X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fustream.py;h=994b60a76b88ef4d7ff8be630c2bafbd989e2c96;hb=cc7fec5818254f4679896823c7de9d17f50201ca;hp=f69b27d4405d719e9c8d87807d2c1392753150eb;hpb=74ac9bdd82b8a625ea9782251258ab7da1463877;p=youtube-dl diff --git a/youtube_dl/extractor/ustream.py b/youtube_dl/extractor/ustream.py index f69b27d44..994b60a76 100644 --- a/youtube_dl/extractor/ustream.py +++ b/youtube_dl/extractor/ustream.py @@ -1,111 +1,102 @@ -import json +from __future__ import unicode_literals + import re from .common import InfoExtractor from ..utils import ( compat_urlparse, - compat_html_parser, + get_meta_content, ) class UstreamIE(InfoExtractor): - _VALID_URL = r'https?://www\.ustream\.tv/recorded/(?P\d+)' - IE_NAME = u'ustream' + _VALID_URL = r'https?://www\.ustream\.tv/(?Precorded|embed|embed/recorded)/(?P\d+)' + IE_NAME = 'ustream' _TEST = { - u'url': u'http://www.ustream.tv/recorded/20274954', - u'file': u'20274954.flv', - u'md5': u'088f151799e8f572f84eb62f17d73e5c', - u'info_dict': { - u"uploader": u"Young Americans for Liberty", - u"title": u"Young Americans for Liberty February 7, 2012 2:28 AM" - } + 'url': 'http://www.ustream.tv/recorded/20274954', + 'md5': '088f151799e8f572f84eb62f17d73e5c', + 'info_dict': { + 'id': '20274954', + 'ext': 'flv', + 'uploader': 'Young Americans for Liberty', + 'title': 'Young Americans for Liberty February 7, 2012 2:28 AM', + }, } def _real_extract(self, url): m = re.match(self._VALID_URL, url) video_id = m.group('videoID') - video_url = u'http://tcdn.ustream.tv/video/%s' % video_id + # some sites use this embed format (see: http://github.com/rg3/youtube-dl/issues/2990) + if m.group('type') == 'embed/recorded': + video_id = m.group('videoID') + desktop_url = 'http://www.ustream.tv/recorded/' + video_id + return self.url_result(desktop_url, 'Ustream') + if m.group('type') == 'embed': + video_id = m.group('videoID') + webpage = self._download_webpage(url, video_id) + desktop_video_id = self._html_search_regex( + r'ContentVideoIds=\["([^"]*?)"\]', webpage, 'desktop_video_id') + desktop_url = 'http://www.ustream.tv/recorded/' + desktop_video_id + return self.url_result(desktop_url, 'Ustream') + + video_url = 'http://tcdn.ustream.tv/video/%s' % video_id webpage = self._download_webpage(url, video_id) self.report_extraction(video_id) video_title = self._html_search_regex(r'data-title="(?P.+)"', - webpage, u'title') + webpage, 'title') uploader = self._html_search_regex(r'data-content-type="channel".*?>(?P<uploader>.*?)</a>', - webpage, u'uploader', fatal=False, flags=re.DOTALL) + webpage, 'uploader', fatal=False, flags=re.DOTALL) thumbnail = self._html_search_regex(r'<link rel="image_src" href="(?P<thumb>.*?)"', - webpage, u'thumbnail', fatal=False) - - info = { - 'id': video_id, - 'url': video_url, - 'ext': 'flv', - 'title': video_title, - 'uploader': uploader, - 'thumbnail': thumbnail, - } - return info - -# More robust than regular expressions - -class ChannelParser(compat_html_parser.HTMLParser): - """ - <meta name="ustream:channel_id" content="1234"> - """ - channel_id = None - - def handle_starttag(self, tag, attrs): - if tag != 'meta': - return - values = dict(attrs) - if values.get('name') != 'ustream:channel_id': - return - value = values.get('content', '') - if value.isdigit(): - self.channel_id = value - -class SocialstreamParser(compat_html_parser.HTMLParser): - """ - <li class="content123 video" data-content-id="123" data-length="1452" - data-href="/recorded/123" data-og-url="/recorded/123"> - """ - def __init__(self): - compat_html_parser.HTMLParser.__init__(self) - self.content_ids = [] - - def handle_starttag(self, tag, attrs): - if tag != 'li': - return - for (attr, value) in attrs: - if attr == 'data-content-id' and value.isdigit(): - self.content_ids.append(value) + webpage, 'thumbnail', fatal=False) + + return { + 'id': video_id, + 'url': video_url, + 'ext': 'flv', + 'title': video_title, + 'uploader': uploader, + 'thumbnail': thumbnail, + } + class UstreamChannelIE(InfoExtractor): _VALID_URL = r'https?://www\.ustream\.tv/channel/(?P<slug>.+)' - IE_NAME = u'ustream:channel' + IE_NAME = 'ustream:channel' + _TEST = { + 'url': 'http://www.ustream.tv/channel/channeljapan', + 'info_dict': { + 'id': '10874166', + }, + 'playlist_mincount': 54, + } def _real_extract(self, url): m = re.match(self._VALID_URL, url) - slug = m.group('slug') + display_id = m.group('slug') + webpage = self._download_webpage(url, display_id) + channel_id = get_meta_content('ustream:channel_id', webpage) - p = ChannelParser() - p.feed(self._download_webpage(url, slug)) - p.close() - channel_id = p.channel_id - - p = SocialstreamParser() BASE = 'http://www.ustream.tv' next_url = '/ajax/socialstream/videos/%s/1.json' % channel_id + video_ids = [] while next_url: - reply = json.loads(self._download_webpage(compat_urlparse.urljoin(BASE, next_url), channel_id)) - p.feed(reply['data']) + reply = self._download_json( + compat_urlparse.urljoin(BASE, next_url), display_id, + note='Downloading video information (next: %d)' % (len(video_ids) + 1)) + video_ids.extend(re.findall(r'data-content-id="(\d.*)"', reply['data'])) next_url = reply['nextUrl'] - p.close() - video_ids = p.content_ids - urls = ['http://www.ustream.tv/recorded/' + vid for vid in video_ids] - url_entries = [self.url_result(eurl, 'Ustream') for eurl in urls] - return self.playlist_result(url_entries, channel_id) + entries = [ + self.url_result('http://www.ustream.tv/recorded/' + vid, 'Ustream') + for vid in video_ids] + return { + '_type': 'playlist', + 'id': channel_id, + 'display_id': display_id, + 'entries': entries, + }