Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/vevo.py

   1 import re
   2 import json
   3 import xml.etree.ElementTree
   4 import datetime
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_HTTPError,
   9     ExtractorError,
  10 )
  11
  12
  13 class VevoIE(InfoExtractor):
  14     """
  15     Accepts urls from vevo.com or in the format 'vevo:{id}'
  16     (currently used by MTVIE)
  17     """
  18     _VALID_URL = r'((http://www.vevo.com/watch/.*?/.*?/)|(vevo:))(?P<id>.*?)(\?|$)'
  19     _TESTS = [{
  20         u'url': u'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
  21         u'file': u'GB1101300280.mp4',
  22         u"md5": u"06bea460acb744eab74a9d7dcb4bfd61",
  23         u'info_dict': {
  24             u"upload_date": u"20130624",
  25             u"uploader": u"Hurts",
  26             u"title": u"Somebody to Die For",
  27             u"duration": 230,
  28             u"width": 1920,
  29             u"height": 1080,
  30         }
  31     }]
  32     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
  33
  34     def _formats_from_json(self, video_info):
  35         last_version = {'version': -1}
  36         for version in video_info['videoVersions']:
  37             # These are the HTTP downloads, other types are for different manifests
  38             if version['sourceType'] == 2:
  39                 if version['version'] > last_version['version']:
  40                     last_version = version
  41         if last_version['version'] == -1:
  42             raise ExtractorError(u'Unable to extract last version of the video')
  43
  44         renditions = xml.etree.ElementTree.fromstring(last_version['data'])
  45         formats = []
  46         # Already sorted from worst to best quality
  47         for rend in renditions.findall('rendition'):
  48             attr = rend.attrib
  49             format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
  50             formats.append({
  51                 'url': attr['url'],
  52                 'format_id': attr['name'],
  53                 'format_note': format_note,
  54                 'height': int(attr['frameheight']),
  55                 'width': int(attr['frameWidth']),
  56             })
  57         return formats
  58
  59     def _formats_from_smil(self, smil_xml):
  60         formats = []
  61         smil_doc = xml.etree.ElementTree.fromstring(smil_xml.encode('utf-8'))
  62         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
  63         for el in els:
  64             src = el.attrib['src']
  65             m = re.match(r'''(?xi)
  66                 (?P<ext>[a-z0-9]+):
  67                 (?P<path>
  68                     [/a-z0-9]+     # The directory and main part of the URL
  69                     _(?P<cbr>[0-9]+)k
  70                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
  71                     _(?P<vcodec>[a-z0-9]+)
  72                     _(?P<vbr>[0-9]+)
  73                     _(?P<acodec>[a-z0-9]+)
  74                     _(?P<abr>[0-9]+)
  75                     \.[a-z0-9]+  # File extension
  76                 )''', src)
  77             if not m:
  78                 continue
  79
  80             format_url = self._SMIL_BASE_URL + m.group('path')
  81             format_note = ('%(vcodec)s@%(vbr)4sk, %(acodec)s@%(abr)3sk' %
  82                            m.groupdict())
  83             formats.append({
  84                 'url': format_url,
  85                 'format_id': u'SMIL_' + m.group('cbr'),
  86                 'format_note': format_note,
  87                 'ext': m.group('ext'),
  88                 'width': int(m.group('width')),
  89                 'height': int(m.group('height')),
  90             })
  91         return formats
  92
  93     def _real_extract(self, url):
  94         mobj = re.match(self._VALID_URL, url)
  95         video_id = mobj.group('id')
  96
  97         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
  98         info_json = self._download_webpage(json_url, video_id, u'Downloading json info')
  99         video_info = json.loads(info_json)['video']
 100
 101         formats = self._formats_from_json(video_info)
 102         try:
 103             smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
 104                 self._SMIL_BASE_URL, video_id, video_id.lower())
 105             smil_xml = self._download_webpage(smil_url, video_id,
 106                                               u'Downloading SMIL info')
 107             formats.extend(self._formats_from_smil(smil_xml))
 108         except ExtractorError as ee:
 109             if not isinstance(ee.cause, compat_HTTPError):
 110                 raise
 111             self._downloader.report_warning(
 112                 u'Cannot download SMIL information, falling back to JSON ..')
 113
 114         timestamp_ms = int(self._search_regex(
 115             r'/Date\((\d+)\)/', video_info['launchDate'], u'launch date'))
 116         upload_date = datetime.datetime.fromtimestamp(timestamp_ms // 1000)
 117         info = {
 118             'id': video_id,
 119             'title': video_info['title'],
 120             'formats': formats,
 121             'thumbnail': video_info['imageUrl'],
 122             'upload_date': upload_date.strftime('%Y%m%d'),
 123             'uploader': video_info['mainArtists'][0]['artistName'],
 124             'duration': video_info['duration'],
 125         }
 126
 127         return info