Fix number formatting

[MrSkwiggs]

I wonder if there are many other places where we use the user's localized number formatting instead of a static en-US format. Will double check the codebase and report back

[MrSkwiggs]

Only places I could find which used the locale-dependant call to .format() where in Slide.swift, LineSpacing.swift & Opacity.swift

[MrSkwiggs]

For now I've implemented static number formatting, but perhaps this could be included in the site configuration and passed along where necessary

[JPToroDev]

Thank you for catching this!

Instead of requiring the use of NumberFormatter, could we create a computed property like oneDecimalPlace in extensions for Percentage, Double, and Float?

[MrSkwiggs]

Instead of requiring the use of NumberFormatter, could we create a computed property like oneDecimalPlace in extensions for Percentage, Double, and Float?

I've implemented it as an extension on FormatStyle directly, much cleaner this way. Also took the liberty of using up to 3 decimal places for opacity, but kept the rest single-decimal precision. Could make the default 3 decimal places always?

[JPToroDev]

I've implemented it as an extension on FormatStyle directly, much cleaner this way.

Awesome!

Could we make the extension a little more concise? nonLocalizedDecimal(_ decimalPlaces: Int = 1) ?

Could make the default 3 decimal places always?

I like the way you have it 🙂

[MrSkwiggs]

Could we make the extension a little more concise? nonLocalizedDecimal(_ decimalPlaces: Int = 1) ?

value.formatted(.nonLocalizedDecimal(3)) isn't a super obvious callsite in my opinion :/
Maybe a good compromise would be value.formatted(.nonLocalizedDecimal(precision: 3)) makes it a bit more concise while still being clear?

Let me know 🙏

[MrSkwiggs]

Also I see that we might be having a race condition since the changes to the PublishingContext were made, as last tests failed with Ignite/PublishingContext.swift:22: Fatal error: "PublishingContext.default accessed before being initialized. Call PublishingContext.initialize() first. Running tests locally a few times I could reproduce this as well. Will make an issue for it 👌

[JPToroDev]

How about length or places then? Would shave off a few more character

[JPToroDev]

Also I see that we might be having a race condition since the changes to the PublishingContext were made, as last tests failed with

Yes, the way we're currently approaching tests doesn't play nice with PublishingContext, but #420 sorts this all out 🙌🏼

[MrSkwiggs]

How about length or places then? Would shave off a few more character

Done, using places 👍

[JPToroDev]

Excellent—thank you!