X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/1931fa5f23ac75ea766517aa1b5ebf267ad3f97d..76d85602f8a22ca3817c3a86f4f0e8969c0b02a9:/youtube_dl/extractor/vimeo.py?ds=inline diff --git a/youtube_dl/extractor/vimeo.py b/youtube_dl/extractor/vimeo.py index 11c7d7e..ca716c8 100644 --- a/youtube_dl/extractor/vimeo.py +++ b/youtube_dl/extractor/vimeo.py @@ -6,25 +6,30 @@ import re import itertools from .common import InfoExtractor -from .subtitles import SubtitlesInfoExtractor -from ..utils import ( +from ..compat import ( compat_HTTPError, - compat_urllib_parse, compat_urllib_request, - clean_html, - get_element_by_attribute, + compat_urlparse, +) +from ..utils import ( + encode_dict, ExtractorError, + InAdvancePagedList, + int_or_none, RegexNotFoundError, + smuggle_url, std_headers, + unified_strdate, unsmuggle_url, urlencode_postdata, - int_or_none, + unescapeHTML, ) class VimeoBaseInfoExtractor(InfoExtractor): _NETRC_MACHINE = 'vimeo' _LOGIN_REQUIRED = False + _LOGIN_URL = 'https://vimeo.com/log_in' def _login(self): (username, password) = self._get_login_info() @@ -33,30 +38,40 @@ class VimeoBaseInfoExtractor(InfoExtractor): raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True) return self.report_login() - login_url = 'https://vimeo.com/log_in' - webpage = self._download_webpage(login_url, None, False) - token = self._search_regex(r'xsrft: \'(.*?)\'', webpage, 'login token') - data = urlencode_postdata({ + webpage = self._download_webpage(self._LOGIN_URL, None, False) + token, vuid = self._extract_xsrft_and_vuid(webpage) + data = urlencode_postdata(encode_dict({ + 'action': 'login', 'email': username, 'password': password, - 'action': 'login', 'service': 'vimeo', 'token': token, - }) - login_request = compat_urllib_request.Request(login_url, data) + })) + login_request = compat_urllib_request.Request(self._LOGIN_URL, data) login_request.add_header('Content-Type', 'application/x-www-form-urlencoded') - login_request.add_header('Cookie', 'xsrft=%s' % token) + login_request.add_header('Cookie', 'vuid=%s' % vuid) + login_request.add_header('Referer', self._LOGIN_URL) self._download_webpage(login_request, None, False, 'Wrong login info') + def _extract_xsrft_and_vuid(self, webpage): + xsrft = self._search_regex( + r'xsrft\s*[=:]\s*(?P["\'])(?P.+?)(?P=q)', + webpage, 'login token', group='xsrft') + vuid = self._search_regex( + r'["\']vuid["\']\s*:\s*(["\'])(?P.+?)\1', + webpage, 'vuid', group='vuid') + return xsrft, vuid + -class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): +class VimeoIE(VimeoBaseInfoExtractor): """Information extractor for vimeo.com.""" # _VALID_URL matches Vimeo URLs _VALID_URL = r'''(?x) - (?P(?:https?:)?//)? + https?:// (?:(?:www|(?Pplayer))\.)? vimeo(?Ppro)?\.com/ + (?!channels/[^/?#]+/?(?:$|[?#])|album/) (?:.*?/)? (?:(?:play_redirect_hls|moogaloop\.swf)\?clip_id=)? (?:videos?/)? @@ -70,12 +85,12 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): 'info_dict': { 'id': '56015672', 'ext': 'mp4', - "upload_date": "20121220", - "description": "This is a test case for youtube-dl.\nFor more information, see github.com/rg3/youtube-dl\nTest chars: \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550", - "uploader_id": "user7108434", - "uploader": "Filippo Valsorda", - "title": "youtube-dl test video - \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550", - "duration": 10, + 'title': "youtube-dl test video - \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550", + 'description': 'md5:2d3305bad981a06ff79f027f19865021', + 'upload_date': '20121220', + 'uploader_id': 'user7108434', + 'uploader': 'Filippo Valsorda', + 'duration': 10, }, }, { @@ -88,6 +103,7 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): 'uploader_id': 'openstreetmapus', 'uploader': 'OpenStreetMap US', 'title': 'Andy Allan - Putting the Carto into OpenStreetMap Cartography', + 'description': 'md5:fd69a7b8d8c34a4e1d2ec2e4afd6ec30', 'duration': 1595, }, }, @@ -102,6 +118,7 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): 'uploader': 'The BLN & Business of Software', 'uploader_id': 'theblnbusinessofsoftware', 'duration': 3610, + 'description': None, }, }, { @@ -116,6 +133,7 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): 'uploader_id': 'user18948128', 'uploader': 'Jaime Marquínez Ferrándiz', 'duration': 10, + 'description': 'This is "youtube-dl password protected test video" by Jaime Marquínez Ferrándiz on Vimeo, the home for high quality videos and the people\u2026', }, 'params': { 'videopassword': 'youtube-dl', @@ -133,12 +151,12 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): 'description': 'md5:8678b246399b070816b12313e8b4eb5c', 'uploader_id': 'atencio', 'uploader': 'Peter Atencio', + 'upload_date': '20130927', 'duration': 187, }, }, { 'url': 'http://vimeo.com/76979871', - 'md5': '3363dd6ffebe3784d56f4132317fd446', 'note': 'Video with subtitles', 'info_dict': { 'id': '76979871', @@ -151,43 +169,65 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): 'duration': 62, } }, + { + # from https://www.ouya.tv/game/Pier-Solar-and-the-Great-Architects/ + 'url': 'https://player.vimeo.com/video/98044508', + 'note': 'The js code contains assignments to the same variable as the config', + 'info_dict': { + 'id': '98044508', + 'ext': 'mp4', + 'title': 'Pier Solar OUYA Official Trailer', + 'uploader': 'Tulio Gonçalves', + 'uploader_id': 'user28849593', + }, + }, + { + 'url': 'https://vimeo.com/109815029', + 'note': 'Video not completely processed, "failed" seed status', + 'only_matching': True, + }, ] - @classmethod - def suitable(cls, url): - if VimeoChannelIE.suitable(url): - # Otherwise channel urls like http://vimeo.com/channels/31259 would - # match - return False - else: - return super(VimeoIE, cls).suitable(url) + @staticmethod + def _extract_vimeo_url(url, webpage): + # Look for embedded (iframe) Vimeo player + mobj = re.search( + r']+?src=(["\'])(?P(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage) + if mobj: + player_url = unescapeHTML(mobj.group('url')) + surl = smuggle_url(player_url, {'Referer': url}) + return surl + # Look for embedded (swf embed) Vimeo player + mobj = re.search( + r']+?src="((?:https?:)?//(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage) + if mobj: + return mobj.group(1) def _verify_video_password(self, url, video_id, webpage): password = self._downloader.params.get('videopassword', None) if password is None: - raise ExtractorError('This video is protected by a password, use the --video-password option') - token = self._search_regex(r'xsrft: \'(.*?)\'', webpage, 'login token') - data = compat_urllib_parse.urlencode({ + raise ExtractorError('This video is protected by a password, use the --video-password option', expected=True) + token, vuid = self._extract_xsrft_and_vuid(webpage) + data = urlencode_postdata(encode_dict({ 'password': password, 'token': token, - }) - # I didn't manage to use the password with https - if url.startswith('https'): - pass_url = url.replace('https', 'http') - else: - pass_url = url - password_request = compat_urllib_request.Request(pass_url + '/password', data) + })) + if url.startswith('http://'): + # vimeo only supports https now, but the user can give an http url + url = url.replace('http://', 'https://') + password_request = compat_urllib_request.Request(url + '/password', data) password_request.add_header('Content-Type', 'application/x-www-form-urlencoded') - password_request.add_header('Cookie', 'xsrft=%s' % token) - self._download_webpage(password_request, video_id, - 'Verifying the password', - 'Wrong password') + password_request.add_header('Cookie', 'clip_test2=1; vuid=%s' % vuid) + password_request.add_header('Referer', url) + return self._download_webpage( + password_request, video_id, + 'Verifying the password', 'Wrong password') def _verify_player_video_password(self, url, video_id): password = self._downloader.params.get('videopassword', None) if password is None: raise ExtractorError('This video is protected by a password, use the --video-password option') - data = compat_urllib_parse.urlencode({'password': password}) + data = urlencode_postdata(encode_dict({'password': password})) pass_url = url + '/check-password' password_request = compat_urllib_request.Request(pass_url, data) password_request.add_header('Content-Type', 'application/x-www-form-urlencoded') @@ -205,12 +245,17 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): if data is not None: headers = headers.copy() headers.update(data) + if 'Referer' not in headers: + headers['Referer'] = url # Extract ID from URL mobj = re.match(self._VALID_URL, url) video_id = mobj.group('id') + orig_url = url if mobj.group('pro') or mobj.group('player'): - url = 'http://player.vimeo.com/video/' + video_id + url = 'https://player.vimeo.com/video/' + video_id + else: + url = 'https://vimeo.com/' + video_id # Retrieve video webpage to extract further information request = compat_urllib_request.Request(url, None, headers) @@ -232,11 +277,31 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): # and latter we extract those that are Vimeo specific. self.report_extraction(video_id) + vimeo_config = self._search_regex( + r'vimeo\.config\s*=\s*(?:({.+?})|_extend\([^,]+,\s+({.+?})\));', webpage, + 'vimeo config', default=None) + if vimeo_config: + seed_status = self._parse_json(vimeo_config, video_id).get('seed_status', {}) + if seed_status.get('state') == 'failed': + raise ExtractorError( + '%s said: %s' % (self.IE_NAME, seed_status['title']), + expected=True) + # Extract the config JSON try: try: config_url = self._html_search_regex( - r' data-config-url="(.+?)"', webpage, 'config URL') + r' data-config-url="(.+?)"', webpage, + 'config URL', default=None) + if not config_url: + # Sometimes new react-based page is served instead of old one that require + # different config URL extraction approach (see + # https://github.com/rg3/youtube-dl/pull/7209) + vimeo_clip_page_config = self._search_regex( + r'vimeo\.clip_page_config\s*=\s*({.+?});', webpage, + 'vimeo clip page config') + config_url = self._parse_json( + vimeo_clip_page_config, video_id)['player']['config_url'] config_json = self._download_webpage(config_url, video_id) config = json.loads(config_json) except RegexNotFoundError: @@ -244,19 +309,22 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): # We try to find out to which variable is assigned the config dic m_variable_name = re.search('(\w)\.video\.id', webpage) if m_variable_name is not None: - config_re = r'%s=({.+?});' % re.escape(m_variable_name.group(1)) + config_re = r'%s=({[^}].+?});' % re.escape(m_variable_name.group(1)) else: config_re = [r' = {config:({.+?}),assets:', r'(?:[abc])=({.+?});'] config = self._search_regex(config_re, webpage, 'info section', - flags=re.DOTALL) + flags=re.DOTALL) config = json.loads(config) except Exception as e: if re.search('The creator of this video has not given you permission to embed it on this domain.', webpage): raise ExtractorError('The author has restricted the access to this video, try with the "--referer" option') - if re.search(']+?id="pw_form"', webpage) is not None: + if re.search(r']+?id="pw_form"', webpage) is not None: + if data and '_video_password_verified' in data: + raise ExtractorError('video password verification failed!') self._verify_video_password(url, video_id, webpage) - return self._real_extract(url) + return self._real_extract( + smuggle_url(url, {'_video_password_verified': 'verified'})) else: raise ExtractorError('Unable to extract info section', cause=e) @@ -279,27 +347,32 @@ class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor): _, video_thumbnail = sorted((int(width if width.isdigit() else 0), t_url) for (width, t_url) in video_thumbs.items())[-1] # Extract video description - video_description = None - try: - video_description = get_element_by_attribute("class", "description_wrapper", webpage) - if video_description: - video_description = clean_html(video_description) - except AssertionError as err: - # On some pages like (http://player.vimeo.com/video/54469442) the - # html tags are not closed, python 2.6 cannot handle it - if err.args[0] == 'we should not get here!': - pass - else: - raise + + video_description = self._html_search_regex( + r'(?s)]*>(.*?)', + webpage, 'description', default=None) + if not video_description: + video_description = self._html_search_meta( + 'description', webpage, default=None) + if not video_description and mobj.group('pro'): + orig_webpage = self._download_webpage( + orig_url, video_id, + note='Downloading webpage for description', + fatal=False) + if orig_webpage: + video_description = self._html_search_meta( + 'description', orig_webpage, default=None) + if not video_description and not mobj.group('player'): + self._downloader.report_warning('Cannot find video description') # Extract video duration video_duration = int_or_none(config["video"].get("duration")) # Extract upload date video_upload_date = None - mobj = re.search(r'[^/]+)/?(\?.*)?$' + _VALID_URL = r'https://vimeo\.com/channels/(?P[^/?#]+)/?(?:$|[?#])' _MORE_PAGES_INDICATOR = r']+?title="(.*?)"' + _TESTS = [{ + 'url': 'https://vimeo.com/channels/tributes', + 'info_dict': { + 'id': 'tributes', + 'title': 'Vimeo Tributes', + }, + 'playlist_mincount': 25, + }] def _page_url(self, base_url, pagenum): return '%s/videos/page:%d/' % (base_url, pagenum) def _extract_list_title(self, webpage): - return self._html_search_regex(self._TITLE_RE, webpage, 'list title') + return self._TITLE or self._html_search_regex(self._TITLE_RE, webpage, 'list title') + + def _login_list_password(self, page_url, list_id, webpage): + login_form = self._search_regex( + r'(?s)]+?id="pw_form"(.*?)', + webpage, 'login form', default=None) + if not login_form: + return webpage + + password = self._downloader.params.get('videopassword', None) + if password is None: + raise ExtractorError('This album is protected by a password, use the --video-password option', expected=True) + fields = self._hidden_inputs(login_form) + token, vuid = self._extract_xsrft_and_vuid(webpage) + fields['token'] = token + fields['password'] = password + post = urlencode_postdata(encode_dict(fields)) + password_path = self._search_regex( + r'action="([^"]+)"', login_form, 'password URL') + password_url = compat_urlparse.urljoin(page_url, password_path) + password_request = compat_urllib_request.Request(password_url, post) + password_request.add_header('Content-type', 'application/x-www-form-urlencoded') + password_request.add_header('Cookie', 'vuid=%s' % vuid) + self._set_cookie('vimeo.com', 'xsrft', token) + + return self._download_webpage( + password_request, list_id, + 'Verifying the password', 'Wrong password') def _extract_videos(self, list_id, base_url): video_ids = [] for pagenum in itertools.count(1): + page_url = self._page_url(base_url, pagenum) webpage = self._download_webpage( - self._page_url(base_url, pagenum), list_id, + page_url, list_id, 'Downloading page %s' % pagenum) + + if pagenum == 1: + webpage = self._login_list_password(page_url, list_id, webpage) + video_ids.extend(re.findall(r'id="clip_(\d+?)"', webpage)) if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None: break - entries = [self.url_result('http://vimeo.com/%s' % video_id, 'Vimeo') + entries = [self.url_result('https://vimeo.com/%s' % video_id, 'Vimeo') for video_id in video_ids] return {'_type': 'playlist', 'id': list_id, @@ -409,43 +527,71 @@ class VimeoChannelIE(InfoExtractor): def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) channel_id = mobj.group('id') - return self._extract_videos(channel_id, 'http://vimeo.com/channels/%s' % channel_id) + return self._extract_videos(channel_id, 'https://vimeo.com/channels/%s' % channel_id) class VimeoUserIE(VimeoChannelIE): IE_NAME = 'vimeo:user' - _VALID_URL = r'(?:https?://)?vimeo\.com/(?P[^/]+)(?:/videos|[#?]|$)' + _VALID_URL = r'https://vimeo\.com/(?!(?:[0-9]+|watchlater)(?:$|[?#/]))(?P[^/]+)(?:/videos|[#?]|$)' _TITLE_RE = r']+?class="user">([^<>]+?)' - - @classmethod - def suitable(cls, url): - if VimeoChannelIE.suitable(url) or VimeoIE.suitable(url) or VimeoAlbumIE.suitable(url) or VimeoGroupsIE.suitable(url): - return False - return super(VimeoUserIE, cls).suitable(url) + _TESTS = [{ + 'url': 'https://vimeo.com/nkistudio/videos', + 'info_dict': { + 'title': 'Nki', + 'id': 'nkistudio', + }, + 'playlist_mincount': 66, + }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) name = mobj.group('name') - return self._extract_videos(name, 'http://vimeo.com/%s' % name) + return self._extract_videos(name, 'https://vimeo.com/%s' % name) class VimeoAlbumIE(VimeoChannelIE): IE_NAME = 'vimeo:album' - _VALID_URL = r'(?:https?://)?vimeo\.com/album/(?P\d+)' + _VALID_URL = r'https://vimeo\.com/album/(?P\d+)' _TITLE_RE = r'
\n\s*

(.*?)

' + _TESTS = [{ + 'url': 'https://vimeo.com/album/2632481', + 'info_dict': { + 'id': '2632481', + 'title': 'Staff Favorites: November 2013', + }, + 'playlist_mincount': 13, + }, { + 'note': 'Password-protected album', + 'url': 'https://vimeo.com/album/3253534', + 'info_dict': { + 'title': 'test', + 'id': '3253534', + }, + 'playlist_count': 1, + 'params': { + 'videopassword': 'youtube-dl', + } + }] def _page_url(self, base_url, pagenum): return '%s/page:%d/' % (base_url, pagenum) def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - album_id = mobj.group('id') - return self._extract_videos(album_id, 'http://vimeo.com/album/%s' % album_id) + album_id = self._match_id(url) + return self._extract_videos(album_id, 'https://vimeo.com/album/%s' % album_id) class VimeoGroupsIE(VimeoAlbumIE): IE_NAME = 'vimeo:group' - _VALID_URL = r'(?:https?://)?vimeo\.com/groups/(?P[^/]+)' + _VALID_URL = r'https://vimeo\.com/groups/(?P[^/]+)' + _TESTS = [{ + 'url': 'https://vimeo.com/groups/rolexawards', + 'info_dict': { + 'id': 'rolexawards', + 'title': 'Rolex Awards for Enterprise', + }, + 'playlist_mincount': 73, + }] def _extract_list_title(self, webpage): return self._og_search_title(webpage) @@ -453,22 +599,35 @@ class VimeoGroupsIE(VimeoAlbumIE): def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) name = mobj.group('name') - return self._extract_videos(name, 'http://vimeo.com/groups/%s' % name) + return self._extract_videos(name, 'https://vimeo.com/groups/%s' % name) class VimeoReviewIE(InfoExtractor): IE_NAME = 'vimeo:review' IE_DESC = 'Review pages on vimeo' - _VALID_URL = r'(?:https?://)?vimeo\.com/[^/]+/review/(?P[^/]+)' - _TEST = { + _VALID_URL = r'https://vimeo\.com/[^/]+/review/(?P[^/]+)' + _TESTS = [{ 'url': 'https://vimeo.com/user21297594/review/75524534/3c257a1b5d', - 'file': '75524534.mp4', 'md5': 'c507a72f780cacc12b2248bb4006d253', 'info_dict': { + 'id': '75524534', + 'ext': 'mp4', 'title': "DICK HARDWICK 'Comedian'", 'uploader': 'Richard Hardwick', } - } + }, { + 'note': 'video player needs Referer', + 'url': 'https://vimeo.com/user22258446/review/91613211/13f927e053', + 'md5': '6295fdab8f4bf6a002d058b2c6dce276', + 'info_dict': { + 'id': '91613211', + 'ext': 'mp4', + 'title': 're:(?i)^Death by dogma versus assembling agile . Sander Hoogendoorn', + 'uploader': 'DevWeek Events', + 'duration': 2773, + 'thumbnail': 're:^https?://.*\.jpg$', + } + }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -477,12 +636,16 @@ class VimeoReviewIE(InfoExtractor): return self.url_result(player_url, 'Vimeo', video_id) -class VimeoWatchLaterIE(VimeoBaseInfoExtractor, VimeoChannelIE): +class VimeoWatchLaterIE(VimeoChannelIE): IE_NAME = 'vimeo:watchlater' IE_DESC = 'Vimeo watch later list, "vimeowatchlater" keyword (requires authentication)' - _VALID_URL = r'https?://vimeo\.com/home/watchlater|:vimeowatchlater' + _VALID_URL = r'https://vimeo\.com/(?:home/)?watchlater|:vimeowatchlater' + _TITLE = 'Watch Later' _LOGIN_REQUIRED = True - _TITLE_RE = r'href="/home/watchlater".*?>(.*?)<' + _TESTS = [{ + 'url': 'https://vimeo.com/watchlater', + 'only_matching': True, + }] def _real_initialize(self): self._login() @@ -496,4 +659,60 @@ class VimeoWatchLaterIE(VimeoBaseInfoExtractor, VimeoChannelIE): return request def _real_extract(self, url): - return self._extract_videos('watchlater', 'https://vimeo.com/home/watchlater') + return self._extract_videos('watchlater', 'https://vimeo.com/watchlater') + + +class VimeoLikesIE(InfoExtractor): + _VALID_URL = r'https://(?:www\.)?vimeo\.com/user(?P[0-9]+)/likes/?(?:$|[?#]|sort:)' + IE_NAME = 'vimeo:likes' + IE_DESC = 'Vimeo user likes' + _TEST = { + 'url': 'https://vimeo.com/user755559/likes/', + 'playlist_mincount': 293, + "info_dict": { + 'id': 'user755559_likes', + "description": "See all the videos urza likes", + "title": 'Videos urza likes', + }, + } + + def _real_extract(self, url): + user_id = self._match_id(url) + webpage = self._download_webpage(url, user_id) + page_count = self._int( + self._search_regex( + r'''(?x)
  • + .*?
    • \s* + ''', webpage, 'page count'), + 'page count', fatal=True) + PAGE_SIZE = 12 + title = self._html_search_regex( + r'(?s)

    (.+?)

    ', webpage, 'title', fatal=False) + description = self._html_search_meta('description', webpage) + + def _get_page(idx): + page_url = 'https://vimeo.com/user%s/likes/page:%d/sort:date' % ( + user_id, idx + 1) + webpage = self._download_webpage( + page_url, user_id, + note='Downloading page %d/%d' % (idx + 1, page_count)) + video_list = self._search_regex( + r'(?s)
      ]*>(.*?)
    ', + webpage, 'video content') + paths = re.findall( + r']*>\s*