How to expose environments created with Snakemake

[HealthyPear]

In Snakemake you normally do something like this

rule my_rule:
    input:
        "some/input"
    output:
        "some/output"
    conda:
        "a/conda/env.yml"
    script:
        "some/script.py"

when the workflow is launched a new conda env will be created under .snakemake/conda/xxx where xxx is a hash created by snakemake (kind of a name)

Now I was hoping that doing this in .vscode/settings.json was enough,

{
    ....
    "python-envs.workspaceSearchPaths": [".snakemake/conda/**"],
    ....
}

But then when I e.g. open a jupyter notebook from VSCode I do not see any of those envs (neither as simple envs nor as kernels as all of them have also ipykernel installed)

Am I missing something?

[edvilme]

Hello, thanks for bringing this to our attention. Just to confirm, can you see the environment outside of jupyter notebook?

[HealthyPear]

I can see them in the Environments Manager view under "conda", so from there I can copy the path using the UI symbol, and put that into Select Interpreter ---> Enter interpreter path and only then I will see it when clicking Select Kernel from the Jupyter notebook, but I remember that at some point all these environments where listed from there without me having to add them manually

I am not sure if this is an issue with this extension or the Jupyter one - I see this in the docs, maybe it is related?

[HealthyPear]

just to make my use-case clearer:

I have a bunch of conda envs under .snakemake/conda relative to the workspace root path.

These conda envs do not have names, they get a hash like .snakemake/conda/ef29ae010085770cfafa71e16de6ead5_

I can see them from the environment manager UI so I guess the extension knows about them

The problem, is that I do not see them when I do "Select interpreter" (in there I see a couple of homebrew python intepreters under /opt/homebrew/bin and some associated to my main conda installation in under my home)

And when I open a Jupyter notebook from VSCode I also do not see them from "Select kernel" ---> "Select another kernel" --> "Python environments", but this last point I do not know if it is a problem of this extension or the jupyter one