github/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-12-12 09:55:51 -05:00
Code Issues Projects Releases Packages Wiki Activity

[ceskatelevize] Fetch iframe from nextJS data (#1904)

Browse source 
Closes #1899
Authored by: mkubecek
This commit is contained in:
Michal Kubeček 2021-12-07 17:44:43 +01:00 committed by GitHub
parent 66f4c04e50
commit 443b21dc4e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 8 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
yt_dlp/extractor/ceskatelevize.py
View file

@ -12,8 +12,7 @@
ExtractorError, ExtractorError,
float_or_none, float_or_none,
sanitized_Request, sanitized_Request,
unescapeHTML, traverse_obj,
update_url_query,
urlencode_postdata, urlencode_postdata,
USER_AGENTS, USER_AGENTS,
) )
@ -99,11 +98,13 @@ def _real_extract(self, url):
playlist_description = playlist_description.replace('\xa0', ' ') playlist_description = playlist_description.replace('\xa0', ' ')
if parsed_url.path.startswith('/porady/'): if parsed_url.path.startswith('/porady/'):
refer_url = update_url_query(unescapeHTML(self._search_regex( next_data = self._search_nextjs_data(webpage, playlist_id)
(r'<span[^>]*\bdata-url=(["\'])(?P<url>(?:(?!\1).)+)\1', idec = traverse_obj(next_data, ('props', 'pageProps', 'data', ('show', 'mediaMeta'), 'idec'), get_all=False)
r'<iframe[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//(?:www\.)?ceskatelevize\.cz/ivysilani/embed/iFramePlayer\.php.*?)\1'), if not idec:
webpage, 'iframe player url', group='url')), query={'autoStart': 'true'}) raise ExtractorError('Failed to find IDEC id')
webpage = self._download_webpage(refer_url, playlist_id) iframe_hash = self._download_webpage('https://www.ceskatelevize.cz/v-api/iframe-hash/', playlist_id)
webpage = self._download_webpage('https://www.ceskatelevize.cz/ivysilani/embed/iFramePlayer.php', playlist_id,
query={'hash': iframe_hash, 'origin': 'iVysilani', 'autoStart': 'true', 'IDEC': idec})
NOT_AVAILABLE_STRING = 'This content is not available at your territory due to limited copyright.' NOT_AVAILABLE_STRING = 'This content is not available at your territory due to limited copyright.'
if '%s</p>' % NOT_AVAILABLE_STRING in webpage: if '%s</p>' % NOT_AVAILABLE_STRING in webpage: