X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/a316b1d93c357b5edf19d5e3100526a94191c029..f1bb69fb5c709917a78f1bb6019781e1f76f40ae:/youtube_dl/extractor/dailymotion.py diff --git a/youtube_dl/extractor/dailymotion.py b/youtube_dl/extractor/dailymotion.py index 1ea449c..21a2d02 100644 --- a/youtube_dl/extractor/dailymotion.py +++ b/youtube_dl/extractor/dailymotion.py @@ -1,110 +1,464 @@ +# coding: utf-8 +from __future__ import unicode_literals + import re import json import itertools from .common import InfoExtractor -from ..utils import ( - compat_urllib_request, - get_element_by_attribute, - get_element_by_id, +from ..utils import ( + determine_ext, + error_to_compat_str, ExtractorError, + int_or_none, + parse_iso8601, + sanitized_Request, + str_to_int, + unescapeHTML, + mimetype2ext, ) -class DailymotionIE(InfoExtractor): - """Information Extractor for Dailymotion""" - - _VALID_URL = r'(?i)(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/video/([^/]+)' - IE_NAME = u'dailymotion' - _TEST = { - u'url': u'http://www.dailymotion.com/video/x33vw9_tutoriel-de-youtubeur-dl-des-video_tech', - u'file': u'x33vw9.mp4', - u'md5': u'392c4b85a60a90dc4792da41ce3144eb', - u'info_dict': { - u"uploader": u"Amphora Alex and Van .", - u"title": u"Tutoriel de Youtubeur\"DL DES VIDEO DE YOUTUBE\"" - } - } + +class DailymotionBaseInfoExtractor(InfoExtractor): + @staticmethod + def _build_request(url): + """Build a request with the family filter disabled""" + request = sanitized_Request(url) + request.add_header('Cookie', 'family_filter=off; ff=off') + return request + + def _download_webpage_handle_no_ff(self, url, *args, **kwargs): + request = self._build_request(url) + return self._download_webpage_handle(request, *args, **kwargs) + + def _download_webpage_no_ff(self, url, *args, **kwargs): + request = self._build_request(url) + return self._download_webpage(request, *args, **kwargs) + + +class DailymotionIE(DailymotionBaseInfoExtractor): + _VALID_URL = r'(?i)https?://(?:(www|touch)\.)?dailymotion\.[a-z]{2,3}/(?:(?:(?:embed|swf|#)/)?video|swf)/(?P[^/?_]+)' + 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': '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', + 'thumbnail': r're:^https?:.*\.(?:jpg|png)$', + 'duration': 187, + 'timestamp': 1493651285, + 'upload_date': '20170501', + 'uploader': 'Deadline', + 'uploader_id': 'x1xm8ri', + 'age_limit': 0, + 'view_count': int, + }, + }, { + 'url': 'https://www.dailymotion.com/video/x2iuewm_steam-machine-models-pricing-listed-on-steam-store-ign-news_videogames', + 'md5': '2137c41a8e78554bb09225b8eb322406', + 'info_dict': { + 'id': 'x2iuewm', + 'ext': 'mp4', + 'title': 'Steam Machine Models, Pricing Listed on Steam Store - IGN News', + 'description': 'Several come bundled with the Steam Controller.', + 'thumbnail': r're:^https?:.*\.(?:jpg|png)$', + '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', + }, { + # 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, + }, + '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, + }] + + @staticmethod + def _extract_urls(webpage): + # Look for embedded Dailymotion player + matches = re.findall( + r'<(?:(?:embed|iframe)[^>]+?src=|input[^>]+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=)(["\'])(?P(?:https?:)?//(?:www\.)?dailymotion\.com/(?:embed|swf)/video/.+?)\1', webpage) + return list(map(lambda m: unescapeHTML(m[1]), matches)) def _real_extract(self, url): - # Extract id and simplified title from URL - mobj = re.match(self._VALID_URL, url) + video_id = self._match_id(url) - video_id = mobj.group(1).split('_')[0].split('?')[0] + webpage = self._download_webpage_no_ff( + 'https://www.dailymotion.com/video/%s' % video_id, video_id) - video_extension = 'mp4' + age_limit = self._rta_search(webpage) - # Retrieve video webpage to extract further information - request = compat_urllib_request.Request(url) - request.add_header('Cookie', 'family_filter=off') - webpage = self._download_webpage(request, video_id) + description = self._og_search_description(webpage) or self._html_search_meta( + 'description', webpage, 'description') - # Extract URL, uploader and title from webpage - self.report_extraction(video_id) + view_count_str = self._search_regex( + (r']+itemprop="interactionCount"[^>]+content="UserPlays:([\s\d,.]+)"', + r'video_views_count[^>]+>\s+([\s\d\,.]+)'), + webpage, 'view count', default=None) + if view_count_str: + view_count_str = re.sub(r'\s', '', view_count_str) + view_count = str_to_int(view_count_str) + comment_count = int_or_none(self._search_regex( + r']+itemprop="interactionCount"[^>]+content="UserComments:(\d+)"', + webpage, 'comment count', default=None)) - video_uploader = self._search_regex([r'(?im)[^<]+?]+?>([^<]+?)', - # Looking for official user - r'<(?:span|a) .*?rel="author".*?>([^<]+?)([0-9]{2})-([0-9]{2})-([0-9]{4})', webpage) - if mobj is not None: - video_upload_date = mobj.group(3) + mobj.group(2) + mobj.group(1) + self._check_error(metadata) - embed_url = 'http://www.dailymotion.com/embed/video/%s' % video_id - embed_page = self._download_webpage(embed_url, video_id, - u'Downloading embed page') - info = self._search_regex(r'var info = ({.*?}),', embed_page, 'video info') - info = json.loads(info) + formats = [] + for quality, media_list in metadata['qualities'].items(): + for media in media_list: + media_url = media.get('url') + if not media_url: + continue + type_ = media.get('type') + if type_ == 'application/vnd.lumberjack.manifest': + continue + ext = mimetype2ext(type_) or determine_ext(media_url) + if ext == 'm3u8': + formats.extend(self._extract_m3u8_formats( + media_url, video_id, 'mp4', preference=-1, + m3u8_id='hls', fatal=False)) + elif ext == 'f4m': + formats.extend(self._extract_f4m_formats( + media_url, video_id, preference=-1, f4m_id='hds', fatal=False)) + else: + f = { + 'url': media_url, + 'format_id': 'http-%s' % quality, + 'ext': ext, + } + m = re.search(r'H264-(?P\d+)x(?P\d+)', media_url) + if m: + f.update({ + 'width': int(m.group('width')), + 'height': int(m.group('height')), + }) + formats.append(f) + self._sort_formats(formats) + + title = metadata['title'] + duration = int_or_none(metadata.get('duration')) + timestamp = int_or_none(metadata.get('created_time')) + thumbnail = metadata.get('poster_url') + uploader = metadata.get('owner', {}).get('screenname') + uploader_id = metadata.get('owner', {}).get('id') + + subtitles = {} + subtitles_data = metadata.get('subtitles', {}).get('data', {}) + if subtitles_data and isinstance(subtitles_data, dict): + for subtitle_lang, subtitle in subtitles_data.items(): + subtitles[subtitle_lang] = [{ + 'ext': determine_ext(subtitle_url), + 'url': subtitle_url, + } for subtitle_url in subtitle.get('urls', [])] + + return { + 'id': video_id, + 'title': title, + 'description': description, + 'thumbnail': thumbnail, + 'duration': duration, + 'timestamp': timestamp, + 'uploader': uploader, + 'uploader_id': uploader_id, + 'age_limit': age_limit, + 'view_count': view_count, + 'comment_count': comment_count, + 'formats': formats, + 'subtitles': subtitles, + } + + # vevo embed + vevo_id = self._search_regex( + r'[\w]*)', + webpage, 'vevo embed', default=None) + if vevo_id: + return self.url_result('vevo:%s' % vevo_id, 'Vevo') + + # fallback old player + embed_page = self._download_webpage_no_ff( + 'https://www.dailymotion.com/embed/video/%s' % video_id, + video_id, 'Downloading embed page') + + timestamp = parse_iso8601(self._html_search_meta( + 'video:release_date', webpage, 'upload date')) + + info = self._parse_json( + self._search_regex( + r'var info = ({.*?}),$', embed_page, + 'video info', flags=re.MULTILINE), + video_id) + + self._check_error(info) + + 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))) + else: + width, height = None, None + formats.append({ + 'url': video_url, + 'ext': 'mp4', + 'format_id': format_id, + 'width': width, + 'height': height, + }) + self._sort_formats(formats) + + # subtitles + video_subtitles = self.extract_subtitles(video_id, webpage) + + title = self._og_search_title(webpage, default=None) + if title is None: + title = self._html_search_regex( + r'(?s)]*>(.*?)', webpage, + 'title') + + return { + 'id': video_id, + 'formats': formats, + 'uploader': info['owner.screenname'], + 'timestamp': timestamp, + 'title': title, + 'description': description, + 'subtitles': video_subtitles, + 'thumbnail': info['thumbnail_url'], + 'age_limit': age_limit, + 'view_count': view_count, + 'duration': info['duration'] + } - # TODO: support choosing qualities + def _check_error(self, info): + error = info.get('error') + if info.get('error') is not None: + title = error['title'] + # See https://developer.dailymotion.com/api#access-error + if error.get('code') == 'DM007': + self.raise_geo_restricted(msg=title) + raise ExtractorError( + '%s said: %s' % (self.IE_NAME, title), expected=True) - for key in ['stream_h264_hd1080_url','stream_h264_hd_url', - 'stream_h264_hq_url','stream_h264_url', - 'stream_h264_ld_url']: - if info.get(key):#key in info and info[key]: - max_quality = key - self.to_screen(u'Using %s' % key) + 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' % error_to_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): + IE_NAME = 'dailymotion:playlist' + _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P[^/?#&]+)' + _MORE_PAGES_INDICATOR = r'(?s)
.*?.+?)/' - _MORE_PAGES_INDICATOR = r'' + if re.search(self._MORE_PAGES_INDICATOR, webpage) is None: + break def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) - playlist_id = mobj.group('id') - video_ids = [] + playlist_id = mobj.group('id') + webpage = self._download_webpage(url, playlist_id) - for pagenum in itertools.count(1): - webpage = self._download_webpage('https://www.dailymotion.com/playlist/%s/%s' % (playlist_id, pagenum), - playlist_id, u'Downloading page %s' % pagenum) + return { + '_type': 'playlist', + 'id': playlist_id, + 'title': self._og_search_title(webpage), + 'entries': self._extract_entries(playlist_id), + } - playlist_el = get_element_by_attribute(u'class', u'video_list', webpage) - video_ids.extend(re.findall(r'data-id="(.+?)" data-ext-id', playlist_el)) - if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None: - break +class DailymotionUserIE(DailymotionPlaylistIE): + IE_NAME = 'dailymotion:user' + _VALID_URL = r'https?://(?:www\.)?dailymotion\.[a-z]{2,3}/(?!(?:embed|swf|#|video|playlist)/)(?:(?:old/)?user/)?(?P[^/]+)' + _PAGE_TEMPLATE = 'http://www.dailymotion.com/user/%s/%s' + _TESTS = [{ + 'url': 'https://www.dailymotion.com/user/nqtv', + 'info_dict': { + 'id': 'nqtv', + 'title': 'Rémi Gaillard', + }, + 'playlist_mincount': 100, + }, { + 'url': 'http://www.dailymotion.com/user/UnderProject', + 'info_dict': { + 'id': 'UnderProject', + 'title': 'UnderProject', + }, + 'playlist_mincount': 1800, + 'expected_warnings': [ + 'Stopped at duplicated page', + ], + 'skip': 'Takes too long time', + }] + + 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), + } - entries = [self.url_result('http://www.dailymotion.com/video/%s' % video_id, 'Dailymotion') - for video_id in video_ids] - return {'_type': 'playlist', - 'id': playlist_id, - 'title': get_element_by_id(u'playlist_name', webpage), - 'entries': entries, - } + +class DailymotionCloudIE(DailymotionBaseInfoExtractor): + _VALID_URL_PREFIX = r'https?://api\.dmcloud\.net/(?:player/)?embed/' + _VALID_URL = r'%s[^/]+/(?P[^/?]+)' % _VALID_URL_PREFIX + _VALID_EMBED_URL = r'%s[^/]+/[^\'"]+' % _VALID_URL_PREFIX + + _TESTS = [{ + # From http://www.francetvinfo.fr/economie/entreprises/les-entreprises-familiales-le-secret-de-la-reussite_933271.html + # Tested at FranceTvInfo_2 + 'url': 'http://api.dmcloud.net/embed/4e7343f894a6f677b10006b4/556e03339473995ee145930c?auth=1464865870-0-jyhsm84b-ead4c701fb750cf9367bf4447167a3db&autoplay=1', + 'only_matching': True, + }, { + # http://www.francetvinfo.fr/societe/larguez-les-amarres-le-cobaturage-se-developpe_980101.html + 'url': 'http://api.dmcloud.net/player/embed/4e7343f894a6f677b10006b4/559545469473996d31429f06?auth=1467430263-0-90tglw2l-a3a4b64ed41efe48d7fccad85b8b8fda&autoplay=1', + 'only_matching': True, + }] + + @classmethod + def _extract_dmcloud_url(cls, webpage): + mobj = re.search(r']+src=[\'"](%s)[\'"]' % cls._VALID_EMBED_URL, webpage) + if mobj: + return mobj.group(1) + + mobj = re.search( + r']+id=[\'"]dmcloudUrlEmissionSelect[\'"][^>]+value=[\'"](%s)[\'"]' % cls._VALID_EMBED_URL, + webpage) + if mobj: + return mobj.group(1) + + def _real_extract(self, url): + video_id = self._match_id(url) + + webpage = self._download_webpage_no_ff(url, video_id) + + title = self._html_search_regex(r'([^>]+)', webpage, 'title') + + video_info = self._parse_json(self._search_regex( + r'var\s+info\s*=\s*([^;]+);', webpage, 'video info'), video_id) + + # TODO: parse ios_url, which is in fact a manifest + video_url = video_info['mp4_url'] + + return { + 'id': video_id, + 'url': video_url, + 'title': title, + 'thumbnail': video_info.get('thumbnail_url'), + }