diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-15 16:49:24 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-15 16:49:24 +0000 |
commit | 2415e66f889f38503b73e8ebc5f43ca342390e5c (patch) | |
tree | ac48ab69d1d96bae3d83756134921e0d90593aa5 /yt_dlp/extractor/clyp.py | |
parent | Initial commit. (diff) | |
download | yt-dlp-2415e66f889f38503b73e8ebc5f43ca342390e5c.tar.xz yt-dlp-2415e66f889f38503b73e8ebc5f43ca342390e5c.zip |
Adding upstream version 2024.03.10.upstream/2024.03.10
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'yt_dlp/extractor/clyp.py')
-rw-r--r-- | yt_dlp/extractor/clyp.py | 99 |
1 files changed, 99 insertions, 0 deletions
diff --git a/yt_dlp/extractor/clyp.py b/yt_dlp/extractor/clyp.py new file mode 100644 index 0000000..273d002 --- /dev/null +++ b/yt_dlp/extractor/clyp.py @@ -0,0 +1,99 @@ +from .common import InfoExtractor +from ..utils import ( + float_or_none, + parse_qs, + unified_timestamp, +) + + +class ClypIE(InfoExtractor): + _VALID_URL = r'https?://(?:www\.)?clyp\.it/(?P<id>[a-z0-9]+)' + _TESTS = [{ + 'url': 'https://clyp.it/iynkjk4b', + 'md5': '4bc6371c65210e7b372097fce4d92441', + 'info_dict': { + 'id': 'iynkjk4b', + 'ext': 'ogg', + 'title': 'research', + 'description': '#Research', + 'duration': 51.278, + 'timestamp': 1435524981, + 'upload_date': '20150628', + }, + }, { + 'url': 'https://clyp.it/b04p1odi?token=b0078e077e15835845c528a44417719d', + 'info_dict': { + 'id': 'b04p1odi', + 'ext': 'ogg', + 'title': 'GJ! (Reward Edit)', + 'description': 'Metal Resistance (THE ONE edition)', + 'duration': 177.789, + 'timestamp': 1528241278, + 'upload_date': '20180605', + }, + 'params': { + 'skip_download': True, + }, + }, { + 'url': 'https://clyp.it/v42214lc', + 'md5': '4aca4dfc3236fb6d6ddc4ea08314f33f', + 'info_dict': { + 'id': 'v42214lc', + 'ext': 'wav', + 'title': 'i dont wanna go (old version)', + 'duration': 113.528, + 'timestamp': 1607348505, + 'upload_date': '20201207', + }, + }] + + def _real_extract(self, url): + audio_id = self._match_id(url) + + qs = parse_qs(url) + token = qs.get('token', [None])[0] + + query = {} + if token: + query['token'] = token + + metadata = self._download_json( + 'https://api.clyp.it/%s' % audio_id, audio_id, query=query) + + formats = [] + for secure in ('', 'Secure'): + for ext in ('Ogg', 'Mp3'): + format_id = '%s%s' % (secure, ext) + format_url = metadata.get('%sUrl' % format_id) + if format_url: + formats.append({ + 'url': format_url, + 'format_id': format_id, + 'vcodec': 'none', + 'acodec': ext.lower(), + }) + + page = self._download_webpage(url, video_id=audio_id) + wav_url = self._html_search_regex( + r'var\s*wavStreamUrl\s*=\s*["\'](?P<url>https?://[^\'"]+)', page, 'url', default=None) + if wav_url: + formats.append({ + 'url': wav_url, + 'format_id': 'wavStreamUrl', + 'vcodec': 'none', + 'acodec': 'wav', + }) + + title = metadata['Title'] + description = metadata.get('Description') + duration = float_or_none(metadata.get('Duration')) + timestamp = unified_timestamp(metadata.get('DateCreated')) + + return { + 'id': audio_id, + 'title': title, + 'description': description, + 'duration': duration, + 'timestamp': timestamp, + 'formats': formats, + } |