w/youtube-dl
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

[spiegel] Add description and modernize

Browse source
This commit is contained in:
Sergey M․ 2014-06-21 16:31:18 +07:00
parent 98aeac6ea9
commit 55c97a03e1
1 changed files with 26 additions and 13 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

39
youtube_dl/extractor/spiegel.py
View file

@ -1,3 +1,4 @@
# encoding: utf-8
from __future__ import unicode_literals from __future__ import unicode_literals
import re import re
@ -9,18 +10,33 @@ class SpiegelIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?spiegel\.de/video/[^/]*-(?P<videoID>[0-9]+)(?:\.html)?(?:#.*)?$' _VALID_URL = r'https?://(?:www\.)?spiegel\.de/video/[^/]*-(?P<videoID>[0-9]+)(?:\.html)?(?:#.*)?$'
_TESTS = [{ _TESTS = [{
'url': 'http://www.spiegel.de/video/vulkan-tungurahua-in-ecuador-ist-wieder-aktiv-video-1259285.html', 'url': 'http://www.spiegel.de/video/vulkan-tungurahua-in-ecuador-ist-wieder-aktiv-video-1259285.html',
'file': '1259285.mp4',
'md5': '2c2754212136f35fb4b19767d242f66e', 'md5': '2c2754212136f35fb4b19767d242f66e',
'info_dict': { 'info_dict': {
'id': '1259285',
'ext': 'mp4',
'title': 'Vulkanausbruch in Ecuador: Der "Feuerschlund" ist wieder aktiv', 'title': 'Vulkanausbruch in Ecuador: Der "Feuerschlund" ist wieder aktiv',
'description': 'md5:8029d8310232196eb235d27575a8b9f4',
'duration': 49,
}, },
}, }, {
{
'url': 'http://www.spiegel.de/video/schach-wm-videoanalyse-des-fuenften-spiels-video-1309159.html', 'url': 'http://www.spiegel.de/video/schach-wm-videoanalyse-des-fuenften-spiels-video-1309159.html',
'file': '1309159.mp4',
'md5': 'f2cdf638d7aa47654e251e1aee360af1', 'md5': 'f2cdf638d7aa47654e251e1aee360af1',
'info_dict': { 'info_dict': {
'id': '1309159',
'ext': 'mp4',
'title': 'Schach-WM in der Videoanalyse: Carlsen nutzt die Fehlgriffe des Titelverteidigers', 'title': 'Schach-WM in der Videoanalyse: Carlsen nutzt die Fehlgriffe des Titelverteidigers',
'description': 'md5:c2322b65e58f385a820c10fa03b2d088',
'duration': 983,
},
}, {
'url': 'http://www.spiegel.de/video/johann-westhauser-videobotschaft-des-hoehlenforschers-video-1502367.html',
'md5': '54f58ba0e752e3c07bc2a26222dd0acf',
'info_dict': {
'id': '1502367',
'ext': 'mp4',
'title': 'Videobotschaft: Höhlenforscher Westhauser dankt seinen Rettern',
'description': 'md5:c6f1ec11413ebd1088b6813943e5fc91',
'duration': 42,
}, },
}] }]
@ -30,19 +46,15 @@ class SpiegelIE(InfoExtractor):
webpage = self._download_webpage(url, video_id) webpage = self._download_webpage(url, video_id)
video_title = self._html_search_regex( title = self._html_search_regex(
r'<div class="module-title">(.*?)</div>', webpage, 'title') r'<div class="module-title">(.*?)</div>', webpage, 'title')
description = self._html_search_meta('description', webpage, 'description')
base_url = self._search_regex( base_url = self._search_regex(
r'var\s+server\s+=\s+\"(http://video\d*\.spiegel\.de/flash/\d+/\d+/)\";', r'var\s+server\s*=\s*"([^"]+)\"', webpage, 'server URL')
webpage,
'base_url',
)
xml_url = base_url + video_id + '.xml' xml_url = base_url + video_id + '.xml'
idoc = self._download_xml( idoc = self._download_xml(xml_url, video_id)
xml_url, video_id,
note='Downloading XML', errnote='Failed to download XML from "{0}"'.format(xml_url))
formats = [ formats = [
{ {
@ -65,7 +77,8 @@ class SpiegelIE(InfoExtractor):
return { return {
'id': video_id, 'id': video_id,
'title': video_title, 'title': title,
'description': description,
'duration': duration, 'duration': duration,
'formats': formats, 'formats': formats,
} }