X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/7df4343e553a9942a290e5e1dffe931261ab844e..f8df414a4abcde0ddd39325dac26ca071d2d15c6:/youtube_dl/extractor/ustream.py diff --git a/youtube_dl/extractor/ustream.py b/youtube_dl/extractor/ustream.py index c39c278..5737d4d 100644 --- a/youtube_dl/extractor/ustream.py +++ b/youtube_dl/extractor/ustream.py @@ -1,17 +1,25 @@ from __future__ import unicode_literals -import json +import random import re from .common import InfoExtractor from ..compat import ( + compat_str, compat_urlparse, ) -from ..utils import ExtractorError +from ..utils import ( + encode_data_uri, + ExtractorError, + int_or_none, + float_or_none, + mimetype2ext, + str_or_none, +) class UstreamIE(InfoExtractor): - _VALID_URL = r'https?://www\.ustream\.tv/(?Precorded|embed|embed/recorded)/(?P\d+)' + _VALID_URL = r'https?://(?:www\.)?ustream\.tv/(?Precorded|embed|embed/recorded)/(?P\d+)' IE_NAME = 'ustream' _TESTS = [{ 'url': 'http://www.ustream.tv/recorded/20274954', @@ -19,8 +27,12 @@ class UstreamIE(InfoExtractor): 'info_dict': { 'id': '20274954', 'ext': 'flv', - 'uploader': 'Young Americans for Liberty', 'title': 'Young Americans for Liberty February 7, 2012 2:28 AM', + 'description': 'Young Americans for Liberty February 7, 2012 2:28 AM', + 'timestamp': 1328577035, + 'upload_date': '20120207', + 'uploader': 'yaliberty', + 'uploader_id': '6780869', }, }, { # From http://sportscanada.tv/canadagames/index.php/week2/figure-skating/444 @@ -32,78 +44,207 @@ class UstreamIE(InfoExtractor): 'ext': 'flv', 'title': '-CG11- Canada Games Figure Skating', 'uploader': 'sportscanadatv', - } + }, + 'skip': 'This Pro Broadcaster has chosen to remove this video from the ustream.tv site.', + }, { + 'url': 'http://www.ustream.tv/embed/10299409', + 'info_dict': { + 'id': '10299409', + }, + 'playlist_count': 3, + }, { + 'url': 'http://www.ustream.tv/recorded/91343263', + 'info_dict': { + 'id': '91343263', + 'ext': 'mp4', + 'title': 'GitHub Universe - General Session - Day 1', + 'upload_date': '20160914', + 'description': 'GitHub Universe - General Session - Day 1', + 'timestamp': 1473872730, + 'uploader': 'wa0dnskeqkr', + 'uploader_id': '38977840', + }, + 'params': { + 'skip_download': True, # m3u8 download + }, }] + @staticmethod + def _extract_url(webpage): + mobj = re.search( + r']+?src=(["\'])(?Phttp://www\.ustream\.tv/embed/.+?)\1', webpage) + if mobj is not None: + return mobj.group('url') + + def _get_stream_info(self, url, video_id, app_id_ver, extra_note=None): + def num_to_hex(n): + return hex(n)[2:] + + rnd = random.randrange + + if not extra_note: + extra_note = '' + + conn_info = self._download_json( + 'http://r%d-1-%s-recorded-lp-live.ums.ustream.tv/1/ustream' % (rnd(1e8), video_id), + video_id, note='Downloading connection info' + extra_note, + query={ + 'type': 'viewer', + 'appId': app_id_ver[0], + 'appVersion': app_id_ver[1], + 'rsid': '%s:%s' % (num_to_hex(rnd(1e8)), num_to_hex(rnd(1e8))), + 'rpin': '_rpin.%d' % rnd(1e15), + 'referrer': url, + 'media': video_id, + 'application': 'recorded', + }) + host = conn_info[0]['args'][0]['host'] + connection_id = conn_info[0]['args'][0]['connectionId'] + + return self._download_json( + 'http://%s/1/ustream?connectionId=%s' % (host, connection_id), + video_id, note='Downloading stream info' + extra_note) + + def _get_streams(self, url, video_id, app_id_ver): + # Sometimes the return dict does not have 'stream' + for trial_count in range(3): + stream_info = self._get_stream_info( + url, video_id, app_id_ver, + extra_note=' (try %d)' % (trial_count + 1) if trial_count > 0 else '') + if 'stream' in stream_info[0]['args'][0]: + return stream_info[0]['args'][0]['stream'] + return [] + + def _parse_segmented_mp4(self, dash_stream_info): + def resolve_dash_template(template, idx, chunk_hash): + return template.replace('%', compat_str(idx), 1).replace('%', chunk_hash) + + formats = [] + for stream in dash_stream_info['streams']: + # Use only one provider to avoid too many formats + provider = dash_stream_info['providers'][0] + fragments = [{ + 'url': resolve_dash_template( + provider['url'] + stream['initUrl'], 0, dash_stream_info['hashes']['0']) + }] + for idx in range(dash_stream_info['videoLength'] // dash_stream_info['chunkTime']): + fragments.append({ + 'url': resolve_dash_template( + provider['url'] + stream['segmentUrl'], idx, + dash_stream_info['hashes'][compat_str(idx // 10 * 10)]) + }) + content_type = stream['contentType'] + kind = content_type.split('/')[0] + f = { + 'format_id': '-'.join(filter(None, [ + 'dash', kind, str_or_none(stream.get('bitrate'))])), + 'protocol': 'http_dash_segments', + # TODO: generate a MPD doc for external players? + 'url': encode_data_uri(b'', 'text/xml'), + 'ext': mimetype2ext(content_type), + 'height': stream.get('height'), + 'width': stream.get('width'), + 'fragments': fragments, + } + if kind == 'video': + f.update({ + 'vcodec': stream.get('codec'), + 'acodec': 'none', + 'vbr': stream.get('bitrate'), + }) + else: + f.update({ + 'vcodec': 'none', + 'acodec': stream.get('codec'), + 'abr': stream.get('bitrate'), + }) + formats.append(f) + return formats + def _real_extract(self, url): m = re.match(self._VALID_URL, url) - video_id = m.group('videoID') + video_id = m.group('id') - # some sites use this embed format (see: http://github.com/rg3/youtube-dl/issues/2990) + # some sites use this embed format (see: https://github.com/rg3/youtube-dl/issues/2990) if m.group('type') == 'embed/recorded': - video_id = m.group('videoID') + video_id = m.group('id') 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') + video_id = m.group('id') 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') + content_video_ids = self._parse_json(self._search_regex( + r'ustream\.vars\.offAirContentVideoIds=([^;]+);', webpage, + 'content video IDs'), video_id) + return self.playlist_result( + map(lambda u: self.url_result('http://www.ustream.tv/recorded/' + u, 'Ustream'), content_video_ids), + video_id) params = self._download_json( - 'http://cdngw.ustream.tv/rgwjson/Viewer.getVideo/' + json.dumps({ - 'brandId': 1, - 'videoId': int(video_id), - 'autoplay': False, - }), video_id) + 'https://api.ustream.tv/videos/%s.json' % video_id, video_id) - if 'error' in params: - raise ExtractorError(params['error']['message'], expected=True) + error = params.get('error') + if error: + raise ExtractorError( + '%s returned error: %s' % (self.IE_NAME, error), expected=True) - video_url = params['flv'] + video = params['video'] - webpage = self._download_webpage(url, video_id) + title = video['title'] + filesize = float_or_none(video.get('file_size')) - self.report_extraction(video_id) + formats = [{ + 'id': video_id, + 'url': video_url, + 'ext': format_id, + 'filesize': filesize, + } for format_id, video_url in video['media_urls'].items() if video_url] - video_title = self._html_search_regex(r'data-title="(?P.+)"', - webpage, 'title', default=None) + if not formats: + hls_streams = self._get_streams(url, video_id, app_id_ver=(11, 2)) + if hls_streams: + # m3u8_native leads to intermittent ContentTooShortError + formats.extend(self._extract_m3u8_formats( + hls_streams[0]['url'], video_id, ext='mp4', m3u8_id='hls')) - if not video_title: - try: - video_title = params['moduleConfig']['meta']['title'] - except KeyError: - pass + ''' + # DASH streams handling is incomplete as 'url' is missing + dash_streams = self._get_streams(url, video_id, app_id_ver=(3, 1)) + if dash_streams: + formats.extend(self._parse_segmented_mp4(dash_streams)) + ''' - if not video_title: - video_title = 'Ustream video ' + video_id + self._sort_formats(formats) - uploader = self._html_search_regex(r'data-content-type="channel".*?>(?P<uploader>.*?)</a>', - webpage, 'uploader', fatal=False, flags=re.DOTALL, default=None) + description = video.get('description') + timestamp = int_or_none(video.get('created_at')) + duration = float_or_none(video.get('length')) + view_count = int_or_none(video.get('views')) - if not uploader: - try: - uploader = params['moduleConfig']['meta']['userName'] - except KeyError: - uploader = None + uploader = video.get('owner', {}).get('username') + uploader_id = video.get('owner', {}).get('id') - thumbnail = self._html_search_regex(r'<link rel="image_src" href="(?P<thumb>.*?)"', - webpage, 'thumbnail', fatal=False) + thumbnails = [{ + 'id': thumbnail_id, + 'url': thumbnail_url, + } for thumbnail_id, thumbnail_url in video.get('thumbnail', {}).items()] return { 'id': video_id, - 'url': video_url, - 'ext': 'flv', - 'title': video_title, + 'title': title, + 'description': description, + 'thumbnails': thumbnails, + 'timestamp': timestamp, + 'duration': duration, + 'view_count': view_count, 'uploader': uploader, - 'thumbnail': thumbnail, + 'uploader_id': uploader_id, + 'formats': formats, } class UstreamChannelIE(InfoExtractor): - _VALID_URL = r'https?://www\.ustream\.tv/channel/(?P<slug>.+)' + _VALID_URL = r'https?://(?:www\.)?ustream\.tv/channel/(?P<slug>.+)' IE_NAME = 'ustream:channel' _TEST = { 'url': 'http://www.ustream.tv/channel/channeljapan',