youtube_dl/extractor/vevo.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import xml.etree.ElementTree
   5 import datetime
   6
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     compat_HTTPError,
  10     ExtractorError,
  11 )
  12
  13
  14 class VevoIE(InfoExtractor):
  15     """
  16     Accepts urls from vevo.com or in the format 'vevo:{id}'
  17     (currently used by MTVIE)
  18     """
  19     _VALID_URL = r'''(?x)
  20         (?:https?://www\.vevo\.com/watch/(?:[^/]+/[^/]+/)?|
  21            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
  22            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
  23            vevo:)
  24         (?P<id>[^&?#]+)'''
  25     _TESTS = [{
  26         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
  27         'file': 'GB1101300280.mp4',
  28         "md5": "06bea460acb744eab74a9d7dcb4bfd61",
  29         'info_dict': {
  30             "upload_date": "20130624",
  31             "uploader": "Hurts",
  32             "title": "Somebody to Die For",
  33             "duration": 230.12,
  34             "width": 1920,
  35             "height": 1080,
  36         }
  37     }]
  38     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
  39
  40     def _formats_from_json(self, video_info):
  41         last_version = {'version': -1}
  42         for version in video_info['videoVersions']:
  43             # These are the HTTP downloads, other types are for different manifests
  44             if version['sourceType'] == 2:
  45                 if version['version'] > last_version['version']:
  46                     last_version = version
  47         if last_version['version'] == -1:
  48             raise ExtractorError('Unable to extract last version of the video')
  49
  50         renditions = xml.etree.ElementTree.fromstring(last_version['data'])
  51         formats = []
  52         # Already sorted from worst to best quality
  53         for rend in renditions.findall('rendition'):
  54             attr = rend.attrib
  55             format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
  56             formats.append({
  57                 'url': attr['url'],
  58                 'format_id': attr['name'],
  59                 'format_note': format_note,
  60                 'height': int(attr['frameheight']),
  61                 'width': int(attr['frameWidth']),
  62             })
  63         return formats
  64
  65     def _formats_from_smil(self, smil_xml):
  66         formats = []
  67         smil_doc = xml.etree.ElementTree.fromstring(smil_xml.encode('utf-8'))
  68         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
  69         for el in els:
  70             src = el.attrib['src']
  71             m = re.match(r'''(?xi)
  72                 (?P<ext>[a-z0-9]+):
  73                 (?P<path>
  74                     [/a-z0-9]+     # The directory and main part of the URL
  75                     _(?P<cbr>[0-9]+)k
  76                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
  77                     _(?P<vcodec>[a-z0-9]+)
  78                     _(?P<vbr>[0-9]+)
  79                     _(?P<acodec>[a-z0-9]+)
  80                     _(?P<abr>[0-9]+)
  81                     \.[a-z0-9]+  # File extension
  82                 )''', src)
  83             if not m:
  84                 continue
  85
  86             format_url = self._SMIL_BASE_URL + m.group('path')
  87             formats.append({
  88                 'url': format_url,
  89                 'format_id': 'SMIL_' + m.group('cbr'),
  90                 'vcodec': m.group('vcodec'),
  91                 'acodec': m.group('acodec'),
  92                 'vbr': int(m.group('vbr')),
  93                 'abr': int(m.group('abr')),
  94                 'ext': m.group('ext'),
  95                 'width': int(m.group('width')),
  96                 'height': int(m.group('height')),
  97             })
  98         return formats
  99
 100     def _real_extract(self, url):
 101         mobj = re.match(self._VALID_URL, url)
 102         video_id = mobj.group('id')
 103
 104         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 105         video_info = self._download_json(json_url, video_id)['video']
 106
 107         formats = self._formats_from_json(video_info)
 108         try:
 109             smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
 110                 self._SMIL_BASE_URL, video_id, video_id.lower())
 111             smil_xml = self._download_webpage(smil_url, video_id,
 112                                               'Downloading SMIL info')
 113             formats.extend(self._formats_from_smil(smil_xml))
 114         except ExtractorError as ee:
 115             if not isinstance(ee.cause, compat_HTTPError):
 116                 raise
 117             self._downloader.report_warning(
 118                 'Cannot download SMIL information, falling back to JSON ..')
 119
 120         timestamp_ms = int(self._search_regex(
 121             r'/Date\((\d+)\)/', video_info['launchDate'], 'launch date'))
 122         upload_date = datetime.datetime.fromtimestamp(timestamp_ms // 1000)
 123         return {
 124             'id': video_id,
 125             'title': video_info['title'],
 126             'formats': formats,
 127             'thumbnail': video_info['imageUrl'],
 128             'upload_date': upload_date.strftime('%Y%m%d'),
 129             'uploader': video_info['mainArtists'][0]['artistName'],
 130             'duration': video_info['duration'],
 131         }