Setting the Python path for local project in VS Code without using the settings.json file

Question

This answer from a Visual Studio Code developer says that it is reasonable to keep a Visual Studio Code project's settings.json file in a project's git repository to enforce code standards across different development environments. One down-side though is that when I select the Python interpreter path at the bottom of the screen (so the linter can find the installed packages), the following entry gets added to the settings.json file:

"python.pythonPath": "/path/to/conda/envs/my-env-name/bin/python",

This path is local to the machine and assumes that conda is being used. I have to avoid adding this setting to the settings.json that I commit to the git repository, which is annoying.

Is there a way to set the Python path locally for a project without writing a setting into the settings.json file?

Answer 1

It currently isn't directly supported, but we have a feature request that you can upvote if you would like to see it prioritized.

Answer 2

One workaround would be to:

• remove that line from the local workspace settings.json
• copy-paste it to your User settings.json

See "VSCode User and Workspace Settings".

That way, your Git codebase can keep a generic settings.json without local path.

Since this would work for only one project, you can instead reference all your projects in several Multi-root workspaces.

Then, regarding settings in that environment, you have three files:

• Preferences: Open User Settings - Open your global User settings
• references: Open Workspace Settings - Open the settings section of your Workspace file.
• Preferences: Open Folder Settings - Open the settings for the active folder.

That means you could switch workspaces, and in each multi-root workspace (each one composed of only one root), you would keep:

• the global pythonPath path in user settings (applies to everything, everywhere)
• the specific pythonPath for a given multi-root workspace in the Workspace setting (outside of the project folder which is the only root for that "multi-root" workspace)
• the public versioned settings.json project-specific settings in the project folder (which is the only root of the workspace)

Again, by switching workspace, you can differentiate between:

• workspace-specific private settings, local to your computer, like a pythonPath
• project specific settings, in the project folder, that you can version and publish.