[Vbox7IE] Improve extraction

* DASH extraction no longer fails with new range support
* but always find combined formats if available
* suppress ineffective XFF geo-bypass (causes time-outs)
* adapted from https://github.com/ytdl-org/youtube-dl/pull/29680
* thx former GH user kikuyan
This commit is contained in:
dirkf 2024-01-27 15:57:25 +00:00
parent 1fd8f802b8
commit bdda6b81df

View File

@ -2,9 +2,19 @@
from __future__ import unicode_literals from __future__ import unicode_literals
import re import re
import time
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import ExtractorError from ..utils import (
determine_ext,
ExtractorError,
float_or_none,
merge_dicts,
T,
traverse_obj,
txt_or_none,
url_or_none,
)
class Vbox7IE(InfoExtractor): class Vbox7IE(InfoExtractor):
@ -20,10 +30,12 @@ class Vbox7IE(InfoExtractor):
) )
(?P<id>[\da-fA-F]+) (?P<id>[\da-fA-F]+)
''' '''
_EMBED_REGEX = [r'<iframe[^>]+src=(?P<q>["\'])(?P<url>(?:https?:)?//vbox7\.com/emb/external\.php.+?)(?P=q)']
_GEO_COUNTRIES = ['BG'] _GEO_COUNTRIES = ['BG']
_GEO_BYPASS = False
_TESTS = [{ _TESTS = [{
'url': 'http://vbox7.com/play:0946fff23c', 'url': 'https://vbox7.com/play:0946fff23c',
'md5': 'a60f9ab3a3a2f013ef9a967d5f7be5bf', 'md5': '50ca1f78345a9c15391af47d8062d074',
'info_dict': { 'info_dict': {
'id': '0946fff23c', 'id': '0946fff23c',
'ext': 'mp4', 'ext': 'mp4',
@ -34,18 +46,21 @@ class Vbox7IE(InfoExtractor):
'upload_date': '20160812', 'upload_date': '20160812',
'uploader': 'zdraveibulgaria', 'uploader': 'zdraveibulgaria',
}, },
'params': { 'expected_warnings': [
'proxy': '127.0.0.1:8118', 'Unable to download webpage',
}, ],
}, { }, {
'url': 'http://vbox7.com/play:249bb972c2', 'url': 'http://vbox7.com/play:249bb972c2',
'md5': '99f65c0c9ef9b682b97313e052734c3f', 'md5': 'aaf19465e37ec0b30b918df83ec32c50',
'info_dict': { 'info_dict': {
'id': '249bb972c2', 'id': '249bb972c2',
'ext': 'mp4', 'ext': 'mp4',
'title': 'Смях! Чудо - чист за секунди - Скрита камера', 'title': 'Смях! Чудо - чист за секунди - Скрита камера',
'description': 'Смях! Чудо - чист за секунди - Скрита камера',
'timestamp': 1360215023,
'upload_date': '20130207',
'uploader': 'svideteliat_ot_varshava',
}, },
'skip': 'georestricted',
}, { }, {
'url': 'http://vbox7.com/emb/external.php?vid=a240d20f9c&autoplay=1', 'url': 'http://vbox7.com/emb/external.php?vid=a240d20f9c&autoplay=1',
'only_matching': True, 'only_matching': True,
@ -54,52 +69,88 @@ class Vbox7IE(InfoExtractor):
'only_matching': True, 'only_matching': True,
}] }]
@staticmethod @classmethod
def _extract_url(webpage): def _extract_url(cls, webpage):
mobj = re.search( mobj = re.search(cls._EMBED_REGEX[0], webpage)
r'<iframe[^>]+src=(?P<q>["\'])(?P<url>(?:https?:)?//vbox7\.com/emb/external\.php.+?)(?P=q)',
webpage)
if mobj: if mobj:
return mobj.group('url') return mobj.group('url')
def _real_extract(self, url): def _real_extract(self, url):
video_id = self._match_id(url) video_id = self._match_id(url)
url = 'https://vbox7.com/play:%s' % (video_id,)
now = time.time()
response = self._download_json( response = self._download_json(
'https://www.vbox7.com/ajax/video/nextvideo.php?vid=%s' % video_id, 'https://www.vbox7.com/aj/player/item/options?vid=%s' % (video_id,),
video_id) video_id, headers={'Referer': url})
# estimate time to which possible `ago` member is relative
now = now + 0.5 * (time.time() - now)
if 'error' in response: if 'error' in response:
raise ExtractorError( raise ExtractorError(
'%s said: %s' % (self.IE_NAME, response['error']), expected=True) '%s said: %s' % (self.IE_NAME, response['error']), expected=True)
video = response['options'] video_url = traverse_obj(response, ('options', 'src', T(url_or_none)))
title = video['title'] if '/na.mp4' in video_url or '':
video_url = video['src']
if '/na.mp4' in video_url:
self.raise_geo_restricted(countries=self._GEO_COUNTRIES) self.raise_geo_restricted(countries=self._GEO_COUNTRIES)
uploader = video.get('uploader') ext = determine_ext(video_url)
if ext == 'mpd':
# In case MPD cannot be parsed, or anyway, get mp4 combined
# formats usually provided to Safari, iOS, and old Windows
try:
formats, subtitles = self._extract_mpd_formats_and_subtitles(
video_url, video_id, 'dash', fatal=False)
except KeyError:
self.report_warning('Failed to parse MPD manifest')
formats, subtitles = [], {}
webpage = self._download_webpage( video = response['options']
'http://vbox7.com/play:%s' % video_id, video_id, fatal=None) resolutions = (1080, 720, 480, 240, 144)
highest_res = traverse_obj(video, ('highestRes', T(int))) or resolutions[0]
for res in traverse_obj(video, ('resolutions', lambda _, r: int(r) > 0)) or resolutions:
if res > highest_res:
continue
formats.append({
'url': video_url.replace('.mpd', '_%d.mp4' % res),
'format_id': '%dp' % res,
'height': res,
})
# if above formats are flaky, enable the line below
# self._check_formats(formats, video_id)
else:
formats = [{
'url': video_url,
}]
subtitles = {}
self._sort_formats(formats)
info = {} webpage = self._download_webpage(url, video_id, fatal=False) or ''
if webpage:
info = self._search_json_ld( info = self._search_json_ld(
webpage.replace('"/*@context"', '"@context"'), video_id, webpage.replace('"/*@context"', '"@context"'), video_id,
fatal=False) fatal=False) if webpage else {}
info.update({ if not info.get('title'):
info['title'] = traverse_obj(response, (
'options', 'title', T(txt_or_none))) or self._og_search_title(webpage)
def if_missing(k):
return lambda x: None if k in info else x
info = merge_dicts(info, {
'id': video_id, 'id': video_id,
'title': title, 'formats': formats,
'url': video_url, 'subtitles': subtitles or None,
'uploader': uploader, }, info, traverse_obj(response, ('options', {
'thumbnail': self._proto_relative_url( 'uploader': ('uploader', T(txt_or_none)),
'timestamp': ('ago', T(if_missing('timestamp')), T(lambda t: int(round((now - t) / 60.0)) * 60)),
'duration': ('duration', T(if_missing('duration')), T(float_or_none)),
})))
if 'thumbnail' not in info:
info['thumbnail'] = self._proto_relative_url(
info.get('thumbnail') or self._og_search_thumbnail(webpage), info.get('thumbnail') or self._og_search_thumbnail(webpage),
'http:'), 'https:'),
})
return info return info