When parser.add_subparsers() is used, Cli2Gui fails with the message

[evan@blackbox picklecast] python /tmp/test.py
Traceback (most recent call last):
  File "/tmp/test.py", line 30, in <module>
    gui()
  File "/home/evan/resources/venv3/lib/python3.10/site-packages/cli2gui/decorators.py", line 286, in inner
    return callingFunction(*args, **kwargs)
  File "/tmp/test.py", line 14, in main
    args = parser.parse_args()
  File "/home/evan/resources/venv3/lib/python3.10/site-packages/cli2gui/decorators.py", line 249, in runCli2Gui
    buildSpec = createFromParser(
  File "/home/evan/resources/venv3/lib/python3.10/site-packages/cli2gui/decorators.py", line 85, in createFromParser
    return FullBuildSpec(**convertMap["self"][parser](selfParser), **buildSpec)
  File "/home/evan/resources/venv3/lib/python3.10/site-packages/cli2gui/tojson/argparse2json.py", line 186, in convert
    for _, subparser in iterParsers(parser):
  File "/home/evan/resources/venv3/lib/python3.10/site-packages/cli2gui/tojson/argparse2json.py", line 42, in iterParsers
    ].choices.arg_items()
AttributeError: 'dict' object has no attribute 'arg_items'

#!/usr/bin/env python3
from cli2gui import Cli2Gui
import argparse

def run(args):
    print(args.arg)

def main():
    parser = argparse.ArgumentParser(description="this is an example parser")
    parser.add_argument("arg", type=str, help="positional arg")

    parser.add_subparsers()

    args = parser.parse_args()
    run(args)

decorator_function = Cli2Gui(
    run_function=run,
    auto_enable=True,
)

gui = decorator_function(main)

if __name__ == "__main__":
    gui()

Ideally Cli2Gui would support these subparsers (perhaps as additional tabs in the GUI), but at least they could be ignored and not cause an exception.

This issue was generated by todo based on a TODO comment in be67588. It's been assigned to @FredHappyface because they committed the code.

Question

• [x ] I have read the comment above and have completed each step

Bug

• [x ] I have read the comment above and have completed each step
• [x ] I have filled out the system info
• [x ] I have described the bug, filled in the expected outcome and the actual
  outcome including screenshots where appropriate

Describe the bug

This is what's shown on the homepage / readme

Expected outcome

Icon should be shown

Traceback:
Traceback (most recent call last):
File "C:\Users\delaplai\git\py\iosp\commands\gui.py", line 26, in
Click2Gui(run_function=hello)
File "C:\Users\delaplai\AppData\Local\Programs\Python\Python39\lib\site-packages\cli2gui\decorators.py", line 141, in Click2Gui
bSpec = BuildSpec(

"""Tests a simple parser
Program from https://click.palletsprojects.com/en/7.x/#documentation
"""

from future import annotations

import sys
from pathlib import Path

import click

THISDIR = str(Path(file).resolve().parent)
sys.path.insert(0, str(Path(THISDIR).parent.parent))
from cli2gui import Click2Gui

@click.command()
@click.option("--count", default=1, help="Number of greetings.")
@click.option("--name", prompt="Your name", help="The person to greet.")
def hello(count, name):
"""Simple program that greets NAME for a total of COUNT times."""
for _index in range(count):
click.echo("Hello %s!" % name)

Click2Gui(run_function=hello)

⬇️ This is how you call the function without a GUI

hello()

Feature

add a virtual terminal widget, so we can show the live stdout and stderr of the run function

Is your feature request related to a problem? Please describe

problem: when running the app without a terminal, then stdout/stderr are not visible

Describe the solution you'd like

challenges:

• capture stdout and stderr → function call to run_function or subprocess
• correctly interleave stdout and stderr
• render terminal control commands → virtual terminal (tty) / pseudo terminal (pty)
  • https://github.com/Saadmairaj/tkterminal
  • https://stackoverflow.com/questions/10239965/giving-a-command-in-a-embedded-terminal
  • https://stackoverflow.com/questions/7253448/how-to-embed-a-terminal-in-a-tkinter-application
  • https://stackoverflow.com/questions/62335989/redirect-terminal-output-to-tkinter

example: capture stdout with io.TextIOBase
https://stackoverflow.com/questions/51527750/capture-print-output-from-multiple-threads-in-python

import threading
import time
import sys
import io

class demo_copy_stdout():

  def __init__(self):

    def run_function():
      for i in range(0, 100):
        print(*sys.argv, i)
        time.sleep(1)

    self.run_function = run_function

    def main_function_wrapper(*argv):
        sys.argv = argv # set arguments
        # note: run_function takes no arguments
        return self.run_function()

    print("main_thread create")
    self.main_thread = threading.Thread(target=main_function_wrapper, args=("toy.py", "hello",))

    class CopyStdout(io.TextIOBase):
        def __init__(self, target):
            self.target = target
            #self.stringio = io.StringIO()
        def write(self, s):
            writecount = self.target.write(s)
            written = s[:writecount]
            #self.stringio.write(written)
            # TODO send to GUI
            with open("stdout.txt", "a") as f:
                f.write(written)
            return writecount
    print("copying stdout to tee.txt")
    original_stdout = sys.stdout
    sys.stdout = CopyStdout(sys.stdout)

    print("main_thread start")
    self.main_thread.start() # run

    if True:
        print("main_thread wait ...")
        self.main_thread.join() # wait
        print("main_thread done")
        sys.stdout = original_stdout # stop copying stdout

x = demo_copy_stdout()

Describe alternatives you've considered

Additional context

on nix, pypandoc pulls in these dependencies:

these 4 paths will be fetched (20.74 MiB download, 164.43 MiB unpacked):
  /nix/store/5gm4rzsdd40i9ph97bxs5bwrfanmg11v-lua-5.3.6
  /nix/store/7iyh4zqkmk8i6bpkdg99zrqhbckmy685-python3.9-pypandoc-1.7.2
  /nix/store/sf07y0qw3bn11l7zj71kpadf65mffipb-pandoc-2.14.0.3-data
  /nix/store/yvbr6knw00jl6wd7afcc8f9knl60h3hb-pandoc-2.14.0.3

feels a bit heavy. do we really need pandoc?

Hello Guys. I am new to Cli2gui and argparser and I am trying to create a GUI for my work. So for fileType (type), how can I set up a predefined dir to open as default rather than select every time?

Currently it looks like Cli2Gui doesn't make use of the default argument for argparse. It would be very nice if these showed up generated GUI.

# main.py

from cli2gui import Cli2Gui
import argparse

def run(args):
    print(args.arg)

def main():
    parser = argparse.ArgumentParser(description="this is an example parser")
    parser.add_argument("--arg", type=str, default="foo", help="keyword arg")
    parser.add_argument("--bool", action='store_true', default=True, help="boolean arg")
    args = parser.parse_args()
    run(args)

decorator_function = Cli2Gui(
    run_function=run,
    auto_enable=True,
)

gui = decorator_function(main)

if __name__ == "__main__":
    gui()

why do we call open here?

the run function just needs the file path
path should be validated with os.path.exists(path)

in my case values[key] == "" so open throws FileNotFoundError(2, 'No such file or directory')

my parser looks like

    parser = argparse.ArgumentParser()

    extra_args.add_argument(
        "--somefile",
        type=argparse.FileType("r"),
        required=False,
    )

I'd really like to use cli2gui in a Jupyterhub project but I'm not sure if you're supporting that environment. I see that "web" is one of the GUI toolkits that's supported but perhaps that doesn't mean using Jupyterhub?

I've got a freshly built TLJH docker image from today's main head of https://github.com/jupyterhub/the-littlest-jupyterhub and have installed cli2gui by running "./pip install cli2gui" from within the /opt/tljh/user/bin directory of the running tljh docker instance. I've copied the contents of https://github.com/FHPythonUtils/Cli2Gui/blob/master/tests/argparse/test_10.py into a jupyterlab notebook cell (and commented out the top few file pathing stuff since cli2gui is already installed) . Upon running the cell, I get a tkinter error which suggests it's not detecting the Jupyter environment. The error is "TclError: no display name and no $DISPLAY environment variable." which makes a lot of sense since I'm not running this as a desktop app.

I'm hoping to get clarification that I can use your project within a Jupyter environment and if so, some pointers on how to get a working example.

I've found that this approach works well for detecting if a script is running under jupyter or not : https://stackoverflow.com/questions/15411967/how-can-i-check-if-code-is-executed-in-the-ipython-notebook

Feature

• I have read the comment above and have completed each step
• I have filled in each heading below

Is your feature request related to a problem? Please describe

Describe the solution you'd like

A possible options it's to add "input ()" build-in function like readable option.

Describe alternatives you've considered

Instead of;
@cli2gui(argparse="argparse"...)
Can be possible put ...?
@cli2gui(argparse="inpu" ...)
Def func():
Var=imput ("...")

Many thanks in advance.

Interesting project, does this support Typer or plans to be supported as one of those CLI packages?