kumi/yt-dlp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

[sportbox] Fix extraction

Browse source
This commit is contained in:
Sergey M․ 2015-05-22 00:45:33 +06:00
parent 0459432d96
commit 77d9cb2f04
1 changed files with 9 additions and 16 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

25
youtube_dl/extractor/sportbox.py
View file

@ -7,7 +7,7 @@ from .common import InfoExtractor
from ..compat import compat_urlparse from ..compat import compat_urlparse
from ..utils import ( from ..utils import (
parse_duration, parse_duration,
parse_iso8601, unified_strdate,
) )
@ -20,11 +20,9 @@ class SportBoxIE(InfoExtractor):
'id': '80822', 'id': '80822',
'ext': 'mp4', 'ext': 'mp4',
'title': 'Гонка 2 заезд ««Объединенный 2000»: классы Туринг и Супер-продакшн', 'title': 'Гонка 2 заезд ««Объединенный 2000»: классы Туринг и Супер-продакшн',
'description': 'md5:81715fa9c4ea3d9e7915dc8180c778ed', 'description': 'md5:3d72dc4a006ab6805d82f037fdc637ad',
'thumbnail': 're:^https?://.*\.jpg$', 'thumbnail': 're:^https?://.*\.jpg$',
'timestamp': 1411896237,
'upload_date': '20140928', 'upload_date': '20140928',
'duration': 4846,
}, },
'params': { 'params': {
# m3u8 download # m3u8 download
@ -48,17 +46,13 @@ class SportBoxIE(InfoExtractor):
r'src="/?(vdl/player/[^"]+)"', webpage, 'player') r'src="/?(vdl/player/[^"]+)"', webpage, 'player')
title = self._html_search_regex( title = self._html_search_regex(
r'<h1 itemprop="name">([^<]+)</h1>', webpage, 'title') [r'"nodetitle"\s*:\s*"([^"]+)"', r'class="node-header_{1,2}title">([^<]+)'],
description = self._html_search_regex( webpage, 'title')
r'(?s)<div itemprop="description">(.+?)</div>', description = self._og_search_description(webpage) or self._html_search_meta(
webpage, 'description', fatal=False) 'description', webpage, 'description')
thumbnail = self._og_search_thumbnail(webpage) thumbnail = self._og_search_thumbnail(webpage)
timestamp = parse_iso8601(self._search_regex( upload_date = unified_strdate(self._html_search_meta(
r'<span itemprop="uploadDate">([^<]+)</span>', 'dateCreated', webpage, 'upload date'))
webpage, 'timestamp', fatal=False))
duration = parse_duration(self._html_search_regex(
r'<meta itemprop="duration" content="PT([^"]+)">',
webpage, 'duration', fatal=False))
return { return {
'_type': 'url_transparent', '_type': 'url_transparent',
@ -67,8 +61,7 @@ class SportBoxIE(InfoExtractor):
'title': title, 'title': title,
'description': description, 'description': description,
'thumbnail': thumbnail, 'thumbnail': thumbnail,
'timestamp': timestamp, 'upload_date': upload_date,
'duration': duration,
} }