mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2024-11-25 00:00:48 +01:00
86e5f3ed2e
Using https://github.com/asottile/pyupgrade 1. `__future__` imports and `coding: utf-8` were removed 2. Files were rewritten with `pyupgrade --py36-plus --keep-percent-format` 3. f-strings were cherry-picked from `pyupgrade --py36-plus` Extractors are left untouched (except removing header) to avoid unnecessary merge conflicts
166 lines
6.7 KiB
Python
166 lines
6.7 KiB
Python
from .theplatform import ThePlatformFeedIE
|
|
from ..utils import (
|
|
ExtractorError,
|
|
int_or_none,
|
|
find_xpath_attr,
|
|
xpath_element,
|
|
xpath_text,
|
|
update_url_query,
|
|
url_or_none,
|
|
)
|
|
|
|
|
|
class CBSBaseIE(ThePlatformFeedIE):
|
|
def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
|
|
subtitles = {}
|
|
for k, ext in [('sMPTE-TTCCURL', 'tt'), ('ClosedCaptionURL', 'ttml'), ('webVTTCaptionURL', 'vtt')]:
|
|
cc_e = find_xpath_attr(smil, self._xpath_ns('.//param', namespace), 'name', k)
|
|
if cc_e is not None:
|
|
cc_url = cc_e.get('value')
|
|
if cc_url:
|
|
subtitles.setdefault(subtitles_lang, []).append({
|
|
'ext': ext,
|
|
'url': cc_url,
|
|
})
|
|
return subtitles
|
|
|
|
def _extract_common_video_info(self, content_id, asset_types, mpx_acc, extra_info):
|
|
tp_path = 'dJ5BDC/media/guid/%d/%s' % (mpx_acc, content_id)
|
|
tp_release_url = f'https://link.theplatform.com/s/{tp_path}'
|
|
info = self._extract_theplatform_metadata(tp_path, content_id)
|
|
|
|
formats, subtitles = [], {}
|
|
last_e = None
|
|
for asset_type, query in asset_types.items():
|
|
try:
|
|
tp_formats, tp_subtitles = self._extract_theplatform_smil(
|
|
update_url_query(tp_release_url, query), content_id,
|
|
'Downloading %s SMIL data' % asset_type)
|
|
except ExtractorError as e:
|
|
last_e = e
|
|
if asset_type != 'fallback':
|
|
continue
|
|
query['formats'] = '' # blank query to check if expired
|
|
try:
|
|
tp_formats, tp_subtitles = self._extract_theplatform_smil(
|
|
update_url_query(tp_release_url, query), content_id,
|
|
'Downloading %s SMIL data, trying again with another format' % asset_type)
|
|
except ExtractorError as e:
|
|
last_e = e
|
|
continue
|
|
formats.extend(tp_formats)
|
|
subtitles = self._merge_subtitles(subtitles, tp_subtitles)
|
|
if last_e and not formats:
|
|
self.raise_no_formats(last_e, True, content_id)
|
|
self._sort_formats(formats)
|
|
|
|
extra_info.update({
|
|
'id': content_id,
|
|
'formats': formats,
|
|
'subtitles': subtitles,
|
|
})
|
|
info.update({k: v for k, v in extra_info.items() if v is not None})
|
|
return info
|
|
|
|
def _extract_video_info(self, *args, **kwargs):
|
|
# Extract assets + metadata and call _extract_common_video_info
|
|
raise NotImplementedError('This method must be implemented by subclasses')
|
|
|
|
def _real_extract(self, url):
|
|
return self._extract_video_info(self._match_id(url))
|
|
|
|
|
|
class CBSIE(CBSBaseIE):
|
|
_VALID_URL = r'''(?x)
|
|
(?:
|
|
cbs:|
|
|
https?://(?:www\.)?(?:
|
|
cbs\.com/(?:shows|movies)/(?:video|[^/]+/video|[^/]+)/|
|
|
colbertlateshow\.com/(?:video|podcasts)/)
|
|
)(?P<id>[\w-]+)'''
|
|
|
|
# All tests are blocked outside US
|
|
_TESTS = [{
|
|
'url': 'https://www.cbs.com/shows/video/xrUyNLtl9wd8D_RWWAg9NU2F_V6QpB3R/',
|
|
'info_dict': {
|
|
'id': 'xrUyNLtl9wd8D_RWWAg9NU2F_V6QpB3R',
|
|
'ext': 'mp4',
|
|
'title': 'Tough As Nails - Dreams Never Die',
|
|
'description': 'md5:a3535a62531cdd52b0364248a2c1ae33',
|
|
'duration': 2588,
|
|
'timestamp': 1639015200,
|
|
'upload_date': '20211209',
|
|
'uploader': 'CBSI-NEW',
|
|
},
|
|
'params': {
|
|
# m3u8 download
|
|
'skip_download': True,
|
|
},
|
|
}, {
|
|
'url': 'https://www.cbs.com/shows/video/sZH1MGgomIosZgxGJ1l263MFq16oMtW1/',
|
|
'info_dict': {
|
|
'id': 'sZH1MGgomIosZgxGJ1l263MFq16oMtW1',
|
|
'title': 'The Late Show - 3/16/22 (Michael Buble, Rose Matafeo)',
|
|
'timestamp': 1647488100,
|
|
'description': 'md5:d0e6ec23c544b7fa8e39a8e6844d2439',
|
|
'uploader': 'CBSI-NEW',
|
|
'upload_date': '20220317',
|
|
},
|
|
'params': {
|
|
'ignore_no_formats_error': True,
|
|
'skip_download': True,
|
|
},
|
|
'expected_warnings': [
|
|
'This content expired on', 'No video formats found', 'Requested format is not available'],
|
|
}, {
|
|
'url': 'http://colbertlateshow.com/video/8GmB0oY0McANFvp2aEffk9jZZZ2YyXxy/the-colbeard/',
|
|
'only_matching': True,
|
|
}, {
|
|
'url': 'http://www.colbertlateshow.com/podcasts/dYSwjqPs_X1tvbV_P2FcPWRa_qT6akTC/in-the-bad-room-with-stephen/',
|
|
'only_matching': True,
|
|
}]
|
|
|
|
def _extract_video_info(self, content_id, site='cbs', mpx_acc=2198311517):
|
|
items_data = self._download_xml(
|
|
'https://can.cbs.com/thunder/player/videoPlayerService.php',
|
|
content_id, query={'partner': site, 'contentId': content_id})
|
|
video_data = xpath_element(items_data, './/item')
|
|
title = xpath_text(video_data, 'videoTitle', 'title') or xpath_text(video_data, 'videotitle', 'title')
|
|
|
|
asset_types = {}
|
|
has_drm = False
|
|
for item in items_data.findall('.//item'):
|
|
asset_type = xpath_text(item, 'assetType')
|
|
query = {
|
|
'mbr': 'true',
|
|
'assetTypes': asset_type,
|
|
}
|
|
if not asset_type:
|
|
# fallback for content_ids that videoPlayerService doesn't return anything for
|
|
asset_type = 'fallback'
|
|
query['formats'] = 'M3U+none,MPEG4,M3U+appleHlsEncryption,MP3'
|
|
del query['assetTypes']
|
|
if asset_type in asset_types:
|
|
continue
|
|
elif any(excluded in asset_type for excluded in ('HLS_FPS', 'DASH_CENC', 'OnceURL')):
|
|
if 'DASH_CENC' in asset_type:
|
|
has_drm = True
|
|
continue
|
|
if asset_type.startswith('HLS') or 'StreamPack' in asset_type:
|
|
query['formats'] = 'MPEG4,M3U'
|
|
elif asset_type in ('RTMP', 'WIFI', '3G'):
|
|
query['formats'] = 'MPEG4,FLV'
|
|
asset_types[asset_type] = query
|
|
|
|
if not asset_types and has_drm:
|
|
self.report_drm(content_id)
|
|
|
|
return self._extract_common_video_info(content_id, asset_types, mpx_acc, extra_info={
|
|
'title': title,
|
|
'series': xpath_text(video_data, 'seriesTitle'),
|
|
'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')),
|
|
'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')),
|
|
'duration': int_or_none(xpath_text(video_data, 'videoLength'), 1000),
|
|
'thumbnail': url_or_none(xpath_text(video_data, 'previewImageURL')),
|
|
})
|