[spankbang] Fix extraction (closes #20023)

This commit is contained in:
Sergey M․ 2019-03-03 06:25:45 +07:00
parent a8f83f0c56
commit 7465e0aee2
No known key found for this signature in database
GPG key ID: 2C393E0F18A9236D

View file

@ -9,6 +9,8 @@ from ..utils import (
parse_duration, parse_duration,
parse_resolution, parse_resolution,
str_to_int, str_to_int,
url_or_none,
urlencode_postdata,
) )
@ -64,16 +66,49 @@ class SpankBangIE(InfoExtractor):
'Video %s is not available' % video_id, expected=True) 'Video %s is not available' % video_id, expected=True)
formats = [] formats = []
for mobj in re.finditer(
r'stream_url_(?P<id>[^\s=]+)\s*=\s*(["\'])(?P<url>(?:(?!\2).)+)\2', def extract_format(format_id, format_url):
webpage): f_url = url_or_none(format_url)
format_id, format_url = mobj.group('id', 'url') if not f_url:
return
f = parse_resolution(format_id) f = parse_resolution(format_id)
f.update({ f.update({
'url': format_url, 'url': f_url,
'format_id': format_id, 'format_id': format_id,
}) })
formats.append(f) formats.append(f)
STREAM_URL_PREFIX = 'stream_url_'
for mobj in re.finditer(
r'%s(?P<id>[^\s=]+)\s*=\s*(["\'])(?P<url>(?:(?!\2).)+)\2'
% STREAM_URL_PREFIX, webpage):
extract_format(mobj.group('id', 'url'))
if not formats:
stream_key = self._search_regex(
r'data-streamkey\s*=\s*(["\'])(?P<value>(?:(?!\1).)+)\1',
webpage, 'stream key', group='value')
sb_csrf_session = self._get_cookies(
'https://spankbang.com')['sb_csrf_session'].value
stream = self._download_json(
'https://spankbang.com/api/videos/stream', video_id,
'Downloading stream JSON', data=urlencode_postdata({
'id': stream_key,
'data': 0,
'sb_csrf_session': sb_csrf_session,
}), headers={
'Referer': url,
'X-CSRFToken': sb_csrf_session,
})
for format_id, format_url in stream.items():
if format_id.startswith(STREAM_URL_PREFIX):
extract_format(
format_id[len(STREAM_URL_PREFIX):], format_url)
self._sort_formats(formats) self._sort_formats(formats)
title = self._html_search_regex( title = self._html_search_regex(