X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/9dc487f48b50767cf540fa36c3de2c386fd74c04..0626de37fa3331c654d4affae363132a0d766ffb:/youtube_dl/extractor/kaltura.py?ds=inline diff --git a/youtube_dl/extractor/kaltura.py b/youtube_dl/extractor/kaltura.py index a65697f..49d1346 100644 --- a/youtube_dl/extractor/kaltura.py +++ b/youtube_dl/extractor/kaltura.py @@ -6,7 +6,6 @@ import base64 from .common import InfoExtractor from ..compat import ( - compat_urllib_parse_urlencode, compat_urlparse, compat_parse_qs, ) @@ -15,6 +14,7 @@ from ..utils import ( ExtractorError, int_or_none, unsmuggle_url, + smuggle_url, ) @@ -23,18 +23,25 @@ class KalturaIE(InfoExtractor): (?: kaltura:(?P\d+):(?P[0-9a-z_]+)| https?:// - (:?(?:www|cdnapi(?:sec)?)\.)?kaltura\.com/ + (:?(?:www|cdnapi(?:sec)?)\.)?kaltura\.com(?::\d+)?/ (?: (?: # flash player - index\.php/kwidget| + index\.php/(?:kwidget|extwidget/preview)| # html5 player html5/html5lib/[^/]+/mwEmbedFrame\.php ) )(?:/(?P[^?]+))?(?:\?(?P.*))? ) ''' - _API_BASE = 'http://cdnapi.kaltura.com/api_v3/index.php?' + _SERVICE_URL = 'http://cdnapi.kaltura.com' + _SERVICE_BASE = '/api_v3/index.php' + # See https://github.com/kaltura/server/blob/master/plugins/content/caption/base/lib/model/enums/CaptionType.php + _CAPTION_TYPES = { + 1: 'srt', + 2: 'ttml', + 3: 'vtt', + } _TESTS = [ { 'url': 'kaltura:269692:1_1jc2y3e4', @@ -61,19 +68,117 @@ class KalturaIE(InfoExtractor): { 'url': 'https://cdnapisec.kaltura.com/html5/html5lib/v2.30.2/mwEmbedFrame.php/p/1337/uiconf_id/20540612/entry_id/1_sf5ovm7u?wid=_243342', 'only_matching': True, + }, + { + # video with subtitles + 'url': 'kaltura:111032:1_cw786r8q', + 'only_matching': True, + }, + { + # video with ttml subtitles (no fileExt) + 'url': 'kaltura:1926081:0_l5ye1133', + 'info_dict': { + 'id': '0_l5ye1133', + 'ext': 'mp4', + 'title': 'What Can You Do With Python?', + 'upload_date': '20160221', + 'uploader_id': 'stork', + 'thumbnail': 're:^https?://.*/thumbnail/.*', + 'timestamp': int, + 'subtitles': { + 'en': [{ + 'ext': 'ttml', + }], + }, + }, + 'skip': 'Gone. Maybe https://www.safaribooksonline.com/library/tutorials/introduction-to-python-anon/3469/', + 'params': { + 'skip_download': True, + }, + }, + { + 'url': 'https://www.kaltura.com/index.php/extwidget/preview/partner_id/1770401/uiconf_id/37307382/entry_id/0_58u8kme7/embed/iframe?&flashvars[streamerType]=auto', + 'only_matching': True, + }, + { + 'url': 'https://www.kaltura.com:443/index.php/extwidget/preview/partner_id/1770401/uiconf_id/37307382/entry_id/0_58u8kme7/embed/iframe?&flashvars[streamerType]=auto', + 'only_matching': True, + }, + { + # unavailable source format + 'url': 'kaltura:513551:1_66x4rg7o', + 'only_matching': True, } ] - def _kaltura_api_call(self, video_id, actions, *args, **kwargs): + @staticmethod + def _extract_url(webpage): + urls = KalturaIE._extract_urls(webpage) + return urls[0] if urls else None + + @staticmethod + def _extract_urls(webpage): + # Embed codes: https://knowledge.kaltura.com/embedding-kaltura-media-players-your-site + finditer = ( + re.finditer( + r"""(?xs) + kWidget\.(?:thumb)?[Ee]mbed\( + \{.*? + (?P['"])wid(?P=q1)\s*:\s* + (?P['"])_?(?P(?:(?!(?P=q2)).)+)(?P=q2),.*? + (?P['"])entry_?[Ii]d(?P=q3)\s*:\s* + (?P['"])(?P(?:(?!(?P=q4)).)+)(?P=q4)(?:,|\s*\}) + """, webpage) + or re.finditer( + r'''(?xs) + (?P["']) + (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com(?::\d+)?/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P\d+)(?:(?!(?P=q1)).)* + (?P=q1).*? + (?: + (?: + entry_?[Ii]d| + (?P["'])entry_?[Ii]d(?P=q2) + )\s*:\s*| + \[\s*(?P["'])entry_?[Ii]d(?P=q2_1)\s*\]\s*=\s* + ) + (?P["'])(?P(?:(?!(?P=q3)).)+)(?P=q3) + ''', webpage) + or re.finditer( + r'''(?xs) + <(?:iframe[^>]+src|meta[^>]+\bcontent)=(?P["']) + (?:https?:)?//(?:(?:www|cdnapi(?:sec)?)\.)?kaltura\.com/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P\d+) + (?:(?!(?P=q1)).)* + [?&;]entry_id=(?P(?:(?!(?P=q1))[^&])+) + (?:(?!(?P=q1)).)* + (?P=q1) + ''', webpage) + ) + urls = [] + for mobj in finditer: + embed_info = mobj.groupdict() + for k, v in embed_info.items(): + if v: + embed_info[k] = v.strip() + url = 'kaltura:%(partner_id)s:%(id)s' % embed_info + escaped_pid = re.escape(embed_info['partner_id']) + service_mobj = re.search( + r']+src=(["\'])(?P(?:https?:)?//(?:(?!\1).)+)/p/%s/sp/%s00/embedIframeJs' % (escaped_pid, escaped_pid), + webpage) + if service_mobj: + url = smuggle_url(url, {'service_url': service_mobj.group('id')}) + urls.append(url) + return urls + + def _kaltura_api_call(self, video_id, actions, service_url=None, *args, **kwargs): params = actions[0] if len(actions) > 1: for i, a in enumerate(actions[1:], start=1): for k, v in a.items(): params['%d:%s' % (i, k)] = v - query = compat_urllib_parse_urlencode(params) - url = self._API_BASE + query - data = self._download_json(url, video_id, *args, **kwargs) + data = self._download_json( + (service_url or self._SERVICE_URL) + self._SERVICE_BASE, + video_id, query=params, *args, **kwargs) status = data if len(actions) == 1 else data[0] if status.get('objectType') == 'KalturaAPIException': @@ -82,20 +187,7 @@ class KalturaIE(InfoExtractor): return data - def _get_kaltura_signature(self, video_id, partner_id): - actions = [{ - 'apiVersion': '3.1', - 'expiry': 86400, - 'format': 1, - 'service': 'session', - 'action': 'startWidgetSession', - 'widgetId': '_%s' % partner_id, - }] - return self._kaltura_api_call( - video_id, actions, note='Downloading Kaltura signature')['ks'] - - def _get_video_info(self, video_id, partner_id): - signature = self._get_kaltura_signature(video_id, partner_id) + def _get_video_info(self, video_id, partner_id, service_url=None): actions = [ { 'action': 'null', @@ -103,22 +195,36 @@ class KalturaIE(InfoExtractor): 'clientTag': 'kdp:v3.8.5', 'format': 1, # JSON, 2 = XML, 3 = PHP 'service': 'multirequest', - 'ks': signature, + }, + { + 'expiry': 86400, + 'service': 'session', + 'action': 'startWidgetSession', + 'widgetId': '_%s' % partner_id, }, { 'action': 'get', 'entryId': video_id, 'service': 'baseentry', - 'version': '-1', + 'ks': '{1:result:ks}', + 'responseProfile:fields': 'createdAt,dataUrl,duration,name,plays,thumbnailUrl,userId', + 'responseProfile:type': 1, }, { 'action': 'getbyentryid', 'entryId': video_id, 'service': 'flavorAsset', + 'ks': '{1:result:ks}', + }, + { + 'action': 'list', + 'filter:entryIdEqual': video_id, + 'service': 'caption_captionasset', + 'ks': '{1:result:ks}', }, ] return self._kaltura_api_call( - video_id, actions, note='Downloading video info JSON') + video_id, actions, service_url, note='Downloading video info JSON') def _real_extract(self, url): url, smuggled_data = unsmuggle_url(url, {}) @@ -126,8 +232,9 @@ class KalturaIE(InfoExtractor): mobj = re.match(self._VALID_URL, url) partner_id, entry_id = mobj.group('partner_id', 'id') ks = None + captions = None if partner_id and entry_id: - info, flavor_assets = self._get_video_info(entry_id, partner_id) + _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id, smuggled_data.get('service_url')) else: path, query = mobj.group('path', 'query') if not path and not query: @@ -142,11 +249,13 @@ class KalturaIE(InfoExtractor): partner_id = params['wid'][0][1:] elif 'p' in params: partner_id = params['p'][0] + elif 'partner_id' in params: + partner_id = params['partner_id'][0] else: raise ExtractorError('Invalid URL', expected=True) if 'entry_id' in params: entry_id = params['entry_id'][0] - info, flavor_assets = self._get_video_info(entry_id, partner_id) + _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id) elif 'uiconf_id' in params and 'flashvars[referenceId]' in params: reference_id = params['flashvars[referenceId]'][0] webpage = self._download_webpage(url, reference_id) @@ -156,6 +265,17 @@ class KalturaIE(InfoExtractor): reference_id)['entryResult'] info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets'] entry_id = info['id'] + # Unfortunately, data returned in kalturaIframePackageData lacks + # captions so we will try requesting the complete data using + # regular approach since we now know the entry_id + try: + _, info, flavor_assets, captions = self._get_video_info( + entry_id, partner_id) + except ExtractorError: + # Regular scenario failed but we already have everything + # extracted apart from captions and can process at least + # with this + pass else: raise ExtractorError('Invalid URL', expected=True) ks = params.get('flashvars[ks]', [None])[0] @@ -175,39 +295,83 @@ class KalturaIE(InfoExtractor): unsigned_url += '?referrer=%s' % referrer return unsigned_url + data_url = info['dataUrl'] + if '/flvclipper/' in data_url: + data_url = re.sub(r'/flvclipper/.*', '/serveFlavor', data_url) + formats = [] for f in flavor_assets: # Continue if asset is not ready - if f['status'] != 2: + if f.get('status') != 2: continue - video_url = sign_url('%s/flavorId/%s' % (info['dataUrl'], f['id'])) + # Original format that's not available (e.g. kaltura:1926081:0_c03e1b5g) + # skip for now. + if f.get('fileExt') == 'chun': + continue + # DRM-protected video, cannot be decrypted + if f.get('fileExt') == 'wvm': + continue + if not f.get('fileExt'): + # QT indicates QuickTime; some videos have broken fileExt + if f.get('containerFormat') == 'qt': + f['fileExt'] = 'mov' + else: + f['fileExt'] = 'mp4' + video_url = sign_url( + '%s/flavorId/%s' % (data_url, f['id'])) + format_id = '%(fileExt)s-%(bitrate)s' % f + # Source format may not be available (e.g. kaltura:513551:1_66x4rg7o) + if f.get('isOriginal') is True and not self._is_valid_url( + video_url, entry_id, format_id): + continue + # audio-only has no videoCodecId (e.g. kaltura:1926081:0_c03e1b5g + # -f mp4-56) + vcodec = 'none' if 'videoCodecId' not in f and f.get( + 'frameRate') == 0 else f.get('videoCodecId') formats.append({ - 'format_id': '%(fileExt)s-%(bitrate)s' % f, + 'format_id': format_id, 'ext': f.get('fileExt'), 'tbr': int_or_none(f['bitrate']), 'fps': int_or_none(f.get('frameRate')), 'filesize_approx': int_or_none(f.get('size'), invscale=1024), 'container': f.get('containerFormat'), - 'vcodec': f.get('videoCodecId'), + 'vcodec': vcodec, 'height': int_or_none(f.get('height')), 'width': int_or_none(f.get('width')), 'url': video_url, }) - m3u8_url = sign_url(info['dataUrl'].replace('format/url', 'format/applehttp')) - formats.extend(self._extract_m3u8_formats( - m3u8_url, entry_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)) + if '/playManifest/' in data_url: + m3u8_url = sign_url(data_url.replace( + 'format/url', 'format/applehttp')) + formats.extend(self._extract_m3u8_formats( + m3u8_url, entry_id, 'mp4', 'm3u8_native', + m3u8_id='hls', fatal=False)) - self._check_formats(formats, entry_id) self._sort_formats(formats) + subtitles = {} + if captions: + for caption in captions.get('objects', []): + # Continue if caption is not ready + if caption.get('status') != 2: + continue + if not caption.get('id'): + continue + caption_format = int_or_none(caption.get('format')) + subtitles.setdefault(caption.get('languageCode') or caption.get('language'), []).append({ + 'url': '%s/api_v3/service/caption_captionasset/action/serve/captionAssetId/%s' % (self._SERVICE_URL, caption['id']), + 'ext': caption.get('fileExt') or self._CAPTION_TYPES.get(caption_format) or 'ttml', + }) + return { 'id': entry_id, 'title': info['name'], 'formats': formats, + 'subtitles': subtitles, 'description': clean_html(info.get('description')), 'thumbnail': info.get('thumbnailUrl'), 'duration': info.get('duration'), 'timestamp': info.get('createdAt'), - 'uploader_id': info.get('userId'), + 'uploader_id': info.get('userId') if info.get('userId') != 'None' else None, 'view_count': info.get('plays'), }