About

This is a basic attempt to understand modern Python package and environment management using Poetry.

Tools used

• pyenv
• poetry
• pytest
• python-dotenv

Python setup (MAC)

Setup Pyenv

brew install openssl readline sqlite3 xz zlib

brew update

brew install pyenv

For ZSH

echo 'eval "$(pyenv init --path)"' >> ~/.zshrc

INSTALLING PYTHON

# this will give the already installed versions of Python on your system
pyenv versions

# to install a specific version
pyenv install 3.11.1

# to **set a specific version** as global from the available list
pyenv global 3.11.1

check in a new terminal

➜  ~ python -V
Python 3.11.1
➜  ~ pyenv versions
  system
* 3.11.1 (set by /Users/subhranshu/.pyenv/version)

INSTALLING POETRY

Reference Video - https://www.youtube.com/watch?v=0f3moPe_bhk

Since we have pyenv , pip is automatically understood to use which version

pip poetry

add this to the .zshrc file

export PATH=$PATH:$HOME/.poetry/bin

then hit

source ~/.zshrc
# best practice is to start working on a new terminal after this step

SETTING UP A NEW PYTHON PROJECT WITH POETRY

if you are using PYENV, do this before anything

poetry config virtualenvs.prefer-active-python true

setup the virtual environment inside the current project directory

poetry config virtualenvs.in-project true

to check the POETRY config in the terminal

poetry config --list

now create a project

poetry init

# go through the interactive CLI to generate the "pyproject.toml" file

poetry install

💡 We will have the following - “.venv” poetry.lock

Running the VIRTUAL ENV in the CLI

poetry shell

We get something like below. If you notice, the VENV is pointing to the current project directory’s .venv directory

➜  mypylib poetry shell
Spawning shell within /Users/subhranshu/Desktop/projects/mypylib/.venv
➜  mypylib emulate bash -c '. /Users/subhranshu/Desktop/projects/mypylib/.venv/bin
/activate'
(mypylib-py3.11) ➜  mypylib

💡 Now you can **run the Python program in this shell** so that your python code runs in this Poetry managed virtual environment.

Sample Code

import requests

print("hello poetry!!")
response = requests.get('https://jsonplaceholder.typicode.com/todos/1')
response_json = response.json()
print(response_json["title"])

IN THE CLI

(mypylib-py3.11) ➜  mypylib python app.py
hello poetry!!
delectus aut autem
(mypylib-py3.11) ➜  mypylib

ANOTHER WAY

poetry new --name pysandbox --src pysandbox

poetry install

This creates the below structure-

➜  pysandbox tree
.
├── README.md
├── poetry.lock
├── pyproject.toml
├── src
│   └── pysandbox
│       └── __init__.py
└── tests
    └── __init__.py

4 directories, 5 files

create a .gitignore file

.venv
.env
**/**/__pycache__/**
*.pyc

💡 Make sure to create this file, else you might have to manually `rm -rf` the pycache directories and then `git commit`

Add pytest for running tests

poetry add pytest --group dev

create a test file test_adder.py for the src/adder.py file

import unittest
from pysandbox.adder import add, addBy100

class AdderTest(unittest.TestCase):

    def test_add_should_add_two_numbers(self):
        a = 20
        b = 40
        self.assertEqual(add(a, b), 60)

    def test_add_by_100_should_add_100(self):
        a = 40
        b = 80
        self.assertEqual(addBy100(a, b), 220)
    

if __name__ == "__main__":
    unittest.main()

we can run pytest in the poetry shell to see the test results.

Using python-dotenv to load ENV variables

poetry add python-dotenv --group dev

create a profile.py

import os
from dotenv import load_dotenv

# load ENV variables using load_dotenv()
load_dotenv()

profile = os.getenv("APP_PROFILE")
print(f"the Python app is for '{profile}'")

create a ENV variable in the .env file

APP_PROFILE=beginner

RUN the program

(pysandbox-py3.11) ➜  pysandbox git:(main) ✗ python src/pysandbox/profile.py

# output
the Python app is for 'beginner'
(pysandbox-py3.11) ➜  pysandbox git:(main) ✗