# encoding: utf-8
+from __future__ import unicode_literals
-import os.path
import re
import json
import hashlib
import uuid
from .common import InfoExtractor
-from ..utils import (
+from ..compat import (
compat_urllib_parse,
compat_urllib_request,
+)
+from ..utils import (
ExtractorError,
- url_basename,
+ int_or_none,
+ unified_strdate,
)
class SmotriIE(InfoExtractor):
- IE_DESC = u'Smotri.com'
- IE_NAME = u'smotri'
- _VALID_URL = r'^https?://(?:www\.)?(?P<url>smotri\.com/video/view/\?id=(?P<videoid>v(?P<realvideoid>[0-9]+)[a-z0-9]{4}))'
+ IE_DESC = 'Smotri.com'
+ IE_NAME = 'smotri'
+ _VALID_URL = r'^https?://(?:www\.)?(?:smotri\.com/video/view/\?id=|pics\.smotri\.com/(?:player|scrubber_custom8)\.swf\?file=)(?P<id>v(?P<realvideoid>[0-9]+)[a-z0-9]{4})'
+ _NETRC_MACHINE = 'smotri'
_TESTS = [
# real video id 2610366
{
- u'url': u'http://smotri.com/video/view/?id=v261036632ab',
- u'file': u'v261036632ab.mp4',
- u'md5': u'2a7b08249e6f5636557579c368040eb9',
- u'info_dict': {
- u'title': u'катастрофа с камер видеонаблюдения',
- u'uploader': u'rbc2008',
- u'uploader_id': u'rbc08',
- u'upload_date': u'20131118',
- u'description': u'катастрофа с камер видеонаблюдения, видео катастрофа с камер видеонаблюдения',
- u'thumbnail': u'http://frame6.loadup.ru/8b/a9/2610366.3.3.jpg',
+ 'url': 'http://smotri.com/video/view/?id=v261036632ab',
+ 'md5': '2a7b08249e6f5636557579c368040eb9',
+ 'info_dict': {
+ 'id': 'v261036632ab',
+ 'ext': 'mp4',
+ 'title': 'катастрофа с камер видеонаблюдения',
+ 'uploader': 'rbc2008',
+ 'uploader_id': 'rbc08',
+ 'upload_date': '20131118',
+ 'thumbnail': 'http://frame6.loadup.ru/8b/a9/2610366.3.3.jpg',
},
},
# real video id 57591
{
- u'url': u'http://smotri.com/video/view/?id=v57591cb20',
- u'file': u'v57591cb20.flv',
- u'md5': u'830266dfc21f077eac5afd1883091bcd',
- u'info_dict': {
- u'title': u'test',
- u'uploader': u'Support Photofile@photofile',
- u'uploader_id': u'support-photofile',
- u'upload_date': u'20070704',
- u'description': u'test, видео test',
- u'thumbnail': u'http://frame4.loadup.ru/03/ed/57591.2.3.jpg',
+ 'url': 'http://smotri.com/video/view/?id=v57591cb20',
+ 'md5': '830266dfc21f077eac5afd1883091bcd',
+ 'info_dict': {
+ 'id': 'v57591cb20',
+ 'ext': 'flv',
+ 'title': 'test',
+ 'uploader': 'Support Photofile@photofile',
+ 'uploader_id': 'support-photofile',
+ 'upload_date': '20070704',
+ 'thumbnail': 'http://frame4.loadup.ru/03/ed/57591.2.3.jpg',
+ },
+ },
+ # video-password, not approved by moderator
+ {
+ 'url': 'http://smotri.com/video/view/?id=v1390466a13c',
+ 'md5': 'f6331cef33cad65a0815ee482a54440b',
+ 'info_dict': {
+ 'id': 'v1390466a13c',
+ 'ext': 'mp4',
+ 'title': 'TOCCA_A_NOI_-_LE_COSE_NON_VANNO_CAMBIAMOLE_ORA-1',
+ 'uploader': 'timoxa40',
+ 'uploader_id': 'timoxa40',
+ 'upload_date': '20100404',
+ 'thumbnail': 'http://frame7.loadup.ru/af/3f/1390466.3.3.jpg',
+ },
+ 'params': {
+ 'videopassword': 'qwerty',
},
+ 'skip': 'Video is not approved by moderator',
},
# video-password
{
- u'url': u'http://smotri.com/video/view/?id=v1390466a13c',
- u'file': u'v1390466a13c.mp4',
- u'md5': u'f6331cef33cad65a0815ee482a54440b',
- u'info_dict': {
- u'title': u'TOCCA_A_NOI_-_LE_COSE_NON_VANNO_CAMBIAMOLE_ORA-1',
- u'uploader': u'timoxa40',
- u'uploader_id': u'timoxa40',
- u'upload_date': u'20100404',
- u'thumbnail': u'http://frame7.loadup.ru/af/3f/1390466.3.3.jpg',
- u'description': u'TOCCA_A_NOI_-_LE_COSE_NON_VANNO_CAMBIAMOLE_ORA-1, видео TOCCA_A_NOI_-_LE_COSE_NON_VANNO_CAMBIAMOLE_ORA-1',
+ 'url': 'http://smotri.com/video/view/?id=v6984858774#',
+ 'md5': 'f11e01d13ac676370fc3b95b9bda11b0',
+ 'info_dict': {
+ 'id': 'v6984858774',
+ 'ext': 'mp4',
+ 'title': 'Дача Солженицина ПАРОЛЬ 223322',
+ 'uploader': 'psavari1',
+ 'uploader_id': 'psavari1',
+ 'upload_date': '20081103',
+ 'thumbnail': 're:^https?://.*\.jpg$',
},
- u'params': {
- u'videopassword': u'qwerty',
+ 'params': {
+ 'videopassword': '223322',
},
},
+ # age limit + video-password, not approved by moderator
+ {
+ 'url': 'http://smotri.com/video/view/?id=v15408898bcf',
+ 'md5': '91e909c9f0521adf5ee86fbe073aad70',
+ 'info_dict': {
+ 'id': 'v15408898bcf',
+ 'ext': 'flv',
+ 'title': 'этот ролик не покажут по ТВ',
+ 'uploader': 'zzxxx',
+ 'uploader_id': 'ueggb',
+ 'upload_date': '20101001',
+ 'thumbnail': 'http://frame3.loadup.ru/75/75/1540889.1.3.jpg',
+ 'age_limit': 18,
+ },
+ 'params': {
+ 'videopassword': '333'
+ },
+ 'skip': 'Video is not approved by moderator',
+ },
# age limit + video-password
{
- u'url': u'http://smotri.com/video/view/?id=v15408898bcf',
- u'file': u'v15408898bcf.flv',
- u'md5': u'91e909c9f0521adf5ee86fbe073aad70',
- u'info_dict': {
- u'title': u'этот ролик не покажут по ТВ',
- u'uploader': u'zzxxx',
- u'uploader_id': u'ueggb',
- u'upload_date': u'20101001',
- u'thumbnail': u'http://frame3.loadup.ru/75/75/1540889.1.3.jpg',
- u'age_limit': 18,
- u'description': u'этот ролик не покажут по ТВ, видео этот ролик не покажут по ТВ',
+ 'url': 'http://smotri.com/video/view/?id=v7780025814',
+ 'md5': 'b4599b068422559374a59300c5337d72',
+ 'info_dict': {
+ 'id': 'v7780025814',
+ 'ext': 'mp4',
+ 'title': 'Sexy Beach (пароль 123)',
+ 'uploader': 'вАся',
+ 'uploader_id': 'asya_prosto',
+ 'upload_date': '20081218',
+ 'thumbnail': 're:^https?://.*\.jpg$',
+ 'age_limit': 18,
},
- u'params': {
- u'videopassword': u'333'
- }
- }
+ 'params': {
+ 'videopassword': '123'
+ },
+ },
+ # swf player
+ {
+ 'url': 'http://pics.smotri.com/scrubber_custom8.swf?file=v9188090500',
+ 'md5': '31099eeb4bc906712c5f40092045108d',
+ 'info_dict': {
+ 'id': 'v9188090500',
+ 'ext': 'mp4',
+ 'title': 'Shakira - Don\'t Bother',
+ 'uploader': 'HannahL',
+ 'uploader_id': 'lisaha95',
+ 'upload_date': '20090331',
+ 'thumbnail': 'http://frame8.loadup.ru/44/0b/918809.7.3.jpg',
+ },
+ },
]
-
- _SUCCESS = 0
- _PASSWORD_NOT_VERIFIED = 1
- _PASSWORD_DETECTED = 2
- _VIDEO_NOT_FOUND = 3
+
+ @classmethod
+ def _extract_url(cls, webpage):
+ mobj = re.search(
+ r'<embed[^>]src=(["\'])(?P<url>http://pics\.smotri\.com/(?:player|scrubber_custom8)\.swf\?file=v.+?\1)',
+ webpage)
+ if mobj is not None:
+ return mobj.group('url')
+
+ mobj = re.search(
+ r'''(?x)<div\s+class="video_file">http://smotri\.com/video/download/file/[^<]+</div>\s*
+ <div\s+class="video_image">[^<]+</div>\s*
+ <div\s+class="video_id">(?P<id>[^<]+)</div>''', webpage)
+ if mobj is not None:
+ return 'http://smotri.com/video/view/?id=%s' % mobj.group('id')
def _search_meta(self, name, html, display_name=None):
if display_name is None:
display_name = name
- return self._html_search_regex(
- r'<meta itemprop="%s" content="([^"]+)" />' % re.escape(name),
- html, display_name, fatal=False)
return self._html_search_meta(name, html, display_name)
def _real_extract(self, url):
- mobj = re.match(self._VALID_URL, url)
- video_id = mobj.group('videoid')
- real_video_id = mobj.group('realvideoid')
-
- # Download video JSON data
- video_json_url = 'http://smotri.com/vt.php?id=%s' % real_video_id
- video_json_page = self._download_webpage(video_json_url, video_id, u'Downloading video JSON')
- video_json = json.loads(video_json_page)
-
- status = video_json['status']
- if status == self._VIDEO_NOT_FOUND:
- raise ExtractorError(u'Video %s does not exist' % video_id, expected=True)
- elif status == self._PASSWORD_DETECTED: # The video is protected by a password, retry with
- # video-password set
- video_password = self._downloader.params.get('videopassword', None)
- if not video_password:
- raise ExtractorError(u'This video is protected by a password, use the --video-password option', expected=True)
- video_json_url += '&md5pass=%s' % hashlib.md5(video_password.encode('utf-8')).hexdigest()
- video_json_page = self._download_webpage(video_json_url, video_id, u'Downloading video JSON (video-password set)')
- video_json = json.loads(video_json_page)
- status = video_json['status']
- if status == self._PASSWORD_NOT_VERIFIED:
- raise ExtractorError(u'Video password is invalid', expected=True)
-
- if status != self._SUCCESS:
- raise ExtractorError(u'Unexpected status value %s' % status)
-
- # Extract the URL of the video
- video_url = video_json['file_data']
-
+ video_id = self._match_id(url)
+
+ video_form = {
+ 'ticket': video_id,
+ 'video_url': '1',
+ 'frame_url': '1',
+ 'devid': 'LoadupFlashPlayer',
+ 'getvideoinfo': '1',
+ }
+
+ video_password = self._downloader.params.get('videopassword', None)
+ if video_password:
+ video_form['pass'] = hashlib.md5(video_password.encode('utf-8')).hexdigest()
+
+ request = compat_urllib_request.Request(
+ 'http://smotri.com/video/view/url/bot/', compat_urllib_parse.urlencode(video_form))
+ request.add_header('Content-Type', 'application/x-www-form-urlencoded')
+
+ video = self._download_json(request, video_id, 'Downloading video JSON')
+
+ video_url = video.get('_vidURL') or video.get('_vidURL_mp4')
+
+ if not video_url:
+ if video.get('_moderate_no'):
+ raise ExtractorError(
+ 'Video %s has not been approved by moderator' % video_id, expected=True)
+
+ if video.get('error'):
+ raise ExtractorError('Video %s does not exist' % video_id, expected=True)
+
+ if video.get('_pass_protected') == 1:
+ msg = ('Invalid video password' if video_password
+ else 'This video is protected by a password, use the --video-password option')
+ raise ExtractorError(msg, expected=True)
+
+ title = video['title']
+ thumbnail = video['_imgURL']
+ upload_date = unified_strdate(video['added'])
+ uploader = video['userNick']
+ uploader_id = video['userLogin']
+ duration = int_or_none(video['duration'])
+
# Video JSON does not provide enough meta data
# We will extract some from the video web page instead
- video_page_url = 'http://' + mobj.group('url')
- video_page = self._download_webpage(video_page_url, video_id, u'Downloading video page')
+ webpage_url = 'http://smotri.com/video/view/?id=%s' % video_id
+ webpage = self._download_webpage(webpage_url, video_id, 'Downloading video page')
# Warning if video is unavailable
warning = self._html_search_regex(
- r'<div class="videoUnModer">(.*?)</div>', video_page,
- u'warning message', default=None)
+ r'<div class="videoUnModer">(.*?)</div>', webpage,
+ 'warning message', default=None)
if warning is not None:
self._downloader.report_warning(
- u'Video %s may not be available; smotri said: %s ' %
+ 'Video %s may not be available; smotri said: %s ' %
(video_id, warning))
# Adult content
- if re.search(u'EroConfirmText">', video_page) is not None:
+ if re.search('EroConfirmText">', webpage) is not None:
self.report_age_confirmation()
confirm_string = self._html_search_regex(
r'<a href="/video/view/\?id=%s&confirm=([^"]+)" title="[^"]+">' % video_id,
- video_page, u'confirm string')
- confirm_url = video_page_url + '&confirm=%s' % confirm_string
- video_page = self._download_webpage(confirm_url, video_id, u'Downloading video page (age confirmed)')
+ webpage, 'confirm string')
+ confirm_url = webpage_url + '&confirm=%s' % confirm_string
+ webpage = self._download_webpage(confirm_url, video_id, 'Downloading video page (age confirmed)')
adult_content = True
else:
adult_content = False
-
- # Extract the rest of meta data
- video_title = self._search_meta(u'name', video_page, u'title')
- if not video_title:
- video_title = os.path.splitext(url_basename(video_url))[0]
-
- video_description = self._search_meta(u'description', video_page)
- END_TEXT = u' на сайте Smotri.com'
- if video_description and video_description.endswith(END_TEXT):
- video_description = video_description[:-len(END_TEXT)]
- START_TEXT = u'Смотреть онлайн ролик '
- if video_description and video_description.startswith(START_TEXT):
- video_description = video_description[len(START_TEXT):]
- video_thumbnail = self._search_meta(u'thumbnail', video_page)
-
- upload_date_str = self._search_meta(u'uploadDate', video_page, u'upload date')
- if upload_date_str:
- upload_date_m = re.search(r'(?P<year>\d{4})\.(?P<month>\d{2})\.(?P<day>\d{2})T', upload_date_str)
- video_upload_date = (
- (
- upload_date_m.group('year') +
- upload_date_m.group('month') +
- upload_date_m.group('day')
- )
- if upload_date_m else None
- )
- else:
- video_upload_date = None
-
- duration_str = self._search_meta(u'duration', video_page)
- if duration_str:
- duration_m = re.search(r'T(?P<hours>[0-9]{2})H(?P<minutes>[0-9]{2})M(?P<seconds>[0-9]{2})S', duration_str)
- video_duration = (
- (
- (int(duration_m.group('hours')) * 60 * 60) +
- (int(duration_m.group('minutes')) * 60) +
- int(duration_m.group('seconds'))
- )
- if duration_m else None
- )
- else:
- video_duration = None
-
- video_uploader = self._html_search_regex(
- u'<div class="DescrUser"><div>Автор.*?onmouseover="popup_user_info[^"]+">(.*?)</a>',
- video_page, u'uploader', fatal=False, flags=re.MULTILINE|re.DOTALL)
-
- video_uploader_id = self._html_search_regex(
- u'<div class="DescrUser"><div>Автор.*?onmouseover="popup_user_info\\(.*?\'([^\']+)\'\\);">',
- video_page, u'uploader id', fatal=False, flags=re.MULTILINE|re.DOTALL)
-
- video_view_count = self._html_search_regex(
- u'Общее количество просмотров.*?<span class="Number">(\\d+)</span>',
- video_page, u'view count', fatal=False, flags=re.MULTILINE|re.DOTALL)
-
+
+ view_count = self._html_search_regex(
+ 'Общее количество просмотров.*?<span class="Number">(\\d+)</span>',
+ webpage, 'view count', fatal=False, flags=re.MULTILINE | re.DOTALL)
+
return {
'id': video_id,
'url': video_url,
- 'title': video_title,
- 'thumbnail': video_thumbnail,
- 'description': video_description,
- 'uploader': video_uploader,
- 'upload_date': video_upload_date,
- 'uploader_id': video_uploader_id,
- 'duration': video_duration,
- 'view_count': video_view_count,
+ 'title': title,
+ 'thumbnail': thumbnail,
+ 'uploader': uploader,
+ 'upload_date': upload_date,
+ 'uploader_id': uploader_id,
+ 'duration': duration,
+ 'view_count': int_or_none(view_count),
'age_limit': 18 if adult_content else 0,
- 'video_page_url': video_page_url
}
class SmotriCommunityIE(InfoExtractor):
- IE_DESC = u'Smotri.com community videos'
- IE_NAME = u'smotri:community'
+ IE_DESC = 'Smotri.com community videos'
+ IE_NAME = 'smotri:community'
_VALID_URL = r'^https?://(?:www\.)?smotri\.com/community/video/(?P<communityid>[0-9A-Za-z_\'-]+)'
-
+ _TEST = {
+ 'url': 'http://smotri.com/community/video/kommuna',
+ 'info_dict': {
+ 'id': 'kommuna',
+ 'title': 'КПРФ',
+ },
+ 'playlist_mincount': 4,
+ }
+
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
community_id = mobj.group('communityid')
url = 'http://smotri.com/export/rss/video/by/community/-/%s/video.xml' % community_id
- rss = self._download_xml(url, community_id, u'Downloading community RSS')
+ rss = self._download_xml(url, community_id, 'Downloading community RSS')
entries = [self.url_result(video_url.text, 'Smotri')
for video_url in rss.findall('./channel/item/link')]
description_text = rss.find('./channel/description').text
community_title = self._html_search_regex(
- u'^Видео сообщества "([^"]+)"$', description_text, u'community title')
+ '^Видео сообщества "([^"]+)"$', description_text, 'community title')
return self.playlist_result(entries, community_id, community_title)
class SmotriUserIE(InfoExtractor):
- IE_DESC = u'Smotri.com user videos'
- IE_NAME = u'smotri:user'
+ IE_DESC = 'Smotri.com user videos'
+ IE_NAME = 'smotri:user'
_VALID_URL = r'^https?://(?:www\.)?smotri\.com/user/(?P<userid>[0-9A-Za-z_\'-]+)'
+ _TESTS = [{
+ 'url': 'http://smotri.com/user/inspector',
+ 'info_dict': {
+ 'id': 'inspector',
+ 'title': 'Inspector',
+ },
+ 'playlist_mincount': 9,
+ }]
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
user_id = mobj.group('userid')
url = 'http://smotri.com/export/rss/user/video/-/%s/video.xml' % user_id
- rss = self._download_xml(url, user_id, u'Downloading user RSS')
+ rss = self._download_xml(url, user_id, 'Downloading user RSS')
entries = [self.url_result(video_url.text, 'Smotri')
for video_url in rss.findall('./channel/item/link')]
description_text = rss.find('./channel/description').text
user_nickname = self._html_search_regex(
- u'^Видео режиссера (.*)$', description_text,
- u'user nickname')
+ '^Видео режиссера (.*)$', description_text,
+ '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'
+ IE_DESC = 'Smotri.com broadcasts'
+ IE_NAME = 'smotri:broadcast'
_VALID_URL = r'^https?://(?:www\.)?(?P<url>smotri\.com/live/(?P<broadcastid>[^/]+))/?.*'
def _real_extract(self, 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')
+ broadcast_page = self._download_webpage(broadcast_url, broadcast_id, '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)
+ if re.search('>Режиссер с логином <br/>"%s"<br/> <span>не существует<' % broadcast_id, broadcast_page) is not None:
+ raise ExtractorError(
+ 'Broadcast %s does not exist' % broadcast_id, expected=True)
# Adult content
- if re.search(u'EroConfirmText">', broadcast_page) is not None:
+ if re.search('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,
+ raise ExtractorError(
+ 'Erotic broadcasts allowed only for registered users, '
+ 'use --username and --password options to provide account credentials.',
+ expected=True)
+
+ login_form = {
+ 'login-hint53': '1',
+ 'confirm_erotic': '1',
+ 'login': username,
+ '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 = compat_urllib_request.Request(
+ broadcast_url + '/?no_redirect=1', compat_urllib_parse.urlencode(login_form))
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')
+ request, broadcast_id, '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)
+ if re.search('>Неверный логин или пароль<', broadcast_page) is not None:
+ raise ExtractorError('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')
+ r"window\.broadcast_control\.addFlashVar\('file'\s*,\s*'([^']+)'\)",
+ broadcast_page, 'broadcast ticket')
url = 'http://smotri.com/broadcast/view/url/?ticket=%s' % ticket
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')
+ broadcast_json_page = self._download_webpage(
+ url, broadcast_id, '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)
+ raise ExtractorError(
+ '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)
+ raise ExtractorError('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')
+ mobj = re.search(r'^rtmp://[^/]+/(?P<app>.+)/?$', rtmp_url)
+ if not mobj:
+ raise ExtractorError('Unexpected broadcast rtmp URL')
broadcast_playpath = broadcast_json['_streamName']
+ broadcast_app = '%s/%s' % (mobj.group('app'), broadcast_json['_vidURL'])
broadcast_thumbnail = broadcast_json['_imgURL']
- broadcast_title = broadcast_json['title']
+ broadcast_title = self._live_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')
+ raise ExtractorError('Bad broadcast password', expected=True)
+ raise ExtractorError('Unexpected broadcast JSON')
return {
'id': broadcast_id,
'age_limit': 18 if adult_content else 0,
'ext': 'flv',
'play_path': broadcast_playpath,
+ 'player_url': 'http://pics.smotri.com/broadcast_play.swf',
+ 'app': broadcast_app,
'rtmp_live': True,
- 'rtmp_conn': rtmp_conn
+ 'rtmp_conn': rtmp_conn,
+ 'is_live': True,
}