uoft-core 1.0.0

an opinionated set of utilities designed to be easily included in any number of projects

About

This package contains a set of utilities useful for building python libraries, scripts, and command-line utilities

It's designed to be easy to include in other projects. all of its mainline dependencies are vendored and all modules which have external un-vendorable dependencies are available as optional extras

Install

pip install uoft_core

to include all optional dependencies, run

pip install uoft_core[all]

Usage

This toolkit makes it really easy to write small, simple, well designed CLI utilities In fact, the aim of this project is to make well-engineered CLIs almost as easy to write and deploy as basic python scripts it leverages a lot of really fantastic modern libraries and tools to do so, like poetry, typer, and loguru

let's assume you've created a new project with poetry new, and you want to add a CLI interface to it. here's one way to do that: create a common.py files with your super handy dandy useful function in it:

from loguru import logger

def my_super_awesome_function():
    logger.debug("I'm running my_package.common.my_super_awesome_function!")

create a __main__.py file in your package like so:

import os

from .common import my_super_awesome_function

from at.utils import configure_logging
import typer
from loguru import logger

app = typer.Typer()


@app.callback()
def callback(verbose: bool = False):
    """
    Here is my app's main help string
    Users will see this when they run `python -m my_package`
    """
    log_level = 'DEBUG' if verbose else 'INFO'
    configure_logging(
        'my_app_name', 
        stderr_level=log_level, 
        logfile_level='DEBUG', 
        sentry_level=None)


@app.command()
def my_command(option: bool):
    """
    Here's the help text my users will see when they run `python -m my_package my-command -h`
    """
    logger.debug("running my-command")  # this will only get printed if the --verbose flag is set
    my_super_awesome_function(option)


if __name__ == "__main__":
    if os.environ.get('PYDEBUG'):
        # we're in a debugger session
        # here we can put whatever debugging code we want
        # we can configure logging so all messages up to DEBUG are logged to stderr, and nothing gets logged to file:
        configure_logging('my_app_name', 'DEBUG', None, None)
        # do debugging stuff here
        logger.debug("I'm a debug message!")
        exit()
    try:
        app()  # cli code goes here
    except KeyboardInterrupt:
        print("Aborted!")
        exit()

the main api (all the stuff directly importable from at_utils) consists of:

• every function defined in the main module
• the configure_logging function from the log module

configure_logging has an option to enable logging to sentry. in order to use it, you need to install at_utils with the sentry extra (ie pip install at-utils[sentry] or poetry add -D at-utils[sentry])

apart from that, there are other modules which can be imported separately:

yaml has a whole bunch of useful and sometimes esoteric utilities for working with yaml files, built on top of ruamel.yaml

dev_utils has commmand-line utilities for working with python projects, specifically made for projects that use poetry