# coding: utf-8
from __future__ import unicode_literals
import functools
import itertools
import operator
import re
from .common import InfoExtractor
from ..compat import (
compat_HTTPError,
compat_str,
compat_urllib_request,
)
from .openload import PhantomJSwrapper
from ..utils import (
clean_html,
determine_ext,
ExtractorError,
int_or_none,
merge_dicts,
NO_DEFAULT,
orderedSet,
remove_quotes,
str_to_int,
update_url_query,
urlencode_postdata,
url_or_none,
)
class PornHubBaseIE(InfoExtractor):
_NETRC_MACHINE = 'pornhub'
def _download_webpage_handle(self, *args, **kwargs):
def dl(*args, **kwargs):
return super(PornHubBaseIE, self)._download_webpage_handle(*args, **kwargs)
ret = dl(*args, **kwargs)
if not ret:
return ret
webpage, urlh = ret
if any(re.search(p, webpage) for p in (
r'
]+\bonload=["\']go\(\)',
r'document\.cookie\s*=\s*["\']RNKEY=',
r'document\.location\.reload\(true\)')):
url_or_request = args[0]
url = (url_or_request.get_full_url()
if isinstance(url_or_request, compat_urllib_request.Request)
else url_or_request)
phantom = PhantomJSwrapper(self, required_version='2.0')
phantom.get(url, html=webpage)
webpage, urlh = dl(*args, **kwargs)
return webpage, urlh
def _real_initialize(self):
self._logged_in = False
def _login(self, host):
if self._logged_in:
return
site = host.split('.')[0]
# Both sites pornhub and pornhubpremium have separate accounts
# so there should be an option to provide credentials for both.
# At the same time some videos are available under the same video id
# on both sites so that we have to identify them as the same video.
# For that purpose we have to keep both in the same extractor
# but under different netrc machines.
username, password = self._get_login_info(netrc_machine=site)
if username is None:
return
login_url = 'https://www.%s/%slogin' % (host, 'premium/' if 'premium' in host else '')
login_page = self._download_webpage(
login_url, None, 'Downloading %s login page' % site)
def is_logged(webpage):
return any(re.search(p, webpage) for p in (
r'class=["\']signOut',
r'>Sign\s+[Oo]ut\s*<'))
if is_logged(login_page):
self._logged_in = True
return
login_form = self._hidden_inputs(login_page)
login_form.update({
'username': username,
'password': password,
})
response = self._download_json(
'https://www.%s/front/authenticate' % host, None,
'Logging in to %s' % site,
data=urlencode_postdata(login_form),
headers={
'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
'Referer': login_url,
'X-Requested-With': 'XMLHttpRequest',
})
if response.get('success') == '1':
self._logged_in = True
return
message = response.get('message')
if message is not None:
raise ExtractorError(
'Unable to login: %s' % message, expected=True)
raise ExtractorError('Unable to log in')
class PornHubIE(PornHubBaseIE):
IE_DESC = 'PornHub and Thumbzilla'
_VALID_URL = r'''(?x)
https?://
(?:
(?:[^/]+\.)?(?Ppornhub(?:premium)?\.(?:com|net|org))/(?:(?:view_video\.php|video/show)\?viewkey=|embed/)|
(?:www\.)?thumbzilla\.com/video/
)
(?P[\da-z]+)
'''
_TESTS = [{
'url': 'http://www.pornhub.com/view_video.php?viewkey=648719015',
'md5': 'a6391306d050e4547f62b3f485dd9ba9',
'info_dict': {
'id': '648719015',
'ext': 'mp4',
'title': 'Seductive Indian beauty strips down and fingers her pink pussy',
'uploader': 'Babes',
'upload_date': '20130628',
'timestamp': 1372447216,
'duration': 361,
'view_count': int,
'like_count': int,
'dislike_count': int,
'comment_count': int,
'age_limit': 18,
'tags': list,
'categories': list,
'cast': list,
},
}, {
# non-ASCII title
'url': 'http://www.pornhub.com/view_video.php?viewkey=1331683002',
'info_dict': {
'id': '1331683002',
'ext': 'mp4',
'title': '重庆婷婷女王足交',
'upload_date': '20150213',
'timestamp': 1423804862,
'duration': 1753,
'view_count': int,
'like_count': int,
'dislike_count': int,
'comment_count': int,
'age_limit': 18,
'tags': list,
'categories': list,
},
'params': {
'skip_download': True,
},
'skip': 'Video has been flagged for verification in accordance with our trust and safety policy',
}, {
# subtitles
'url': 'https://www.pornhub.com/view_video.php?viewkey=ph5af5fef7c2aa7',
'info_dict': {
'id': 'ph5af5fef7c2aa7',
'ext': 'mp4',
'title': 'BFFS - Cute Teen Girls Share Cock On the Floor',
'uploader': 'BFFs',
'duration': 622,
'view_count': int,
'like_count': int,
'dislike_count': int,
'comment_count': int,
'age_limit': 18,
'tags': list,
'categories': list,
'subtitles': {
'en': [{
"ext": 'srt'
}]
},
},
'params': {
'skip_download': True,
},
'skip': 'This video has been disabled',
}, {
'url': 'http://www.pornhub.com/view_video.php?viewkey=ph557bbb6676d2d',
'only_matching': True,
}, {
# removed at the request of cam4.com
'url': 'http://fr.pornhub.com/view_video.php?viewkey=ph55ca2f9760862',
'only_matching': True,
}, {
# removed at the request of the copyright owner
'url': 'http://www.pornhub.com/view_video.php?viewkey=788152859',
'only_matching': True,
}, {
# removed by uploader
'url': 'http://www.pornhub.com/view_video.php?viewkey=ph572716d15a111',
'only_matching': True,
}, {
# private video
'url': 'http://www.pornhub.com/view_video.php?viewkey=ph56fd731fce6b7',
'only_matching': True,
}, {
'url': 'https://www.thumbzilla.com/video/ph56c6114abd99a/horny-girlfriend-sex',
'only_matching': True,
}, {
'url': 'http://www.pornhub.com/video/show?viewkey=648719015',
'only_matching': True,
}, {
'url': 'https://www.pornhub.net/view_video.php?viewkey=203640933',
'only_matching': True,
}, {
'url': 'https://www.pornhub.org/view_video.php?viewkey=203640933',
'only_matching': True,
}, {
'url': 'https://www.pornhubpremium.com/view_video.php?viewkey=ph5e4acdae54a82',
'only_matching': True,
}, {
# Some videos are available with the same id on both premium
# and non-premium sites (e.g. this and the following test)
'url': 'https://www.pornhub.com/view_video.php?viewkey=ph5f75b0f4b18e3',
'only_matching': True,
}, {
'url': 'https://www.pornhubpremium.com/view_video.php?viewkey=ph5f75b0f4b18e3',
'only_matching': True,
}]
@staticmethod
def _extract_urls(webpage):
return re.findall(
r'