Rule Request: Enforce Double for function arguments instead of CGFloat

[sindresorhus]

New Issue Checklist

• Updated SwiftLint to the latest version
• I searched for existing GitHub issues

New rule request

1. Why should this rule be added? Share links to existing discussion about what the community thinks about this.

I prefer to only use Double for floats in my own functions, for consistency and ease of use, never CGFloat. I only support 64-bit systems, so CGFloat has no practical benefit for me. However, I often get PRs on my open source projects where they use CGFloat. Would be nice to be able to enforce using Double.

2. Provide several examples of what would and wouldn't trigger violations.

Would trigger:

func foo(number: CGFloat) {}

Would not trigger:

func foo(number: Double) {}

3. Should the rule be configurable, if so what parameters should be configurable?

No

4. Should the rule be opt-in or enabled by default? Why?
   See README.md for guidelines on when to mark a rule as opt-in.

Opt-in

[emmavray]

@sindresorhus is there a reason to target function arguments specifically, and not a CGFloat used elsewhere?

[sindresorhus]

It's the easiest case to target that will never have false-positives. Let's say the rule also supports variable declarations. What if a variable is passed into an Apple API that requires CGFloat? You would have to do appleAPI(number: CGFloat(yourNumber)) at every callsite. While I personally don't mind that for the purity of having Double's everywhere, that's more controversial, and I wanted to get the least controversial thing added first.