Installation

Anaconda

clu-spacy is written for Python >= v3.8. One option to develop is to install all virtual environment (ex. conda, venv, poetry, etc.). Using conda, the library can be installed interactively with a compatible environment using the following commands:

conda create --name clu-spacy python=3.8 ipython
source activate clu-spacy
# execute the following command from the project root:
pip install -e ".[dev]"

[dev] will include dependencies for running tests and generating the documentation.

Docker

For those familiar with Docker, another option is to use a container with bind mounts as a development environment. Note that the instructions below assume you're developing using a Linux-based environment (they've also been tested on MacOS Catalina).

First, you'll need to build the docker image:

docker build -f Dockerfile -t "parsertongue/clu-spacy:latest" .

Launch a container using this image and connect to it:

docker run -it -v $PWD:/app "parsertongue/clu-spacy:latest /bin/bash"

Thanks to the bind mount, changes made to files locally (i.e., outside of the container) will be reflected inside the running container. The parsertongue/clu-spacy includes Jupyter and iPython:

ipython

from clu.spacy.utils import ConverterUtils

converter = ConverterUtils()

Removing old docker containers, images, etc.

If you want to save some space on your machine by removing images and containers you're no longer using, see the instructions here. As always, use caution when deleting things.