Provision SwiftLint via SwiftPM instead of CocoaPods

[mokagio]

See https://linear.app/a8c/issue/AINFRA-449

Test

Checkout locally and run bundle install && bundle exec pod install && bundle exec rake lint

You should see something like, possibly with additional logs such as removing SwiftLint from your pods.

Note

The UI tests failure is a long standing issue and unrelated with the changes in this PR.

Review

(Required) Add instructions for reviewers. For example:

Only one developer required to review these changes, but anyone can perform the review.

Release

These changes do not require release notes.

[wpmobilebot]

📲 You can test the changes from this Pull Request in Simplenote Prototype Build by scanning the QR code below to install the corresponding build.

	App Name	Simplenote Prototype Build
	Build Number	1285
	Version	PR #1712
	Bundle ID	com.codality.NotationalFlow.Alpha
	Commit	70de778
	Installation URL	6qvpgbvottb8g

Automatticians: You can use our internal self-serve MC tool to give yourself access to those builds if needed.

[AliSoftware]

Approving to unblock and because it worked on my machine, but left a note about running rake lint from the Xcode shell script build phase and its potential issues depending on developers' setup

[AliSoftware]

I was wary of if running rake from an Xcode build phase would work as expected, given that:

• When Xcode runs shell scripts it runs them in a separate shell that might not have the same env / settings as your Terminal.app's shell (e.g. $PATH and where rake is installed)
• That shell script runs in /bin/sh, while macOS users are using zsh, so all their customization and setup of their shell is in .zshrc and not .bashrc. So again, any custom $PATH or rbenv and bundle-related config etc might be set up properly in .zshrc but not in .bashrc for the /bin/sh shell to pick up, which could lead to this rake lint command to fail
• Notice how you run rake lint, not bundle exec rake lint. Which means one might accidentally run this command with a different version of rake than the one in the Gemfile.lock bundle of this repo. Probably not a big deal (the most important thing is for swiftlint to be using the right version, and rake doesn't have that many updates anyway, and if it does, nothing likely breaking anything) but still inconsistent.

---

It still worked when I tested building the project with Xcode on my Mac:

But I think that might be a happy coincidence in how my Mac is set up, and that might not be the case for every developer? I wonder if switching to shellPath = /bin/zsh; and using bundle exec rake lint wouldn't be better (maybe with a need to source ~/.zshrc explicitly first, in case Xcode running this shell in non-interactive mode would prevent that from loading… not sure, always confused with the rc files and when they're loaded wrt to interactive vs non-interactive shells etc)

[mokagio]

I remember looking into this with Jeremy a while back and coming to the conclusion that calling rake from Xcode works because of a number of happy coincidences in how Apple set up Ruby in macOS.

In the long run, consistently with our goal of removing Ruby from day-to-day DX, I think we'd be better off using make which we know would be consistent without depending on Ruby setup.

[mokagio]

Moved this discussion internally to paaHJt-8yZ-p2