403 on user page

[pablouser1]

403 on user page

[ligix]

Unrelated but the redirector pattern looks wrong (?).

https://vm.tiktok.com/ZMLja28XQ/ gets redirected to https://proxitok.herokuapp.com/ZMLja28XQ/ which is a 404.
It should be https://proxitok.herokuapp.com/@placeholder/video/ZMLja28XQ instead.

[pablouser1]

The user page shows 403 instead of something

Example: https://proxitok.herokuapp.com/@roro.cat

That is because the official instance is being rate limited by TikTok. The only "fix" to this is using another ProxiTok instance or waiting some hours. I also added a robots.txt to the project, almost all of the requests made to the instance were just bots.

Unrelated but the redirector pattern looks wrong (?).

https://vm.tiktok.com/ZMLja28XQ/ gets redirected to https://proxitok.herokuapp.com/ZMLja28XQ/ which is a 404. It should be https://proxitok.herokuapp.com/@placeholder/video/ZMLja28XQ instead.

Probably because Include pattern: (.*//.*)(tiktok.com)(.*) redirects all subdomains to proxitok.herokuapp.com, changing it to only www would do the trick. You would have to make at least 1 connection to the TikTok servers though. As far as I'm aware, Libredirect also uses this method.

[ligix]

You would have to make at least 1 connection to the TikTok servers though

I see. Multiple patterns might be needed then

almost all of the requests made to the instance were just bots

Maybe adding support for http auth might be a good idea (?) You could just use a as both user and password to block all bots (mainly for heroku, with a reverse proxy adding it is simple)

[pablouser1]

Maybe adding support for http auth might be a good idea (?) You could just use a as both user and password to block all bots (mainly for heroku, with a reverse proxy adding it is simple)

The crawlers that were using the instance respect robots.txt so I don't think this is needed, at least for now

[pablouser1]

I'm closing this issue for now... It looks like TikTok doesn't return a 403 anymore when the rate limit is triggered