Add deletion of stack after destroy (downsert/remove)

[Moon1706]

Hi folks! We've already discussed this feature previously, but I just didn't have time to continue. So, here I am. As I can see, action wasn't super reorganized and appending delete stack flag is relevant. Thus, what do you think about the new field?

[RobbieMcKinstry]

Hi again @Moon1706 . Good to hear from you!
Thanks for this PR. It looks more manageable than the last one which had both downsert and version 👍🏻 (since it was from your upstream fork 😄).

I'm in favor of adding a downsert option. Personally, I don't like the name downsert since it's not clear if it refers to the stack or the state. Others have suggested --rm for the CLI.

However, given that we already have upsert, I think we should go ahead with downsert. We can always deprecate the name later if the CLI supports a different term. I asked around internally to see what other people think.

[RobbieMcKinstry]

Before merging the PR, I'd love it if we could add two tests:

1. Demonstrate that stack.removeWorkspace is called when downsert: true.
2. Demonstrate that stack.removeWorkspace isn't called when downsert is not provided.

Lastly, it would be great if we could log a warning if the user has downsert: true but command is set to anything other than destroy.

[RobbieMcKinstry]

Ah, TIL we recently added support for --remove in the CLI, serving the same function as downsert. Can we stick with that naming?
pulumi/pulumi#10484

[Moon1706]

Ah, TIL we recently added support for --remove in the CLI, serving the same function as downsert. Can we stick with that naming? pulumi/pulumi#10484

Yeah, I've checked the discussion. Mmm, I'm not sure. This option doesn't exist in @pulumi/pulumi/automation.

[RobbieMcKinstry]

@Moon1706 You can still keep the code / logic the same, but could you rename the input? I.e. change downsert in the config to remove.

[Moon1706]

@Moon1706 You can still keep the code / logic the same, but could you rename the input? I.e. change downsert in the config to remove.

Sure) I'm just in UTC+4 time zone, and gonna apply this suggestion tomorrow.

[Moon1706]

@RobbieMcKinstry
I'm wondering maybe we can open an issue in the official Pulumi repo about adding remove argument to the stack file. What do you think about it?

[RobbieMcKinstry]

Hey @Moon1706 sorry for the slow reply, once again.

I think we should move forward with this PR with or without the related PR to automation API. We can always switch to the AutoAPI implementation later once it lands, but there's no need to wait for that.

The only blocker is to rename downsert to remove to be consistent with the CLI.

[RobbieMcKinstry]

Here's a changeset that renames downsert to remove.

[RobbieMcKinstry]

@Moon1706 If these changes look good, I'll commit them and merge this PR.

[Moon1706]

@RobbieMcKinstry
I've added. Check please

[Moon1706]

But yeah, changes LGTM. Will we wait for applying changes to the main Pulumi repo?

[RobbieMcKinstry]

LGTM! I think status checks are failing because there's no changelog entry. That's the last step! (Sorry, I should have mentioned that before...)

[RobbieMcKinstry]

@Moon1706

Will we wait for applying changes to the main Pulumi repo?

No, I think this is good to go! :) Could you add a changelog entry?

[Moon1706]

@RobbieMcKinstry Aahah) My bad) I thought you were talking about the main changelogs of the Pulumi repository. Okay, give me 5 minutes, I'll add it.

[RobbieMcKinstry]

Thanks for another contribution Moon! :)