X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/ced7488f6d3a519b2c1b1cbd31048743fb8285bd..1918d37666ea9185c49ca84e64f58e4ff2a422c9:/youtube_dl/extractor/vimeo.py diff --git a/youtube_dl/extractor/vimeo.py b/youtube_dl/extractor/vimeo.py index 7e854f3..6af7056 100644 --- a/youtube_dl/extractor/vimeo.py +++ b/youtube_dl/extractor/vimeo.py @@ -1,4 +1,4 @@ -# encoding: utf-8 +# coding: utf-8 from __future__ import unicode_literals import json @@ -21,12 +21,12 @@ from ..utils import ( sanitized_Request, smuggle_url, std_headers, - unified_strdate, + try_get, + unified_timestamp, unsmuggle_url, urlencode_postdata, unescapeHTML, parse_filesize, - try_get, ) @@ -92,29 +92,30 @@ class VimeoBaseInfoExtractor(InfoExtractor): def _vimeo_sort_formats(self, formats): # Bitrates are completely broken. Single m3u8 may contain entries in kbps and bps # at the same time without actual units specified. This lead to wrong sorting. - self._sort_formats(formats, field_preference=('preference', 'height', 'width', 'fps', 'format_id')) + self._sort_formats(formats, field_preference=('preference', 'height', 'width', 'fps', 'tbr', 'format_id')) def _parse_config(self, config, video_id): + video_data = config['video'] # Extract title - video_title = config['video']['title'] + video_title = video_data['title'] # Extract uploader, uploader_url and uploader_id - video_uploader = config['video'].get('owner', {}).get('name') - video_uploader_url = config['video'].get('owner', {}).get('url') + video_uploader = video_data.get('owner', {}).get('name') + video_uploader_url = video_data.get('owner', {}).get('url') video_uploader_id = video_uploader_url.split('/')[-1] if video_uploader_url else None # Extract video thumbnail - video_thumbnail = config['video'].get('thumbnail') + video_thumbnail = video_data.get('thumbnail') if video_thumbnail is None: - video_thumbs = config['video'].get('thumbs') + video_thumbs = video_data.get('thumbs') if video_thumbs and isinstance(video_thumbs, dict): _, video_thumbnail = sorted((int(width if width.isdigit() else 0), t_url) for (width, t_url) in video_thumbs.items())[-1] # Extract video duration - video_duration = int_or_none(config['video'].get('duration')) + video_duration = int_or_none(video_data.get('duration')) formats = [] - config_files = config['video'].get('files') or config['request'].get('files', {}) + config_files = video_data.get('files') or config['request'].get('files', {}) for f in config_files.get('progressive', []): video_url = f.get('url') if not video_url: @@ -127,10 +128,39 @@ class VimeoBaseInfoExtractor(InfoExtractor): 'fps': int_or_none(f.get('fps')), 'tbr': int_or_none(f.get('bitrate')), }) - m3u8_url = config_files.get('hls', {}).get('url') - if m3u8_url: - formats.extend(self._extract_m3u8_formats( - m3u8_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)) + + for files_type in ('hls', 'dash'): + for cdn_name, cdn_data in config_files.get(files_type, {}).get('cdns', {}).items(): + manifest_url = cdn_data.get('url') + if not manifest_url: + continue + format_id = '%s-%s' % (files_type, cdn_name) + if files_type == 'hls': + formats.extend(self._extract_m3u8_formats( + manifest_url, video_id, 'mp4', + 'm3u8_native', m3u8_id=format_id, + note='Downloading %s m3u8 information' % cdn_name, + fatal=False)) + elif files_type == 'dash': + mpd_pattern = r'/%s/(?:sep/)?video/' % video_id + mpd_manifest_urls = [] + if re.search(mpd_pattern, manifest_url): + for suffix, repl in (('', 'video'), ('_sep', 'sep/video')): + mpd_manifest_urls.append((format_id + suffix, re.sub( + mpd_pattern, '/%s/%s/' % (video_id, repl), manifest_url))) + else: + mpd_manifest_urls = [(format_id, manifest_url)] + for f_id, m_url in mpd_manifest_urls: + mpd_formats = self._extract_mpd_formats( + m_url.replace('/master.json', '/master.mpd'), video_id, f_id, + 'Downloading %s MPD information' % cdn_name, + fatal=False) + for f in mpd_formats: + if f.get('vcodec') == 'none': + f['preference'] = -50 + elif f.get('acodec') == 'none': + f['preference'] = -40 + formats.extend(mpd_formats) subtitles = {} text_tracks = config['request'].get('text_tracks') @@ -189,11 +219,13 @@ class VimeoIE(VimeoBaseInfoExtractor): 'ext': 'mp4', 'title': "youtube-dl test video - \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550", 'description': 'md5:2d3305bad981a06ff79f027f19865021', + 'timestamp': 1355990239, 'upload_date': '20121220', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/user7108434', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/user7108434', 'uploader_id': 'user7108434', 'uploader': 'Filippo Valsorda', 'duration': 10, + 'license': 'by-sa', }, }, { @@ -203,7 +235,7 @@ class VimeoIE(VimeoBaseInfoExtractor): 'info_dict': { 'id': '68093876', 'ext': 'mp4', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/openstreetmapus', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/openstreetmapus', 'uploader_id': 'openstreetmapus', 'uploader': 'OpenStreetMap US', 'title': 'Andy Allan - Putting the Carto into OpenStreetMap Cartography', @@ -220,7 +252,7 @@ class VimeoIE(VimeoBaseInfoExtractor): 'ext': 'mp4', 'title': 'Kathy Sierra: Building the minimum Badass User, Business of Software 2012', 'uploader': 'The BLN & Business of Software', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/theblnbusinessofsoftware', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/theblnbusinessofsoftware', 'uploader_id': 'theblnbusinessofsoftware', 'duration': 3610, 'description': None, @@ -234,12 +266,13 @@ class VimeoIE(VimeoBaseInfoExtractor): 'id': '68375962', 'ext': 'mp4', 'title': 'youtube-dl password protected test video', + 'timestamp': 1371200155, 'upload_date': '20130614', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/user18948128', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/user18948128', 'uploader_id': 'user18948128', 'uploader': 'Jaime Marquínez Ferrándiz', 'duration': 10, - 'description': 'This is "youtube-dl password protected test video" by on Vimeo, the home for high quality videos and the people who love them.', + 'description': 'md5:dca3ea23adb29ee387127bc4ddfce63f', }, 'params': { 'videopassword': 'youtube-dl', @@ -253,10 +286,11 @@ class VimeoIE(VimeoBaseInfoExtractor): 'ext': 'mp4', 'title': 'Key & Peele: Terrorist Interrogation', 'description': 'md5:8678b246399b070816b12313e8b4eb5c', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/atencio', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/atencio', 'uploader_id': 'atencio', 'uploader': 'Peter Atencio', - 'upload_date': '20130927', + 'timestamp': 1380339469, + 'upload_date': '20130928', 'duration': 187, }, }, @@ -268,8 +302,9 @@ class VimeoIE(VimeoBaseInfoExtractor): 'ext': 'mp4', 'title': 'The New Vimeo Player (You Know, For Videos)', 'description': 'md5:2ec900bf97c3f389378a96aee11260ea', + 'timestamp': 1381846109, 'upload_date': '20131015', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/staff', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/staff', 'uploader_id': 'staff', 'uploader': 'Vimeo Staff', 'duration': 62, @@ -284,21 +319,22 @@ class VimeoIE(VimeoBaseInfoExtractor): 'ext': 'mp4', 'title': 'Pier Solar OUYA Official Trailer', 'uploader': 'Tulio Gonçalves', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/user28849593', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/user28849593', 'uploader_id': 'user28849593', }, }, { # contains original format 'url': 'https://vimeo.com/33951933', - 'md5': '2d9f5475e0537f013d0073e812ab89e6', + 'md5': '53c688fa95a55bf4b7293d37a89c5c53', 'info_dict': { 'id': '33951933', 'ext': 'mp4', 'title': 'FOX CLASSICS - Forever Classic ID - A Full Minute', 'uploader': 'The DMCI', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/dmci', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/dmci', 'uploader_id': 'dmci', + 'timestamp': 1324343742, 'upload_date': '20111220', 'description': 'md5:ae23671e82d05415868f7ad1aec21147', }, @@ -309,11 +345,12 @@ class VimeoIE(VimeoBaseInfoExtractor): 'url': 'https://vimeo.com/channels/tributes/6213729', 'info_dict': { 'id': '6213729', - 'ext': 'mp4', + 'ext': 'mov', 'title': 'Vimeo Tribute: The Shining', 'uploader': 'Casey Donahue', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/caseydonahue', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/caseydonahue', 'uploader_id': 'caseydonahue', + 'timestamp': 1250886430, 'upload_date': '20090821', 'description': 'md5:bdbf314014e58713e6e5b66eb252f4a6', }, @@ -322,6 +359,22 @@ class VimeoIE(VimeoBaseInfoExtractor): }, 'expected_warnings': ['Unable to download JSON metadata'], }, + { + # redirects to ondemand extractor and should be passed through it + # for successful extraction + 'url': 'https://vimeo.com/73445910', + 'info_dict': { + 'id': '73445910', + 'ext': 'mp4', + 'title': 'The Reluctant Revolutionary', + 'uploader': '10Ft Films', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/tenfootfilms', + 'uploader_id': 'tenfootfilms', + }, + 'params': { + 'skip_download': True, + }, + }, { 'url': 'http://vimeo.com/moogaloop.swf?clip_id=2539741', 'only_matching': True, @@ -351,24 +404,32 @@ class VimeoIE(VimeoBaseInfoExtractor): ] @staticmethod - def _extract_vimeo_url(url, webpage): + def _smuggle_referrer(url, referrer_url): + return smuggle_url(url, {'http_headers': {'Referer': referrer_url}}) + + @staticmethod + def _extract_urls(url, webpage): + urls = [] # 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, {'http_headers': {'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) - # Look more for non-standard embedded Vimeo player - mobj = re.search( - r']+src=(?P[\'"])(?P(?:https?:)?//(?:www\.)?vimeo\.com/[0-9]+)(?P=q1)', webpage) - if mobj: - return mobj.group('url') + for mobj in re.finditer( + r']+?src=(["\'])(?P(?:https?:)?//player\.vimeo\.com/video/\d+.*?)\1', + webpage): + urls.append(VimeoIE._smuggle_referrer(unescapeHTML(mobj.group('url')), url)) + PLAIN_EMBED_RE = ( + # Look for embedded (swf embed) Vimeo player + r']+?src=(["\'])(?P(?:https?:)?//(?:www\.)?vimeo\.com/moogaloop\.swf.+?)\1', + # Look more for non-standard embedded Vimeo player + r']+src=(["\'])(?P(?:https?:)?//(?:www\.)?vimeo\.com/[0-9]+)\1', + ) + for embed_re in PLAIN_EMBED_RE: + for mobj in re.finditer(embed_re, webpage): + urls.append(mobj.group('url')) + return urls + + @staticmethod + def _extract_url(url, webpage): + urls = VimeoIE._extract_urls(url, webpage) + return urls[0] if urls else None def _verify_player_video_password(self, url, video_id): password = self._downloader.params.get('videopassword') @@ -406,7 +467,13 @@ class VimeoIE(VimeoBaseInfoExtractor): # Retrieve video webpage to extract further information request = sanitized_Request(url, headers=headers) try: - webpage = self._download_webpage(request, video_id) + webpage, urlh = self._download_webpage_handle(request, video_id) + redirect_url = compat_str(urlh.geturl()) + # Some URLs redirect to ondemand can't be extracted with + # this extractor right away thus should be passed through + # ondemand extractor (e.g. https://vimeo.com/73445910) + if VimeoOndemandIE.suitable(redirect_url): + return self.url_result(redirect_url, VimeoOndemandIE.ie_key()) except ExtractorError as ee: if isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 403: errmsg = ee.cause.read() @@ -433,6 +500,9 @@ class VimeoIE(VimeoBaseInfoExtractor): '%s said: %s' % (self.IE_NAME, seed_status['title']), expected=True) + cc_license = None + timestamp = None + # Extract the config JSON try: try: @@ -446,14 +516,18 @@ class VimeoIE(VimeoBaseInfoExtractor): 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'] + page_config = self._parse_json(vimeo_clip_page_config, video_id) + config_url = page_config['player']['config_url'] + cc_license = page_config.get('cc_license') + timestamp = try_get( + page_config, lambda x: x['clip']['uploaded_on'], + compat_str) config_json = self._download_webpage(config_url, video_id) config = json.loads(config_json) except RegexNotFoundError: # For pro videos or player.vimeo.com urls # We try to find out to which variable is assigned the config dic - m_variable_name = re.search('(\w)\.video\.id', webpage) + m_variable_name = re.search(r'(\w)\.video\.id', webpage) if m_variable_name is not None: config_re = r'%s=({[^}].+?});' % re.escape(m_variable_name.group(1)) else: @@ -468,15 +542,15 @@ class VimeoIE(VimeoBaseInfoExtractor): if re.search(r']+?id="pw_form"', webpage) is not None: if '_video_password_verified' in data: raise ExtractorError('video password verification failed!') - self._verify_video_password(url, video_id, webpage) + self._verify_video_password(redirect_url, video_id, webpage) return self._real_extract( - smuggle_url(url, {'_video_password_verified': 'verified'})) + smuggle_url(redirect_url, {'_video_password_verified': 'verified'})) else: raise ExtractorError('Unable to extract info section', cause=e) else: if config.get('view') == 4: - config = self._verify_player_video_password(url, video_id) + config = self._verify_player_video_password(redirect_url, video_id) def is_rented(): if '>You rented this title.<' in webpage: @@ -516,10 +590,10 @@ class VimeoIE(VimeoBaseInfoExtractor): self._downloader.report_warning('Cannot find video description') # Extract upload date - video_upload_date = None - mobj = re.search(r']+datetime="([^"]+)"', webpage) - if mobj is not None: - video_upload_date = unified_strdate(mobj.group(1)) + if not timestamp: + timestamp = self._search_regex( + r']+datetime="([^"]+)"', webpage, + 'timestamp', default=None) try: view_count = int(self._search_regex(r'UserPlays:(\d+)', webpage, 'view count')) @@ -542,7 +616,10 @@ class VimeoIE(VimeoBaseInfoExtractor): if download_url and not source_file.get('is_cold') and not source_file.get('is_defrosting'): source_name = source_file.get('public_name', 'Original') if self._is_valid_url(download_url, video_id, '%s video' % source_name): - ext = source_file.get('extension', determine_ext(download_url)).lower() + ext = (try_get( + source_file, lambda x: x['extension'], + compat_str) or determine_ext( + download_url, None) or 'mp4').lower() formats.append({ 'url': download_url, 'ext': ext, @@ -556,15 +633,22 @@ class VimeoIE(VimeoBaseInfoExtractor): info_dict = self._parse_config(config, video_id) formats.extend(info_dict['formats']) self._vimeo_sort_formats(formats) + + if not cc_license: + cc_license = self._search_regex( + r']+rel=["\']license["\'][^>]+href=(["\'])(?P(?:(?!\1).)+)\1', + webpage, 'license', default=None, group='license') + info_dict.update({ 'id': video_id, 'formats': formats, - 'upload_date': video_upload_date, + 'timestamp': unified_timestamp(timestamp), 'description': video_description, 'webpage_url': url, 'view_count': view_count, 'like_count': like_count, 'comment_count': comment_count, + 'license': cc_license, }) return info_dict @@ -582,9 +666,26 @@ class VimeoOndemandIE(VimeoBaseInfoExtractor): 'ext': 'mp4', 'title': 'המעבדה - במאי יותם פלדמן', 'uploader': 'גם סרטים', - 'uploader_url': 're:https?://(?:www\.)?vimeo\.com/gumfilms', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/gumfilms', 'uploader_id': 'gumfilms', }, + 'params': { + 'format': 'best[protocol=https]', + }, + }, { + # requires Referer to be passed along with og:video:url + 'url': 'https://vimeo.com/ondemand/36938/126682985', + 'info_dict': { + 'id': '126682985', + 'ext': 'mp4', + 'title': 'Rävlock, rätt läte på rätt plats', + 'uploader': 'Lindroth & Norin', + 'uploader_url': r're:https?://(?:www\.)?vimeo\.com/user14430847', + 'uploader_id': 'user14430847', + }, + 'params': { + 'skip_download': True, + }, }, { 'url': 'https://vimeo.com/ondemand/nazmaalik', 'only_matching': True, @@ -599,7 +700,12 @@ class VimeoOndemandIE(VimeoBaseInfoExtractor): def _real_extract(self, url): video_id = self._match_id(url) webpage = self._download_webpage(url, video_id) - return self.url_result(self._og_search_video_url(webpage), VimeoIE.ie_key()) + return self.url_result( + # Some videos require Referer to be passed along with og:video:url + # similarly to generic vimeo embeds (e.g. + # https://vimeo.com/ondemand/36938/126682985). + VimeoIE._smuggle_referrer(self._og_search_video_url(webpage), url), + VimeoIE.ie_key()) class VimeoChannelIE(VimeoBaseInfoExtractor): @@ -664,12 +770,12 @@ class VimeoChannelIE(VimeoBaseInfoExtractor): # Try extracting href first since not all videos are available via # short https://vimeo.com/id URL (e.g. https://vimeo.com/channels/tributes/6213729) clips = re.findall( - r'id="clip_(\d+)"[^>]*>\s*]+href="(/(?:[^/]+/)*\1)', webpage) + r'id="clip_(\d+)"[^>]*>\s*]+href="(/(?:[^/]+/)*\1)(?:[^>]+\btitle="([^"]+)")?', webpage) if clips: - for video_id, video_url in clips: + for video_id, video_url, video_title in clips: yield self.url_result( compat_urlparse.urljoin(base_url, video_url), - VimeoIE.ie_key(), video_id=video_id) + VimeoIE.ie_key(), video_id=video_id, video_title=video_title) # More relaxed fallback else: for video_id in re.findall(r'id=["\']clip_(\d+)', webpage): @@ -794,7 +900,7 @@ class VimeoReviewIE(VimeoBaseInfoExtractor): 'title': 're:(?i)^Death by dogma versus assembling agile . Sander Hoogendoorn', 'uploader': 'DevWeek Events', 'duration': 2773, - 'thumbnail': 're:^https?://.*\.jpg$', + 'thumbnail': r're:^https?://.*\.jpg$', 'uploader_id': 'user22258446', } }, { @@ -810,6 +916,7 @@ class VimeoReviewIE(VimeoBaseInfoExtractor): 'params': { 'videopassword': 'holygrail', }, + 'skip': 'video gone', }] def _real_initialize(self): @@ -818,8 +925,13 @@ class VimeoReviewIE(VimeoBaseInfoExtractor): def _get_config_url(self, webpage_url, video_id, video_password_verified=False): webpage = self._download_webpage(webpage_url, video_id) config_url = self._html_search_regex( - r'data-config-url="([^"]+)"', webpage, 'config URL', - default=NO_DEFAULT if video_password_verified else None) + r'data-config-url=(["\'])(?P(?:(?!\1).)+)\1', webpage, + 'config URL', default=None, group='url') + if not config_url: + data = self._parse_json(self._search_regex( + r'window\s*=\s*_extend\(window,\s*({.+?})\);', webpage, 'data', + default=NO_DEFAULT if video_password_verified else '{}'), video_id) + config_url = data.get('vimeo_esi', {}).get('config', {}).get('configUrl') if config_url is None: self._verify_video_password(webpage_url, video_id, webpage) config_url = self._get_config_url(