yt-dlp/youtube_dl/extractor/vevo.py

76 lines
2.7 KiB
Python
Raw Normal View History

2013-06-24 12:31:41 +02:00
import re
import json
import xml.etree.ElementTree
import datetime
2013-06-24 12:31:41 +02:00
from .common import InfoExtractor
from ..utils import (
determine_ext,
2013-06-24 12:31:41 +02:00
ExtractorError,
)
2013-06-24 12:31:41 +02:00
class VevoIE(InfoExtractor):
"""
2013-08-11 07:12:38 +02:00
Accepts urls from vevo.com or in the format 'vevo:{id}'
(currently used by MTVIE)
"""
2013-08-21 18:20:03 +02:00
_VALID_URL = r'((http://www.vevo.com/watch/.*?/.*?/)|(vevo:))(?P<id>.*?)(\?|$)'
2013-06-27 20:46:46 +02:00
_TEST = {
u'url': u'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
u'file': u'GB1101300280.mp4',
u'info_dict': {
2013-08-21 18:20:03 +02:00
u"upload_date": u"20130624",
u"uploader": u"Hurts",
u"title": u"Somebody to Die For",
u'duration': 230,
2013-06-27 20:46:46 +02:00
}
}
2013-06-24 12:31:41 +02:00
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
video_id = mobj.group('id')
json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
2013-06-24 12:31:41 +02:00
info_json = self._download_webpage(json_url, video_id, u'Downloading json info')
self.report_extraction(video_id)
video_info = json.loads(info_json)['video']
last_version = {'version': -1}
for version in video_info['videoVersions']:
# These are the HTTP downloads, other types are for different manifests
if version['sourceType'] == 2:
if version['version'] > last_version['version']:
last_version = version
if last_version['version'] == -1:
raise ExtractorError(u'Unable to extract last version of the video')
renditions = xml.etree.ElementTree.fromstring(last_version['data'])
formats = []
# Already sorted from worst to best quality
for rend in renditions.findall('rendition'):
attr = rend.attrib
2013-10-29 15:10:09 +01:00
format_note = '%(videoCodec)s@%(videoBitrate)4sK, %(audioCodec)s@%(audioBitrate)3sK' % attr
formats.append({
2013-10-29 15:10:09 +01:00
'url': attr['url'],
'format_id': attr['name'],
'format_note': format_note,
'height': int(attr['frameheight']),
'width': int(attr['frameWidth']),
})
timestamp_ms = int(self._search_regex(
r'/Date\((\d+)\)/', video_info['launchDate'], u'launch date'))
upload_date = datetime.datetime.fromtimestamp(timestamp_ms // 1000)
info = {
'id': video_id,
'title': video_info['title'],
'formats': formats,
'thumbnail': video_info['imageUrl'],
'upload_date': upload_date.strftime('%Y%m%d'),
'uploader': video_info['mainArtists'][0]['artistName'],
'duration': video_info['duration'],
}
return info