Poetry-demo
Basic repository exploring poetry for packaging
To install this repo
- Create a virtual environment (conda or otherwise)
- Install poetry into virtual environment
- (Activate environment)
poetry install
python -m pip install -e .
To use poetry on your own
- Create a virtual environment (conda or otherwise)
- Install poetry into virtual environment
- (Activate environment)
poetry add <packages you need>
poetry lock
to pin versionspoetry build
to package source distribution and wheel
Mixing and matching conda and poetry
It looks like you can use poetry commands to install packages into a conda environment, but you cannot use conda to install packages and expect poetry to update the lockfile/toml file accordingly.
The least painful method of converting a conda yaml file may be to poetry add <package
one at a time from a bare python environment that only contains poetry, then letting poetry resolve dependencies, updating the toml/lockfiles along the way.
Installing packages locally
Some cryptic error messages when using pip to install a package locally.
This discussion has some information, but the gist is that most problems with python -m pip install -e .
may stem from your setup.py
, which can be resolved by actually building your package with poetry build
, and extracting the source distribution for a formatted setup.py
.
Using with docker
- Clone repo, change directory into repo
docker build -t poetry .
docker run poetry
- For interactive seession:
docker run -it poetry /bin/bash
- Some basic imports and printing will be performed to test if the package (and environment) are being installed correctly