User pages not loading: HTTP Code: 200, API error code 0 (OK)

[glottisfaun0000]

On any instance I try, I can load individual videos but user pages return:

There was an error processing your request!
HTTP Code: 200
API error code 0 (OK)

Individual video example: https://proxitok.pabloferreiro.es/%40california.shelling/video/7276537102018333994
User page example: https://proxitok.pabloferreiro.es/@california.shelling

And it looks like RSS is broken by this too.

[dajix350]

I am also getting this issue, and I've tried about 10 different instances

[wedcf120]

same issue here
@pablouser1

[dlabaja]

I'm also experiencing this issue

[kpopdev]

i think its broken

[raashnachand]

Am also experiencing this issue, and can confirm RSS does not work.

[pogue]

Same. Can't find a single working instance, but it does appear trending is accessible.

[petrospap]

Nothing working probably, seems like tiktok has change their API

[telmer-adomix]

Have you always managed to see comments to single posts or is it a feature of the new API?

Another note: while Discover and Trending work regularly, searching posts by tags is broken as well.

[kitten7280]

Maybe tiktok changed it's site and now scraping doesn't work?

[wedcf120]

tiktok changed some code i think.
the user main page didn't work
but video link still working
example:
https://proxitok.pabloferreiro.es/@ {user name} /video/ {video id}
https://proxitok.pabloferreiro.es/@cassiilva/video/7281411677680045318

and user main page still need fixed anyway

[glottisfaun0000]

As a temporary workaround for RSS, RSSHub is currently working with Tiktok: https://docs.rsshub.app/routes/social-media#tiktok

[pablouser1]

Hi, looks like SIGI_STATE does not longer exist, which was the main source of information for ProxiTok. I'm looking for other methods to get data.

Discover and Trending are working properly because it does not scrape HTML, it gets all the data from the API.

Edit: Looks like there is now an equivalent called __UNIVERSAL_DATA_FOR_REHYDRATION__

[pablouser1]

Okay, there's a lot to unpack here so let's go:

• Now sometimes TikTok may give you a simple JS challenge before redirecting you to the actual page. The scraper solves it using this method:
  https://github.com/pablouser1/TikScraperPHP/blob/b97a478ceee1a03feb1d293fcbdc49308a504279/src/Helpers/Algorithm.php#L16
  • This has brought a new side-effect, if you try to get an HTML page with a user-agent bastly different from the user-agent you used for the challenge you will get a 403.
  • For example, changing from Chrome 104 to Chrome 105 is OK but changing from Windows to iOS is not.
  • So, if you change your user agent and get this error you should delete /tmp/tiktok.json
• The scraper usually used SIGI_STATE to get data from the HTML without having to use the UI itself, but now we have both SIGI_STATE and __UNIVERSAL_DATA_FOR_REHYDRATION__. Depending on the page, the data is either on SIGI, REHYDRATION or both (maybe this is temporary and SIGI_STATE will be deleted???)

The scraping works again for all endpoints but we still have to deal with #109 (probably a side-effect of #94???)

Also, the first page of user works but the next one we have a missing required fields error (maybe also a side-effect of #94???)

I think that's it, v2.4.9.3 is already up with those changes ready and the official instance is already updated.

[pablouser1]

@glottisfaun0000 Please confirm if this fixes the issue on your side

[glottisfaun0000]

Looks good, user pages loading (and have more entries than before?). Next page gives HTTP Code: 400\ API error code 10201 (missing required fields...) like you mentioned it would. RSS working great (my main purpose) and gives the full first page of results. Thanks for your work on this!

[petrospap]

RSS links leads to Error 403
i.e

https://proxitok.pabloferreiro.es/stream?url=https://p77-sign-va-lite.tiktokcdn.com/obj/tos-maliva-p-0068/ce479ff83e1749bc8da4b4052750f52d_1612642889?x-expires=1688641200&x-signature=H%2BInhWh8E54Kmpoveo6YsyUf8Z0%3D

[kitten7280]

Still doesn't work for me

[glottisfaun0000]

Looks like the same issue has come back today :-(

[Gexi0619]

Still doesn't work for me

[rafalohaki]

temp. Fixed my instance with pablouser1/TikScraperPHP#16 patch

[misterpyrrhuloxia]

temp. Fixed my instance with pablouser1/TikScraperPHP#16 patch

@rafalohaki, I'm not a developer, don't know programming, and don't really know scripting of any kind. So with my extremely limited knowledge, is it possible for you to briefly explain to me how to implement that patch on my own instance? Or would it just be too involved to explain it?

[rafalohaki]

@rafalohaki, I'm not a developer, don't know programming, and don't really know scripting of any kind. So with my extremely limited knowledge, is it possible for you to briefly explain to me how to implement that patch on my own instance? Or would it just be too involved to explain it?

replace those 2 composer files

then i did built image using docker, then replaced image within existing docker-compose.yml

composer.lock.txt
composer.json
changed composer.lock to composer.lock.txt, because of file extension name not supported on github upload

[nobhead112]

I have updated my docker on unraid using the docker compose plugin but still can't get the tag search or username to work. Trrend/ url are working but nothing else. I used the latest :master version for the repo. What am I doing wrong or is this still broken?

[seascape]

Just discovered this really cool-seeming project. Unfortunately I got this show-stopping error on all four instances I tried, so it seems broken right now.