X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/33cd347759d6d999325ebf3c69b7ed5692c343b2..eed74133841b43ce901ee0dd31c7dad5a234bd48:/youtube_dl/extractor/nba.py diff --git a/youtube_dl/extractor/nba.py b/youtube_dl/extractor/nba.py index 0f17890..78e650b 100644 --- a/youtube_dl/extractor/nba.py +++ b/youtube_dl/extractor/nba.py @@ -1,48 +1,49 @@ +from __future__ import unicode_literals + import re from .common import InfoExtractor from ..utils import ( - ExtractorError, + remove_end, + parse_duration, ) class NBAIE(InfoExtractor): - _VALID_URL = r'^(?:https?://)?(?:watch\.|www\.)?nba\.com/(?:nba/)?video(/[^?]*?)(?:/index\.html)?(?:\?.*)?$' + _VALID_URL = r'https?://(?:watch\.|www\.)?nba\.com/(?:nba/)?video(?P/[^?]*?)(?:/index\.html)?(?:\?.*)?$' _TEST = { - u'url': u'http://www.nba.com/video/games/nets/2012/12/04/0021200253-okc-bkn-recap.nba/index.html', - u'file': u'0021200253-okc-bkn-recap.nba.mp4', - u'md5': u'c0edcfc37607344e2ff8f13c378c88a4', - u'info_dict': { - u"description": u"Kevin Durant scores 32 points and dishes out six assists as the Thunder beat the Nets in Brooklyn.", - u"title": u"Thunder vs. Nets" - } + 'url': 'http://www.nba.com/video/games/nets/2012/12/04/0021200253-okc-bkn-recap.nba/index.html', + 'md5': 'c0edcfc37607344e2ff8f13c378c88a4', + 'info_dict': { + 'id': '0021200253-okc-bkn-recap.nba', + 'ext': 'mp4', + 'title': 'Thunder vs. Nets', + 'description': 'Kevin Durant scores 32 points and dishes out six assists as the Thunder beat the Nets in Brooklyn.', + 'duration': 181, + }, } 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) + video_id = mobj.group('id') webpage = self._download_webpage(url, video_id) - video_url = u'http://ht-mobile.cdn.turner.com/nba/big' + video_id + '_nba_1280x720.mp4' + video_url = 'http://ht-mobile.cdn.turner.com/nba/big' + video_id + '_nba_1280x720.mp4' shortened_video_id = video_id.rpartition('/')[2] - title = self._og_search_title(webpage, default=shortened_video_id).replace('NBA.com: ', '') + title = remove_end( + self._og_search_title(webpage, default=shortened_video_id), ' : NBA.com') - # It isn't there in the HTML it returns to us - # uploader_date = self._html_search_regex(r'Date: (.*?)', webpage, 'upload_date', fatal=False) + description = self._og_search_description(webpage) + duration = parse_duration( + self._html_search_meta('duration', webpage, 'duration', fatal=False)) - description = self._html_search_regex(r'', webpage, 'description', fatal=False) - info = { + return { 'id': shortened_video_id, 'url': video_url, - 'ext': 'mp4', 'title': title, - # 'uploader_date': uploader_date, 'description': description, + 'duration': duration, } - return [info]