Whenever I try to view replies to any comment a 500 Uncaught exception would come up instead of the replies.

Steps to reproduce:

1. Go to any video (Here's an example)
2. Click on the replies of any comment

youtube-local version: 2.5.4
OS: macOS 11.2.1

Traceback

Traceback (most recent call last):
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 2447, in wsgi_app
    response = self.full_dispatch_request()
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 1952, in full_dispatch_request
    rv = self.handle_user_exception(e)
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 1821, in handle_user_exception
    reraise(exc_type, exc_value, tb)
  File "/usr/local/lib/python3.9/site-packages/flask/_compat.py", line 39, in reraise
    raise value
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 1950, in full_dispatch_request
    rv = self.dispatch_request()
  File "/usr/local/lib/python3.9/site-packages/flask/app.py", line 1936, in dispatch_request
    return self.view_functions[rule.endpoint](**req.view_args)
  File "/Users/akbarfadi01/program_files/youtube-local/youtube/comments.py", line 195, in get_comments_page
    comments_info = yt_data_extract.extract_comments_info(request_comments(ctoken, replies))
  File "/Users/akbarfadi01/program_files/youtube-local/youtube/comments.py", line 69, in request_comments
    content = util.fetch_url(
  File "/Users/akbarfadi01/program_files/youtube-local/youtube/util.py", line 308, in fetch_url
    raise FetchError(str(response.status), reason=response.reason,
youtube.util.FetchError: HTTP error during request: 500 Internal Server Error

youtube: https://www.youtube.com/results?search_query=my+search+terms
youtube-local: http://youtube.local/youtube.com/search?query=my+search+terms

Maybe it would be a good idea to also use results?search_query= instead of search?query= ?

Hi, would it be hard to re-create something like or identical to 2008~ youtube UI?

That would be a great addition to this awesome project.

Hi,
I am recently getting the following Error: Could not find player (see screenshot)

There are no error-messages in the console. The error is not persistent so after a couple of Tor-restarts (I am routing through tor) everything works fine again (until it doesn't)
If you need any more information (like the youtube-local-configuration), I can of course provide you with that.
If you know how to fix that would be awesome, I really enjoy using this project :)

Whenever I try to change pages in a channel, show me an error a 500 Uncaught exception

Steps to reproduce:

• Go to any channel (Here's an example)
• Click on the replies of any comment

youtube-local version: 2.5.4
OS: Debian Buster

Traceback

Traceback (most recent call last):
File "/home/liberate/venv/lib/python3.6/site-packages/flask/app.py", line 2447, in wsgi_app response = self.full_dispatch_request()
File "/home/liberate/venv/lib/python3.6/site-packages/flask/app.py", line 1952, in full_dispatch_request rv = self.handle_user_exception(e)
File "/home/liberate/venv/lib/python3.6/site-packages/flask/app.py", line 1821, in handle_user_exception reraise(exc_type, exc_value, tb)
File "/home/liberate/venv/lib/python3.6/site-packages/flask/_compat.py", line 39, in reraise raise value
File "/home/liberate/venv/lib/python3.6/site-packages/flask/app.py", line 1950, in full_dispatch_request rv = self.dispatch_request()
File "/home/liberate/venv/lib/python3.6/site-packages/flask/app.py", line 1936, in dispatch_request return self.view_functions[rule.endpoint](**req.view_args)
File "/home/liberate/youtube/channel.py", line 273, in get_channel_page return get_channel_page_general_url('https://www.youtube.com/channel/' + channel_id, tab, request, channel_id)
File "/home/liberate/youtube/channel.py", line 224, in get_channel_page_general_url util.check_gevent_exceptions(*tasks)
File "/home/liberate/youtube/util.py", line 542, in check_gevent_exceptions raise task.exception
File "src/gevent/greenlet.py", line 854, in gevent._gevent_cgreenlet.Greenlet.run
File "/home/liberate/youtube/channel.py", line 127, in get_channel_tab debug_name='channel_tab', report_text=message)
File "/home/liberate/youtube/util.py", line 318, in fetch_url ip=None) youtube.util.FetchError: HTTP error during request: 500 Internal Server Error 

I was trying to run youtube-local on Docker. While the container starts and there are some signs that the app is running internally, I couldn't access it through localhost:8080. I'm not quite sure if there's something I did wrong (I'm very new to docker) or it's just something from youtube-local itself.

Dockerfile:

FROM python:3.8-alpine
RUN apk add musl-dev build-base gcc libffi-dev python3-dev
COPY . .
RUN pip3 install -r requirements.txt
EXPOSE 8080
CMD ["python3", "server.py"]

Run commands : docker run -dp 8080:8080 youtube-local

I recently stumbled upon youtube-local and I think the project is neat and I would like to contribute.
Whoever I have absolutely no idea how the code works (the structure). It would be nice if I could get a little explanation on how the project's code is structure and where can I start reading the code to understand it.

Just as background info that may help I'm a computer science college student. I'm familiar with python. I've also been doing projects and coding assignments from school but it seems that understanding code written by someone is harder than expected especially when I wasn't there to see how the code was put together.

Here is my method for adding a video to a playlist, perhaps I am doing something wrong. Lets say I am currently watching a video (so I am at the /watch endpoint). I type the name of the playlist I want to add the current video to into the bar next to the "add to playlist" button. A message appears that says the video has been added to the playlist, but when I go to view the playlist it is empty. The file used for storing playlist date (ex music.txt) is created, but is empty.

Hi.

I downloaded the latest version (2.6.0) last night and extracted it to its own folder (youtube-local-2.6.0). After that, I created the settings.txt file to make it portable and ran pip3 install -r requirements.txt. After this, I ran python3 server.py. Everything worked. Perfect! I stopped the service and killed the port.

But after a good night of sleep, I ran python3 server.py again and got sqlite3.OperationalError: unable to open database file as an error. I ran pip3 install -r requirements.txt again and then python3 server.py. Worked perfectly.

So, now to my question. Why do I have to run pip3 install -r requirements.txt command more than once?

This issue were not present when I used an older version of youtube-local. I didn't use portable mode back then because it didn't work despite I had created the settings.txt file after I had used youtube-local a few times before (I suspect this has to be due to caching).

Sometimes googlevideo.com/videoplayback URLs fail with 503 error code.

• This only happens on a subset of videos, but for these it's reproducible for several days (or longer).
• I've encountered the issue at 720p, but haven't seen it happen at 360p.

I looked into it a bit, and found that different browsers report the failure differently. Waterfox doesn't give any errors in the network console, just says the video format isn't supported. Firefox displays a 503 on one of the requests. Ungoogled chromium puts (failed) in place of the status, with a tooltip of "net::ERR_CONTENT_LENGTH_MISMATCH".

For a sample video with the problem, when I wget the URL, it gives "Read error at byte 111496/138135408 (The TLS connection was non-properly terminated.)" and then retries. I doubt the server is actually returning a 503 code, I think the browser/tools are just putting that there as a placeholder. With wget, I noticed for that video that it always fails twice at the same byte positions: 111496 and 372692, which are very early into the file, even with different exit nodes or not using tor. Inspecting these positions in a hex editor, there are repeating units of 00 00 01 ** where ** is some arbitrary byte. Further into the video file is presumably the video data which is much more random looking. I'll have to investigate the mp4 format more deeply. Since this is a TLS error, it will also be worth taking a look at what happens in Wireshark.

As a temporary workaround, I'll try getting video and audio stream muxing with MediaSource working when I get some time, since the error doesn't seem to happen with the video-only urls.

Hello and first let me thank you for making youtube usable again (it became bloated and laggy after they retired the old interface). But, nowadays, certain content creators abuse in the video ads, so i find the sponsorblock extension a must to watch youtube videos.
So, is it possible to implement the sponsorblock ad skipping in the embedded video player?
Kodi has an addon for this, but i don't know how easy it would be to implement in your app:
https://github.com/ajayyy/SponsorBlock/wiki/3rd-Party-Ports#kodi

My current workaround is to use invidious with the sponsorblock extension (which supports invidious instances), but I much prefer your app, because it's much more reliable.
If it's too hard or too much work, thank you very much nonetheless, because it's an excellent tool that made youtube bearable for me (and I'm sure for a lot of other people).

The arrow hotkeys don't work correctly, because the native

EDIT 1: As it turns out, there is not a new identity being acquired. I was seeing this on random auto-checks for subscriptions, whereas now I'm actually using the search feature, and catching this error, and the exit node is not changed.

EDIT 2: So I had my CookieAuthentication set to 1 instead of 0 in my torrc file. I believe this was the cause of my issue. I'm going to leave this open for a bit just to be sure that this was actually the root of the problem.

According to the log, it appears that there is a new identity acquired, but~~ there is a permissions error happening, and I don't know if it matters....

Error: Youtube blocked the request because the Tor exit node is overutilized. Exit node IP address: 89.163.154.91
new_identity: new_identity called
new_identity: New identity lock acquired
Traceback (most recent call last):
  File "/home/ryan/.local/lib/python3.8/site-packages/stem/connection.py", line 1103, in _read_cookie
    with open(cookie_path, 'rb', 0) as f:
PermissionError: [Errno 13] Permission denied: '/run/tor/control.authcookie'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "src/gevent/greenlet.py", line 906, in gevent._gevent_cgreenlet.Greenlet.run
  File "/home/ryan/Downloads/youtube-local-2.5.8/youtube/channel.py", line 229, in get_channel_first_page
    return util.fetch_url(base_url + '/videos?pbj=1&view=0', headers_desktop,
  File "/home/ryan/Downloads/youtube-local-2.5.8/youtube/util.py", line 309, in fetch_url
    error = tor_manager.new_identity(start_time)
  File "/home/ryan/Downloads/youtube-local-2.5.8/youtube/util.py", line 130, in new_identity
    controller.authenticate('')
  File "/home/ryan/.local/lib/python3.8/site-packages/stem/control.py", line 1112, in authenticate
    stem.connection.authenticate(self, *args, **kwargs)
  File "/home/ryan/.local/lib/python3.8/site-packages/stem/connection.py", line 629, in authenticate
    raise auth_exc
  File "/home/ryan/.local/lib/python3.8/site-packages/stem/connection.py", line 591, in authenticate
    authenticate_safecookie(controller, cookie_path, False)
  File "/home/ryan/.local/lib/python3.8/site-packages/stem/connection.py", line 904, in authenticate_safecookie
    cookie_data = _read_cookie(cookie_path, True)
  File "/home/ryan/.local/lib/python3.8/site-packages/stem/connection.py", line 1107, in _read_cookie
    raise UnreadableCookieFile(exc_msg, cookie_path, is_safecookie)
stem.connection.UnreadableCookieFile: Authentication failed: unable to read '/run/tor/control.authcookie' ([Errno 13] Permission denied: '/run/tor/control.authcookie')
2021-07-22T03:09:42Z <Greenlet at 0x7fabcbe3bbf0: get_channel_first_page(channel_id='UCRE3NFNtdjR96-H4QG4U1Fg')> failed with UnreadableCookieFile

Again, I don't know if this matters if the new identity is created/acquired, but I figured I'd bring it up.

The new Youtube layout is slow as snails on older computers. This looks like a really nice alternative, so wondering if there is any way to make this run on 32 bit Windows? Or maybe I could compile it myself?

Hi,

I recently discovered this app and so far it's been awesome. I recently tried to access it on my phone and noticed that the front-end was not mobile friendly at all.

I think it would great to add a bit of CSS/HTML code for the app to be responsive.

Thank you for the app,

Since some days ago, I'm having an annoying problem: a few videos that are shown before newer ones in my subscriptions list
I think a screenshot would be better than a long and boring explanation of the problem

I'm curious as to how you reverse engineered the youtube API, since I'd like to build a similar version of this project for other sites.

I just noticed some shortcuts don't work anymore after 1e65153.

So I looked at it and saw hotkeys.js is disabled in favor of plyr builtin(?) shortcuts.
Why is that? Why not always rely on hotkeys.js, and disable plyr's shortcuts?

In case that's not possible one could expose settings.video_player as js variable, which is read in hotkeys.js and disables some (but not all!) hotkeys.

As for now you can only choose and watch other video qualities if you click on download and the link.
It would be much more convenient if you can directly choose the quality on the video page itself, or atleast increase the default quality in the settings

If this feature is added this project would completely replace invidious for me.

It would be nice to have a way to download a complete playlist like we can do with one video, you choose the quality, audio only, video only or both, and then it downloads everything. I have my music playlist in youtube-local but it's long as hell to download one music by one. There are some online downloaders for youtube playlist, but there is no way for youtube-local playlists, which are not on youtube.

YouTube rolled out very aggressive IP-blocking this month and most exit nodes are blocked most of the time, making Tor-routing nearly unusable. However, the new type of blocking (which doesn't include the IP address in the page) only applies to the watch page. I can't devote much time to the issue at the moment but here's some ideas, if anyone wants to help with them,

• Reverse-engineer the API that the YouTube app uses and pretend to be the YouTube app when requesting the video data (since this blocking apparently doesn't apply to the app).
• Use some sort of automated captcha solver such as buster. I imagine this would require a headless browser to work which might defeat the purpose
• Some kind of paid captcha service like invidious and similar use. I'm not sure how this would work with having independent clients rather than a central service or how the funds would be managed.

Oh and thanks for implementing watching videos through Tor, this has replaced Invidious for me now.

I think that it's redundant, as you can already use middleclick on the "n replies" summary-element.
95f2f02

Hey! I've seen your project and it is pretty nice! I would like to talk to you about joining skills and forces to develop a solution together. Would you like to contribute to Yotter? We could use our codes and join parts together so we can offer a good and robust solution. What do you say?

Hello could you integrate this commit where the theme of all youtube-local is changed

Except the README.md

Please, this integration would be great

Subscriptions are unique per user. They should be done in a separate circuit so they cannot be associated with what videos are being watched. Need to figure out how to do this, preferably without bundling Tor ourselves.

Research:

• Are any special APIs provided by the Tor Socks5 proxy? Does "Tor RC" have anything to do with it or is that strictly the configuration file?
• If not, what libraries/APIs are available and what is the recommended way to integrate with them?

I know it can easily be done with JavaScript in the console but it's a hassle, so I think it'd be nice to have a playback speed adjuster for convenience

Searching videos in the channel page doesn't work but only if the channel url is in the https://www.youtube.com/channel/UCxxxxx format. If the URL is something like https://www.youtube.com/c/SomeChannelName then the search works as expected.

Example:
http://localhost:8080/https://www.youtube.com/channel/UC5I2hjZYiW9gZPVkvzM8_Cw/search?query=minidisc - doesn't work
http://localhost:8080/https://www.youtube.com/c/Techmoan/search?query=minidisc - works

Traceback:

Traceback (most recent call last):
  File "C:\Users\akbar\bin\youtube-local\python\lib\site-packages\flask\app.py", line 2447, in wsgi_app
    response = self.full_dispatch_request()
  File "C:\Users\akbar\bin\youtube-local\python\lib\site-packages\flask\app.py", line 1952, in full_dispatch_request
    rv = self.handle_user_exception(e)
  File "C:\Users\akbar\bin\youtube-local\python\lib\site-packages\flask\app.py", line 1821, in handle_user_exception
    reraise(exc_type, exc_value, tb)
  File "C:\Users\akbar\bin\youtube-local\python\lib\site-packages\flask\_compat.py", line 39, in reraise
    raise value
  File "C:\Users\akbar\bin\youtube-local\python\lib\site-packages\flask\app.py", line 1950, in full_dispatch_request
    rv = self.dispatch_request()
  File "C:\Users\akbar\bin\youtube-local\python\lib\site-packages\flask\app.py", line 1936, in dispatch_request
    return self.view_functions[rule.endpoint](**req.view_args)
  File "C:\Users\akbar\bin\youtube-local\youtube\channel.py", line 289, in get_channel_page
    return get_channel_page_general_url('https://www.youtube.com/channel/' + channel_id, tab, request, channel_id)
  File "C:\Users\akbar\bin\youtube-local\youtube\channel.py", line 256, in get_channel_page_general_url
    polymer_json = get_channel_search_json(channel_id, query, page_number)
  File "C:\Users\akbar\bin\youtube-local\youtube\channel.py", line 195, in get_channel_search_json
    polymer_json = util.fetch_url("https://www.youtube.com/browse_ajax?ctoken=" + ctoken, headers_desktop, debug_name='channel_search')
  File "C:\Users\akbar\bin\youtube-local\youtube\util.py", line 319, in fetch_url
    ip=None)
youtube.util.FetchError: HTTP error during request: 404 Not Found

127.0.0.1 - - [2018-07-31 12:36:35] "GET /https://www.youtube.com/watch?v=vBgulDeV2RU HTTP/1.1" 500 134 1.615836

Hello! I was testing this app on Google's Cloud Shell which runs web apps by proxying their ports to port 80 into a URL:

https://8080-8be694f9-38c9-4de2-8d54-cf17a257f2cd.europe-west4.cloudshell.dev/https://youtube.com

I noticed that double slashes get replaced with a single slash and I am not sure where the fault is -- Google or youtube-local

Suggestion: Can we skip the protocol part altogether in future versions as I don't think it's essential to the functioning of the app?

URLs without the protocol somewhat work, although no resources (images, videos) load.

Bunch of 404s in the console:

In the description of a video open in youtube-local, any hyperlinks are displayed as is, even links to other YouTube videos. This means that clicking on one of these links will drag you into the actual YouTube frontend.
It would be nice to have the href changed so you stay on youtube-local.

That serif font looks hideous for this purpose. What's an easy way to change it?

Alternatively, just officially change the program to use a sans-serif font, it will look much better,

I may have not explained the #49 issue with the right words, but I have the same videos stuck above all others for weeks now. Every new video gets below 6 videos that I have already seen in my subscription list

Broken. Putting this here to remind myself.

Hello dear developer, how can I configure the application with gunicorn or uwsgi?

This would be cool so you could just replace all requests to youtube to youtube-local with an extension like Redirector and still have embedded videos work.

It would be nice if there's a mode where it'll stream only the audio instead of with the video. It saves huge amounts of bandwidth when you're just trying to listen to music. Currently the workaround is to open one of the download links in a new tab and play it from the tab.

One way to implement this is something like in Invidious where there's a button to switch between video and audio mode and an option to force audio only by default.

Example video: https://invidious.snopyta.org/watch?v=gDHJ6nBdLD8&listen=1

Audio mode OFF (the headphone button is used to switch between audio and video mode)

Audio mode ON (the player is just showing the thumbnail)

I'll have some time mid-december to work more on youtube-local. Comment here with what bugfixes or features you think I should prioritize. If another comment suggests what you want, just add a thumbs up to it. I'm most interested in addressing things that are daily annoyances or that create friction, such as blocked exit nodes

I really like this project.
But I think a little javascript could improve "QoL"..
I think of the following:

• ability to jump to timestamps (e.g. 14:31)
• lazy loading of comments (-> faster /watch page load)
• load more comments while staying on the /watch site.
• hotkeys (e.g. 'k' for pause etc.)

In case you want to preserve "Works without Javascript" as a feature, all this could be made optionally.
What do you think?

The autoplay function currently seems to be available only for playlists

Hello there! I just found your project on Github and I have some questions, could you please answer it? :)
I just trying to make a software which will use browser cookies and leave comments under youtube videos and I cant figure out how to do it using http requests. So I founded your project and I cant authorize youtube accounts.
Please if you have free time, could you add me in skype and help me with making post requests for adding comments on youtube videos?
I want to comment videos using Node JS or Http requests using browser cookies without official youtube api

I want to import all my subscription list from YouTube or Newpipe but there is no easy way to do that I have to manually subscribe to the individual channel. Please Help

Cloned repo today, got this issue:
127.0.0.1 - - [2018-07-31 16:10:08] "GET /https://www.youtube.com/ HTTP/1.1" 404 114 0.000179

I tried the regular way from the address bar, I also tried via Redirector. Same issue.

yt.com seems to allow up to 5000 characters in the video description.
I see the following variations:

• a "show more" / "show less" feature
  • implemented fully in .js
  • OR: define the "teaser" section via jinja.
• OR: just use another <details> element

Maybe after the description is less long, the Downloads <details> Element can be moved below the description.

Having problems playing THIS video on archlinux

using Error: Error decrypting url signatures: Could not find player name

https://www.youtube.com/watch?v=gRhAPXJmm2c

if it matters, im using port 8081

hgeres error log from when it happens

127.0.0.1 - - [2021-02-02 08:18:48] "GET /https://r5---sn-hp57yne7.googlevideo.com/videoplayback?expire=1612293516&ei=LFEZYL7iIouUhwbJg7iAAw&ip=[XXX.XXX.XXX.XXX]&id=o-AMF7ULSizyPhQgnSrTKsYMDumgH_0nEQEUN-GYXRe4KB&itag=22&source=youtube&requiressl=yes&mh=iT&mm=31%2C29&mn=sn-hp57yne7%2Csn-hp57knz6&ms=au%2Crdu&mv=m&mvi=5&pl=16&gcr=us&initcwndbps=2151250&vprv=1&mime=video%2Fmp4&ns=FLcVac1YP6PtNfSRyfpg7o4F&cnr=14&ratebypass=yes&dur=245.968&lmt=1472212170500621&mt=1612271575&fvip=6&c=MWEB&n=qBikUebZeTGjxndqc&sparams=expire%2Cei%2Cip%2Cid%2Citag%2Csource%2Crequiressl%2Cgcr%2Cvprv%2Cmime%2Cns%2Ccnr%2Cratebypass%2Cdur%2Clmt&sig=AOq0QJ8wRQIhALR1nfFLR0u-PJaJXkzFeVQbl3T5RuCpRGipTF1wffYBAiBdDKTXMVELLa5X03LqnrIdQMtNLKkIukEouAXYxiaM9A%3D%3D&lsparams=mh%2Cmm%2Cmn%2Cms%2Cmv%2Cmvi%2Cpl%2Cinitcwndbps&lsig=AG3C_xAwRQIgLCJ0Qt6EyPrrcoRHiqtjouOYArO4vMND3OwBxbA3lrsCIQCmwKIAgLlVayqb3ghWVm-hYF1q4DX8B-vm5iBXBMtV5A%3D%3D HTTP/1.1" 206 47531332 10.450338
Retrieved comments Latency: 0.757 Read time: 0.0
127.0.0.1 - - [2021-02-02 08:35:11] "GET /https://www.youtube.com/watch?v=-rUaC5G9uY8 HTTP/1.1" 200 86607 0.840325
Retrieved comments Latency: 0.395 Read time: 0.004
127.0.0.1 - - [2021-02-02 08:35:25] "GET /https://www.youtube.com/watch?v=-rUaC5G9uY8 HTTP/1.1" 200 86782 0.688076
127.0.0.1 - - [2021-02-02 08:35:25] "GET /youtube.com/static/favicon.ico HTTP/1.1" 200 5979 0.000980

Is it possible to make youtube-local accessible to the local network?

It would be great if instead of just being a "videos folder" we could play the entire playlist, shuffle it and loop it (to make the playlist restart when it reaches its end).