X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Flivestream.py;h=2467f8bdd35304a57cc6bd951fa14fd32a25cd4c;hb=674fb0fcc54c72448f80a0573f7fd116f220827e;hp=1a3e0ae6b94ea2b59f2065f8641827d29b39fc37;hpb=c66d2baa9cb33327a3318f49cbb89f9ac559c978;p=youtube-dl diff --git a/youtube_dl/extractor/livestream.py b/youtube_dl/extractor/livestream.py index 1a3e0ae6b..2467f8bdd 100644 --- a/youtube_dl/extractor/livestream.py +++ b/youtube_dl/extractor/livestream.py @@ -1,39 +1,153 @@ +from __future__ import unicode_literals + import re import json -import xml.etree.ElementTree +import itertools from .common import InfoExtractor -from ..utils import ( +from ..compat import ( + compat_str, compat_urllib_parse_urlparse, compat_urlparse, - get_meta_content, - xpath_with_ns, +) +from ..utils import ( ExtractorError, + find_xpath_attr, + int_or_none, + orderedSet, + xpath_with_ns, ) class LivestreamIE(InfoExtractor): - IE_NAME = u'livestream' - _VALID_URL = r'http://new.livestream.com/.*?/(?P.*?)(/videos/(?P\d+))?/?$' - _TEST = { - u'url': u'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370', - u'file': u'4719370.mp4', - u'md5': u'0d2186e3187d185a04b3cdd02b828836', - u'info_dict': { - u'title': u'Live from Webster Hall NYC', - u'upload_date': u'20121012', + IE_NAME = 'livestream' + _VALID_URL = r'https?://new\.livestream\.com/.*?/(?P.*?)(/videos/(?P[0-9]+)(?:/player)?)?/?(?:$|[?#])' + _TESTS = [{ + 'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370', + 'md5': '53274c76ba7754fb0e8d072716f2292b', + 'info_dict': { + 'id': '4719370', + 'ext': 'mp4', + 'title': 'Live from Webster Hall NYC', + 'upload_date': '20121012', + 'like_count': int, + 'view_count': int, + 'thumbnail': 're:^http://.*\.jpg$' } - } + }, { + 'url': 'http://new.livestream.com/tedx/cityenglish', + 'info_dict': { + 'title': 'TEDCity2.0 (English)', + 'id': '2245590', + }, + 'playlist_mincount': 4, + }, { + 'url': 'http://new.livestream.com/chess24/tatasteelchess', + 'info_dict': { + 'title': 'Tata Steel Chess', + 'id': '3705884', + }, + 'playlist_mincount': 60, + }, { + 'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640', + 'only_matching': True, + }] + + def _parse_smil(self, video_id, smil_url): + formats = [] + _SWITCH_XPATH = ( + './/{http://www.w3.org/2001/SMIL20/Language}body/' + '{http://www.w3.org/2001/SMIL20/Language}switch') + smil_doc = self._download_xml( + smil_url, video_id, + note='Downloading SMIL information', + errnote='Unable to download SMIL information', + fatal=False) + if smil_doc is False: # Download failed + return formats + title_node = find_xpath_attr( + smil_doc, './/{http://www.w3.org/2001/SMIL20/Language}meta', + 'name', 'title') + if title_node is None: + self.report_warning('Cannot find SMIL id') + switch_node = smil_doc.find(_SWITCH_XPATH) + else: + title_id = title_node.attrib['content'] + switch_node = find_xpath_attr( + smil_doc, _SWITCH_XPATH, 'id', title_id) + if switch_node is None: + raise ExtractorError('Cannot find switch node') + video_nodes = switch_node.findall( + '{http://www.w3.org/2001/SMIL20/Language}video') + + for vn in video_nodes: + tbr = int_or_none(vn.attrib.get('system-bitrate')) + furl = ( + 'http://livestream-f.akamaihd.net/%s?v=3.0.3&fp=WIN%%2014,0,0,145' % + (vn.attrib['src'])) + if 'clipBegin' in vn.attrib: + furl += '&ssek=' + vn.attrib['clipBegin'] + formats.append({ + 'url': furl, + 'format_id': 'smil_%d' % tbr, + 'ext': 'flv', + 'tbr': tbr, + 'preference': -1000, + }) + return formats def _extract_video_info(self, video_data): - video_url = video_data.get('progressive_url_hd') or video_data.get('progressive_url') - return {'id': video_data['id'], - 'url': video_url, - 'ext': 'mp4', - 'title': video_data['caption'], - 'thumbnail': video_data['thumbnail_url'], - 'upload_date': video_data['updated_at'].replace('-','')[:8], - } + video_id = compat_str(video_data['id']) + + FORMAT_KEYS = ( + ('sd', 'progressive_url'), + ('hd', 'progressive_url_hd'), + ) + formats = [{ + 'format_id': format_id, + 'url': video_data[key], + 'quality': i + 1, + } for i, (format_id, key) in enumerate(FORMAT_KEYS) + if video_data.get(key)] + + smil_url = video_data.get('smil_url') + if smil_url: + formats.extend(self._parse_smil(video_id, smil_url)) + self._sort_formats(formats) + + return { + 'id': video_id, + 'formats': formats, + 'title': video_data['caption'], + 'thumbnail': video_data.get('thumbnail_url'), + 'upload_date': video_data['updated_at'].replace('-', '')[:8], + 'like_count': video_data.get('likes', {}).get('total'), + 'view_count': video_data.get('views'), + } + + def _extract_event(self, info): + event_id = compat_str(info['id']) + account = compat_str(info['owner_account_id']) + root_url = ( + 'https://new.livestream.com/api/accounts/{account}/events/{event}/' + 'feed.json'.format(account=account, event=event_id)) + + def _extract_videos(): + last_video = None + for i in itertools.count(1): + if last_video is None: + info_url = root_url + else: + info_url = '{root}?&id={id}&newer=-1&type=video'.format( + root=root_url, id=last_video) + videos_info = self._download_json(info_url, event_id, 'Downloading page {0}'.format(i))['data'] + videos_info = [v['data'] for v in videos_info if v['type'] == 'video'] + if not videos_info: + break + for v in videos_info: + yield self._extract_video_info(v) + last_video = videos_info[-1]['id'] + return self.playlist_result(_extract_videos(), event_id, info['full_name']) def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -41,60 +155,126 @@ class LivestreamIE(InfoExtractor): event_name = mobj.group('event_name') webpage = self._download_webpage(url, video_id or event_name) + og_video = self._og_search_video_url( + webpage, 'player url', fatal=False, default=None) + if og_video is not None: + query_str = compat_urllib_parse_urlparse(og_video).query + query = compat_urlparse.parse_qs(query_str) + if 'play_url' in query: + api_url = query['play_url'][0].replace('.smil', '') + info = json.loads(self._download_webpage( + api_url, video_id, 'Downloading video info')) + return self._extract_video_info(info) + + config_json = self._search_regex( + r'window.config = ({.*?});', webpage, 'window config') + info = json.loads(config_json)['event'] + + def is_relevant(vdata, vid): + result = vdata['type'] == 'video' + if video_id is not None: + result = result and compat_str(vdata['data']['id']) == vid + return result + if video_id is None: # This is an event page: - config_json = self._search_regex(r'window.config = ({.*?});', - webpage, u'window config') - info = json.loads(config_json)['event'] - videos = [self._extract_video_info(video_data['data']) - for video_data in info['feed']['data'] if video_data['type'] == u'video'] - return self.playlist_result(videos, info['id'], info['full_name']) + return self._extract_event(info) else: - og_video = self._og_search_video_url(webpage, name=u'player url') - query_str = compat_urllib_parse_urlparse(og_video).query - query = compat_urlparse.parse_qs(query_str) - api_url = query['play_url'][0].replace('.smil', '') - info = json.loads(self._download_webpage(api_url, video_id, - u'Downloading video info')) - return self._extract_video_info(info) + videos = [self._extract_video_info(video_data['data']) + for video_data in info['feed']['data'] + if is_relevant(video_data, video_id)] + if not videos: + raise ExtractorError('Cannot find video %s' % video_id) + return videos[0] # The original version of Livestream uses a different system class LivestreamOriginalIE(InfoExtractor): - IE_NAME = u'livestream:original' - _VALID_URL = r'https?://www\.livestream\.com/(?P[^/]+)/video\?.*?clipId=(?P.*?)(&|$)' - _TEST = { - u'url': u'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb', - u'info_dict': { - u'id': u'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb', - u'ext': u'flv', - u'title': u'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital', + IE_NAME = 'livestream:original' + _VALID_URL = r'''(?x)https?://www\.livestream\.com/ + (?P[^/]+)/(?Pvideo|folder) + (?:\?.*?Id=|/)(?P.*?)(&|$) + ''' + _TESTS = [{ + 'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb', + 'info_dict': { + 'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb', + 'ext': 'flv', + 'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital', }, - u'params': { + 'params': { # rtmp - u'skip_download': True, + 'skip_download': True, + }, + }, { + 'url': 'https://www.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3', + 'info_dict': { + 'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3', }, - } + 'playlist_mincount': 4, + }] - def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - video_id = mobj.group('id') - user = mobj.group('user') + def _extract_video(self, user, video_id): api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id) - api_response = self._download_webpage(api_url, video_id) - info = xml.etree.ElementTree.fromstring(api_response.encode('utf-8')) + info = self._download_xml(api_url, video_id) item = info.find('channel').find('item') ns = {'media': 'http://search.yahoo.com/mrss'} thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url'] # Remove the extension and number from the path (like 1.jpg) - path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, u'path') + path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, 'path') return { 'id': video_id, 'title': item.find('title').text, 'url': 'rtmp://extondemand.livestream.com/ondemand', - 'play_path': 'mp4:trans/dv15/mogulus-{0}.mp4'.format(path), + 'play_path': 'trans/dv15/mogulus-{0}'.format(path), + 'player_url': 'http://static.livestream.com/chromelessPlayer/v21/playerapi.swf?hash=5uetk&v=0803&classid=D27CDB6E-AE6D-11cf-96B8-444553540000&jsEnabled=false&wmode=opaque', 'ext': 'flv', 'thumbnail': thumbnail_url, } + + def _extract_folder(self, url, folder_id): + webpage = self._download_webpage(url, folder_id) + paths = orderedSet(re.findall( + r'''(?x)(?: + \s*.+)' + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + id = mobj.group('id') + webpage = self._download_webpage(url, id) + + return { + '_type': 'url', + 'url': self._og_search_url(webpage), + }