Miroirs/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp synced 2025-01-16 03:40:50 +01:00
Code Issues Projects Releases Packages Wiki Activity

asdsadf

Browse source
This commit is contained in:
Jesse Bannon 2024-05-10 00:24:47 -07:00
parent 8cc70b5e4d
commit 0f167c960d
1 changed files with 18 additions and 19 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

19
yt_dlp/extractor/pbs.py
View file

@ -5,7 +5,7 @@ from .common import InfoExtractor
from ..compat import compat_str from ..compat import compat_str
from ..utils import ( from ..utils import (
ExtractorError, ExtractorError,
OnDemandPagedList, LazyList,
determine_ext, determine_ext,
int_or_none, int_or_none,
float_or_none, float_or_none,
@ -782,18 +782,19 @@ class PBSShowIE(InfoExtractor):
_JSON_SEARCH = r'<script[^>]+id="content-strip-data" type="application/json">' _JSON_SEARCH = r'<script[^>]+id="content-strip-data" type="application/json">'
_SHOW_JSON_SEARCH = r'GTMDataLayer\.push\(' _SHOW_JSON_SEARCH = r'GTMDataLayer\.push\('
PAGE_SIZE = 40
@staticmethod @staticmethod
def _make_url(playlist_id): def _make_url(playlist_id):
# pbs does not show metadata, use a different station that does
return f'https://video.ksps.org/show/{playlist_id}' return f'https://video.ksps.org/show/{playlist_id}'
def _fetch_season_page(self, playlist_id, season_indices, page_num): def _fetch_seasons(self, playlist_id, season_indices):
playlist_url = self._make_url(playlist_id) playlist_url = self._make_url(playlist_id)
page_as_season = season_indices[page_num]
season_id = f'{playlist_id}-{page_as_season}'
season_page = self._download_webpage(f'{playlist_url}/episodes/season/{page_as_season}', video_id=season_id) for season_idx in season_indices:
season_id = f'{playlist_id}-{season_idx}'
season_page = self._download_webpage(f'{playlist_url}/episodes/season/{season_idx}', video_id=season_id)
episodes_metadata = [extract_attributes(elem) for elem in get_elements_html_by_class("video-summary", season_page)] episodes_metadata = [extract_attributes(elem) for elem in get_elements_html_by_class("video-summary", season_page)]
for episode_metadata in episodes_metadata: for episode_metadata in episodes_metadata:
yield self.url_result( yield self.url_result(
@ -815,13 +816,11 @@ class PBSShowIE(InfoExtractor):
playlist_title = show_metadata['data-gtm-label'] playlist_title = show_metadata['data-gtm-label']
clean_html(playlist_description[0]) clean_html(playlist_description[0])
# iterate seasons in reverse to get newest vids first
season_indices = list(sorted([x['ordinal'] for x in show_data['episodes_data']['seasons'] if x.get('ordinal', 0) != 0], reverse=True)) season_indices = list(sorted([x['ordinal'] for x in show_data['episodes_data']['seasons'] if x.get('ordinal', 0) != 0], reverse=True))
return self.playlist_result( return self.playlist_result(
OnDemandPagedList( LazyList(self._fetch_seasons(playlist_id, season_indices)),
pagefunc=functools.partial(self._fetch_season_page, playlist_id, season_indices),
pagesize=self.PAGE_SIZE
),
playlist_id=playlist_id, playlist_id=playlist_id,
playlist_title=playlist_title, playlist_title=playlist_title,
playlist_description=playlist_description, playlist_description=playlist_description,