dbrun
/
youtube-dl
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

TEDIE: download the best quality video and use the new _search_regex functions 

Also extracts the description.
This commit is contained in:
Jaime Marquínez Ferrándiz 2013-06-20 20:51:20 +02:00
parent 038a3a1a61
commit f4c8bbcfc2
2 changed files with 22 additions and 26 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
test/tests.json
View File

@ -363,10 +363,10 @@
"name": "TED", "name": "TED",
"url": "http://www.ted.com/talks/dan_dennett_on_our_consciousness.html", "url": "http://www.ted.com/talks/dan_dennett_on_our_consciousness.html",
"file": "102.mp4", "file": "102.mp4",
"md5": "7bc087e71d16f18f9b8ab9fa62a8a031", "md5": "8cd9dfa41ee000ce658fd48fb5d89a61",
"info_dict": { "info_dict": {
"title": "Dan Dennett: The illusion of consciousness", "title": "Dan Dennett: The illusion of consciousness",
"thumbnail": "http://images.ted.com/images/ted/488_389x292.jpg" "description": "md5:c6fa72e6eedbd938c9caf6b2702f5922"
} }
}, },
{ {

44
youtube_dl/InfoExtractors.py
View File

@ -3792,10 +3792,6 @@ class TEDIE(InfoExtractor):
self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name)) self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name))
return [self._playlist_videos_info(url,name,playlist_id)] return [self._playlist_videos_info(url,name,playlist_id)]
def _talk_video_link(self,mediaSlug):
'''Returns the video link for that mediaSlug'''
return 'http://download.ted.com/talks/%s.mp4' % mediaSlug
def _playlist_videos_info(self,url,name,playlist_id=0): def _playlist_videos_info(self,url,name,playlist_id=0):
'''Returns the videos of the playlist''' '''Returns the videos of the playlist'''
video_RE=r''' video_RE=r'''
@ -3808,9 +3804,8 @@ class TEDIE(InfoExtractor):
m_videos=re.finditer(video_RE,webpage,re.VERBOSE) m_videos=re.finditer(video_RE,webpage,re.VERBOSE)
m_names=re.finditer(video_name_RE,webpage) m_names=re.finditer(video_name_RE,webpage)
playlist_RE = r'div class="headline">(\s*?)<h1>(\s*?)<span>(?P<playlist_title>.*?)</span>' playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
m_playlist = re.search(playlist_RE, webpage) webpage, 'playlist title')
playlist_title = m_playlist.group('playlist_title')
playlist_entries = [] playlist_entries = []
for m_video, m_name in zip(m_videos,m_names): for m_video, m_name in zip(m_videos,m_names):
@ -3821,27 +3816,28 @@ class TEDIE(InfoExtractor):
def _talk_info(self, url, video_id=0): def _talk_info(self, url, video_id=0):
"""Return the video for the talk in the url""" """Return the video for the talk in the url"""
m=re.match(self._VALID_URL, url,re.VERBOSE) m = re.match(self._VALID_URL, url,re.VERBOSE)
videoName=m.group('name') video_name = m.group('name')
webpage=self._download_webpage(url, video_id, 'Downloading \"%s\" page' % videoName) webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
self.report_extraction(video_name)
# If the url includes the language we get the title translated # If the url includes the language we get the title translated
title_RE=r'<span id="altHeadline" >(?P<title>.*)</span>' title = self._html_search_regex(r'<span id="altHeadline" >(?P<title>.*)</span>',
title=re.search(title_RE, webpage).group('title') webpage, 'title')
info_RE=r'''<script\ type="text/javascript">var\ talkDetails\ =(.*?) json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
"id":(?P<videoID>[\d]+).*? webpage, 'json data')
"mediaSlug":"(?P<mediaSlug>[\w\d]+?)"''' info = json.loads(json_data)
thumb_RE=r'</span>[\s.]*</div>[\s.]*<img src="(?P<thumbnail>.*?)"' desc = self._html_search_regex(r'<div class="talk-intro">.*?<p.*?>(.*?)</p>',
thumb_match=re.search(thumb_RE,webpage) webpage, 'description', flags = re.DOTALL)
info_match=re.search(info_RE,webpage,re.VERBOSE)
video_id=info_match.group('videoID') thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
mediaSlug=info_match.group('mediaSlug') webpage, 'thumbnail')
video_url=self._talk_video_link(mediaSlug)
info = { info = {
'id': video_id, 'id': info['id'],
'url': video_url, 'url': info['htmlStreams'][-1]['file'],
'ext': 'mp4', 'ext': 'mp4',
'title': title, 'title': title,
'thumbnail': thumb_match.group('thumbnail') 'thumbnail': thumbnail,
'description': desc,
} }
return info return info