[twitter] Don't fail if the description doesn't contain an URL (fixes #7489)

This commit is contained in:
Jaime Marquínez Ferrándiz 2015-11-13 19:09:42 +01:00
parent b84a5f0337
commit b703ebeeaf

View File

@ -147,6 +147,17 @@ class TwitterIE(InfoExtractor):
'uploader': 'Gifs', 'uploader': 'Gifs',
'uploader_id': 'giphz', 'uploader_id': 'giphz',
}, },
}, {
'url': 'https://twitter.com/starwars/status/665052190608723968',
'md5': '39b7199856dee6cd4432e72c74bc69d4',
'info_dict': {
'id': '665052190608723968',
'ext': 'mp4',
'title': 'Star Wars - A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens.',
'description': 'Star Wars on Twitter: "A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens."',
'uploader_id': 'starwars',
'uploader': 'Star Wars',
},
}] }]
def _real_extract(self, url): def _real_extract(self, url):
@ -158,17 +169,16 @@ class TwitterIE(InfoExtractor):
username = remove_end(self._og_search_title(webpage), ' on Twitter') username = remove_end(self._og_search_title(webpage), ' on Twitter')
title = self._og_search_description(webpage).strip('').replace('\n', ' ') title = description = self._og_search_description(webpage).strip('').replace('\n', ' ').strip('“”')
# strip 'https -_t.co_BJYgOjSeGA' junk from filenames # strip 'https -_t.co_BJYgOjSeGA' junk from filenames
mobj = re.match(r'“(.*)\s+(https?://[^ ]+)”', title) title = re.sub(r'\s+(https?://[^ ]+)', '', title)
title, short_url = mobj.groups()
info = { info = {
'uploader_id': user_id, 'uploader_id': user_id,
'uploader': username, 'uploader': username,
'webpage_url': url, 'webpage_url': url,
'description': '%s on Twitter: "%s %s"' % (username, title, short_url), 'description': '%s on Twitter: "%s"' % (username, description),
'title': username + ' - ' + title, 'title': username + ' - ' + title,
} }