[smotri] Add an extractor for live rtmp broadcasts

youtube_dl/FileDownloader.py

@@ -227,6 +227,22 @@ class FileDownloader(object):
         self.to_console_title(u'youtube-dl - %s of %s at %s ETA %s' %
                 (percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip()))
         
+    def report_progress_live_stream(self, downloaded_data_len, speed, elapsed):
+        if self.params.get('noprogress', False):
+            return
+        clear_line = (u'\x1b[K' if sys.stderr.isatty() and os.name != 'nt' else u'')
+        downloaded_str = format_bytes(downloaded_data_len)
+        speed_str = self.format_speed(speed)
+        elapsed_str = FileDownloader.format_seconds(elapsed)
+        if self.params.get('progress_with_newline', False):
+            self.to_screen(u'[download] %s at %s' %
+                (downloaded_str, speed_str))
+        else:
+            self.to_screen(u'\r%s[download] %s at %s ET %s' %
+                (clear_line, downloaded_str, speed_str, elapsed_str), skip_eol=True)
+        self.to_console_title(u'youtube-dl - %s at %s ET %s' %
+                (downloaded_str.strip(), speed_str.strip(), elapsed_str.strip())) 
+
     def report_resuming_byte(self, resume_len):
         """Report attempt to resume at given byte."""
         self.to_screen(u'[download] Resuming download at byte %s' % resume_len)
@@ -255,7 +271,7 @@ class FileDownloader(object):
             self.to_screen(u'\r%s[download] 100%% of %s in %s' %
                 (clear_line, data_len_str, self.format_seconds(tot_time)))
 
-    def _download_with_rtmpdump(self, filename, url, player_url, page_url, play_path, tc_url, live):
+    def _download_with_rtmpdump(self, filename, url, player_url, page_url, play_path, tc_url, live, conn):
         def run_rtmpdump(args):
             start = time.time()
             resume_percent = None
@@ -301,6 +317,22 @@ class FileDownloader(object):
                         'eta': eta,
                         'speed': speed,
                     })
+                else:
+                    # no percent for live streams
+                    mobj = re.search(r'([0-9]+\.[0-9]{3}) kB / [0-9]+\.[0-9]{2} sec', line)
+                    if mobj:
+                        downloaded_data_len = int(float(mobj.group(1))*1024)
+                        time_now = time.time()
+                        speed = self.calc_speed(start, time_now, downloaded_data_len)
+                        self.report_progress_live_stream(downloaded_data_len, speed, time_now - start)
+                        cursor_in_new_line = False
+                        self._hook_progress({
+                            'downloaded_bytes': downloaded_data_len,
+                            'tmpfilename': tmpfilename,
+                            'filename': filename,
+                            'status': 'downloading',
+                            'speed': speed,
+                        })
                     elif self.params.get('verbose', False):
                         if not cursor_in_new_line:
                             self.to_screen(u'')
@@ -338,6 +370,8 @@ class FileDownloader(object):
             basic_args += ['--stop', '1']
         if live:
             basic_args += ['--live']
+        if conn:
+            basic_args += ['--conn', conn]
         args = basic_args + [[], ['--resume', '--skip', '1']][self.params.get('continuedl', False)]
 
         if sys.platform == 'win32' and sys.version_info < (3, 0):
@@ -479,7 +513,8 @@ class FileDownloader(object):
                                                 info_dict.get('page_url', None),
                                                 info_dict.get('play_path', None),
                                                 info_dict.get('tc_url', None),
-                                                info_dict.get('rtmp_live', False))
+                                                info_dict.get('rtmp_live', False),
+                                                info_dict.get('rtmp_conn', None))
 
         # Attempt to download using mplayer
         if url.startswith('mms') or url.startswith('rtsp'):

youtube_dl/extractor/__init__.py

@@ -125,6 +125,7 @@ from .smotri import (
     SmotriIE,
     SmotriCommunityIE,
     SmotriUserIE,
+    SmotriBroadcastIE,
 )
 from .sohu import SohuIE
 from .soundcloud import SoundcloudIE, SoundcloudSetIE, SoundcloudUserIE

youtube_dl/extractor/smotri.py

@@ -3,10 +3,12 @@
 import re
 import json
 import hashlib
+import uuid
 
 from .common import InfoExtractor
 from ..utils import (
-    determine_ext,
+    compat_urllib_parse,
+    compat_urllib_request,
     ExtractorError
 )
 
@@ -250,3 +252,105 @@ class SmotriUserIE(InfoExtractor):
             u'user nickname')
 
         return self.playlist_result(entries, user_id, user_nickname)
+
+
+class SmotriBroadcastIE(InfoExtractor):
+    IE_DESC = u'Smotri.com broadcasts'
+    IE_NAME = u'smotri:broadcast'
+    _VALID_URL = r'^https?://(?:www\.)?(?P<url>smotri\.com/live/(?P<broadcastid>[^/]+))/?.*'
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        broadcast_id = mobj.group('broadcastid')
+
+        broadcast_url = 'http://' + mobj.group('url')
+        broadcast_page = self._download_webpage(broadcast_url, broadcast_id, u'Downloading broadcast page')
+
+        if re.search(u'>Режиссер с логином <br/>"%s"<br/> <span>не существует<' % broadcast_id, broadcast_page) is not None:
+            raise ExtractorError(u'Broadcast %s does not exist' % broadcast_id, expected=True)
+
+        # Adult content
+        if re.search(u'EroConfirmText">', broadcast_page) is not None:
+
+            (username, password) = self._get_login_info()
+            if username is None:
+                raise ExtractorError(u'Erotic broadcasts allowed only for registered users, '
+                    u'use --username and --password options to provide account credentials.', expected=True)
+
+            # Log in
+            login_form_strs = {
+                u'login-hint53': '1',
+                u'confirm_erotic': '1',
+                u'login': username,
+                u'password': password,
+            }
+            # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
+            # chokes on unicode
+            login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
+            login_data = compat_urllib_parse.urlencode(login_form).encode('utf-8')
+            login_url = broadcast_url + '/?no_redirect=1'
+            request = compat_urllib_request.Request(login_url, login_data)
+            request.add_header('Content-Type', 'application/x-www-form-urlencoded')
+            broadcast_page = self._download_webpage(
+                request, broadcast_id, note=u'Logging in and confirming age')
+
+            if re.search(u'>Неверный логин или пароль<', broadcast_page) is not None:
+                raise ExtractorError(u'Unable to log in: bad username or password', expected=True)
+
+            adult_content = True
+        else:
+            adult_content = False
+
+        ticket = self._html_search_regex(
+            u'window.broadcast_control.addFlashVar\\(\'file\', \'([^\']+)\'\\);',
+            broadcast_page, u'broadcast ticket')
+
+        url = 'http://smotri.com/broadcast/view/url/?ticket=%s' % ticket
+
+        broadcast_password = self._downloader.params.get('videopassword', None)
+        if broadcast_password:
+            url += '&pass=%s' % hashlib.md5(broadcast_password.encode('utf-8')).hexdigest()
+
+        broadcast_json_page = self._download_webpage(url, broadcast_id, u'Downloading broadcast JSON')
+
+        try:
+            broadcast_json = json.loads(broadcast_json_page)
+
+            protected_broadcast = broadcast_json['_pass_protected'] == 1
+            if protected_broadcast and not broadcast_password:
+                raise ExtractorError(u'This broadcast is protected by a password, use the --video-password option', expected=True)
+
+            broadcast_offline = broadcast_json['is_play'] == 0
+            if broadcast_offline:
+                raise ExtractorError(u'Broadcast %s is offline' % broadcast_id, expected=True)
+
+            rtmp_url = broadcast_json['_server']
+            if not rtmp_url.startswith('rtmp://'):
+                raise ExtractorError(u'Unexpected broadcast rtmp URL')
+
+            broadcast_playpath = broadcast_json['_streamName']
+            broadcast_thumbnail = broadcast_json['_imgURL']
+            broadcast_title = broadcast_json['title']
+            broadcast_description = broadcast_json['description']
+            broadcaster_nick = broadcast_json['nick']
+            broadcaster_login = broadcast_json['login']
+            rtmp_conn = 'S:%s' % uuid.uuid4().hex
+        except KeyError:
+            if protected_broadcast:
+                raise ExtractorError(u'Bad broadcast password', expected=True)
+            raise ExtractorError(u'Unexpected broadcast JSON')
+
+        return {
+            'id': broadcast_id,
+            'url': rtmp_url,
+            'title': broadcast_title,
+            'thumbnail': broadcast_thumbnail,
+            'description': broadcast_description,
+            'uploader': broadcaster_nick,
+            'uploader_id': broadcaster_login,
+            'age_limit': 18 if adult_content else 0,
+            'ext': 'flv',
+            'play_path': broadcast_playpath,
+            'rtmp_live': True,
+            'rtmp_conn': rtmp_conn
+        }