'giphz', }, @@ -214,7 +312,6 @@ class TwitterIE(InfoExtractor): 'skip': 'Account suspended', }, { 'url': 'https://twitter.com/starwars/status/665052190608723968', - 'md5': '39b7199856dee6cd4432e72c74bc69d4', 'info_dict': { 'id': '665052190608723968', 'ext': 'mp4', @@ -240,18 +337,15 @@ class TwitterIE(InfoExtractor): }, }, { 'url': 'https://twitter.com/jaydingeer/status/700207533655363584', - 'md5': '', 'info_dict': { 'id': '700207533655363584', 'ext': 'mp4', - 'title': 'Donte The Dumbass - BEAT PROD: @suhmeduh #Damndaniel', - 'description': 'Donte The Dumbass on Twitter: "BEAT PROD: @suhmeduh https://t.co/HBrQ4AfpvZ #Damndaniel https://t.co/byBooq2ejZ"', - 'thumbnail': 're:^https?://.*\.jpg', - 'uploader': 'Donte The Dumbass', + 'title': 'JG - BEAT PROD: @suhmeduh #Damndaniel', + 'description': 'JG on Twitter: "BEAT PROD: @suhmeduh https://t.co/HBrQ4AfpvZ #Damndaniel https://t.co/byBooq2ejZ"', + 'thumbnail': r're:^https?://.*\.jpg', + 'uploader': 'JG', 'uploader_id': 'jaydingeer', - }, - 'params': { - 'skip_download': True, # requires ffmpeg + 'duration': 30.0, }, }, { 'url': 'https://twitter.com/Filmdrunk/status/713801302971588609', @@ -259,9 +353,11 @@ class TwitterIE(InfoExtractor): 'info_dict': { 'id': 'MIOxnrUteUd', 'ext': 'mp4', - 'title': 'Dr.Pepperの飲み方 #japanese #バカ #ドクペ #電動ガン', - 'uploader': 'TAKUMA', - 'uploader_id': '1004126642786242560', + 'title': 'Vince Mancini - Vine of the day', + 'description': 'Vince Mancini on Twitter: "Vine of the day https://t.co/xmTvRdqxWf"', + 'uploader': 'Vince Mancini', + 'uploader_id': 'Filmdrunk', + 'timestamp': 1402826626, 'upload_date': '20140615', }, 'add_ie': ['Vine'], @@ -274,6 +370,60 @@ class TwitterIE(InfoExtractor): 'description': 'Captain America on Twitter: "@King0fNerd Are you sure you made the right choice? Find out in theaters. https://t.co/GpgYi9xMJI"', 'uploader_id': 'captainamerica', 'uploader': 'Captain America', + 'duration': 3.17, + }, + }, { + 'url': 'https://twitter.com/OPP_HSD/status/779210622571536384', + 'info_dict': { + 'id': '1zqKVVlkqLaKB', + 'ext': 'mp4', + 'title': 'Sgt Kerry Schmidt - LIVE on #Periscope: Road rage, mischief, assault, rollover and fire in one occurrence', + 'description': 'Sgt Kerry Schmidt on Twitter: "LIVE on #Periscope: Road rage, mischief, assault, rollover and fire in one occurrence https://t.co/EKrVgIXF3s"', + 'upload_date': '20160923', + 'uploader_id': 'OPP_HSD', + 'uploader': 'Sgt Kerry Schmidt', + 'timestamp': 1474613214, + }, + 'add_ie': ['Periscope'], + }, { + # has mp4 formats via mobile API + 'url': 'https://twitter.com/news_al3alm/status/852138619213144067', + 'info_dict': { + 'id': '852138619213144067', + 'ext': 'mp4', + 'title': 'عالم الأخبار - كلمة تاريخية بجلسة الجناسي التاريخية.. النائب خالد مؤنس العتيبي للمعارضين : اتقوا الله .. الظلم ظلمات يوم القيامة', + 'description': 'عالم الأخبار on Twitter: "كلمة تاريخية بجلسة الجناسي التاريخية.. النائب خالد مؤنس العتيبي للمعارضين : اتقوا الله .. الظلم ظلمات يوم القيامة https://t.co/xg6OhpyKfN"', + 'uploader': 'عالم الأخبار', + 'uploader_id': 'news_al3alm', + 'duration': 277.4, + }, + }, { + 'url': 'https://twitter.com/i/web/status/910031516746514432', + 'info_dict': { + 'id': '910031516746514432', + 'ext': 'mp4', + 'title': 'Préfet de Guadeloupe - [Direct] #Maria Le centre se trouve actuellement au sud de Basse-Terre. Restez confinés. Réfugiez-vous dans la pièce la + sûre.', + 'thumbnail': r're:^https?://.*\.jpg', + 'description': 'Préfet de Guadeloupe on Twitter: "[Direct] #Maria Le centre se trouve actuellement au sud de Basse-Terre. Restez confinés. Réfugiez-vous dans la pièce la + sûre. https://t.co/mwx01Rs4lo"', + 'uploader': 'Préfet de Guadeloupe', + 'uploader_id': 'Prefet971', + 'duration': 47.48, + }, + 'params': { + 'skip_download': True, # requires ffmpeg + }, + }, { + # card via api.twitter.com/1.1/videos/tweet/config + 'url': 'https://twitter.com/LisPower1/status/1001551623938805763', + 'info_dict': { + 'id': '1001551623938805763', + 'ext': 'mp4', + 'title': 're:.*?Shep is on a roll today.*?', + 'thumbnail': r're:^https?://.*\.jpg', + 'description': 'md5:63b036c228772523ae1924d5f8e5ed6b', + 'uploader': 'Lis Power', + 'uploader_id': 'LisPower1', + 'duration': 111.278, }, 'params': { 'skip_download': True, # requires ffmpeg @@ -286,11 +436,15 @@ class TwitterIE(InfoExtractor): twid = mobj.group('id') webpage, urlh = self._download_webpage_handle( - self._TEMPLATE_URL % (user_id, twid), twid) + self._TEMPLATE_STATUSES_URL % twid, twid) if 'twitter.com/account/suspended' in urlh.geturl(): raise ExtractorError('Account suspended by Twitter.', expected=True) + if user_id is None: + mobj = re.match(self._VALID_URL, urlh.geturl()) + user_id = mobj.group('user_id') + username = remove_end(self._og_search_title(webpage), ' on Twitter') title = description = self._og_search_description(webpage).strip('').replace('\n', ' ').strip('“”') @@ -328,13 +482,22 @@ class TwitterIE(InfoExtractor): }) return info + twitter_card_url = None if 'class="PlayableMedia' in webpage: + twitter_card_url = '%s//twitter.com/i/videos/tweet/%s' % (self.http_scheme(), twid) + else: + twitter_card_iframe_url = self._search_regex( + r'data-full-card-iframe-url=([\'"])(?P(?:(?!\1).)+)\1', + webpage, 'Twitter card iframe URL', default=None, group='url') + if twitter_card_iframe_url: + twitter_card_url = compat_urlparse.urljoin(url, twitter_card_iframe_url) + + if twitter_card_url: info.update({ '_type': 'url_transparent', 'ie_key': 'TwitterCard', - 'url': '%s//twitter.com/i/videos/tweet/%s' % (self.http_scheme(), twid), + 'url': twitter_card_url, }) - return info raise ExtractorError('There\'s no video in this tweet.') @@ -342,7 +505,7 @@ class TwitterIE(InfoExtractor): class TwitterAmplifyIE(TwitterBaseIE): IE_NAME = 'twitter:amplify' - _VALID_URL = 'https?://amp\.twimg\.com/v/(?P[0-9a-f\-]{36})' + _VALID_URL = r'https?://amp\.twimg\.com/v/(?P[0-9a-f\-]{36})' _TEST = { 'url': 'https://amp.twimg.com/v/0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951', @@ -361,7 +524,7 @@ class TwitterAmplifyIE(TwitterBaseIE): vmap_url = self._html_search_meta( 'twitter:amplify:vmap', webpage, 'vmap url') - video_url = self._get_vmap_video_url(vmap_url, video_id) + formats = self._extract_formats_from_vmap_url(vmap_url, video_id) thumbnails = [] thumbnail = self._html_search_meta( @@ -383,11 +546,10 @@ class TwitterAmplifyIE(TwitterBaseIE): }) video_w, video_h = _find_dimension('player') - formats = [{ - 'url': video_url, + formats[0].update({ 'width': video_w, 'height': video_h, - }] + }) return { 'id': video_id,