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     mimetype2ext,
   9     try_get,
  10     urljoin,
  11 )
  12
  13
  14 class YandexDiskIE(InfoExtractor):
  15     _VALID_URL = r'''(?x)https?://
  16         (?P<domain>
  17             yadi\.sk|
  18             disk\.yandex\.
  19                 (?:
  20                     az|
  21                     by|
  22                     co(?:m(?:\.(?:am|ge|tr))?|\.il)|
  23                     ee|
  24                     fr|
  25                     k[gz]|
  26                     l[tv]|
  27                     md|
  28                     t[jm]|
  29                     u[az]|
  30                     ru
  31                 )
  32         )/(?:[di]/|public.*?\bhash=)(?P<id>[^/?#&]+)'''
  33
  34     _TESTS = [{
  35         'url': 'https://yadi.sk/i/VdOeDou8eZs6Y',
  36         'md5': 'a4a8d52958c8fddcf9845935070402ae',
  37         'info_dict': {
  38             'id': 'VdOeDou8eZs6Y',
  39             'ext': 'mp4',
  40             'title': '4.mp4',
  41             'duration': 168.6,
  42             'uploader': 'y.botova',
  43             'uploader_id': '300043621',
  44             'view_count': int,
  45         },
  46         'expected_warnings': ['Unable to download JSON metadata'],
  47     }, {
  48         'url': 'https://yadi.sk/d/h3WAXvDS3Li3Ce',
  49         'only_matching': True,
  50     }, {
  51         'url': 'https://yadi.sk/public?hash=5DZ296JK9GWCLp02f6jrObjnctjRxMs8L6%2B%2FuhNqk38%3D',
  52         'only_matching': True,
  53     }]
  54
  55     def _real_extract(self, url):
  56         domain, video_id = self._match_valid_url(url).groups()
  57
  58         webpage = self._download_webpage(url, video_id)
  59         store = self._parse_json(self._search_regex(
  60             r'<script[^>]+id="store-prefetch"[^>]*>\s*({.+?})\s*</script>',
  61             webpage, 'store'), video_id)
  62         resource = store['resources'][store['rootResourceId']]
  63
  64         title = resource['name']
  65         meta = resource.get('meta') or {}
  66
  67         public_url = meta.get('short_url')
  68         if public_url:
  69             video_id = self._match_id(public_url)
  70
  71         source_url = (self._download_json(
  72             'https://cloud-api.yandex.net/v1/disk/public/resources/download',
  73             video_id, query={'public_key': url}, fatal=False) or {}).get('href')
  74         video_streams = resource.get('videoStreams') or {}
  75         video_hash = resource.get('hash') or url
  76         environment = store.get('environment') or {}
  77         sk = environment.get('sk')
  78         yandexuid = environment.get('yandexuid')
  79         if sk and yandexuid and not (source_url and video_streams):
  80             self._set_cookie(domain, 'yandexuid', yandexuid)
  81
  82             def call_api(action):
  83                 return (self._download_json(
  84                     urljoin(url, '/public/api/') + action, video_id, data=json.dumps({
  85                         'hash': video_hash,
  86                         'sk': sk,
  87                     }).encode(), headers={
  88                         'Content-Type': 'text/plain',
  89                     }, fatal=False) or {}).get('data') or {}
  90             if not source_url:
  91                 # TODO: figure out how to detect if download limit has
  92                 # been reached and then avoid unnecessary source format
  93                 # extraction requests
  94                 source_url = call_api('download-url').get('url')
  95             if not video_streams:
  96                 video_streams = call_api('get-video-streams')
  97
  98         formats = []
  99         if source_url:
 100             formats.append({
 101                 'url': source_url,
 102                 'format_id': 'source',
 103                 'ext': determine_ext(title, meta.get('ext') or mimetype2ext(meta.get('mime_type')) or 'mp4'),
 104                 'quality': 1,
 105                 'filesize': int_or_none(meta.get('size')),
 106             })
 107
 108         for video in (video_streams.get('videos') or []):
 109             format_url = video.get('url')
 110             if not format_url:
 111                 continue
 112             if video.get('dimension') == 'adaptive':
 113                 formats.extend(self._extract_m3u8_formats(
 114                     format_url, video_id, 'mp4', 'm3u8_native',
 115                     m3u8_id='hls', fatal=False))
 116             else:
 117                 size = video.get('size') or {}
 118                 height = int_or_none(size.get('height'))
 119                 format_id = 'hls'
 120                 if height:
 121                     format_id += f'-{height}p'
 122                 formats.append({
 123                     'ext': 'mp4',
 124                     'format_id': format_id,
 125                     'height': height,
 126                     'protocol': 'm3u8_native',
 127                     'url': format_url,
 128                     'width': int_or_none(size.get('width')),
 129                 })
 130
 131         uid = resource.get('uid')
 132         display_name = try_get(store, lambda x: x['users'][uid]['displayName'])
 133
 134         return {
 135             'id': video_id,
 136             'title': title,
 137             'duration': float_or_none(video_streams.get('duration'), 1000),
 138             'uploader': display_name,
 139             'uploader_id': uid,
 140             'view_count': int_or_none(meta.get('views_counter')),
 141             'formats': formats,
 142         }