yt_dlp/extractor/yandexdisk.py

   1 import json
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     determine_ext,
   6     float_or_none,
   7     int_or_none,
   8     join_nonempty,
   9     mimetype2ext,
  10     try_get,
  11     urljoin,
  12 )
  13
  14
  15 class YandexDiskIE(InfoExtractor):
  16     _VALID_URL = r'''(?x)https?://
  17         (?P<domain>
  18             yadi\.sk|
  19             disk\.yandex\.
  20                 (?:
  21                     az|
  22                     by|
  23                     co(?:m(?:\.(?:am|ge|tr))?|\.il)|
  24                     ee|
  25                     fr|
  26                     k[gz]|
  27                     l[tv]|
  28                     md|
  29                     t[jm]|
  30                     u[az]|
  31                     ru
  32                 )
  33         )/(?:[di]/|public.*?\bhash=)(?P<id>[^/?#&]+)'''
  34
  35     _TESTS = [{
  36         'url': 'https://yadi.sk/i/VdOeDou8eZs6Y',
  37         'md5': 'a4a8d52958c8fddcf9845935070402ae',
  38         'info_dict': {
  39             'id': 'VdOeDou8eZs6Y',
  40             'ext': 'mp4',
  41             'title': '4.mp4',
  42             'duration': 168.6,
  43             'uploader': 'y.botova',
  44             'uploader_id': '300043621',
  45             'view_count': int,
  46         },
  47         'expected_warnings': ['Unable to download JSON metadata'],
  48     }, {
  49         'url': 'https://yadi.sk/d/h3WAXvDS3Li3Ce',
  50         'only_matching': True,
  51     }, {
  52         'url': 'https://yadi.sk/public?hash=5DZ296JK9GWCLp02f6jrObjnctjRxMs8L6%2B%2FuhNqk38%3D',
  53         'only_matching': True,
  54     }]
  55
  56     def _real_extract(self, url):
  57         domain, video_id = self._match_valid_url(url).groups()
  58
  59         webpage = self._download_webpage(url, video_id)
  60         store = self._parse_json(self._search_regex(
  61             r'<script[^>]+id="store-prefetch"[^>]*>\s*({.+?})\s*</script>',
  62             webpage, 'store'), video_id)
  63         resource = store['resources'][store['rootResourceId']]
  64
  65         title = resource['name']
  66         meta = resource.get('meta') or {}
  67
  68         public_url = meta.get('short_url')
  69         if public_url:
  70             video_id = self._match_id(public_url)
  71
  72         source_url = (self._download_json(
  73             'https://cloud-api.yandex.net/v1/disk/public/resources/download',
  74             video_id, query={'public_key': url}, fatal=False) or {}).get('href')
  75         video_streams = resource.get('videoStreams') or {}
  76         video_hash = resource.get('hash') or url
  77         environment = store.get('environment') or {}
  78         sk = environment.get('sk')
  79         yandexuid = environment.get('yandexuid')
  80         if sk and yandexuid and not (source_url and video_streams):
  81             self._set_cookie(domain, 'yandexuid', yandexuid)
  82
  83             def call_api(action):
  84                 return (self._download_json(
  85                     urljoin(url, '/public/api/') + action, video_id, data=json.dumps({
  86                         'hash': video_hash,
  87                         'sk': sk,
  88                     }).encode(), headers={
  89                         'Content-Type': 'text/plain',
  90                     }, fatal=False) or {}).get('data') or {}
  91             if not source_url:
  92                 # TODO: figure out how to detect if download limit has
  93                 # been reached and then avoid unnecessary source format
  94                 # extraction requests
  95                 source_url = call_api('download-url').get('url')
  96             if not video_streams:
  97                 video_streams = call_api('get-video-streams')
  98
  99         formats = []
 100         if source_url:
 101             formats.append({
 102                 'url': source_url,
 103                 'format_id': 'source',
 104                 'ext': determine_ext(title, meta.get('ext') or mimetype2ext(meta.get('mime_type')) or 'mp4'),
 105                 'quality': 1,
 106                 'filesize': int_or_none(meta.get('size')),
 107             })
 108
 109         for video in (video_streams.get('videos') or []):
 110             format_url = video.get('url')
 111             if not format_url:
 112                 continue
 113             if video.get('dimension') == 'adaptive':
 114                 formats.extend(self._extract_m3u8_formats(
 115                     format_url, video_id, 'mp4', 'm3u8_native',
 116                     m3u8_id='hls', fatal=False))
 117             else:
 118                 size = video.get('size') or {}
 119                 height = int_or_none(size.get('height'))
 120                 formats.append({
 121                     'ext': 'mp4',
 122                     'format_id': join_nonempty('hls', height and f'{height}p'),
 123                     'height': height,
 124                     'protocol': 'm3u8_native',
 125                     'url': format_url,
 126                     'width': int_or_none(size.get('width')),
 127                 })
 128
 129         uid = resource.get('uid')
 130         display_name = try_get(store, lambda x: x['users'][uid]['displayName'])
 131
 132         return {
 133             'id': video_id,
 134             'title': title,
 135             'duration': float_or_none(video_streams.get('duration'), 1000),
 136             'uploader': display_name,
 137             'uploader_id': uid,
 138             'view_count': int_or_none(meta.get('views_counter')),
 139             'formats': formats,
 140         }