Prevent CLI from loading on import of robyn module

[LilyRose2798]

It currently looks to be impossible to implement your own argument parser when using Robyn, as just importing the robyn module creates its own ArgumentParser. I do not think that simply importing a module should have this behavior, and that you should have to pass a flag to the Robyn constructor in order for it to create the ArgumentParser instance. You should also be able to create your own Config object containing all the values like processes, workers, etc when initializing Robyn, so that if you have your own args or other way of getting these values, then you can pass them through. I'm currently having to do this in the following way:

app = Robyn(__file__, config=argparse.Namespace(**{
    "processes": 1,
    "workers": 1,
    "dev": False,
    "create": False,
    "docs": False,
    "open_browser": False,
    "version": False,
    "compile_rust_path": None,
    "create_rust_file": None,
    "log_level": "INFO",
}))

So a better way of doing this would be desirable.

[sansyrox]

Hey @LilyRose2798 👋

Thanks for creating this issues and bringing this to my attention.

It currently looks to be impossible to implement your own argument parser when using Robyn, as just importing the robyn module creates its own ArgumentParser

Let me dive a bit deeper and come back with an approach to tackle this.

[Mr-Sunglasses]

Hey @LilyRose2798 👋

Thanks for creating this issues and bringing this to my attention.

It currently looks to be impossible to implement your own argument parser when using Robyn, as just importing the robyn module creates its own ArgumentParser

Let me dive a bit deeper and come back with an approach to tackle this.

IMO, The one way we can tackle this is to rewrite Robyn whole CLI api, where we expose the functions like workers, debug etc. to the Robyn() class and the seprate logic for CLI. Something similar to uvicorn or flask

[LilyRose2798]

A way around the issue I found that isn't ideal is to move the Robyn import to after I've declared my own ArgumentParser:

from argparse import ArgumentParser, Namespace

parser = ArgumentParser("myapi")
parser.add_argument("-a", "--address", default = "127.0.0.1", help = "the address to use for the api (default: %(default)s)")
parser.add_argument("-p", "--port", default = 8000, type = int, help = "the port to use for the api (default: %(default)s)")
parser.add_argument("-l", "--limit", default = 3600, type = int, help = "the number of requests per hour to limit usage of the api to (default: %(default)s)")
parser.add_argument("--processes", default = 1, type = int, help = "the number of processes to use for the api (default: %(default)s)")
parser.add_argument("--workers", default = 1, type = int, help = "the number of workers to use for the api (default: %(default)s)")
parser.add_argument("--log-level", default = "INFO", help = "the log level use for the api (default: %(default)s)")
args = parser.parse_args()

from robyn import Robyn

app = Robyn(__file__, Namespace(**{
    "processes": args.processes,
    "workers": args.workers,
    "dev": False,
    "create": False,
    "docs": False,
    "open_browser": False,
    "version": False,
    "compile_rust_path": None,
    "create_rust_file": None,
    "log_level": args.log_level,
}))

# API routes declared here...

app.start(host = args.address, port = args.port)

[sansyrox]

@LilyRose2798 , this seems a bit too complicated. I am thinking of just adding an override args flag and then add the ability the ability to pass through the app.start