mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2024-11-30 20:30:27 +00:00
41 lines
1.4 KiB
Python
41 lines
1.4 KiB
Python
import re
|
|
|
|
from .common import InfoExtractor
|
|
from ..utils import (
|
|
ExtractorError,
|
|
)
|
|
|
|
|
|
class NBAIE(InfoExtractor):
|
|
_VALID_URL = r'^(?:https?://)?(?:watch\.|www\.)?nba\.com/(?:nba/)?video(/[^?]*?)(?:/index\.html)?(?:\?.*)?$'
|
|
|
|
def _real_extract(self, url):
|
|
mobj = re.match(self._VALID_URL, url)
|
|
if mobj is None:
|
|
raise ExtractorError(u'Invalid URL: %s' % url)
|
|
|
|
video_id = mobj.group(1)
|
|
|
|
webpage = self._download_webpage(url, video_id)
|
|
|
|
video_url = u'http://ht-mobile.cdn.turner.com/nba/big' + video_id + '_nba_1280x720.mp4'
|
|
|
|
shortened_video_id = video_id.rpartition('/')[2]
|
|
title = self._html_search_regex(r'<meta property="og:title" content="(.*?)"',
|
|
webpage, 'title', default=shortened_video_id).replace('NBA.com: ', '')
|
|
|
|
# It isn't there in the HTML it returns to us
|
|
# uploader_date = self._html_search_regex(r'<b>Date:</b> (.*?)</div>', webpage, 'upload_date', fatal=False)
|
|
|
|
description = self._html_search_regex(r'<meta name="description" (?:content|value)="(.*?)" />', webpage, 'description', fatal=False)
|
|
|
|
info = {
|
|
'id': shortened_video_id,
|
|
'url': video_url,
|
|
'ext': 'mp4',
|
|
'title': title,
|
|
# 'uploader_date': uploader_date,
|
|
'description': description,
|
|
}
|
|
return [info]
|