]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/livestream.py
Rapsys Git / youtubedl / blobdiff
? search:
summary | shortlog | log | commit | commitdiff | tree
raw | inline | side by side
Merge tag 'upstream/2017.03.07'
[youtubedl] / youtube_dl / extractor / livestream.py
diff --git a/youtube_dl/extractor/livestream.py b/youtube_dl/extractor/livestream.py
index 2c100d424650fed5d98330b1e5124df117296c75..c863413bf008baa6baf0233b8185a10ea119d091 100644 (file)
--- a/youtube_dl/extractor/livestream.py
+++ b/youtube_dl/extractor/livestream.py
@@ -1,128 +1,353 @@
 from __future__ import unicode_literals
 
 import re
 from __future__ import unicode_literals
 
 import re
-import json
+import itertools
 
 from .common import InfoExtractor
 
 from .common import InfoExtractor
-from ..utils import (
-    compat_urllib_parse_urlparse,
+from ..compat import (
+    compat_str,
     compat_urlparse,
     compat_urlparse,
+)
+from ..utils import (
+    find_xpath_attr,
+    xpath_attr,
     xpath_with_ns,
     xpath_with_ns,
-    compat_str,
+    xpath_text,
     orderedSet,
     orderedSet,
+    update_url_query,
+    int_or_none,
+    float_or_none,
+    parse_iso8601,
+    determine_ext,
 )
 
 
 class LivestreamIE(InfoExtractor):
     IE_NAME = 'livestream'
 )
 
 
 class LivestreamIE(InfoExtractor):
     IE_NAME = 'livestream'
-    _VALID_URL = r'http://new\.livestream\.com/.*?/(?P<event_name>.*?)(/videos/(?P<id>\d+))?/?$'
-    _TEST = {
+    _VALID_URL = r'https?://(?:new\.)?livestream\.com/(?:accounts/(?P<account_id>\d+)|(?P<account_name>[^/]+))/(?:events/(?P<event_id>\d+)|(?P<event_name>[^/]+))(?:/videos/(?P<id>\d+))?'
+    _TESTS = [{
         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
         'md5': '53274c76ba7754fb0e8d072716f2292b',
         'info_dict': {
             'id': '4719370',
             'ext': 'mp4',
             'title': 'Live from Webster Hall NYC',
         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
         'md5': '53274c76ba7754fb0e8d072716f2292b',
         'info_dict': {
             'id': '4719370',
             'ext': 'mp4',
             'title': 'Live from Webster Hall NYC',
+            'timestamp': 1350008072,
             'upload_date': '20121012',
             'upload_date': '20121012',
+            'duration': 5968.0,
+            'like_count': int,
+            'view_count': int,
+            'thumbnail': r're:^http://.*\.jpg$'
         }
         }
-    }
+    }, {
+        '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,
+    }, {
+        'url': 'http://livestream.com/bsww/concacafbeachsoccercampeonato2015',
+        'only_matching': True,
+    }]
+    _API_URL_TEMPLATE = 'http://livestream.com/api/accounts/%s/events/%s'
+
+    def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
+        base_ele = find_xpath_attr(
+            smil, self._xpath_ns('.//meta', namespace), 'name', 'httpBase')
+        base = base_ele.get('content') if base_ele is not None else 'http://livestreamvod-f.akamaihd.net/'
+
+        formats = []
+        video_nodes = smil.findall(self._xpath_ns('.//video', namespace))
+
+        for vn in video_nodes:
+            tbr = int_or_none(vn.attrib.get('system-bitrate'), 1000)
+            furl = (
+                update_url_query(compat_urlparse.urljoin(base, vn.attrib['src']), {
+                    'v': '3.0.3',
+                    'fp': 'WIN% 14,0,0,145',
+                }))
+            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):
 
     def _extract_video_info(self, video_data):
-        video_url = video_data.get('progressive_url_hd') or video_data.get('progressive_url')
+        video_id = compat_str(video_data['id'])
+
+        FORMAT_KEYS = (
+            ('sd', 'progressive_url'),
+            ('hd', 'progressive_url_hd'),
+        )
+
+        formats = []
+        for format_id, key in FORMAT_KEYS:
+            video_url = video_data.get(key)
+            if video_url:
+                ext = determine_ext(video_url)
+                if ext == 'm3u8':
+                    continue
+                bitrate = int_or_none(self._search_regex(
+                    r'(\d+)\.%s' % ext, video_url, 'bitrate', default=None))
+                formats.append({
+                    'url': video_url,
+                    'format_id': format_id,
+                    'tbr': bitrate,
+                    'ext': ext,
+                })
+
+        smil_url = video_data.get('smil_url')
+        if smil_url:
+            formats.extend(self._extract_smil_formats(smil_url, video_id))
+
+        m3u8_url = video_data.get('m3u8_url')
+        if m3u8_url:
+            formats.extend(self._extract_m3u8_formats(
+                m3u8_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
+
+        f4m_url = video_data.get('f4m_url')
+        if f4m_url:
+            formats.extend(self._extract_f4m_formats(
+                f4m_url, video_id, f4m_id='hds', fatal=False))
+        self._sort_formats(formats)
+
+        comments = [{
+            'author_id': comment.get('author_id'),
+            'author': comment.get('author', {}).get('full_name'),
+            'id': comment.get('id'),
+            'text': comment['text'],
+            'timestamp': parse_iso8601(comment.get('created_at')),
+        } for comment in video_data.get('comments', {}).get('data', [])]
+
         return {
         return {
-            'id': compat_str(video_data['id']),
-            'url': video_url,
-            'ext': 'mp4',
+            'id': video_id,
+            'formats': formats,
             'title': video_data['caption'],
             'title': video_data['caption'],
-            'thumbnail': video_data['thumbnail_url'],
-            'upload_date': video_data['updated_at'].replace('-', '')[:8],
+            'description': video_data.get('description'),
+            'thumbnail': video_data.get('thumbnail_url'),
+            'duration': float_or_none(video_data.get('duration'), 1000),
+            'timestamp': parse_iso8601(video_data.get('publish_at')),
+            'like_count': video_data.get('likes', {}).get('total'),
+            'comment_count': video_data.get('comments', {}).get('total'),
+            'view_count': video_data.get('views'),
+            'comments': comments,
         }
 
         }
 
+    def _extract_stream_info(self, stream_info):
+        broadcast_id = compat_str(stream_info['broadcast_id'])
+        is_live = stream_info.get('is_live')
+
+        formats = []
+        smil_url = stream_info.get('play_url')
+        if smil_url:
+            formats.extend(self._extract_smil_formats(smil_url, broadcast_id))
+
+        entry_protocol = 'm3u8' if is_live else 'm3u8_native'
+        m3u8_url = stream_info.get('m3u8_url')
+        if m3u8_url:
+            formats.extend(self._extract_m3u8_formats(
+                m3u8_url, broadcast_id, 'mp4', entry_protocol, m3u8_id='hls', fatal=False))
+
+        rtsp_url = stream_info.get('rtsp_url')
+        if rtsp_url:
+            formats.append({
+                'url': rtsp_url,
+                'format_id': 'rtsp',
+            })
+        self._sort_formats(formats)
+
+        return {
+            'id': broadcast_id,
+            'formats': formats,
+            'title': self._live_title(stream_info['stream_title']) if is_live else stream_info['stream_title'],
+            'thumbnail': stream_info.get('thumbnail_url'),
+            'is_live': is_live,
+        }
+
+    def _extract_event(self, event_data):
+        event_id = compat_str(event_data['id'])
+        account_id = compat_str(event_data['owner_account_id'])
+        feed_root_url = self._API_URL_TEMPLATE % (account_id, event_id) + '/feed.json'
+
+        stream_info = event_data.get('stream_info')
+        if stream_info:
+            return self._extract_stream_info(stream_info)
+
+        last_video = None
+        entries = []
+        for i in itertools.count(1):
+            if last_video is None:
+                info_url = feed_root_url
+            else:
+                info_url = '{root}?&id={id}&newer=-1&type=video'.format(
+                    root=feed_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:
+                v_id = compat_str(v['id'])
+                entries.append(self.url_result(
+                    'http://livestream.com/accounts/%s/events/%s/videos/%s' % (account_id, event_id, v_id),
+                    'Livestream', v_id, v.get('caption')))
+            last_video = videos_info[-1]['id']
+        return self.playlist_result(entries, event_id, event_data['full_name'])
+
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
-        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, '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'] == 'video']
-            return self.playlist_result(videos, info['id'], info['full_name'])
+        event = mobj.group('event_id') or mobj.group('event_name')
+        account = mobj.group('account_id') or mobj.group('account_name')
+        api_url = self._API_URL_TEMPLATE % (account, event)
+        if video_id:
+            video_data = self._download_json(
+                api_url + '/videos/%s' % video_id, video_id)
+            return self._extract_video_info(video_data)
         else:
         else:
-            og_video = self._og_search_video_url(webpage, '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, 'Downloading video info'))
-            return self._extract_video_info(info)
+            event_data = self._download_json(api_url, video_id)
+            return self._extract_event(event_data)
 
 
 # The original version of Livestream uses a different system
 class LivestreamOriginalIE(InfoExtractor):
     IE_NAME = 'livestream:original'
 
 
 # The original version of Livestream uses a different system
 class LivestreamOriginalIE(InfoExtractor):
     IE_NAME = 'livestream:original'
-    _VALID_URL = r'''(?x)https?://www\.livestream\.com/
-        (?P<user>[^/]+)/(?P<type>video|folder)
-        (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
+    _VALID_URL = r'''(?x)https?://original\.livestream\.com/
+        (?P<user>[^/\?#]+)(?:/(?P<type>video|folder)
+        (?:(?:\?.*?Id=|/)(?P<id>.*?)(&|$))?)?
         '''
         '''
-    _TEST = {
-        'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
+    _TESTS = [{
+        'url': 'http://original.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
         'info_dict': {
             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
         'info_dict': {
             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
-            'ext': 'flv',
+            'ext': 'mp4',
             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
+            'duration': 771.301,
+            'view_count': int,
         },
         },
-        'params': {
-            # rtmp
-            'skip_download': True,
+    }, {
+        'url': 'https://original.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
+        'info_dict': {
+            'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
         },
         },
-    }
-
-    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)
+        'playlist_mincount': 4,
+    }, {
+        # live stream
+        'url': 'http://original.livestream.com/znsbahamas',
+        'only_matching': True,
+    }]
 
 
+    def _extract_video_info(self, user, video_id):
+        api_url = 'http://x%sx.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id=%s' % (user, video_id)
         info = self._download_xml(api_url, video_id)
         info = self._download_xml(api_url, video_id)
+
         item = info.find('channel').find('item')
         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, 'path')
+        title = xpath_text(item, 'title')
+        media_ns = {'media': 'http://search.yahoo.com/mrss'}
+        thumbnail_url = xpath_attr(
+            item, xpath_with_ns('media:thumbnail', media_ns), 'url')
+        duration = float_or_none(xpath_attr(
+            item, xpath_with_ns('media:content', media_ns), 'duration'))
+        ls_ns = {'ls': 'http://api.channel.livestream.com/2.0'}
+        view_count = int_or_none(xpath_text(
+            item, xpath_with_ns('ls:viewsCount', ls_ns)))
 
         return {
             'id': video_id,
 
         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),
-            'ext': 'flv',
+            'title': title,
             'thumbnail': thumbnail_url,
             'thumbnail': thumbnail_url,
+            'duration': duration,
+            'view_count': view_count,
         }
 
         }
 
+    def _extract_video_formats(self, video_data, video_id, entry_protocol):
+        formats = []
+
+        progressive_url = video_data.get('progressiveUrl')
+        if progressive_url:
+            formats.append({
+                'url': progressive_url,
+                'format_id': 'http',
+            })
+
+        m3u8_url = video_data.get('httpUrl')
+        if m3u8_url:
+            formats.extend(self._extract_m3u8_formats(
+                m3u8_url, video_id, 'mp4', entry_protocol, m3u8_id='hls', fatal=False))
+
+        rtsp_url = video_data.get('rtspUrl')
+        if rtsp_url:
+            formats.append({
+                'url': rtsp_url,
+                'format_id': 'rtsp',
+            })
+
+        self._sort_formats(formats)
+        return formats
+
     def _extract_folder(self, url, folder_id):
         webpage = self._download_webpage(url, folder_id)
     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))
+        paths = orderedSet(re.findall(
+            r'''(?x)(?:
+                <li\s+class="folder">\s*<a\s+href="|
+                <a\s+href="(?=https?://livestre\.am/)
+            )([^"]+)"''', webpage))
 
 
-        return {
-            '_type': 'playlist',
-            'id': folder_id,
-            'entries': [{
-                '_type': 'url',
-                'url': video_url,
-            } for video_url in urls],
-        }
+        entries = [{
+            '_type': 'url',
+            'url': compat_urlparse.urljoin(url, p),
+        } for p in paths]
+
+        return self.playlist_result(entries, folder_id)
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
 
     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')
         user = mobj.group('user')
         url_type = mobj.group('type')
+        content_id = mobj.group('id')
         if url_type == 'folder':
         if url_type == 'folder':
-            return self._extract_folder(url, id)
+            return self._extract_folder(url, content_id)
         else:
         else:
-            return self._extract_video(user, id)
+            # this url is used on mobile devices
+            stream_url = 'http://x%sx.api.channel.livestream.com/3.0/getstream.json' % user
+            info = {}
+            if content_id:
+                stream_url += '?id=%s' % content_id
+                info = self._extract_video_info(user, content_id)
+            else:
+                content_id = user
+                webpage = self._download_webpage(url, content_id)
+                info = {
+                    'title': self._og_search_title(webpage),
+                    'description': self._og_search_description(webpage),
+                    'thumbnail': self._search_regex(r'channelLogo.src\s*=\s*"([^"]+)"', webpage, 'thumbnail', None),
+                }
+            video_data = self._download_json(stream_url, content_id)
+            is_live = video_data.get('isLive')
+            entry_protocol = 'm3u8' if is_live else 'm3u8_native'
+            info.update({
+                'id': content_id,
+                'title': self._live_title(info['title']) if is_live else info['title'],
+                'formats': self._extract_video_formats(video_data, content_id, entry_protocol),
+                'is_live': is_live,
+            })
+            return info
 
 
 # The server doesn't support HEAD request, the generic extractor can't detect
 
 
 # The server doesn't support HEAD request, the generic extractor can't detect
Backup clone of https://github.com/ytdl-org unavailable since RIAA fucking raid...