Diagnose invalid characters in bundle identifiers

[franklinsch]

DocC should emit a diagnostic for bundle identifiers that contain characters that aren't valid in URL host (e.g., spaces).

Steps to Reproduce

Create a DocC catalog with a space in its bundle identifier. DocC doesn't produce a diagnostic, even though spaces aren't allowed in bundle identifiers because they're URL hosts.

Swift-DocC Version Information

Swift-DocC version: All

[sevki]

this is also true for other unicode code points, for instance I have a function decl which looks like

public func 𝄞(@PitchBuilder _ makeAbsolutePitches: () -> [Pitch]) -> Track {
    Track(makeAbsolutePitches())
}

it doesn't work if you use

``𝄞(_:)``

docc will complain

/Users/sevki/src/Beato/Sources/Beato/Documentation.docc/Documentation.md:7:45: warning: Link destination '𝄞(_:)' is not a valid URL

which is a bit hand wavy, cause it can resolve that name to https://beato.sevki.dev/documentation/beato/_(_:) when it's enumerating functions.

TD;DR

[franklinsch]

Thanks for pointing this out @sevki! I believe this is an unrelated issue. @d-ronnqvist I believe you investigated something like this in the past?

[d-ronnqvist]

Yes, the (unrelated) issue with unicode code points in symbol links is resolved on the 5.9 branch (which is included in the Xcode 15 betas).

[agisilaos]

@franklinsch @d-ronnqvist I would love to work on this. Can you please give me some first pointers on where to look so I can try and fix the bug?

[agisilaos]

@franklinsch @d-ronnqvist Actually I decided to give it a go! Please let me know what you think: #1012

[d-ronnqvist]

@agisilaos I'm happy to hear that you want to work on this. I didn't have time to provide any first pointers beforehand, but I tried to give some suggestions about where to start in #1012 (review). Don't hesitate to reach out if you have any questions.

[agisilaos]

@d-ronnqvist Thank you so much. I will take a look and fix all the issues you mentioned. Appreciate the thorough review.

[d-ronnqvist]

This is fixed by #1069