X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/5ea61d104de9ca8a7d19d63d83173eb7391081d2..c8c68895a8818fe160e81ce9c6f8baea38fe3051:/youtube_dl/extractor/dailymotion.py diff --git a/youtube_dl/extractor/dailymotion.py b/youtube_dl/extractor/dailymotion.py index 70aa433..b852905 100644 --- a/youtube_dl/extractor/dailymotion.py +++ b/youtube_dl/extractor/dailymotion.py @@ -1,253 +1,393 @@ # coding: utf-8 from __future__ import unicode_literals -import re +import functools import json -import itertools +import re from .common import InfoExtractor - -from ..compat import ( - compat_str, - compat_urllib_request, -) +from ..compat import compat_HTTPError from ..utils import ( + age_restricted, + clean_html, ExtractorError, int_or_none, - orderedSet, - str_to_int, + OnDemandPagedList, + try_get, unescapeHTML, + urlencode_postdata, ) class DailymotionBaseInfoExtractor(InfoExtractor): + _FAMILY_FILTER = None + _HEADERS = { + 'Content-Type': 'application/json', + 'Origin': 'https://www.dailymotion.com', + } + _NETRC_MACHINE = 'dailymotion' + + def _get_dailymotion_cookies(self): + return self._get_cookies('https://www.dailymotion.com/') + @staticmethod - def _build_request(url): - """Build a request with the family filter disabled""" - request = compat_urllib_request.Request(url) - request.add_header('Cookie', 'family_filter=off; ff=off') - return request + def _get_cookie_value(cookies, name): + cookie = cookies.get(name) + if cookie: + return cookie.value + + def _set_dailymotion_cookie(self, name, value): + self._set_cookie('www.dailymotion.com', name, value) + + def _real_initialize(self): + cookies = self._get_dailymotion_cookies() + ff = self._get_cookie_value(cookies, 'ff') + self._FAMILY_FILTER = ff == 'on' if ff else age_restricted(18, self._downloader.params.get('age_limit')) + self._set_dailymotion_cookie('ff', 'on' if self._FAMILY_FILTER else 'off') + + def _call_api(self, object_type, xid, object_fields, note, filter_extra=None): + if not self._HEADERS.get('Authorization'): + cookies = self._get_dailymotion_cookies() + token = self._get_cookie_value(cookies, 'access_token') or self._get_cookie_value(cookies, 'client_token') + if not token: + data = { + 'client_id': 'f1a362d288c1b98099c7', + 'client_secret': 'eea605b96e01c796ff369935357eca920c5da4c5', + } + username, password = self._get_login_info() + if username: + data.update({ + 'grant_type': 'password', + 'password': password, + 'username': username, + }) + else: + data['grant_type'] = 'client_credentials' + try: + token = self._download_json( + 'https://graphql.api.dailymotion.com/oauth/token', + None, 'Downloading Access Token', + data=urlencode_postdata(data))['access_token'] + except ExtractorError as e: + if isinstance(e.cause, compat_HTTPError) and e.cause.code == 400: + raise ExtractorError(self._parse_json( + e.cause.read().decode(), xid)['error_description'], expected=True) + raise + self._set_dailymotion_cookie('access_token' if username else 'client_token', token) + self._HEADERS['Authorization'] = 'Bearer ' + token + + resp = self._download_json( + 'https://graphql.api.dailymotion.com/', xid, note, data=json.dumps({ + 'query': '''{ + %s(xid: "%s"%s) { + %s + } +}''' % (object_type, xid, ', ' + filter_extra if filter_extra else '', object_fields), + }).encode(), headers=self._HEADERS) + obj = resp['data'][object_type] + if not obj: + raise ExtractorError(resp['errors'][0]['message'], expected=True) + return obj class DailymotionIE(DailymotionBaseInfoExtractor): - """Information Extractor for Dailymotion""" - - _VALID_URL = r'(?i)(?:https?://)?(?:(www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(embed|#)/)?video/(?P[^/?_]+)' + _VALID_URL = r'''(?ix) + https?:// + (?: + (?:(?:www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(?:(?:embed|swf|\#)/)?video|swf)| + (?:www\.)?lequipe\.fr/video + ) + /(?P[^/?_]+)(?:.+?\bplaylist=(?Px[0-9a-z]+))? + ''' IE_NAME = 'dailymotion' - - _FORMATS = [ - ('stream_h264_ld_url', 'ld'), - ('stream_h264_url', 'standard'), - ('stream_h264_hq_url', 'hq'), - ('stream_h264_hd_url', 'hd'), - ('stream_h264_hd1080_url', 'hd180'), - ] - - _TESTS = [ - { - 'url': 'https://www.dailymotion.com/video/x2iuewm_steam-machine-models-pricing-listed-on-steam-store-ign-news_videogames', - 'md5': '2137c41a8e78554bb09225b8eb322406', - 'info_dict': { - 'id': 'x2iuewm', - 'ext': 'mp4', - 'uploader': 'IGN', - 'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News', - 'upload_date': '20150306', - } + _TESTS = [{ + 'url': 'http://www.dailymotion.com/video/x5kesuj_office-christmas-party-review-jason-bateman-olivia-munn-t-j-miller_news', + 'md5': '074b95bdee76b9e3654137aee9c79dfe', + 'info_dict': { + 'id': 'x5kesuj', + 'ext': 'mp4', + 'title': 'Office Christmas Party Review – Jason Bateman, Olivia Munn, T.J. Miller', + 'description': 'Office Christmas Party Review - Jason Bateman, Olivia Munn, T.J. Miller', + 'duration': 187, + 'timestamp': 1493651285, + 'upload_date': '20170501', + 'uploader': 'Deadline', + 'uploader_id': 'x1xm8ri', + 'age_limit': 0, + }, + }, { + 'url': 'https://www.dailymotion.com/video/x2iuewm_steam-machine-models-pricing-listed-on-steam-store-ign-news_videogames', + 'md5': '2137c41a8e78554bb09225b8eb322406', + 'info_dict': { + 'id': 'x2iuewm', + 'ext': 'mp4', + 'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News', + 'description': 'Several come bundled with the Steam Controller.', + 'thumbnail': r're:^https?:.*\.(?:jpg|png)$', + 'duration': 74, + 'timestamp': 1425657362, + 'upload_date': '20150306', + 'uploader': 'IGN', + 'uploader_id': 'xijv66', + 'age_limit': 0, + 'view_count': int, }, + 'skip': 'video gone', + }, { # Vevo video - { - 'url': 'http://www.dailymotion.com/video/x149uew_katy-perry-roar-official_musi', - 'info_dict': { - 'title': 'Roar (Official)', - 'id': 'USUV71301934', - 'ext': 'mp4', - 'uploader': 'Katy Perry', - 'upload_date': '20130905', - }, - 'params': { - 'skip_download': True, - }, - 'skip': 'VEVO is only available in some countries', + 'url': 'http://www.dailymotion.com/video/x149uew_katy-perry-roar-official_musi', + 'info_dict': { + 'title': 'Roar (Official)', + 'id': 'USUV71301934', + 'ext': 'mp4', + 'uploader': 'Katy Perry', + 'upload_date': '20130905', + }, + 'params': { + 'skip_download': True, }, + 'skip': 'VEVO is only available in some countries', + }, { # age-restricted video - { - 'url': 'http://www.dailymotion.com/video/xyh2zz_leanna-decker-cyber-girl-of-the-year-desires-nude-playboy-plus_redband', - 'md5': '0d667a7b9cebecc3c89ee93099c4159d', - 'info_dict': { - 'id': 'xyh2zz', - 'ext': 'mp4', - 'title': 'Leanna Decker - Cyber Girl Of The Year Desires Nude [Playboy Plus]', - 'uploader': 'HotWaves1012', - 'age_limit': 18, - } - } - ] + 'url': 'http://www.dailymotion.com/video/xyh2zz_leanna-decker-cyber-girl-of-the-year-desires-nude-playboy-plus_redband', + 'md5': '0d667a7b9cebecc3c89ee93099c4159d', + 'info_dict': { + 'id': 'xyh2zz', + 'ext': 'mp4', + 'title': 'Leanna Decker - Cyber Girl Of The Year Desires Nude [Playboy Plus]', + 'uploader': 'HotWaves1012', + 'age_limit': 18, + }, + 'skip': 'video gone', + }, { + # geo-restricted, player v5 + 'url': 'http://www.dailymotion.com/video/xhza0o', + 'only_matching': True, + }, { + # with subtitles + 'url': 'http://www.dailymotion.com/video/x20su5f_the-power-of-nightmares-1-the-rise-of-the-politics-of-fear-bbc-2004_news', + 'only_matching': True, + }, { + 'url': 'http://www.dailymotion.com/swf/video/x3n92nf', + 'only_matching': True, + }, { + 'url': 'http://www.dailymotion.com/swf/x3ss1m_funny-magic-trick-barry-and-stuart_fun', + 'only_matching': True, + }, { + 'url': 'https://www.lequipe.fr/video/x791mem', + 'only_matching': True, + }, { + 'url': 'https://www.lequipe.fr/video/k7MtHciueyTcrFtFKA2', + 'only_matching': True, + }, { + 'url': 'https://www.dailymotion.com/video/x3z49k?playlist=xv4bw', + 'only_matching': True, + }] + _GEO_BYPASS = False + _COMMON_MEDIA_FIELDS = '''description + geoblockedCountries { + allowed + } + xid''' - def _real_extract(self, url): - video_id = self._match_id(url) - url = 'https://www.dailymotion.com/video/%s' % video_id - - # Retrieve video webpage to extract further information - request = self._build_request(url) - webpage = self._download_webpage(request, video_id) - - # Extract URL, uploader and title from webpage - self.report_extraction(video_id) - - # It may just embed a vevo video: - m_vevo = re.search( - r'[\w]*)', - webpage) - if m_vevo is not None: - vevo_id = m_vevo.group('id') - self.to_screen('Vevo video detected: %s' % vevo_id) - return self.url_result('vevo:%s' % vevo_id, ie='Vevo') - - age_limit = self._rta_search(webpage) - - video_upload_date = None - mobj = re.search(r'', webpage) - if mobj is not None: - video_upload_date = mobj.group(1) + mobj.group(2) + mobj.group(3) - - embed_url = 'https://www.dailymotion.com/embed/video/%s' % video_id - embed_request = self._build_request(embed_url) - embed_page = self._download_webpage( - embed_request, video_id, 'Downloading embed page') - info = self._search_regex(r'var info = ({.*?}),$', embed_page, - 'video info', flags=re.MULTILINE) - info = json.loads(info) - if info.get('error') is not None: - msg = 'Couldn\'t get video, Dailymotion says: %s' % info['error']['title'] - raise ExtractorError(msg, expected=True) + @staticmethod + def _extract_urls(webpage): + urls = [] + # Look for embedded Dailymotion player + # https://developer.dailymotion.com/player#player-parameters + for mobj in re.finditer( + r'<(?:(?:embed|iframe)[^>]+?src=|input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=)(["\'])(?P(?:https?:)?//(?:www\.)?dailymotion\.com/(?:embed|swf)/video/.+?)\1', webpage): + urls.append(unescapeHTML(mobj.group('url'))) + for mobj in re.finditer( + r'(?s)DM\.player\([^,]+,\s*{.*?video[\'"]?\s*:\s*["\']?(?P[0-9a-zA-Z]+).+?}\s*\);', webpage): + urls.append('https://www.dailymotion.com/embed/video/' + mobj.group('id')) + return urls + def _real_extract(self, url): + video_id, playlist_id = re.match(self._VALID_URL, url).groups() + + if playlist_id: + if not self._downloader.params.get('noplaylist'): + self.to_screen('Downloading playlist %s - add --no-playlist to just download video' % playlist_id) + return self.url_result( + 'http://www.dailymotion.com/playlist/' + playlist_id, + 'DailymotionPlaylist', playlist_id) + self.to_screen('Downloading just video %s because of --no-playlist' % video_id) + + password = self._downloader.params.get('videopassword') + media = self._call_api( + 'media', video_id, '''... on Video { + %s + stats { + likes { + total + } + views { + total + } + } + } + ... on Live { + %s + audienceCount + isOnAir + }''' % (self._COMMON_MEDIA_FIELDS, self._COMMON_MEDIA_FIELDS), 'Downloading media JSON metadata', + 'password: "%s"' % self._downloader.params.get('videopassword') if password else None) + xid = media['xid'] + + metadata = self._download_json( + 'https://www.dailymotion.com/player/metadata/video/' + xid, + xid, 'Downloading metadata JSON', + query={'app': 'com.dailymotion.neon'}) + + error = metadata.get('error') + if error: + title = error.get('title') or error['raw_message'] + # See https://developer.dailymotion.com/api#access-error + if error.get('code') == 'DM007': + allowed_countries = try_get(media, lambda x: x['geoblockedCountries']['allowed'], list) + self.raise_geo_restricted(msg=title, countries=allowed_countries) + raise ExtractorError( + '%s said: %s' % (self.IE_NAME, title), expected=True) + + title = metadata['title'] + is_live = media.get('isOnAir') formats = [] - for (key, format_id) in self._FORMATS: - video_url = info.get(key) - if video_url is not None: - m_size = re.search(r'H264-(\d+)x(\d+)', video_url) - if m_size is not None: - width, height = map(int_or_none, (m_size.group(1), m_size.group(2))) + for quality, media_list in metadata['qualities'].items(): + for m in media_list: + media_url = m.get('url') + media_type = m.get('type') + if not media_url or media_type == 'application/vnd.lumberjack.manifest': + continue + if media_type == 'application/x-mpegURL': + formats.extend(self._extract_m3u8_formats( + media_url, video_id, 'mp4', + 'm3u8' if is_live else 'm3u8_native', + m3u8_id='hls', fatal=False)) else: - width, height = None, None - formats.append({ - 'url': video_url, - 'ext': 'mp4', - 'format_id': format_id, - 'width': width, - 'height': height, - }) - if not formats: - raise ExtractorError('Unable to extract video URL') - - # subtitles - video_subtitles = self.extract_subtitles(video_id, webpage) - - view_count = str_to_int(self._search_regex( - r'video_views_count[^>]+>\s+([\d\.,]+)', - webpage, 'view count', fatal=False)) - - title = self._og_search_title(webpage, default=None) - if title is None: - title = self._html_search_regex( - r'(?s)]*>(.*?)', webpage, - 'title') + f = { + 'url': media_url, + 'format_id': 'http-' + quality, + } + m = re.search(r'/H264-(\d+)x(\d+)(?:-(60)/)?', media_url) + if m: + width, height, fps = map(int_or_none, m.groups()) + f.update({ + 'fps': fps, + 'height': height, + 'width': width, + }) + formats.append(f) + for f in formats: + f['url'] = f['url'].split('#')[0] + if not f.get('fps') and f['format_id'].endswith('@60'): + f['fps'] = 60 + self._sort_formats(formats) + + subtitles = {} + subtitles_data = try_get(metadata, lambda x: x['subtitles']['data'], dict) or {} + for subtitle_lang, subtitle in subtitles_data.items(): + subtitles[subtitle_lang] = [{ + 'url': subtitle_url, + } for subtitle_url in subtitle.get('urls', [])] + + thumbnails = [] + for height, poster_url in metadata.get('posters', {}).items(): + thumbnails.append({ + 'height': int_or_none(height), + 'id': height, + 'url': poster_url, + }) + + owner = metadata.get('owner') or {} + stats = media.get('stats') or {} + get_count = lambda x: int_or_none(try_get(stats, lambda y: y[x + 's']['total'])) return { 'id': video_id, + 'title': self._live_title(title) if is_live else title, + 'description': clean_html(media.get('description')), + 'thumbnails': thumbnails, + 'duration': int_or_none(metadata.get('duration')) or None, + 'timestamp': int_or_none(metadata.get('created_time')), + 'uploader': owner.get('screenname'), + 'uploader_id': owner.get('id') or metadata.get('screenname'), + 'age_limit': 18 if metadata.get('explicit') else 0, + 'tags': metadata.get('tags'), + 'view_count': get_count('view') or int_or_none(media.get('audienceCount')), + 'like_count': get_count('like'), 'formats': formats, - 'uploader': info['owner.screenname'], - 'upload_date': video_upload_date, - 'title': title, - 'subtitles': video_subtitles, - 'thumbnail': info['thumbnail_url'], - 'age_limit': age_limit, - 'view_count': view_count, + 'subtitles': subtitles, + 'is_live': is_live, } - def _get_subtitles(self, video_id, webpage): - try: - sub_list = self._download_webpage( - 'https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id, - video_id, note=False) - except ExtractorError as err: - self._downloader.report_warning('unable to download video subtitles: %s' % compat_str(err)) - return {} - info = json.loads(sub_list) - if (info['total'] > 0): - sub_lang_list = dict((l['language'], [{'url': l['url'], 'ext': 'srt'}]) for l in info['list']) - return sub_lang_list - self._downloader.report_warning('video doesn\'t have subtitles') - return {} - - -class DailymotionPlaylistIE(DailymotionBaseInfoExtractor): + +class DailymotionPlaylistBaseIE(DailymotionBaseInfoExtractor): + _PAGE_SIZE = 100 + + def _fetch_page(self, playlist_id, page): + page += 1 + videos = self._call_api( + self._OBJECT_TYPE, playlist_id, + '''videos(allowExplicit: %s, first: %d, page: %d) { + edges { + node { + xid + url + } + } + }''' % ('false' if self._FAMILY_FILTER else 'true', self._PAGE_SIZE, page), + 'Downloading page %d' % page)['videos'] + for edge in videos['edges']: + node = edge['node'] + yield self.url_result( + node['url'], DailymotionIE.ie_key(), node['xid']) + + def _real_extract(self, url): + playlist_id = self._match_id(url) + entries = OnDemandPagedList(functools.partial( + self._fetch_page, playlist_id), self._PAGE_SIZE) + return self.playlist_result( + entries, playlist_id) + + +class DailymotionPlaylistIE(DailymotionPlaylistBaseIE): IE_NAME = 'dailymotion:playlist' - _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P.+?)/' - _MORE_PAGES_INDICATOR = r'(?s)
.*?x[0-9a-z]+)' _TESTS = [{ 'url': 'http://www.dailymotion.com/playlist/xv4bw_nqtv_sport/1#video=xl8v3q', 'info_dict': { - 'title': 'SPORT', - 'id': 'xv4bw_nqtv_sport', + 'id': 'xv4bw', }, 'playlist_mincount': 20, }] - - def _extract_entries(self, id): - video_ids = [] - for pagenum in itertools.count(1): - request = self._build_request(self._PAGE_TEMPLATE % (id, pagenum)) - webpage = self._download_webpage(request, - id, 'Downloading page %s' % pagenum) - - video_ids.extend(re.findall(r'data-xid="(.+?)"', webpage)) - - if re.search(self._MORE_PAGES_INDICATOR, webpage) is None: - break - return [self.url_result('http://www.dailymotion.com/video/%s' % video_id, 'Dailymotion') - for video_id in orderedSet(video_ids)] - - def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - playlist_id = mobj.group('id') - webpage = self._download_webpage(url, playlist_id) - - return { - '_type': 'playlist', - 'id': playlist_id, - 'title': self._og_search_title(webpage), - 'entries': self._extract_entries(playlist_id), - } + _OBJECT_TYPE = 'collection' -class DailymotionUserIE(DailymotionPlaylistIE): +class DailymotionUserIE(DailymotionPlaylistBaseIE): IE_NAME = 'dailymotion:user' - _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?:(?:old/)?user/)?(?P[^/]+)$' - _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s' + _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|swf|#|video|playlist)/)(?:(?:old/)?user/)?(?P[^/]+)' _TESTS = [{ 'url': 'https://www.dailymotion.com/user/nqtv', 'info_dict': { 'id': 'nqtv', - 'title': 'Rémi Gaillard', }, - 'playlist_mincount': 100, + 'playlist_mincount': 152, + }, { + 'url': 'http://www.dailymotion.com/user/UnderProject', + 'info_dict': { + 'id': 'UnderProject', + }, + 'playlist_mincount': 1000, + 'skip': 'Takes too long time', + }, { + 'url': 'https://www.dailymotion.com/user/nqtv', + 'info_dict': { + 'id': 'nqtv', + }, + 'playlist_mincount': 148, + 'params': { + 'age_limit': 0, + }, }] - - def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - user = mobj.group('user') - webpage = self._download_webpage( - 'https://www.dailymotion.com/user/%s' % user, user) - full_user = unescapeHTML(self._html_search_regex( - r'' % re.escape(user), - webpage, 'user')) - - return { - '_type': 'playlist', - 'id': user, - 'title': full_user, - 'entries': self._extract_entries(user), - } + _OBJECT_TYPE = 'channel'