Conda Export Tool

A simple tool that produces environment files for easily sharing a conda environment. This script uses the PyYAML package. Additionally, it was tested on Python 3.9 but I think it should work for any version that supports f-strings.

This project was originally hosted here. I decided to open a separate repository after fixing the bug in this issue.

The three main issues I wanted to solve about the built-in conda env export tool are the following:

- Not including pip installations when using flag `--from-history`
- When using flag `--from-history`, package versions are often not
included (depending on whether the user manually typed them)
- When using --from-history flag, third-party channels are not included
in the file.

I want to be able to do the following:

conda env export --from-history > environment.yml

and have a file that contains strictly the packages I installed (but including those with pip). Additionally, those packages should have the version number that was installed in the environment. This script solves that.

Run the script inside the conda environment you want to export:

python conda_export.py --from-history --use-versions -o environment.yml

to get an output file with only manual imports but include version numbers. If no output file name is provided, the output will be printed to the terminal.

The script also maintains normal Conda's env export functionality:

python conda_export.py -o environment.yml

The command above is equivalent to running:

conda env export > environment.yml

Then:

python conda_export.py --from-history -o environment.yml

is equivalent almost like using:

conda env export --from-history > environment.yml

but including pip packages as well as conda channel.

Credits:

This script uses part of the gist you can find here: https://gist.github.com/gwerbin/dab3cf5f8db07611c6e0aeec177916d8

In particular, I copied the export_env function from the gist.

TODO's

I want to use a newer implementation of the YAML standard so I will look into other packages. Or even better, completely remove the need for a yaml library. Since this script needs to be run inside a conda environment in order to get the information, it would be great if there were no extra libraries needed to be installed. This would prevent polluting the environment with unwanted libraries.

Another option could be to write the functions in a way that the script can be run from any environment, even the base, and still extract dependency information from other environments. Then, one could just keep a separate conda environment just for extracting other environments' info.