ueshin / conda-charmer

conda-charmer

conda-charmer is a package of utilities to manage Python development environment using conda.

Prerequisite

These utilities are using conda>=4.4.

https://conda.io/projects/conda/en/latest/user-guide/install/index.html

• No need to set up environment variables here, which will be handled by these utilities.

Prepare for the Python development

The following example command is assuming the project name is py-proj. Please replace it to your project name.

Download conda-charmer

Download ZIP file for the conda-charmer package and extend as py-proj.

https://github.com/ueshin/conda-charmer/archive/main.zip

or git clone:

git clone [e-mail]:ueshin/conda-charmer.git py-proj

Place the project files

Place your project files into SOURCE_PATH to be set later. The default path is python.

cd py-proj
git clone /path/to/py-proj.git python

Edit _env.sh

Copy _env.sh-template to _env.sh and edit the file corresponding to your local environment.

• CONDA_PROFILE: the conda profile file path.
• CONDA_ENVS: the path to store conda environments for the project. You can share the envs with other projects by the same path.
• SOURCE_PATH: the path to the Python project to be developed. The default is python.
• CONDA_ENV_PREFIX: the prefix for each conda environment.
• ADDITIONAL_PACKAGES: additional packages to be installed by conda/create_env.sh.
• REQUIREMENTS_FILE: the requirement file path relative to SOURCE_PATH.

Commands

conda/list_envs.sh

List all the conda environments.

./conda/list_envs.sh

conda/create_env.sh

Create a new conda environment.

./conda/create_env.sh [python_version] [env_suffix]

• python_version: the default is 3.10. the python version will be used as the suffix of the new conda environment.
• env_suffix: If env_suffix is not specified, python_version will be used.

conda/clone_env.sh

Clone an existing conda environment.

./conda/clone_env.sh from_suffix to_suffix

conda/remove_env.sh

Remove the conda environment.

./conda/remove_env.sh [env_suffix]

• env_suffix: If env_suffix is not specified, 3.10 will be used.

conda/update_conda.sh

Update the conda command.

./conda/update_conda.sh

activate.sh

Activate and run child shell the conda environment and cd to the worktree path.

./activate.sh [-w worktree_path] [env_suffix]

• worktree_path: If -w worktree_path is not specified, the SOURCE_PATH will be used.
• env_suffix: If env_suffix is not specified, 3.10 will be used.

To disable the environment, just exit from the shell.

jupyter.sh

Run jupyter with activating the conda environment.

./jupyter.sh [-w worktree_path] [env_suffix]

• worktree_path: If -w worktree_path is not specified, the SOURCE_PATH will be used.
• env_suffix: If env_suffix is not specified, 3.10 will be used.

Work with git worktree

If you want to use git worktree, I'd recommend to use main for the source path and place the worktrees in worktrees.

cd py-proj
vi _env.sh

export SOURCE_PATH='main'

git clone /path/to/py-proj.git main

(cd main && git worktree add ../worktrees/wt -b wt main)
./activate.sh -w worktrees/wt

For PySpark development

If you are PySpark developer, please use pyspark branch.

https://github.com/ueshin/conda-charmer/tree/pyspark