Streamlinkproxy Python3 Only discussion & support

Quote from rpl

Unfortunately it doesn’t read streamlink configuration files

It's not hard to add. In principle, I thought to make the main styreamlink settings (session options) in the form of a GUI accessible from the plugins menu ... But the parameters for specific add-on plugins can be passed element-wise to url.fragment and this does not require streamlink configuration files

Example for https://streamlink.github.io/plugins.html#bbciplayer add-on

http://localhost:port/https://www.bbc.co.uk/iplayer/episodes/p0frq6zd/champion?seriesId=p0frq798?&bbciplayer-username=xxxx&bbciplayer-password=xxxxx&bbciplayer-hd=1

parameters passed to url.query are simply parsed and can be passed in proxy "HandleRequest" to the streamlink add-on plugin used to get a link to the video ... It's not difficult at all...

Quote from rpl

Some minor things like loglevel cannot be changed

The server code is so simple that I can't even imagine what to log there =)

Quote from rpl

an alternative to streamlinksrv

The difference is in more concise code =) and more or less correct implementation of chunked HTTP streaming - https://www.rfc-editor.org/rfc…2#field.transfer-encoding compare with how it is completely wrong implemented in streamlinksrv

UPDATE

Added the ability to use parameters and their values for streamlink add-on plugins ... Download the plugin from the first post and reinstall using the --force-reinstall option

A small instruction with examples on how to pass parameters and their values has been added to the first post

Quote from pepsik

Quote from rpl

Unfortunately it doesn’t read streamlink configuration files

It's not hard to add. In principle, I thought to make the main styreamlink settings (session options) in the form of a GUI accessible from the plugins menu ... But the parameters for specific add-on plugins can be passed element-wise to url.fragment and this does not require streamlink configuration files

Example for https://streamlink.github.io/plugins.html#bbciplayer add-on

Code

http://localhost:port/https://www.bbc.co.uk/iplayer/episodes/p0frq6zd/champion?seriesId=p0frq798#bbciplayer-username=xxxx&bbciplayer-password=xxxxx&bbciplayer-hd=1

parameters passed to url.fragment are simply parsed and can be passed in proxy "HandleRequest" to the streamlink add-on plugin used to get a link to the video ... It's not difficult at all...

It is not about GUI and anything related to that. You missing the crucial info. Streamlink by design supports cli commands and configuration files and that is described in documentation. I will direct you to documentation as well https://streamlink.github.io/cli/config.html, but to the page where is described what I’m talking about. You don't need to pass parameters separately to each stream individual, that’s why streamlink supports configuration file for each plugin individual , or globally for all.

An example
In case of example e2 when you have root user boxes path looks like this
/home/root/.config/streamlink/config.bbciplayer
This path is home directory, and that why is included in settings backup in basically every image.

That just works with streamlinksrv since many years with accordance to documentation. I actually created my configuration files five years ago and they still works, because this format didn’t changed. Adding login, password and for example proxy to each link is unnecessary, when you can do that collectivity in a file.

Passing parameters also works with streamlinsrv and that can be used when specific streams need individual parameters becouse cli command can be handled by http HTTPServer. So those two methods were implemented. Don't see iprovment here as streamlinkproxy only one of those cases, but both are equqly important.

Practical Example.

You have 5 boxes, and want to use on them the same channel list prepared on PC editor. But due to device/streams limitations on service provider side you have to use separate account in each box. Then with passing parameters with url, you have to prepare 5 separate lists on PC editor, or create a script that will replace credentials each time the channel list is updated. With configuration files, no extra steps are necessary, you just sending list to those 5 boxes in the same time, and everything works because box have credentials stored seperatly.

Quote

I checked on OpenPli dev it works without problems .. it’s not at all clear why it wouldn’t work ...

I also don’t know, only if I had more detailed log, but

Quote

The server code is so simple that I can't even imagine what to log there =)

Maybe you see now the problem.

Quote

The difference is in more concise code =) and more or less correct implementation of chunked HTTP streaming -

Less correct, more correct, the jury is always on. Debate never ends, ultimately whats matter is if it works or if it doesn’t work. Anyway could you post any practical example when improvements can be observed ?

Quote from rpl

Maybe you see now the problem.

In general, there are no problems if you are at least at the minimum level able to read the python code from 30-40 lines .. what is there to debug? Judging by your comments about streamllink CLI - you are far from fluently Python ... I do not want to argue with you if you superficially own the subject of the dispute

As for everything else - I posted an update where you can fully (at 100%) use any of the plugins offered by the "library" stramlink ... When I wrote about the GUI menu, I talked about the possibility of changing the GLOBAL settings for the streamlink session, which have NOTHING to do with the specific local parameters of this or that of the stramlink project plugins add-on.

By the way ... I modified the code a little more, limiting the selection to stream types that do not require ffmpeg "processing" and now VOD broadcasts from youtube work fine and without any "gags" + threw out the extra request-files library from the streamlink/plugon/api/http_session.py code ... it's not at all clear why we need so much extra text from external library, if class FileAdapter can be easily implemented by overriding the BaseAdapter class of the requests library in the 40 lines

class FileAdapter(BaseAdapter):

    @staticmethod
    def _chkpath(method, path):
        """Return an HTTP status for the given filesystem path."""
        if method.lower() in ('put', 'delete'):
            return 501, "Not Implemented"                                                                                                                      
        elif method.lower() not in ('get', 'head'):                                                                                                            
            return 405, "Method Not Allowed"                                                                                                                   
        elif os.path.isdir(path):                                                                                                                              
            return 400, "Path Not A File"                                                                                                                      
        elif not os.path.isfile(path):                                                                                                                         
            return 404, "File Not Found"                                                                                                                       
        elif not os.access(path, os.R_OK):                                                                                                                     
            return 403, "Access Denied"                                                                                                                        
        else:                                                                                                                                                  
            return 200, "OK"

    def send(self, req, **kwargs):
        """Return the file specified by the given request
        """
        path = os.path.normcase(os.path.normpath(url2pathname(req.path_url)))
        response = requests.Response()
        response.encoding = 'utf-8'
        response.status_code, response.reason = self._chkpath(req.method, path)
        if response.status_code == 200 and req.method.lower() != 'head':
            try:
                response.raw = open(path, 'rb')
            except (OSError, IOError) as err:
                response.status_code = 500
                response.reason = str(err)
        if isinstance(req.url, bytes):
            response.url = req.url.decode('utf-8')
        else:
            response.url = req.url

        response.request = req
        response.connection = self
        return response

Compare (provided you are able to understand python code) with requests_file.py code - https://github.com/streamlink/…packages/requests_file.py

I will update streamlinkproxy tomorrow ... at the same time I will commit the solution to the steamlink branch on GitHub

Quote from jeepcook

How to know which plugins are installed. Alle the ones available here Plugins - Streamlink 5.5.1 documentation ? or do have I to add some in a specific directory ?

I don't understand your question. The installation package of the plugin includes the latest "library" streamlink from GitHub ... It includes all add-on plugins described on the official web page - https://streamlink.github.io/plugins.html

Quote from rpl

Less correct, more correct, the jury is always on. Debate never ends, ultimately whats matter is if it works or if it doesn’t work. Anyway could you post any practical example when improvements can be observed ?

The quality of the code is in its conciseness ... it is fashionable to write code that does the same thing in 140 and 40 lines ... From the compiler's point of view, 40 lines will always be executed faster than 140 ... an example a few posts above ... but you "measure" - "by eye" by clicking on the channels ... and do not use a disassembler to compare the speed of the code ... It's like putting your hand to your forehead instead of using a thermometer to take your temperature...

p.s. "Less correct, more correct,...." - RFC 9112 written and approved as mandatory in order to adhere to it just like that if only it were?

export PYTHON_VERSION=`python -c 'import sys; print(".".join(map(str, sys.version_info[0:2])))'`

export PYTHON_VERSION_FULL=`python -c 'import sys; print(".".join(map(str, sys.version_info[0:3])))'`

Quote from Captain

export PYTHON_VERSION=`python -c 'import sys; print(".".join(map(str, sys.version_info[0:2])))'`

export PYTHON_VERSION_FULL=`python -c 'import sys; print(".".join(map(str, sys.version_info[0:3])))'`

With all due respect ... I understand these two lines ))) but I don't understand what they have to do with streamlinkporxy or streamlink "library"?

Quote from jeepcook

if possible to exploit.

Nothing is impossible ... it takes time and knowledge to write a streamlink add-on plugin

that was only because of python recognition in ipk script,

Quote from Captain

that was only because of python recognition in ipk script,

anyway I call python binary in bash script .. what to parse python version from stdout, what to run python script .... in terms of hardware costs - comparable ... although 'import sys;' and executing a python script ... I think it "eats" more resources than awk

p.s. I'm just a "perfectionist" and this is my problem I used to be an assembler programmer for the Motorola 68000 platform a very long time ago And everything that is object-oriented is horror, in my "understanding"

p.s.s. Yes! I'm oldskool, but not useless! (C) Terminator: Genesys

Quote from jeepcook

example:

"http://localhost:8088/https://www.youtube.com/watch?v=gxG3pdKvlIs"

Снимок экрана 2023-07-03 в 07.19.29.pngСнимок экрана 2023-07-03 в 07.19.43.png

Снимок экрана 2023-07-03 в 07.19.58.png

What am I doing wrong and why does the same link work for me? Use exteplayer3 with streamproxy, I also wrote to you about this, but not on this forum thread ...

Quote from jeepcook

So for YT videos , had to install streamlinkwrapper and modify the url

streamwrapper is not a proxy, it is a code that returns a link to the player and the player is engaged in HLS parsing,

streamproxy - gives the player a ready "stream of bytes" and uses for it streamlink library which "rakes" HLS, reads chunks, has its own "ring buffer", etc.

Quote from pepsik

By the way ... I modified the code a little more, limiting the selection to stream types that do not require ffmpeg "processing" and now VOD broadcasts from youtube work fine and without any "gags"

I wrote this yesterday, did you read it? But this applies to VOD broadcasts youtube, live - as it worked, it still works

Quote from jeepcook

Test with another url:

"http://localhost:8088/https://www.france.tv/france-2/direct.html"

Снимок экрана 2023-07-03 в 07.45.39.pngСнимок экрана 2023-07-03 в 07.45.51.pngСнимок экрана 2023-07-03 в 07.46.03.png

???

Quote from jeepcook

root 3173 3164 5 01:16 ? 00:00:00 /usr/bin/ffmpeg -nostats -y -i /tmp/streamlinkpipe-3164-1-5710 -i /tmp/streamlinkpipe-3164-2-6910 -c:v copy -c:a copy -map 0:v? -map 0:a? -map 1:a -f mpegts pipe:1

In this case, with a probability of 90%, there was an hls-multi stream with several audio tracks ... There is nothing wrong with this. You simply have access to the choice of audio tracks by the "yellow" button on the RC ...