X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/3477c644417600d9ec8f8d2a44f82da0a4b15eb5..0433503f78928c1f0c76a8d66ef236dc35ff23df:/youtube_dl/extractor/twitter.py?ds=inline diff --git a/youtube_dl/extractor/twitter.py b/youtube_dl/extractor/twitter.py index 9d3e46b..1b0b963 100644 --- a/youtube_dl/extractor/twitter.py +++ b/youtube_dl/extractor/twitter.py @@ -4,28 +4,58 @@ from __future__ import unicode_literals import re from .common import InfoExtractor -from ..compat import compat_urllib_request +from ..compat import compat_urlparse from ..utils import ( + determine_ext, + dict_get, + ExtractorError, float_or_none, - xpath_text, + int_or_none, remove_end, + try_get, + xpath_text, ) +from .periscope import PeriscopeIE + + +class TwitterBaseIE(InfoExtractor): + def _extract_formats_from_vmap_url(self, vmap_url, video_id): + vmap_data = self._download_xml(vmap_url, video_id) + 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, + }] -class TwitterCardIE(InfoExtractor): + @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/(?P\d+)' + _VALID_URL = r'https?://(?:www\.)?twitter\.com/i/(?:cards/tfw/v1|videos(?:/tweet)?)/(?P\d+)' _TESTS = [ { 'url': 'https://twitter.com/i/cards/tfw/v1/560070183650213889', - 'md5': '7d2f6b4d2eb841a7ccc893d479bfceb4', + # MD5 checksums are different in different places 'info_dict': { 'id': '560070183650213889', 'ext': 'mp4', - 'title': 'TwitterCard', - 'thumbnail': 're:^https?://.*\.jpg$', + 'title': 'Twitter Card', + 'thumbnail': r're:^https?://.*\.jpg$', 'duration': 30.033, - } + }, + 'skip': 'Video gone', }, { 'url': 'https://twitter.com/i/cards/tfw/v1/623160978427936768', @@ -33,82 +63,177 @@ class TwitterCardIE(InfoExtractor): 'info_dict': { 'id': '623160978427936768', 'ext': 'mp4', - 'title': 'TwitterCard', - 'thumbnail': 're:^https?://.*\.jpg', + 'title': 'Twitter Card', + 'thumbnail': r're:^https?://.*\.jpg', 'duration': 80.155, }, + 'skip': 'Video gone', }, { 'url': 'https://twitter.com/i/cards/tfw/v1/654001591733886977', - 'md5': 'b6f35e8b08a0bec6c8af77a2f4b3a814', + '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-things-to-do-after-installing-ubuntu-11-10/', + 'description': 'md5:a831e97fa384863d6e26ce48d1c43376', 'upload_date': '20111013', 'uploader': 'OMG! Ubuntu!', 'uploader_id': 'omgubuntu', }, - } + 'add_ie': ['Youtube'], + }, + { + 'url': 'https://twitter.com/i/cards/tfw/v1/665289828897005568', + 'md5': '6dabeaca9e68cbb71c99c322a4b42a11', + 'info_dict': { + 'id': 'iBb2x00UVlv', + 'ext': 'mp4', + 'upload_date': '20151113', + 'uploader_id': '1189339351084113920', + 'uploader': 'ArsenalTerje', + 'title': 'Vine by ArsenalTerje', + 'timestamp': 1447451307, + }, + 'add_ie': ['Vine'], + }, { + 'url': 'https://twitter.com/i/videos/tweet/705235433198714880', + 'md5': '884812a2adc8aaf6fe52b15ccbfa3b88', + 'info_dict': { + 'id': '705235433198714880', + 'ext': 'mp4', + 'title': 'Twitter web player', + 'thumbnail': r're:^https?://.*', + }, + }, { + 'url': 'https://twitter.com/i/videos/752274308186120192', + 'only_matching': True, + }, ] + 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: + vbr = int_or_none(dict_get(media_variant, ('bitRate', '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'] + + 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') + guest_token = self._search_regex( + r'document\.cookie\s*=\s*decodeURIComponent\("gt=(\d+)', + webpage, 'guest token') + api_data = self._download_json( + 'https://api.twitter.com/2/timeline/conversation/%s.json' % video_id, + video_id, 'Downloading mobile API data', + headers={ + 'Authorization': 'Bearer ' + bearer_token, + 'x-guest-token': guest_token, + }) + media_info = try_get(api_data, lambda o: o['globalObjects']['tweets'][video_id] + ['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) - # Different formats served for different User-Agents - USER_AGENTS = [ - 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)', # mp4 - 'Mozilla/5.0 (Windows NT 5.2; WOW64; rv:38.0) Gecko/20100101 Firefox/38.0', # webm - ] - config = None formats = [] - for user_agent in USER_AGENTS: - request = compat_urllib_request.Request(url) - request.add_header('User-Agent', user_agent) - webpage = self._download_webpage(request, video_id) - - youtube_url = self._html_search_regex( - r']+src="((?:https?:)?//www.youtube.com/embed/[^"]+)"', - webpage, 'youtube iframe', default=None) - if youtube_url: - return self.url_result(youtube_url, 'Youtube') - - config = self._parse_json(self._html_search_regex( - r'data-player-config="([^"]+)"', webpage, 'data player config'), - video_id) - if 'playlist' not in config: - if 'vmapUrl' in config: - vmap_data = self._download_xml(config['vmapUrl'], video_id) - video_url = xpath_text(vmap_data, './/MediaFile').strip() - formats.append({ - 'url': video_url, - }) - break # same video regardless of UA - continue - - video_url = config['playlist'][0]['source'] - - f = { - 'url': video_url, - } - - m = re.search(r'/(?P\d+)x(?P\d+)/', video_url) - if m: - f.update({ - 'width': int(m.group('width')), - 'height': int(m.group('height')), - }) - formats.append(f) + duration = None + + webpage = self._download_webpage(url, 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) + + config = self._parse_json(self._html_search_regex( + r'data-(?:player-)?config="([^"]+)"', webpage, + 'data player config', default='{}'), + video_id) + + if config.get('source_type') == 'vine': + return self.url_result(config['player_url'], 'Vine') + + periscope_url = PeriscopeIE._extract_url(webpage) + if periscope_url: + return self.url_result(periscope_url, PeriscopeIE.ie_key()) + + video_url = config.get('video_url') or config.get('playlist', [{}])[0].get('source') + + 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, + } + + self._search_dimensions_in_video_url(f, video_url) + + formats.append(f) + + 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)) + + media_info = None + + for entity in config.get('status', {}).get('entities', []): + if 'mediaInfo' in entity: + media_info = entity['mediaInfo'] + + 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) + + username = config.get('user', {}).get('screen_name') + if username: + formats.extend(self._extract_mobile_formats(username, video_id)) + + self._remove_duplicate_formats(formats) self._sort_formats(formats) - thumbnail = config.get('posterImageUrl') - duration = float_or_none(config.get('duration')) + 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 return { 'id': video_id, - 'title': 'TwitterCard', + 'title': title, 'thumbnail': thumbnail, 'duration': duration, 'formats': formats, @@ -117,50 +242,278 @@ class TwitterCardIE(InfoExtractor): 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' - _TEST = { + _TESTS = [{ 'url': 'https://twitter.com/freethenipple/status/643211948184596480', - 'md5': '31cd83a116fc41f99ae3d909d4caf6a0', 'info_dict': { 'id': '643211948184596480', 'ext': 'mp4', 'title': 'FREE THE NIPPLE - FTN supporters on Hollywood Blvd today!', - 'thumbnail': 're:^https?://.*\.jpg', - 'duration': 12.922, + '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', + 'duration': 12.922, }, - } + 'params': { + 'skip_download': True, # requires ffmpeg + }, + }, { + 'url': 'https://twitter.com/giphz/status/657991469417025536/photo/1', + 'md5': 'f36dcd5fb92bf7057f155e7d927eeb42', + 'info_dict': { + 'id': '657991469417025536', + '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': r're:^https?://.*\.png', + 'uploader': 'Gifs', + 'uploader_id': 'giphz', + }, + 'expected_warnings': ['height', 'width'], + 'skip': 'Account suspended', + }, { + 'url': 'https://twitter.com/starwars/status/665052190608723968', + 'md5': '39b7199856dee6cd4432e72c74bc69d4', + 'info_dict': { + 'id': '665052190608723968', + 'ext': 'mp4', + 'title': 'Star Wars - A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens.', + 'description': 'Star Wars on Twitter: "A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens."', + 'uploader_id': 'starwars', + 'uploader': 'Star Wars', + }, + }, { + 'url': 'https://twitter.com/BTNBrentYarina/status/705235433198714880', + 'info_dict': { + 'id': '705235433198714880', + 'ext': 'mp4', + 'title': 'Brent Yarina - Khalil Iverson\'s missed highlight dunk. And made highlight dunk. In one highlight.', + 'description': 'Brent Yarina on Twitter: "Khalil Iverson\'s missed highlight dunk. And made highlight dunk. In one highlight."', + 'uploader_id': 'BTNBrentYarina', + 'uploader': 'Brent Yarina', + }, + 'params': { + # The same video as https://twitter.com/i/videos/tweet/705235433198714880 + # Test case of TwitterCardIE + 'skip_download': True, + }, + }, { + 'url': 'https://twitter.com/jaydingeer/status/700207533655363584', + 'md5': '', + 'info_dict': { + 'id': '700207533655363584', + 'ext': 'mp4', + 'title': 'あかさ - BEAT PROD: @suhmeduh #Damndaniel', + 'description': 'あかさ on Twitter: "BEAT PROD: @suhmeduh https://t.co/HBrQ4AfpvZ #Damndaniel https://t.co/byBooq2ejZ"', + 'thumbnail': r're:^https?://.*\.jpg', + 'uploader': 'あかさ', + 'uploader_id': 'jaydingeer', + 'duration': 30.0, + }, + 'params': { + 'skip_download': True, # requires ffmpeg + }, + }, { + 'url': 'https://twitter.com/Filmdrunk/status/713801302971588609', + 'md5': '89a15ed345d13b86e9a5a5e051fa308a', + 'info_dict': { + 'id': 'MIOxnrUteUd', + 'ext': 'mp4', + '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'], + }, { + 'url': 'https://twitter.com/captainamerica/status/719944021058060289', + 'info_dict': { + 'id': '719944021058060289', + 'ext': 'mp4', + 'title': 'Captain America - @King0fNerd Are you sure you made the right choice? Find out in theaters.', + '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, + }, + 'params': { + 'skip_download': True, # requires ffmpeg + }, + }, { + '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, + }, + 'params': { + 'format': 'best[format_id^=http-]', + }, + }, { + '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 + }, + }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) user_id = mobj.group('user_id') twid = mobj.group('id') - webpage = self._download_webpage(self._TEMPLATE_URL % (user_id, twid), twid) + webpage, urlh = self._download_webpage_handle( + 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 = self._og_search_description(webpage).strip('').replace('\n', ' ') + title = description = self._og_search_description(webpage).strip('').replace('\n', ' ').strip('“”') # strip 'https -_t.co_BJYgOjSeGA' junk from filenames - mobj = re.match(r'“(.*)\s+(https?://[^ ]+)”', title) - title, short_url = mobj.groups() - - card_id = self._search_regex( - r'["\']/i/cards/tfw/v1/(\d+)', webpage, 'twitter card url') - card_url = 'https://twitter.com/i/cards/tfw/v1/' + card_id + title = re.sub(r'\s+(https?://[^ ]+)', '', title) - return { - '_type': 'url_transparent', - 'ie_key': 'TwitterCard', + info = { 'uploader_id': user_id, 'uploader': username, - 'url': card_url, 'webpage_url': url, - 'description': '%s on Twitter: "%s %s"' % (username, title, short_url), + 'description': '%s on Twitter: "%s"' % (username, description), 'title': username + ' - ' + title, } + + mobj = re.search(r'''(?x) + ]+class="animated-gif"(?P[^>]+)>\s* + ]+video-src="(?P[^"]+)" + ''', webpage) + + if mobj: + more_info = mobj.group('more_info') + height = int_or_none(self._search_regex( + r'data-height="(\d+)"', more_info, 'height', fatal=False)) + width = int_or_none(self._search_regex( + r'data-width="(\d+)"', more_info, 'width', fatal=False)) + thumbnail = self._search_regex( + r'poster="([^"]+)"', more_info, 'poster', fatal=False) + info.update({ + 'id': twid, + 'url': mobj.group('url'), + 'height': height, + 'width': width, + 'thumbnail': thumbnail, + }) + 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': twitter_card_url, + }) + return info + + raise ExtractorError('There\'s no video in this tweet.') + + +class TwitterAmplifyIE(TwitterBaseIE): + IE_NAME = 'twitter:amplify' + _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', + 'md5': '7df102d0b9fd7066b86f3159f8e81bf6', + 'info_dict': { + 'id': '0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951', + 'ext': 'mp4', + 'title': 'Twitter Video', + 'thumbnail': 're:^https?://.*', + }, + } + + def _real_extract(self, url): + video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) + + vmap_url = self._html_search_meta( + 'twitter:amplify:vmap', webpage, 'vmap url') + formats = self._extract_formats_from_vmap_url(vmap_url, video_id) + + thumbnails = [] + thumbnail = self._html_search_meta( + 'twitter:image:src', webpage, 'thumbnail', fatal=False) + + def _find_dimension(target): + w = int_or_none(self._html_search_meta( + 'twitter:%s:width' % target, webpage, fatal=False)) + h = int_or_none(self._html_search_meta( + 'twitter:%s:height' % target, webpage, fatal=False)) + return w, h + + if thumbnail: + thumbnail_w, thumbnail_h = _find_dimension('image') + thumbnails.append({ + 'url': thumbnail, + 'width': thumbnail_w, + 'height': thumbnail_h, + }) + + video_w, video_h = _find_dimension('player') + formats[0].update({ + 'width': video_w, + 'height': video_h, + }) + + return { + 'id': video_id, + 'title': 'Twitter Video', + 'formats': formats, + 'thumbnails': thumbnails, + }