Is there a quick how to on getting autocompletion to work for Python in VSCode?

You have to enable indexing and then you can configure the index depth if you want to increase that to include nested submodules – see relevant

settings.json

configuration values below. However, depending on the Pulumi provider SDK and the amount of symbols, VScode might hit the 4GB memory limit and blow up. For that reason I haven't been able to increase the indexing depth for

pulumi_azure_native

above 2 which in that case excludes versioned

vYYYYMMDD

API submodules. I also want to set the

diagnosticMode

to analyze the whole

workspace

and not just

openFilesOnly asdf

to show typing errors in my whole project. And even with indexing depth set to 2 levels this will make VScode on my M1 Mac really sweat with noticeable load surges and I have to take care to be specific with imports (always use

from pulumi_x import y

instead of

import pulumi_x

) so it won't trigger a major indexing. Either the IDE indexers need to become much more clever/performant or Pulumi provider SDKs need to be split up.

{
    "python.analysis.indexing": true,
    "python.analysis.packageIndexDepths": [
    {
      "name": "pulumi",
      "depth": 2
    },
    {
      "name": "pulumi_azure_native",
      "depth": 2
    },
    {
      "name": "pulumi_azuread",
      "depth": 2
    },
    "python.analysis.diagnosticMode": "workspace",
}