[udn] fix extraction

This commit is contained in:
Remita Amine 2017-10-12 16:04:41 +00:00
parent 414e709405
commit 81ce479f4d

View file

@ -1,7 +1,6 @@
# coding: utf-8 # coding: utf-8
from __future__ import unicode_literals from __future__ import unicode_literals
import json
import re import re
from .common import InfoExtractor from .common import InfoExtractor
@ -29,6 +28,7 @@ class UDNEmbedIE(InfoExtractor):
# m3u8 download # m3u8 download
'skip_download': True, 'skip_download': True,
}, },
'expected_warnings': ['Failed to parse JSON Expecting value'],
}, { }, {
'url': 'https://video.udn.com/embed/news/300040', 'url': 'https://video.udn.com/embed/news/300040',
'only_matching': True, 'only_matching': True,
@ -43,10 +43,21 @@ class UDNEmbedIE(InfoExtractor):
page = self._download_webpage(url, video_id) page = self._download_webpage(url, video_id)
options = json.loads(js_to_json(self._html_search_regex( options_str = self._html_search_regex(
r'var\s+options\s*=\s*([^;]+);', page, 'video urls dictionary'))) r'var\s+options\s*=\s*([^;]+);', page, 'options')
trans_options_str = js_to_json(options_str)
video_urls = options['video'] options = self._parse_json(trans_options_str, 'options', fatal=False) or {}
if options:
video_urls = options['video']
title = options['title']
poster = options.get('poster')
else:
video_urls = self._parse_json(self._html_search_regex(
r'"video"\s*:\s*({.+?})\s*,', trans_options_str, 'video urls'), 'video urls')
title = self._html_search_regex(
r"title\s*:\s*'(.+?)'\s*,", options_str, 'title')
poster = self._html_search_regex(
r"poster\s*:\s*'(.+?)'\s*,", options_str, 'poster', default=None)
if video_urls.get('youtube'): if video_urls.get('youtube'):
return self.url_result(video_urls.get('youtube'), 'Youtube') return self.url_result(video_urls.get('youtube'), 'Youtube')
@ -68,7 +79,7 @@ class UDNEmbedIE(InfoExtractor):
formats.extend(self._extract_f4m_formats( formats.extend(self._extract_f4m_formats(
video_url, video_id, f4m_id='hds')) video_url, video_id, f4m_id='hds'))
else: else:
mobj = re.search(r'_(?P<height>\d+)p_(?P<tbr>\d+).mp4', video_url) mobj = re.search(r'_(?P<height>\d+)p_(?P<tbr>\d+)\.mp4', video_url)
a_format = { a_format = {
'url': video_url, 'url': video_url,
# video_type may be 'mp4', which confuses YoutubeDL # video_type may be 'mp4', which confuses YoutubeDL
@ -83,14 +94,9 @@ class UDNEmbedIE(InfoExtractor):
self._sort_formats(formats) self._sort_formats(formats)
thumbnails = [{
'url': img_url,
'id': img_type,
} for img_type, img_url in options.get('gallery', [{}])[0].items() if img_url]
return { return {
'id': video_id, 'id': video_id,
'formats': formats, 'formats': formats,
'title': options['title'], 'title': title,
'thumbnails': thumbnails, 'thumbnail': poster,
} }