|
|
@ -3,7 +3,7 @@ from __future__ import unicode_literals
|
|
|
|
|
|
|
|
|
|
|
|
import re
|
|
|
|
import re
|
|
|
|
|
|
|
|
|
|
|
|
from .common import InfoExtractor
|
|
|
|
from .adobepass import AdobePassIE
|
|
|
|
from ..compat import compat_str
|
|
|
|
from ..compat import compat_str
|
|
|
|
from ..utils import (
|
|
|
|
from ..utils import (
|
|
|
|
xpath_text,
|
|
|
|
xpath_text,
|
|
|
@ -16,11 +16,11 @@ from ..utils import (
|
|
|
|
)
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class TurnerBaseIE(InfoExtractor):
|
|
|
|
class TurnerBaseIE(AdobePassIE):
|
|
|
|
def _extract_timestamp(self, video_data):
|
|
|
|
def _extract_timestamp(self, video_data):
|
|
|
|
return int_or_none(xpath_attr(video_data, 'dateCreated', 'uts'))
|
|
|
|
return int_or_none(xpath_attr(video_data, 'dateCreated', 'uts'))
|
|
|
|
|
|
|
|
|
|
|
|
def _extract_cvp_info(self, data_src, video_id, path_data={}):
|
|
|
|
def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}):
|
|
|
|
video_data = self._download_xml(data_src, video_id)
|
|
|
|
video_data = self._download_xml(data_src, video_id)
|
|
|
|
video_id = video_data.attrib['id']
|
|
|
|
video_id = video_data.attrib['id']
|
|
|
|
title = xpath_text(video_data, 'headline', fatal=True)
|
|
|
|
title = xpath_text(video_data, 'headline', fatal=True)
|
|
|
@ -70,11 +70,14 @@ class TurnerBaseIE(InfoExtractor):
|
|
|
|
secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
|
|
|
|
secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
|
|
|
|
token = tokens.get(secure_path)
|
|
|
|
token = tokens.get(secure_path)
|
|
|
|
if not token:
|
|
|
|
if not token:
|
|
|
|
|
|
|
|
query = {
|
|
|
|
|
|
|
|
'path': secure_path,
|
|
|
|
|
|
|
|
'videoId': content_id,
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
if ap_data.get('auth_required'):
|
|
|
|
|
|
|
|
query['accessToken'] = self._extract_mvpd_auth(ap_data['url'], video_id, ap_data['site_name'], ap_data['site_name'])
|
|
|
|
auth = self._download_xml(
|
|
|
|
auth = self._download_xml(
|
|
|
|
secure_path_data['tokenizer_src'], video_id, query={
|
|
|
|
secure_path_data['tokenizer_src'], video_id, query=query)
|
|
|
|
'path': secure_path,
|
|
|
|
|
|
|
|
'videoId': content_id,
|
|
|
|
|
|
|
|
})
|
|
|
|
|
|
|
|
error_msg = xpath_text(auth, 'error/msg')
|
|
|
|
error_msg = xpath_text(auth, 'error/msg')
|
|
|
|
if error_msg:
|
|
|
|
if error_msg:
|
|
|
|
raise ExtractorError(error_msg, expected=True)
|
|
|
|
raise ExtractorError(error_msg, expected=True)
|
|
|
|