github/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-07 20:30:41 -05:00
Code Issues Projects Releases Packages Wiki Activity

[twitter] Fix metadata extraction and test_Twitter_1

Browse source
This commit is contained in:
Yen Chi Hsuan 2016-02-21 17:29:28 +08:00
parent 132e3b74bd
commit 7efc1c2b49
1 changed files with 12 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

19
youtube_dl/extractor/twitter.py
View file

@ -165,6 +165,7 @@ class TwitterIE(InfoExtractor):
'uploader': 'Gifs',
'uploader_id': 'giphz',
},
'expected_warnings': ['height', 'width'],
}, {
'url': 'https://twitter.com/starwars/status/665052190608723968',
'md5': '39b7199856dee6cd4432e72c74bc69d4',
@ -212,20 +213,24 @@ def _real_extract(self, url):
return info
mobj = re.search(r'''(?x)
<video[^>]+class="animated-gif"[^>]+
(?:data-height="(?P<height>\d+)")?[^>]+
(?:data-width="(?P<width>\d+)")?[^>]+
(?:poster="(?P<poster>[^"]+)")?[^>]*>\s*
<video[^>]+class="animated-gif"(?P<more_info>[^>]+)>\s*
<source[^>]+video-src="(?P<url>[^"]+)"
''', webpage)
if mobj:
more_info = mobj.group('more_info')
height = int_or_none(self._search_regex(
r'data-height="(\d+)"', more_info, 'height', fatal=False))
width = int_or_none(self._search_regex(
r'data-width="(\d+)"', more_info, 'width', fatal=False))
thumbnail = self._search_regex(
r'poster="([^"]+)"', more_info, 'poster', fatal=False)
info.update({
'id': twid,
'url': mobj.group('url'),
'height': int_or_none(mobj.group('height')),
'width': int_or_none(mobj.group('width')),
'thumbnail': mobj.group('poster'),
'height': height,
'width': width,
'thumbnail': thumbnail,
})
return info