Twitch VODs fail with parsing of the thumbnail data

[Salz]

Checklist

• I'm reporting a broken site support
• I've verified that I'm running youtube-dl version 2020.05.29
  I also have applied the diff from by previous bug #25528, otherwise I would getthe 410 error.
• I've checked that all provided URLs are alive and playable in a browser
  I have my browser configured to not play videos, but the videos are alive and the attached patch fixes playback of the videos.
• I've checked that all URLs and arguments with special characters are properly quoted or escaped
• I've searched the bugtracker for similar issues including closed ones

Verbose log

[debug] System config: []
[debug] User config: []
[debug] Custom config: []
[debug] Command-line args: ['--verbose', 'https://www.twitch.tv/videos/640872506']
[debug] Encodings: locale UTF-8, fs utf-8, out utf-8, pref UTF-8
[debug] youtube-dl version 2020.05.29
[debug] Python version 3.8.3 (CPython) - Linux-5.6.9-pinguin20200502-x86_64-with-glibc2.29
[debug] exe versions: ffmpeg 4.2.2-1, ffprobe 4.2.2-1, rtmpdump 2.4
[debug] Proxy map: {}
[twitch:vod] 640872506: Downloading vod info JSON
[twitch:vod] 640872506: Downloading vod access token
[twitch:vod] 640872506: Downloading m3u8 information
Traceback (most recent call last):
File "/usr/bin/youtube-dl", line 11, in
load_entry_point('youtube-dl==2020.5.29', 'console_scripts', 'youtube-dl')()
File "/usr/lib/python3/dist-packages/youtube_dl/init.py", line 478, in main
_real_main(argv)
File "/usr/lib/python3/dist-packages/youtube_dl/init.py", line 468, in _real_main
retcode = ydl.download(all_urls)
File "/usr/lib/python3/dist-packages/youtube_dl/YoutubeDL.py", line 2018, in download
res = self.extract_info(
File "/usr/lib/python3/dist-packages/youtube_dl/YoutubeDL.py", line 808, in extract_info
return self.process_ie_result(ie_result, download, extra_info)
File "/usr/lib/python3/dist-packages/youtube_dl/YoutubeDL.py", line 863, in process_ie_result
return self.process_video_result(ie_result, download=download)
File "/usr/lib/python3/dist-packages/youtube_dl/YoutubeDL.py", line 1460, in process_video_result
t['url'] = sanitize_url(t['url'])
File "/usr/lib/python3/dist-packages/youtube_dl/utils.py", line 2135, in sanitize_url
if url.startswith('//'):
AttributeError: 'dict' object has no attribute 'startswith'

Description

It seems Twitch is returning a dictionary of several thumbnails now with info.get('preview') (extractor/twitch.py line 194). On the above URL i get the following json data:

{"small": "https://static-cdn.jtvnw.net/cf_vods/d2nvs31859zcd8/dfa4f0f421b0286b0cb0_rahmschnitzel_1840886609_115117682//thumb/thumb0-80x45.jpg",
 "medium": "https://static-cdn.jtvnw.net/cf_vods/d2nvs31859zcd8/dfa4f0f421b0286b0cb0_rahmschnitzel_1840886609_115117682//thumb/thumb0-320x180.jpg",
 "large": "https://static-cdn.jtvnw.net/cf_vods/d2nvs31859zcd8/dfa4f0f421b0286b0cb0_rahmschnitzel_1840886609_115117682//thumb/thumb0-640x360.jpg",
 "template": "https://static-cdn.jtvnw.net/cf_vods/d2nvs31859zcd8/dfa4f0f421b0286b0cb0_rahmschnitzel_1840886609_115117682//thumb/thumb0-{width}x{height}.jpg"}

Parsing of the thumbnail(s) only supports a string or an array of dictionaries. Arbitrarily selecting one seems to work (see attached patch), otherwise the preview data must be reparsed into the expected format for "thumbnails".

thumbnail.diff.txt

[FREEZEball]

Can confirm the same issue, fixed with the applied patch.

[seboss666]

Thanks for this quick patch, hope to see it in the next release :)