diff options
Diffstat (limited to 'youtube_dl/extractor/livestream.py')
| -rw-r--r-- | youtube_dl/extractor/livestream.py | 202 |
1 files changed, 164 insertions, 38 deletions
diff --git a/youtube_dl/extractor/livestream.py b/youtube_dl/extractor/livestream.py index 1dcd1fb2d..281a0ce40 100644 --- a/youtube_dl/extractor/livestream.py +++ b/youtube_dl/extractor/livestream.py @@ -1,36 +1,109 @@ +from __future__ import unicode_literals + import re import json from .common import InfoExtractor from ..utils import ( + compat_str, compat_urllib_parse_urlparse, compat_urlparse, + ExtractorError, + find_xpath_attr, + int_or_none, + orderedSet, xpath_with_ns, ) class LivestreamIE(InfoExtractor): - IE_NAME = u'livestream' + IE_NAME = 'livestream' _VALID_URL = r'http://new\.livestream\.com/.*?/(?P<event_name>.*?)(/videos/(?P<id>\d+))?/?$' _TEST = { - u'url': u'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370', - u'file': u'4719370.mp4', - u'md5': u'0d2186e3187d185a04b3cdd02b828836', - u'info_dict': { - u'title': u'Live from Webster Hall NYC', - u'upload_date': u'20121012', + 'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370', + 'md5': '53274c76ba7754fb0e8d072716f2292b', + 'info_dict': { + 'id': '4719370', + 'ext': 'mp4', + 'title': 'Live from Webster Hall NYC', + 'upload_date': '20121012', + 'like_count': int, + 'view_count': int, + 'thumbnail': 're:^http://.*\.jpg$' } } + 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 _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -38,45 +111,57 @@ class LivestreamIE(InfoExtractor): event_name = mobj.group('event_name') webpage = self._download_webpage(url, video_id or event_name) - if video_id is None: - # This is an event page: - config_json = self._search_regex(r'window.config = ({.*?});', - webpage, u'window config') + og_video = self._og_search_video_url(webpage, 'player url', fatal=False, default=None) + if og_video is None: + 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 + 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']) + for video_data in info['feed']['data'] + if is_relevant(video_data, video_id)] + if video_id is None: + # This is an event page: + return self.playlist_result(videos, info['id'], info['full_name']) + else: + if videos: + return videos[0] 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')) + info = json.loads(self._download_webpage( + api_url, video_id, 'Downloading video info')) return self._extract_video_info(info) # 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<user>[^/]+)/video\?.*?clipId=(?P<id>.*?)(&|$)' + IE_NAME = 'livestream:original' + _VALID_URL = r'''(?x)https?://www\.livestream\.com/ + (?P<user>[^/]+)/(?P<type>video|folder) + (?:\?.*?Id=|/)(?P<id>.*?)(&|$) + ''' _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', + '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, }, } - 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) info = self._download_xml(api_url, video_id) @@ -84,7 +169,7 @@ class LivestreamOriginalIE(InfoExtractor): 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, @@ -94,3 +179,44 @@ class LivestreamOriginalIE(InfoExtractor): 'ext': 'flv', 'thumbnail': thumbnail_url, } + + def _extract_folder(self, url, folder_id): + webpage = self._download_webpage(url, folder_id) + urls = orderedSet(re.findall(r'<a href="(https?://livestre\.am/.*?)"', webpage)) + + return { + '_type': 'playlist', + 'id': folder_id, + 'entries': [{ + '_type': 'url', + 'url': video_url, + } for video_url in urls], + } + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + id = mobj.group('id') + user = mobj.group('user') + url_type = mobj.group('type') + if url_type == 'folder': + return self._extract_folder(url, id) + else: + return self._extract_video(user, id) + + +# The server doesn't support HEAD request, the generic extractor can't detect +# the redirection +class LivestreamShortenerIE(InfoExtractor): + IE_NAME = 'livestream:shortener' + IE_DESC = False # Do not list + _VALID_URL = r'https?://livestre\.am/(?P<id>.+)' + + 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), + } |