Proposal: Add tomledit to conda-forge

[fmerinocasallo]

Summary

I would like to propose submitting the tomledit R package to conda-forge to enable easy installation via conda/mamba/Pixi for users working in conda-based environments.

Motivation

• tomledit is currently available on CRAN and can be installed via install.packages("tomledit") and remotes::install_github("extendr/tomledit")
• However, for users working in conda-based environments (e.g., using Pixi, mamba, or conda), having tomledit available through conda-forge would:
  • Provide better integration with conda environment management
  • Enable reproducible environments that include tomledit as a dependency
  • Allow for easier dependency resolution within conda ecosystems
  • Support HPC and scientific computing workflows that rely on conda

Submission Process

The submission would follow the standard conda-forge process for R packages:

1. Generate the recipe using conda_r_skeleton_helper (recommended for CRAN packages)
2. Fork the conda-forge/staged-recipes repository
3. Create a new branch and add the recipe to the recipes directory
4. Open a pull request with the recipe
5. After review and merge, the package will be available via conda install -c conda-forge r-tomledit

For a detailed step-by-step walkthrough of this process, see this guide.

Dependencies

The package may have dependencies that need to be checked:

• R base (required)
• Rust toolchain (if needed for building from source)
• Any R package dependencies listed in the package DESCRIPTION

What's Involved

As a conda-forge maintainer, you would be responsible for:

• Updating the recipe when new versions are released (usually straightforward - just version number and SHA256)
• Responding to occasional issues in the feedstock repository
• Ensuring the package continues to build correctly (conda-forge CI handles most of this automatically)

References

• conda-forge documentation on adding packages
• conda-forge staged-recipes repository
• conda_r_skeleton_helper documentation
• Step-by-step guide: Building a conda-forge package from an R CRAN package

[JosiahParry]

Thanks @fmerinocasallo! I personally have never use conda-forge and likely will not. If you would like to make PRs to this repo to enable the installation of toml-edit in conda-forge, I'm happy to review and approve them. But this added workload is not something I have bandwidth for.