diff options
author | Sergey M․ <[email protected]> | 2017-06-25 04:54:56 +0700 |
---|---|---|
committer | Sergey M․ <[email protected]> | 2017-06-25 05:03:17 +0700 |
commit | 0c7a631b613fe8ec443e6b17ea8f7a17fb8abb5b (patch) | |
tree | b4d56a5bd668f74d28fdad1852f61299523463ed | |
parent | fd9ee4de8c3aa640d51c5ff42f5d8241970e894d (diff) | |
download | youtube-dl-0c7a631b613fe8ec443e6b17ea8f7a17fb8abb5b.tar.gz youtube-dl-0c7a631b613fe8ec443e6b17ea8f7a17fb8abb5b.zip |
[adobepass] Add support for ATTOTT MSO (DIRECTV NOW) (closes #13472)
-rw-r--r-- | youtube_dl/extractor/adobepass.py | 43 |
1 files changed, 37 insertions, 6 deletions
diff --git a/youtube_dl/extractor/adobepass.py b/youtube_dl/extractor/adobepass.py index 3dfc632e7..b83b51efb 100644 --- a/youtube_dl/extractor/adobepass.py +++ b/youtube_dl/extractor/adobepass.py @@ -15,6 +15,7 @@ from ..utils import ( urlencode_postdata, unified_timestamp, ExtractorError, + NO_DEFAULT, ) @@ -24,6 +25,11 @@ MSO_INFO = { 'username_field': 'username', 'password_field': 'password', }, + 'ATTOTT': { + 'name': 'DIRECTV NOW', + 'username_field': 'email', + 'password_field': 'loginpassword', + }, 'Rogers': { 'name': 'Rogers', 'username_field': 'UserName', @@ -1316,6 +1322,8 @@ class AdobePassIE(InfoExtractor): _USER_AGENT = 'Mozilla/5.0 (X11; Linux i686; rv:47.0) Gecko/20100101 Firefox/47.0' _MVPD_CACHE = 'ap-mvpd' + _DOWNLOADING_LOGIN_PAGE = 'Downloading Provider Login Page' + def _download_webpage_handle(self, *args, **kwargs): headers = kwargs.get('headers', {}) headers.update(self.geo_verification_headers()) @@ -1365,6 +1373,21 @@ class AdobePassIE(InfoExtractor): 'Use --ap-mso to specify Adobe Pass Multiple-system operator Identifier ' 'and --ap-username and --ap-password or --netrc to provide account credentials.', expected=True) + def extract_redirect_url(html, url=None, fatal=False): + # TODO: eliminate code duplication with generic extractor and move + # redirection code into _download_webpage_handle + REDIRECT_REGEX = r'[0-9]{,2};\s*(?:URL|url)=\'?([^\'"]+)' + redirect_url = self._search_regex( + r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")' + r'(?:[a-z-]+="[^"]+"\s+)*?content="%s' % REDIRECT_REGEX, + html, 'meta refresh redirect', + default=NO_DEFAULT if fatal else None, fatal=fatal) + if not redirect_url: + return None + if url: + redirect_url = compat_urlparse.urljoin(url, unescapeHTML(redirect_url)) + return redirect_url + mvpd_headers = { 'ap_42': 'anonymous', 'ap_11': 'Linux i686', @@ -1414,16 +1437,15 @@ class AdobePassIE(InfoExtractor): if '<form name="signin"' in provider_redirect_page: provider_login_page_res = provider_redirect_page_res elif 'http-equiv="refresh"' in provider_redirect_page: - oauth_redirect_url = self._html_search_regex( - r'content="0;\s*url=([^\'"]+)', - provider_redirect_page, 'meta refresh redirect') + oauth_redirect_url = extract_redirect_url( + provider_redirect_page, fatal=True) provider_login_page_res = self._download_webpage_handle( oauth_redirect_url, video_id, - 'Downloading Provider Login Page') + self._DOWNLOADING_LOGIN_PAGE) else: provider_login_page_res = post_form( provider_redirect_page_res, - 'Downloading Provider Login Page') + self._DOWNLOADING_LOGIN_PAGE) mvpd_confirm_page_res = post_form( provider_login_page_res, 'Logging in', { @@ -1470,8 +1492,17 @@ class AdobePassIE(InfoExtractor): 'Content-Type': 'application/x-www-form-urlencoded' }) else: + # Some providers (e.g. DIRECTV NOW) have another meta refresh + # based redirect that should be followed. + provider_redirect_page, urlh = provider_redirect_page_res + provider_refresh_redirect_url = extract_redirect_url( + provider_redirect_page, url=urlh.geturl()) + if provider_refresh_redirect_url: + provider_redirect_page_res = self._download_webpage_handle( + provider_refresh_redirect_url, video_id, + 'Downloading Provider Redirect Page (meta refresh)') provider_login_page_res = post_form( - provider_redirect_page_res, 'Downloading Provider Login Page') + provider_redirect_page_res, self._DOWNLOADING_LOGIN_PAGE) mvpd_confirm_page_res = post_form(provider_login_page_res, 'Logging in', { mso_info.get('username_field', 'username'): username, mso_info.get('password_field', 'password'): password, |