Hatch Party Tricks

Setting Up

Add hatch autocomplete for oh-my-zsh

1. Run this command

mkdir -p ${ZSH_CUSTOM:=~/.oh-my-zsh/custom}/plugins/hatch-zsh-completion && _HATCH_COMPLETE=zsh_source hatch > ${ZSH_CUSTOM}/plugins/hatch-zsh-completion/hatch-zsh-completion.plugin.zsh

2. Add hatch-zsh-completion to your .zshrc file at plugins= like so

plugins=(git hatch-zsh-completion)

Using environments

Run a script from a non-default environment

hatch run <env-name>:<script-alias>

Deactivate an environment

exit

Enter an environment

hatch -e <env-name> shell

Adding environment variables

[tool.hatch.envs.<env-name>.env-vars]
NAME = "value"

Run with uv as the virtual env backend.
Link to discussion

Using Environment Variables in pyproject.toml with Hatch

When configuring Hatch to use environment variables in pyproject.toml, note that TOML doesn’t directly support variable substitution within strings. To work around this, we can leverage shell commands to access environment variables.

Here's an example setup to pull data from Kaggle, unzip it, and launch a notebook using custom paths:

[tool.hatch.envs.default.scripts]
pull-data = [
  "sh -c 'kaggle datasets download -d $KAGGLE_PATH -p src/demo_supply_chain/data/raw'",
  "unzip src/demo_supply_chain/data/raw/supply-chain-analysis.zip -d src/demo_supply_chain/data/raw",
]
notebook = ["marimo edit src/demo_supply_chain/notebooks/demo.py"]

[tool.hatch.envs.default.env-vars]
DATA_PATH = "src/demo_supply_chain/data/raw/supply_chain_data.csv"
KAGGLE_PATH = "fayez1/inventory-management"