streamlit-process-manager 0.0.1a0

Manage the heck out of long-running processes in streamlit

Streamlit Process Manager

Let's be honest, trying to manage external processes or subprocesses from a Streamlit app is tricky. The execution model makes it difficult to keep track of which processes are supposed to be running and which ones aren't. Half the time you end up spawning twenty different processes all trying to access the same file and stomping all over one another.

Streamlit Process Manager attempts to make a that a bit easier. It looks something like this:

import streamlit_process_manager as spm

# Get the global ProcessManager
pm = spm.get_manager()

# Configure the Process and start it
my_program = ["python", "my_program.py"]
output_file = "program_output.txt"
process = pm.single(spm.Process(my_program, output_file))
process.start_safe()

# Run the process and monitor it in streamlit!
spm.st_process_monitor(process).loop_until_finished()

⚠️ DISCLAIMER ⚠️

Runing processes on your machine from a public-facing application like a Streamlit app can be incredibly risky and comes with all sorts of security considerations. Use this project on public apps at your own risk, and never inject user-provided values or arguments into your processes without serious consideration of all the possible injection attacks you might be opening your server up to.

Installation

pip install streamlit-process-monitor

Requires

• Python 3.8.* or above
• Streamlit 1.30.* or above
• psutil 5.9.8 or above

Development

This project uses pdm for most of its development tooling. Ensure you have python3.8 installed, then run

pip install pdm
pdm venv create 3.8
pdm install --dev --group test --group lint

to get started, then run

pdm check

to verify that everything is set up and ready to go.

If you must use Python 3.12 or greater, you may have issues with numpy 1.24 needing to be built from source since it only supports python versions up to 3.11.

Testing with tox

If you have multiple versions of Python available on your system you can use tox to run tests on all of them at once. tox should be installed when you install the --dev dependencies so you can just run it. Any Python's not available on your system will be skipped.

pdm run tox