X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/f46044c66663049e286c20ee015db99d47d9dd8a..82a01b147e1b7241e1040c446b10b58b658c70cf:/youtube_dl/extractor/kaltura.py diff --git a/youtube_dl/extractor/kaltura.py b/youtube_dl/extractor/kaltura.py index d287304..a65697f 100644 --- a/youtube_dl/extractor/kaltura.py +++ b/youtube_dl/extractor/kaltura.py @@ -2,23 +2,38 @@ from __future__ import unicode_literals import re +import base64 from .common import InfoExtractor -from ..compat import compat_urllib_parse +from ..compat import ( + compat_urllib_parse_urlencode, + compat_urlparse, + compat_parse_qs, +) from ..utils import ( + clean_html, ExtractorError, int_or_none, + unsmuggle_url, ) class KalturaIE(InfoExtractor): _VALID_URL = r'''(?x) - (?:kaltura:| - https?://(:?(?:www|cdnapisec)\.)?kaltura\.com/index\.php/kwidget/(?:[^/]+/)*?wid/_ - )(?P\d+) - (?::| - /(?:[^/]+/)*?entry_id/ - )(?P[0-9a-z_]+)''' + (?: + kaltura:(?P\d+):(?P[0-9a-z_]+)| + https?:// + (:?(?:www|cdnapi(?:sec)?)\.)?kaltura\.com/ + (?: + (?: + # flash player + index\.php/kwidget| + # html5 player + html5/html5lib/[^/]+/mwEmbedFrame\.php + ) + )(?:/(?P[^?]+))?(?:\?(?P.*))? + ) + ''' _API_BASE = 'http://cdnapi.kaltura.com/api_v3/index.php?' _TESTS = [ { @@ -27,7 +42,7 @@ class KalturaIE(InfoExtractor): 'info_dict': { 'id': '1_1jc2y3e4', 'ext': 'mp4', - 'title': 'Track 4', + 'title': 'Straight from the Heart', 'upload_date': '20131219', 'uploader_id': 'mlundberg@wolfgangsvault.com', 'description': 'The Allman Brothers Band, 12/16/1981', @@ -43,6 +58,10 @@ class KalturaIE(InfoExtractor): 'url': 'https://cdnapisec.kaltura.com/index.php/kwidget/wid/_557781/uiconf_id/22845202/entry_id/1_plr1syf3', 'only_matching': True, }, + { + '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, + } ] def _kaltura_api_call(self, video_id, actions, *args, **kwargs): @@ -52,7 +71,7 @@ class KalturaIE(InfoExtractor): for k, v in a.items(): params['%d:%s' % (i, k)] = v - query = compat_urllib_parse.urlencode(params) + query = compat_urllib_parse_urlencode(params) url = self._API_BASE + query data = self._download_json(url, video_id, *args, **kwargs) @@ -93,43 +112,99 @@ class KalturaIE(InfoExtractor): 'version': '-1', }, { - 'action': 'getContextData', - 'contextDataParams:objectType': 'KalturaEntryContextDataParams', - 'contextDataParams:referrer': 'http://www.kaltura.com/', - 'contextDataParams:streamerType': 'http', + 'action': 'getbyentryid', 'entryId': video_id, - 'service': 'baseentry', + 'service': 'flavorAsset', }, ] return self._kaltura_api_call( video_id, actions, note='Downloading video info JSON') def _real_extract(self, url): - video_id = self._match_id(url) + url, smuggled_data = unsmuggle_url(url, {}) + mobj = re.match(self._VALID_URL, url) - partner_id, entry_id = mobj.group('partner_id'), mobj.group('id') - - info, source_data = self._get_video_info(entry_id, partner_id) - - formats = [{ - 'format_id': '%(fileExt)s-%(bitrate)s' % f, - 'ext': f['fileExt'], - 'tbr': f['bitrate'], - 'fps': f.get('frameRate'), - 'filesize_approx': int_or_none(f.get('size'), invscale=1024), - 'container': f.get('containerFormat'), - 'vcodec': f.get('videoCodecId'), - 'height': f.get('height'), - 'width': f.get('width'), - 'url': '%s/flavorId/%s' % (info['dataUrl'], f['id']), - } for f in source_data['flavorAssets']] + partner_id, entry_id = mobj.group('partner_id', 'id') + ks = None + if partner_id and entry_id: + info, flavor_assets = self._get_video_info(entry_id, partner_id) + else: + path, query = mobj.group('path', 'query') + if not path and not query: + raise ExtractorError('Invalid URL', expected=True) + params = {} + if query: + params = compat_parse_qs(query) + if path: + splitted_path = path.split('/') + params.update(dict((zip(splitted_path[::2], [[v] for v in splitted_path[1::2]])))) + if 'wid' in params: + partner_id = params['wid'][0][1:] + elif 'p' in params: + partner_id = params['p'][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) + elif 'uiconf_id' in params and 'flashvars[referenceId]' in params: + reference_id = params['flashvars[referenceId]'][0] + webpage = self._download_webpage(url, reference_id) + entry_data = self._parse_json(self._search_regex( + r'window\.kalturaIframePackageData\s*=\s*({.*});', + webpage, 'kalturaIframePackageData'), + reference_id)['entryResult'] + info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets'] + entry_id = info['id'] + else: + raise ExtractorError('Invalid URL', expected=True) + ks = params.get('flashvars[ks]', [None])[0] + + source_url = smuggled_data.get('source_url') + if source_url: + referrer = base64.b64encode( + '://'.join(compat_urlparse.urlparse(source_url)[:2]) + .encode('utf-8')).decode('utf-8') + else: + referrer = None + + def sign_url(unsigned_url): + if ks: + unsigned_url += '/ks/%s' % ks + if referrer: + unsigned_url += '?referrer=%s' % referrer + return unsigned_url + + formats = [] + for f in flavor_assets: + # Continue if asset is not ready + if f['status'] != 2: + continue + video_url = sign_url('%s/flavorId/%s' % (info['dataUrl'], f['id'])) + formats.append({ + 'format_id': '%(fileExt)s-%(bitrate)s' % f, + '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'), + '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)) + + self._check_formats(formats, entry_id) self._sort_formats(formats) return { - 'id': video_id, + 'id': entry_id, 'title': info['name'], 'formats': formats, - 'description': info.get('description'), + 'description': clean_html(info.get('description')), 'thumbnail': info.get('thumbnailUrl'), 'duration': info.get('duration'), 'timestamp': info.get('createdAt'),