mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2024-11-27 06:10:12 +01:00
parent
046cab3915
commit
d6bc443bde
|
@ -1,18 +1,16 @@
|
|||
# coding: utf-8
|
||||
from __future__ import unicode_literals
|
||||
|
||||
import hashlib
|
||||
|
||||
from .common import InfoExtractor
|
||||
from ..compat import (
|
||||
compat_parse_qs,
|
||||
compat_urllib_request,
|
||||
compat_urlparse,
|
||||
)
|
||||
from ..utils import (
|
||||
ExtractorError,
|
||||
sanitized_Request,
|
||||
traverse_obj,
|
||||
urlencode_postdata,
|
||||
urljoin,
|
||||
)
|
||||
|
||||
|
||||
|
@ -82,41 +80,32 @@ def _real_extract(self, url):
|
|||
self._downloader.cookiejar.clear_session_cookies() # must clear
|
||||
self._login()
|
||||
|
||||
title = 'FC2 video %s' % video_id
|
||||
thumbnail = None
|
||||
title, thumbnail, description = None, None, None
|
||||
if webpage is not None:
|
||||
title = self._og_search_title(webpage)
|
||||
title = self._html_search_regex(
|
||||
(r'<h2\s+class="videoCnt_title">([^<]+?)</h2>',
|
||||
r'\s+href="[^"]+"\s*title="([^"]+?)"\s*rel="nofollow">\s*<img',
|
||||
# there's two matches in the webpage
|
||||
r'\s+href="[^"]+"\s*title="([^"]+?)"\s*rel="nofollow">\s*\1'),
|
||||
webpage,
|
||||
'title', fatal=False)
|
||||
thumbnail = self._og_search_thumbnail(webpage)
|
||||
refer = url.replace('/content/', '/a/content/') if '/a/content/' not in url else url
|
||||
description = self._og_search_description(webpage)
|
||||
|
||||
mimi = hashlib.md5((video_id + '_gGddgPfeaf_gzyr').encode('utf-8')).hexdigest()
|
||||
|
||||
info_url = (
|
||||
'http://video.fc2.com/ginfo.php?mimi={1:s}&href={2:s}&v={0:s}&fversion=WIN%2011%2C6%2C602%2C180&from=2&otag=0&upid={0:s}&tk=null&'.
|
||||
format(video_id, mimi, compat_urllib_request.quote(refer, safe=b'').replace('.', '%2E')))
|
||||
|
||||
info_webpage = self._download_webpage(
|
||||
info_url, video_id, note='Downloading info page')
|
||||
info = compat_urlparse.parse_qs(info_webpage)
|
||||
|
||||
if 'err_code' in info:
|
||||
# most of the time we can still download wideo even if err_code is 403 or 602
|
||||
self.report_warning(
|
||||
'Error code was: %s... but still trying' % info['err_code'][0])
|
||||
|
||||
if 'filepath' not in info:
|
||||
raise ExtractorError('Cannot download file. Are you logged in?')
|
||||
|
||||
video_url = info['filepath'][0] + '?mid=' + info['mid'][0]
|
||||
title_info = info.get('title')
|
||||
if title_info:
|
||||
title = title_info[0]
|
||||
vidplaylist = self._download_json(
|
||||
'https://video.fc2.com/api/v3/videoplaylist/%s?sh=1&fs=0' % video_id, video_id,
|
||||
note='Downloading info page')
|
||||
vid_url = traverse_obj(vidplaylist, ('playlist', 'nq'))
|
||||
if not vid_url:
|
||||
raise ExtractorError('Unable to extract video URL')
|
||||
vid_url = urljoin('https://video.fc2.com/', vid_url)
|
||||
|
||||
return {
|
||||
'id': video_id,
|
||||
'title': title,
|
||||
'url': video_url,
|
||||
'ext': 'flv',
|
||||
'url': vid_url,
|
||||
'ext': 'mp4',
|
||||
'description': description,
|
||||
'thumbnail': thumbnail,
|
||||
}
|
||||
|
||||
|
|
Loading…
Reference in New Issue
Block a user