[vine] Extract more metadata and support low format

This commit is contained in:
Sergey M. 2014-02-23 19:01:15 +07:00
parent 7ff5d5c2e2
commit f919201ecc
1 changed files with 31 additions and 15 deletions

View File

@ -1,8 +1,10 @@
from __future__ import unicode_literals from __future__ import unicode_literals
import re import re
import json
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import unified_strdate
class VineIE(InfoExtractor): class VineIE(InfoExtractor):
@ -13,32 +15,46 @@ class VineIE(InfoExtractor):
'info_dict': { 'info_dict': {
'id': 'b9KOOWX7HUx', 'id': 'b9KOOWX7HUx',
'ext': 'mp4', 'ext': 'mp4',
'uploader': 'Jack Dorsey',
'title': 'Chicken.', 'title': 'Chicken.',
'description': 'Chicken.',
'upload_date': '20130519',
'uploader': 'Jack Dorsey',
'uploader_id': '76',
}, },
} }
def _real_extract(self, url): def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url) mobj = re.match(self._VALID_URL, url)
video_id = mobj.group('id') video_id = mobj.group('id')
webpage_url = 'https://vine.co/v/' + video_id
webpage = self._download_webpage(webpage_url, video_id)
self.report_extraction(video_id) webpage = self._download_webpage('https://vine.co/v/' + video_id, video_id)
video_url = self._html_search_meta('twitter:player:stream', webpage, data = json.loads(self._html_search_regex(
'video URL') r'window\.POST_DATA = { %s: ({.+?}) }' % video_id, webpage, 'vine data'))
twitter_title = self._html_search_meta('twitter:title', webpage, formats = [
'twitter title') {
uploader = re.sub('\'s post on Vine', '', twitter_title) 'url': data['videoLowURL'],
'ext': 'mp4',
'format_id': 'low',
},
{
'url': data['videoUrl'],
'ext': 'mp4',
'format_id': 'standard',
}
]
return { return {
'id': video_id, 'id': video_id,
'url': video_url,
'ext': 'mp4',
'title': self._og_search_title(webpage), 'title': self._og_search_title(webpage),
'thumbnail': self._og_search_thumbnail(webpage), 'description': data['description'],
'uploader': uploader, 'thumbnail': data['thumbnailUrl'],
} 'upload_date': unified_strdate(data['created']),
'uploader': data['username'],
'uploader_id': data['userIdStr'],
'like_count': data['likes']['count'],
'comment_count': data['comments']['count'],
'repost_count': data['reposts']['count'],
'formats': formats,
}