Concurrent programming languages

Taking control of Ruff checks on Python scripts

22^nd October 2025

Positron is becoming my tool of choice for developing Python code. Along from using a Python console like a REPL environment, it also includes Ruff for checking code compliance. One of its rules is that Python modules must be declared at the top. However, I want to use some code that checks for the present of any modules used in a script, installing those that are missing. This means that import statements appear later in a script that Ruff recommends, making me wish for a way to turn off that check since things run well anyway. The chosen solution is to create a file called pyproject.toml in the directory where my scripts are store and add the following lines in there to accomplish what I want:

[tool.ruff] ignore = ["E402"]

Here, it helps if you open a folder in Positron, achieving the same outcome as you would in the VSCode on which the IDE is based. While I have only listed one check here, you also can have a comma-delimited list of quoted strings if you need to switch off more than one rule at once.

Avoiding Python missing package errors with automatic installation checks

20^th October 2025

Though some may not like having something preceding package import statements in Python scripts, I prefer the added robustness of an extra piece of code checking for package presence and installing anything that is missing in place getting an error. In what follows, I define the list of packages that need to be present for everything to work:

required_packages = ["pandas", "tqdm", "progressbar2", "sqlalchemy", "pymysql"]

Then, I declare the inbuilt modules in advance of looping through the list that was already defined (adding special handling for a case where there has been a name change):

import subprocess
import sys
for package in required_packages:
    try:
        __import__(package if package != "progressbar2" else "progressbar")
        print(f"{package} is already installed.")
    except ImportError:
        print(f"{package} not found. Installing...")
        subprocess.check_call([sys.executable, "-m", "pip", "install", package])

The above code tries importing the package and catches the error to do the required installation. While a stable environment may be a better way around all of this, I find that this way of working adds valuable robustness to a script and automates what you would need to do anyway. Though the use of requirements files and even the Poetry tool for dependency management may be next steps, this approach suffices for my simpler needs, at least when it comes to personal projects.